report no. 39736-ni nicaragua poverty assessment · 2016. 7. 12. · 2 figure 1.2. headcount...

May 30, 2008

Document of the World Bank

Report No. 39736-NI

NicaraguaPoverty Assessment

(In Three Volumes) Volume II: Background Paper

Central America Country Management UnitPoverty Reduction and Economic Management SectorLatin America and the Caribbean Region

Report N

o. 39736-NI N

icaragua Poverty A

ssessment

Vol. II

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

TABLE OF CONTENTS

1. Poverty Profile Gabriel Demombynes

2. Qualitative Poverty – Voices

Ximena del Carpio and Vanessa Castro 3. Labor

Catalina Gutierrez and Marco Ranzani 4. Migration

Catalina Herrera y Edmundo Murrugarra 5. Attaining the MDGs

Leopoldo Lopez 6. Inequality of Opportunity in Education

Diego Angel-Urdinola and Jose Ramon Laguna 7. Inequality of Opportunity in Preventative Health

Diego Angel-Urdinola and Kimie Tanabe 8. Inequality of Opportunity in Health – Water and Sanitation

Simon Zbinden and Diego Angel-Urdinola 9. Inequality of Opportunity in Health – Malnutrition

Janet Irene Picado, Rafael Flores and Jose Ramon Laguna 10. Inequality of Opportunity in Income Generation

Diego Angel-Urdinola,Ezequiel Molina and Maria Victoria Fazio

1

1. POVERTY PROFILE OF NICARAGUA

By Gabriel Demombynes*

This paper presents a profile of poverty in 2005, analyzes changes in poverty and inequality from 1993-2005, examines the relationship between poverty and growth over the same period, and explores the dynamics of poverty changes at the household level. Given the rich nature of survey data available for several recent years in Nicaragua, the focus is on changes over time rather than the static picture of poverty. The main emphasis is on changes between 1998-2005; period for which three comparable household surveys are available. Basic poverty numbers from 1993 are also presented, although these figures are not strictly comparable to those from later years.

THE EVOLUTION OF POVERTY AND INEQUALITY 1993-2005

Nicaragua presents a classic case where looking only at the most commonly used poverty indicator—the fraction of the population living below the poverty line, i.e. the headcount—would offer a highly incomplete understanding of changes that have taken place over time.

The country as a whole saw essentially no change between 1998 and 2005 in the percentage of Nicaraguans living in moderate poverty, as measured using the standard consumption-based general poverty line. Figure 1.1 shows that while rural dropped between 1993 and 1998, for both urban and rural areas, the estimates of the headcount declined only slightly between 1998 and 2005. None of the 1998-2005 changes are statistically significant. In 2005, 46 percent of Nicaraguans were living in poverty, including 68 percent of those in rural areas and 29 percent of urban Nicaraguans.

* The author is with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. I wish to thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), Jaime Saavedra (Sector Manager, LCSPP) and Norman Hicks (Consultant) for their valuable comments and suggestions. The views expressed here are those of the author and need not reflect those of the World Bank, its Executive Directors, or the countries they represent.

Figure 1.1. Headcount Poverty Rates by Area 1993-2005 (General Poverty Line)

50

32

76

48

31

69

46

30

68

46

29

68

0%

10%

20%

30%

40%

50%

60%

70%

80%

All Nicaragua Urban Rural

1993 1998 2001 2005

Source: Own analysis of EMNV data.

2

Figure 1.2. Headcount Poverty Rates by Area 1993-2005 (Extreme Poverty Line)

19

7

36

17

8

29

15

6

27

15

5

27

0%

5%

10%

15%

20%

25%

30%

35%

40%


1993 1998 2001 2005


At the national level, since 1998 there has been a statistically significant drop1 in the fraction of Nicaraguans living in extreme poverty, from 17 percent to 15 percent (see Figure 1.2.) Although there have been declines in the point estimates of rural and urban extreme headcount, these changes are not statistically significant.

Despite the minimal movement in overall poverty as measured by the headcount, there have been more substantial declines in inequality. The Gini coefficient for the country as a whole has dropped from 0.45 to 0.40 over the period 1998-2005, continuing its fall from 0.49 in 1993, and urban and rural areas separately have seen similar drops (see Figure 1.3.)2 The sources of this change in inequality are discussed later in this paper.

Paired with the declines in inequality, there have been large declines in the poverty gap and particularly the extreme poverty gap. The poverty gap is an index which measures the average distance, or “gap,” between the

consumption level of the poor and the poverty line. The index averages the gap over the entire population and takes it as a percentage of the poverty line. Declines in the poverty gap can be driven by a drop in the fraction of the population that is poor (the headcount) and also by increases in the average level of consumption among those who are poor. The extreme poverty gap is simply the poverty gap using the extreme poverty line.

1 At the 10% level. 2 Note that because the 1993 consumption aggregate is not identical to that used in later surveys, comparisons between 1993 and later years should be taken as only suggestive, particularly for measures like the Gini coefficient that are sensitive to the entire distribution of consumption.

Figure 1.3. Inequality 1993-2005

0.490.45

0.430.45 0.44

0.37

0.430.41

0.35

0.400.38

0.34

0.00

0.10

0.20

0.30

0.40

0.50

0.60

National Urban Rural

Gin

i coe

ffici

ent

1993 1998 2001 2005


3

Figure 1.4 shows the large declines in the extreme poverty gap. For the nation as a whole, the poverty gap has declined from 4.8% to 3.4%. This represents a decline of nearly 30 percent. Likewise, the extreme poverty gap for urban areas dropped almost by half from its already low level of 1.9% to just 1.0%. What this indicates is that while 15 percent of Nicaraguans still live in extreme poverty, the depth of their poverty is notably less than it was in 1998.

Next we return to the most commonly cited poverty measure, the headcount, to consider the regional distribution of poverty using both the general and extreme poverty lines. Overall poverty patterns for the country are

summarized in Table 1. As in the country as a whole, the headcount using the general poverty line has not declined substantially in most regions. Exceptions are the Pacific Rural and Atlantic Urban regions. The extreme poverty headcount has fallen in a wider set of places; the fraction living below the extreme poverty line has dropped in rural and urban areas of both the Pacific and Atlantic regions. Similar patterns by region are observed for the poverty gap (estimates by region of the poverty gap and also the poverty severity index can be found in Annex 3 to the Main Report).

Table 1.1. Poverty Patterns in Nicaragua in 2005

% of NationalPopulation All Poor Extreme Poor All Poor Extreme Poor

All Nicaragua 100.0 46.2 14.9 100.0 100.0By Geographic Area

Urban 55.8 29.1 5.4 35.1 20.1Rural 44.2 67.9 26.9 64.9 79.9

By RegionManagua 24.5 19.5 3.4 10.3 5.5Pacific Urban 16.9 35.9 4.8 13.2 5.4Pacific Rural 12.4 58.2 17.0 15.6 14.2Central Urban 12.3 37.9 10.5 10.1 8.6Central Rural 19.8 74.4 32.9 31.9 43.9Atlantic Urban 4.4 34.8 7.4 3.3 2.2Atlantic Rural 9.6 74.9 31.2 15.6 20.2

By Gender of Household HeadFemale 29.6 39.9 11.8 25.5 23.5Male 70.4 49.0 16.2 74.5 76.5

Poverty Headcount Index (% of Pop.)

Contribution to National Poverty (% of Category)


Figure 1.4. Poverty Gaps by Region 1998-2005 (Extreme Poverty Line)

4.8

1.9

8.3

4.1

1.5

7.8

1.0

6.4

3.4

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%


Pov

erty

gap

1998 2001 2005

Source: Own analysis of EMNV data

4

How do drops in poverty in particular regions affect the national poverty rate? The answer depends on the relative sizes of different population groups. Table 1.2 shows a decomposition of the changes in the extreme poverty headcount along with the shares of the population in each region, in 1998 and 2005.3 This decomposition splits the total drop in the extreme poverty headcount into components due 1) to poverty changes within each region, 2) shifts of population between richer and poorer regions, and 3) a residual “interaction effect.” With this sort of decomposition, it is possible to pinpoint which regional changes in poverty were most important for the overall drop. Recall that this decomposition concerns the drop in the national extreme poverty headcount from 17.3 percent to 14.9 percent, a drop of 2.4 percent points.

The figures in the table show that declines in poverty in the Pacific Urban and Pacific rural regions explained 35 and 45 percent, respectively, of the national drop in extreme poverty. Declines in the Atlantic region were also important for the national poverty rate, although the region is home to only a small fraction of the country’s population. Due to their large drops in extreme poverty, declines in the Atlantic Urban and Atlantic Rural regions accounted for 20 and 22 percent, respectively, of the national extreme poverty decline.

The total changes in extreme poverty within-region account for 124 percent of the drop in poverty at the national level. This is because the “population-shift effect” tended to increase extreme poverty. The share of the population living in the Atlantic Rural region, one of the poorest parts of the country, increased. Consequently, even though extreme poverty dropped in the region, the increase in the region’s population share worked to increase national poverty.

Growth and Poverty 1998-2005

This section examines the relationship between growth and poverty over time. Growth incidence curves (GIC) are a useful tool for examining graphically the impact of growth on poverty. A GIC is a plot of the growth rate for each decile (or other quantile division) of the distribution of per capita consumption. The curves are constructed based on the two household surveys, and the annual growth rates reflect average changes over the period. Note that the vertical scales differ for the different curves.

We examine changes over the 1998-2005 period. The horizontal scale shows percentiles within each area—national, urban, rural, Managua—so points at the same percentile different sectors correspond to different levels of consumption. To make this clear, the position of the poverty line in 2005 is shown in each graph as a dashed line at each figure. As a result, the portion of the figure to the left of the dashed

3 This follows the procedure of Huppi and Ravallion (1991).

Table 1.2. Decomposition of the Change in the Extreme Poverty Headcount, 1998-2005: Percentages of National Change Due to Changes Within Region and Population Shifts Between Regions

% of Change in National Headcount

Within-region changesManagua -3%Pacific Urban 35%Pacific Rural 45%Central Urban 8%Central Rural -2%Atlantic Urban 20%Atlantic Rural 22%

Total within-region change 124%

Population-shift effect -33%Interaction effect 8%Total 100%


5

line shows the pattern of growth among the poor. The extreme poverty line is shown as a dotted line in the national and rural GICs. (The extreme poverty line is not shown on the Managua and urban plots. Because extreme poverty is very low in those areas, the extreme poverty line would appear at the far left edge of the graphs.) The figures also show 95 percent confidence intervals for the curves.

Figure 1.5. Growth Incidence Curve 1998-2005: National

Figure 1.6. Growth Incidence Curve 1998-2005: Urban

-4-2

02

% a

nnua

l cha

nge

in p

er c

apita

con

sum

ptio

n

0 10 20 30 40 50 60 70 80 90 100Percentiles of consumption per capita

-4-2

02

% a

nnua

l cha

nge

in p

er c

apita

con

sum

ptio

n


Source: Own analysis of EMNV data. Note: The dashed vertical line indicates the normal poverty line in 2005. The extreme poverty line is not shown.


For urban and rural areas separately and for the nation as a whole, the GICs, shown in Figures 1.5, 1.6, and 1.7, are downward sloping, which corresponds to decreases in inequality, as seen in the previous section. The pattern in the GICs is the result of declines in consumption among the middle and upper quintiles and gains for the poorest quintiles. Note that at the national level, gains in consumption took place only for the population below the extreme poverty line.

Figure 1.7. Growth Incidence Curve 1998-2005: Rural Figure 1.8. Growth Incidence Curve 1998-2005: Managua only

-4-2

02

4

% a

nnua

l cha

nge

in p

er c

apita

con

sum

ptio

n


-10

-50

5

% a

nnua

l cha

nge

in p

er c

apita

con

sum

ptio

n


Source: Own analysis of EMNV data. Note: The dashed vertical line indicates the normal poverty line in 2005, and the dotted vertical line is at the point of the extreme poverty line in 2005.


6

The pattern for Managua (Figure 1.8) is different: the growth incidence curve is essentially flat and negative. In other words, the best estimate provided by the survey data is that consumption dropped slightly for the typical household in Managua, with similar percentage declines across the distribution.

As the national growth incidence curve, the decline in inequality has been driven by two factors: an increase in consumption levels of the poor and particularly the extreme poor, and a sharp fall in consumption at the top of the distribution. Of these two factors, the drop at the top is by far the most important to the drop of the Gini coefficient. This can be illustrated by the fall in the “trimmed Gini” estimated by dropping the top 10 percent of the population. This trimmed Gini fell from just 0.31 to 0.30 between 1998 and 2005.

It is important to recognize that the GICs are based on cross-sectional data, not panel data which tracks individuals over time. As a result the GICs do not reflect changes in consumption for particular households. In a later section of this paper, we consider the pattern of changes using the panel data and compare those to the cross-sectional changes shown by the GICs.

Growth and Poverty Elasticities and Reaching the Millennium Development Goal for Extreme Poverty Reduction

Figure 9 shows the pattern of slow declines in poverty and the annual growth rate of GDP since 1995. Although growth per capita has been somewhat erratic, it has been positive in every year except 2002. This might lead one to wonder why poverty has not declined more, particularly in light of the substantial decline in inequality.

Figure 1.9. Poverty Declines and GDP Growth in Nicaragua

Nicaragua: GDP growth (left axis) and Poverty 1993, 1998, 2001 y 2005 (right axis)

-4%

-2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

GDP Growth

Per Capita GDP Growth

Total Poverty

Headcount

Extreme Poverty

Headcount


As Figure 1.10 shows, however, the experience in Nicaragua is broadly similar to that observed in other Latin American countries in recent years. Growth of roughly one percent of GDP, like that Nicaragua has experienced since 2001, has been associated with no change in poverty. Only higher levels of growth, like Nicaragua experienced over the longer period 1993-2005, have resulted in drops in poverty.

7

Figure 1.10. Poverty Declines and GDP Growth in Latin America

ARG

CHL

COL

CRI

ECU

HND

JAM

MEX

NICARAGUA 1993-2005

PRY

URY

NICARAGUA 2001-2005

-5-4

-3-2

-10

12

34

56

78

910

Ann

ual %

Cha

nge

in H

eadc

ount

-1 0 1 2 3 4

Annual % Change in GDP per Capita

(Official Moderate Poverty)

Latin America: Growth and Poverty 1993-2005

Source: World Development Indicators and SEDLAC database.

We can also quantify the simple relationship between growth and poverty using a poverty-growth elasticity. This is calculated by examining changes between two points in time. It is simply the annual percentage change in the poverty rate divided by the annual growth rate of GDP per capita. The elasticity can be interpreted as the change in the poverty rate that takes place for each 1 percent of growth. Poverty-growth elasticity estimates calculated in this way do not offer a clear prediction as to the future relationship between poverty and growth. Instead, they offer a concise description of how the two have been related in the past. Note that in general we expect such elasticities to be negative, as an increase in GDP per capita is typically associated with a decrease in poverty rates.

These elasticities are most informative when calculated over a longer period. Elasticities calculated over short periods may be substantially driven by short-term variations in output and poverty rates and as a consequence are less helpful as a guide to the likely future path of poverty reduction.

Table 1.3 shows elasticities calculated for both general and extreme headcount poverty over various periods. The most important figure is the elasticity measured over 1993-2005, the longest period for which poverty data is available. Over this period, the general poverty elasticity was -0.4 and the extreme poverty elasticity was -1.1. Given 2005 poverty headcount of 46.2% , the general poverty elasticity of -0.4 implies that for each one percentage point of growth of GDP per capita, overall poverty would decline by 1% * 0.4 * 0.462 = 0.18 percentage point. Likewise, given the 2005 extreme poverty level of 14.9%, the extreme poverty elasticity of -1.1 indicates that for each point of growth of GDP per capita, extreme poverty would drop by 1% * 1.1 * 0.149 = 0.16 percentage point. Even though poverty showed little change over the most recent short period (2001-2005)

Table 1.3. Poverty Headcount Elasticities with Respect to Growth: 1998-2005

Period

General poverty elasticity

Extreme poverty elasticity

Short Term1998-2001 -0.5 -1.52001-2005 0.2 -0.3

Long Term1993-2001 -0.5 -1.31998-2005 -0.2 -1.01993-2005 -0.4 -1.1


8

these long-term elasticities indicate that Nicaragua has the potential to substantially reduce poverty over the long term if it can maintain high rates of growth.

How does the effectiveness of growth in reducing poverty in Nicaragua compare with the experience of other countries in the region? Table 1.4 shows estimates of the poverty-growth elasticities for all countries in the region for which multiple years of poverty estimates are available. Note that these estimates employ all available poverty data from household survey. Thus they represent the best long-term estimates of growth’s effect on poverty in each country.4 On average in the region (not weighting countries by population), the elasticity of moderate poverty with respect to growth has been -0.9, while for extreme poverty the elasticity has been -1.5. This indicates that the poverty-reducing power of Nicaragua’s growth has been relatively low compared to other countries in the region.

Keeping in mind that elasticities do not provide a clear guide to the future, we can consider what the likely future path of poverty reduction will be if such elasticities continue to hold. Specifically, what level of growth would Nicaragua need to achieve its Millennium Development Goal for extreme poverty, reducing its level to 9.7 percent by 2015? Assuming population growth and the extreme poverty elasticity remains at its long-term value of -1.1 (observed over 1993-2005), Nicaragua will need GDP growth averaging 5.5 percent per year between 2005 and 2015 to reach its extreme poverty MDG.5

Correlates of Poverty

A useful way to examine changes in the salient features of poverty is through regression analysis. Specifically, we seek to explain variation in household consumption per capita with a variety of variables. This type of analysis is best understood as one of correlates of economic welfare, although we may also cautiously interpret observed relationships as stemming from causal relationships. We restrict the set of explanatory variables to a fairly limited set: geographic location, education and demographics of the household head, employment status and sector of household head, household composition, and access to infrastructure.

Table 1.5 displays results from this analysis. The regional variables shown are all with reference to Managua. The results show that as a whole, consumption levels of other regions and Managua have moved towards convergence. In other words, the gaps between other regions and Managua have declined. Most strikingly, controlling for other variables, consumption levels in the Rural Atlantic region, which were 30 percent those of Managua in 1998, were equal to those of the capital in 2005.

Overall, the relationship between consumption and the main explanatory variables has remained remarkably constant over time. In all three years, female-headed households were no poorer than male-

4 In the case of some countries, poverty estimates from only two different years are available. Consequently the length of the “long-term” varies by country. 5 This calculation is for overall (not per capita) GDP growth, assuming continued population growth of 1.7 per year.

Table 1.4. Poverty Elasticities for Countries in Latin America and the Caribbean

Moderate Poverty

Extreme Poverty

Argentina -1.0 -2.1Bolivia -0.5 -Brazil -1.7 -2.0Chile -1.3 -1.9Colombia -0.4 -1.7Costa Rica -1.1 -1.5Ecuador 0.6 -1.6Honduras -1.4 -1.7Jamaica -1.2 -Mexico -0.6 -1.0Nicaragua -0.4 -1.2Panama -0.1 -1.4Peru 1.0 0.7Paraguay -2.5 0.3El Salvador -2.3 -3.1Uruguay -2.4 -3.6Venezuela -0.9 -1.5

Average -0.9 -1.5 Source: Own analysis based on SEDLAC database poverty figures and World Development Indicator GDP per capita numbers.

9

headed households, while those with younger household heads (under age 35) were 9-13 percent poorer. Education levels are strongly and consistently associated with higher household consumption. Completion of primary and secondary education for the household head is associated with consumption gains of 17 and 36 percent, respectively, over a household with a head who has not completed primary.

The correlation between consumption and the household head’s sector of activity is surprisingly weak. Only having a head employed in commerce, transport, or financial services was strongly associated with higher consumption. These relationships changed only slightly over time.

Household access to services is consistently associated with higher consumption. Households with piped water, electricity, and paved roads are significantly better off. Note that these may not reflect the effects of access to services but rather the fact that better off households are more likely to be able to afford utilities and to be located closer to paved roads. The link between consumption levels and paved roads has actually weakened over time. Households with paved roads were on average 22 percent wealthier in 1998 and only 11 percent wealthier in 2005. This probably is due to the massive expansion of paved roads that took place after 2001, which expanded paved roads into areas that are not as well off.

The results for household composition show that larger households are less well off in general, particularly those that have more children and babies. This is unsurprising, because young children consume household resources but are not productive themselves. However, even households with more seniors and adults—who are both consumers and producers—have lower consumption levels on a per capita basis.

10

Table 1.5. Correlates of Consumption in Nicaragua: 1998-2005 1998 2001 2005

RegionUrban Pacific -0.28 -0.19 -0.20Rural Pacific -0.27 -0.11 -0.15Urban Central -0.22 -0.16 -0.15Rural Central -0.33 -0.23 -0.22Urban Atlantic -0.03 0.03 0.07Rural Atlantic -0.30 -0.04 -0.03

Household headFemale -0.03 -0.03 -0.01Under age 35 -0.13 -0.13 -0.09Primary education 0.14 0.14 0.17Secondary education 0.37 0.37 0.36More than sec. education 0.86 0.82 0.87Not in labor force 0.07 0.09 0.10

Household head sectorAgriculture 0.09 0.08 0.06Mining 0.08 -0.04 -0.08Manufacturing 0.02 0.04 0.03Gas, Elec, Water 0.08 0.11 0.10Construction 0.04 0.01 0.00Commerce 0.17 0.18 0.18Transport 0.30 0.27 0.17Financial Services 0.22 0.24 0.14Community Services 0.04 0.02 0.00

Household servicesPiped Water 0.17 0.18 0.19Electricity 0.22 0.23 0.21Paved Road 0.22 0.19 0.11

Household composition# babies (under 5) -0.17 -0.15 -0.16# children (5-14) -0.14 -0.14 -0.14# adults -0.05 -0.06 -0.07# seniors -0.10 -0.04 -0.06Constant 9.34 9.10 9.10

Number of observations 3827 4165 6856R-squared 0.56 0.57 0.55 Source: Own analysis of EMNV data. Notes: Results shown are coefficient estimates from regressions with log per capita consumption as the dependent variable. Observations are at the household level, and household weights were used for the analysis. Estimates significant at the 5% level are shown in bold (robust standard errors were calculated taking into account the two-stage sampling for the surveys.) Omitted dummy categories correspond to a household in the Rural Central region with an unemployed head with no education.

11

Income

Nicaragua follows international best practice and estimates poverty levels based on consumption data rather than income data. Consumption data is preferred over income data for this purpose for a variety of reasons (see Box 1.1 for discussion.) While consumption and income are correlated, the two do not necessarily move together, either for individual households or for the country as a whole. Figures 1.11 and 1.12 show scatter plots of consumption versus income in 1998 and 2005. For points above the straight, 45 degree line, consumption is higher than income, while income exceeds consumption for points below the line. The curved line is a quadratic best-fit, showing the average relationship. The figures show that on average poorer households have consumption that exceeds their income, while wealthier households have greater income than consumption. This makes sense because households with low income are often those that are experiencing transitory hardship. Consequently, they are using up their savings and consuming more than they have in income.

Box 1.1. Why Measure Poverty with Consumption Instead of Income?

Consumption is preferred over income as a measure of household welfare for several reasons. First, consumption tends to be less variable than income over the course of time (due to consumption smoothing) and thus provides a better measure of long-term welfare. Second, household surveys in developing countries typically measure consumption more accurately than income. Third, consumption of the household’s own production, which is often a large portion of consumption for agricultural households, is usually not captured well (if at all) in income data. Ignoring home-produced food would greatly understate the consumption levels of rural households.

Figure 1.11. Scatter Plot of Consumption vs.

Income for Households in 1998 Figure 1.12. Scatter Plot of Consumption vs.

Income for Households in 2005

68

1012

Log

cons

umpt

ion

per

capi

ta

6 8 10 12Log income per capita

68

1012

Log

cons

umpt

ion

per

capi

ta

6 8 10 12Log income per capita

Consumption and income show different patterns over time in Nicaragua. Average levels of consumption have fallen in household surveys since 1998. Between 1998-2001, consumption fell by 3.4 percent annually, and it continued to fall at a 0.4 percent annual rate between 2001 and 2005. During the same period, the country experienced modest but substantial growth in GDP per capita. While consumption and GDP per capita diverged, mean income (in survey data) grew by 2.1 percent between 1998-2001 and 2.3 percent 2001-2005. These growth rates are summarized in Figure 1.13.

12

Figure 1.13. Growth Rates of Consumption, Income, and GDP per Capita

Real Growth Rates of Consumption, Income, and GDP

3.0%

1.4%

3.0%

1.6%2.1% 2.3%

-3.4%

-0.4%

-4%

-3%

-2%

-1%

0%

1%

2%

3%

4%

1998-2001 2001-2005

Ann

ual G

row

th R

ate

(%)

GDP pc (National Accounts) Consumption pc (National Accounts) Income pc (Survey) Consumption pc (Survey)


The divide between income and consumption also holds for all regions of the country, as shown in Figure 1.14. Between 1998 and 2005, mean income grew in every region. Consumption grew slightly in the Rural Pacific region, was stagnant in both the Urban and Rural Atlantic regions, and fell everywhere else.

How can income be increasing while consumption is declining? As Figure 1.15 shows, this pattern is consistent across the distribution. Figure 1.15 displays the same consumption-based growth incidence curve shown earlier in the chapter, along with a similar curve calculated using income data. For all deciles of the income distribution, income grew between 1998 and 2005. This means that

income has increasingly been channeled into non-consumption activities, which include cash savings and investment.

Figure 1.14. Annual Growth Rates of Income and Consumption, 1998-2005

-4%

-3%

-2%

-1%

0%

1%

2%

3%

4%

5%

6%

Managua UrbanPacific

RuralPacific

UrbanCentral

RuralCentral

UrbanAtlantic

RuralAtlantic

Income GrowthConsumption Growth

Source: Own analysis of EMNV data. Note: Growth rates shown are for mean per capita income and consumption, as measured in the household surveys.

13

How can income be increasing while consumption is declining? As Figure 1.15 shows, this pattern is consistent across the distribution. Figure 1.15 displays the same consumption-based growth incidence curve shown earlier in the chapter, along with a similar curve calculated using income data. For all deciles of the income distribution, income grew between 1998 and 2005. This means that income has increasingly been channeled into non-consumption activities, which include cash savings and investment.

While the nature of the relationship between income and consumption has evidently shifted over time, the composition of income at the

household income has remained fairly stable. Figure 1.16 shows a breakdown of household income by major sources in the three survey years. On average, the profiles of household income sources look very similar in 2005 and 1998. Of the five major categories, non-agricultural wages is the largest source of income for the average household, followed by agriculture and a broad category of “other” sources, followed by non-agricultural self-employment and finally remittances.

Figure 1.16. Sources of Household Income by Year

020

4060

8010

0%

1998 2001 2005

Remittances

Other

Non-ag self-employment

Non-ag wages

Agriculture

Source: Own analysis of EMNV data. Note: Percentages shown are averages of percentages across households, weighted using household sampling weights.

Figure 1.15. Growth Rates of Consumption, Income, and GDP per Capita, 1998-2005

-.05

0.0

5.1

% a

nnua

l cha

nge

in p

c in

com

e/co

nsum

ptio

n

0 2 4 6 8 10Deciles of income/consumption per capita

Consumption Income

14

In Figure 1.17, we look at household income by quintile in 2005. Among the poorest twenty percent of households—those in the bottom quintile—the largest source of income is agriculture. For wealthier households, wages and self-employment income are more important. As a percentage of household income, remittances are roughly of equal importance for households across the distribution of income.

Figure 1.17. Sources of Household Income by Quintile, 2005

020

4060

8010

0%

1 2 3 4 5Consumption Quintile

Remittances

Other

Self-employment, non-ag.

Wages, non-ag.

Agriculture


Next, we consider more closely changes in the composition of income for the very poorest Nicaraguans, those in the bottom 10 percent of the distribution of consumption. Figure 1.18 shows the composition of household income for this group. Note that agriculture has grown in importance for this group, rising from 50 to 60 percent of income. This suggests that the rise in income, and thus consumption, of the poorest may be due in part to an increase in agricultural earnings.

15

Figure 1.18. Sources of Household Income for Poorest 10%: 1998, 2001, and 2005

020

4060

8010

0%

1998 2001 2005Year

Remittances

Other


Wages, non-ag.

Agriculture


Figure 1.19. Sources of Household Income for 2nd Poorest 10%: 1998, 2001, and 2005


0

20

40

60

80

100

%

1998 2001 2005Year

Remittances

Other


Wages, non-ag.

Agriculture

16

SOURCES OF CHANGES IN INEQUALITY

A useful way to consider the sources of changes in inequality is by decomposing the Gini Coefficient of income inequality by income source. This decomposition is based on Shorrocks (1982). Full details of the decomposition are shown in Annex 3. Note that the analysis presented at the beginning of this paper focuses on changes in the Gini using consumption data, while this decomposition is necessarily using income data. The Gini coefficient based on income was 0.54 in 1998, 0.55 in 2001, and 0.2005. Although unlike the consumption-based Gini, the income Gini showed a slight (statistically insignificant) increase between 1998 and 2001, the overall pattern of a substantial drop in the Gini 1998-2005 is the same for both consumption and inequality.

With the decomposition, we can consider the role of each source of income in overall equality. For this decomposition, income is broken down into the finest categories available in the household survey. First, we see that inequality is determined mostly by non-agricultural wages, non-agricultural self-employment. Figures in the Annex show that between 1998 and 2005, income inequality declined for these two sources, which contributed to an overall decline in inequality. In other words, income from these sources became less concentrated among the better-off. At the same time, the share of each of these sources in overall inequality dropped (as shown in the last column.) This happened because the relative share of agriculture in inequality increased.

At the same time, agriculture's share in total national income grew (from 15-21 percent). Because agricultural income goes to the poor more than other main income sources, it is strongly inequality decreasing (on average). So the overall growth in agricultural income explains much of the decline in the Gini.

Table 1.6. Decomposition of Gini Coefficient by Year

1998 2001 2005Change

1998-2001Change

2001-2005Change

1998-2005Income SourceAg wages 3% 1% 2% -2% 1% -1%Non-ag wages 42% 39% 38% -3% -1% -4%Non-ag self-employment 27% 37% 24% 9% -13% -3%Ag self-employment 4% 7% 14% 3% 7% 10%Imputed value of housing 12% 10% 11% -2% 1% 0%Education transfers 0% 0% 0% 0% 0% 0%Gifts of food 0% 0% 0% 0% 0% 0%Remittances 7% 2% 6% -5% 4% -1%Institutional donations 0% 0% 0% 0% 0% 0%Returns to capital 3% 2% 2% -1% 0% -1%Pensions 1% 2% 2% 1% 1% 1%Other 0% 1% 1% 0% 0% 1%Undefined sources 1% 0% 0% -1% 0% -1%Total 100% 100% 100% 0% 0% 0% Source: Own analysis of EMNV data

Basic Needs Indicator

Basic Needs Indicators (BNI) is sometimes used as an alternative to income or consumption-based measures to examine changes in welfare over time. Such indicators typically aggregate several subcomponents. The Nicaraguan government has a BNI based on five individual measures, related to household economic dependency, crowding, water and sanitation, housing materials, and school

17

attendance. A household is deemed to have its basic needs “satisfied” in a particular area if its conditions meet a particular set of criteria; in other words, each subcomponent has a value of either one or zero for each household. In Nicaragua, the five subcomponents are sometimes they are added together to produce a single index. The single index, however, is less informative, than the five individual measures.

Figure 1.20 shows changes over time in four of the subcomponents over time. (The fifth subcomponent, economic dependency, is not shown; its value has remained essentially constant at 90 percent over time.) Note that the figure shows values calculated from both censuses and surveys, and due to differences in the wording of questions, the values are not fully comparable between the two data sources. All four subcomponents show steady improvement over time.

Figure 1.20. Basic Needs Index Components, 1998-2005 (National Level)

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Crowding Water andSanitation

HousingMaterials

SchoolAttendance

% w

ith N

eed

Uns

atis

fied

1995 Census 1998 Survey 2001 Survey 2005 Survey 2005 Census


18

Dynamic of Economic Welfare

Finally, we consider the dynamics of welfare using household panel data. A portion of the households visited in the 1998 survey were reinterviewed in 2001 and 2005, which makes it possible to track their welfare over time. It is important to recognize that the households in the panel have different characteristics, on average, than those in the population as a whole. In particular, because the survey is one of dwellings, not households per se, households were only followed over time if they did not move during the 1998-2005 period.6 This is evident from Figure 1.21, which shows a growth incidence curve (GIC), like that displayed earlier, but using only the panel data. For all parts of the distribution, the GIC shows positive growth. This is in stark contrast to the GIC using the full data, which shows positive growth only for the extreme poor. What this indicates is that households in the panel on average saw improvements in their consumption, which is not true for households in the population as a whole.

As highlighted earlier, growth incidence curves are useful tools, but they are sometimes misinterpreted. Growth incidence curves are based on cross-sectional data and consequently, they do not reveal the experience of individual households. Figure 1.22 displays the equivalent of a growth incidence curve generated with the actual panel data. The figure shows, by quintiles, the average growth in consumption experienced by individual households tracked over time. The figure shows that the largest gains took place for households in the bottom quintile, and the smallest gains were for those at the top. This supports the general pattern seen in the GIC using all the data and shows that this pattern represents the experience of individual households.

6 A small number of households that moved a small distance (within the enumeration area) were re-interviewed, in those cases that survey enumerators were able to locate them.

Figure 1.21. Growth Incidence Curve 1998-2005: National, Using Panel Data Only (in Cross-Section)

-20

24

6

% a

nnua

l cha

nge

in p

er c

apita

con

sum

ptio

n



Figure 22. Pseudo-Growth Incidence Curve 1998-2005: National, Using Panel Data, Based on 2001 Quintiles

.05

.1.1

5.2

Log

diffe

renc

e in

con

sum

ptio

n pe

r ca

pita

199

8-20

05

1 2 3 4 5Quintiles of consumption per capita in 2001

19

Annex 1. Definitions of Basic Needs Index Components

Urbano Rural

Alojamiento Mínimo

Adecuado para la Familia

Hogares donde Habiten Cuatro o Más Personas por Cuarto

Hogares donde Habiten Cinco o Más Personas por Cuarto

Acceso a Servicios

Básicos que Asegure un

Nivel Higiénico

Adecuado

Hogares que no Posean Inodoro o Letrina o Agua Conectada a la Red

Pública (dentro o fuera de la vivienda)

Hogares que no Posean Inodoro o Letrina o que Acarrean el Agua de un Río, Manantial u Ojo de Agua[2]

Pared

No Contiene: Bloque de cemento o concreto, Concreto reforzado,

Loseta de concreto o Paneles tipo covintec o Gypsum o Lamina tipo

plycem, nicalit o Concreto y Madera (Minifalda) o Concreto y outro

material o Ladrillo o Bloque de Barro o Adobe o Taquezal

No Contiene: Bloque de cemento o concreto, Concreto reforzado,

Loseta de concreto o Paneles tipo covintec o Gypsum o Lamina tipo

plycem, nicalit o Concreto y Madera (Minifalda) o Concreto y u otro

material o Ladrillo o Bloque de Barro o Adobe o Taquezal o Madera

o Piedra Cantera.

TechoNo Contiene: Zinc, Teja de Barro o Cemento o Lámina Plycem o Nicalit

o Loseta de Concreto reforzado.

No Contiene: Zinc, Teja de Barro o Cemento o Lámina Plycem o Nicalit o Loseta de Concreto reforzado o

Paja Palma o Similares.

PisoNo Contiene: Embaldosado o

Ladrillo de Barro o de Cemento o de Mosaico o de Terrazo

No Contiene: Embaldosado o Ladrillo de Barro o de Cemento o de

Mosaico o de Terrazo o Madera

Educación del Jefe del

Hogar y Acceso al

Empleo de los Miembros

del Hogar

Hogares con Más de Dos Personas por Cada Ocupado y con un Jefe de Hogar con Escolaridad de Primaria

Incompleta como Máximo

Hogares con Más de Tres Personas por Cada Ocupado y con un Jefe de Hogar con Escolaridad de Primaria

Incompleta como Máximo

Acceso a la Educación

Básica de los Niños en

Edad Escolar

Definición de Indicadores de Necesidades Básicas InsatisfechasCuadro xxxx

[1] Si por razones de características de la vivienda en que residía el hogar, no existían en ella dormitorios, se consideró por defecto a la vivienda como un dormitorio.

Materiales de la Pared, el

Techo y el Piso de la

Vivienda

Vivienda Inadecuada

Indicador de Insuficiencia por Área de ResidenciaNecesidades Básicas

Insatisfechas en el HogarDimensiones

Hacinamiento[1]

Servicios Insuficientes en la Vivienda

[2] Una excepción a este criterio, se estableció para los hogares urbanos y rurales del Atlántico (RAAN, RAAS y Río San Juan), donde se acostumbra utilizar agua de pozo para el consumo humano.[3] Si no existe jefe del hogar, se reemplaza por el conyuge.

Dependencia Económica[3]

Baja EducaciónHogares con al Menos un Niño de 7-14 Años que actualmente no Asista a

la Escuela.

20

Annex 2. Issues Regarding the $1-a-Day and $2-a-Day Calculations

The World Bank calculates US$1-a-day and US$2-a-day poverty headcount figures which are published each year as part of its World Development Indicators (WDI). The figures are calculated on the basis of purchasing power parity (PPP) exchange rates, which account for the fact that in a developing country like Nicaragua, US$1 can typically purchase more in terms of goods and services in a basic consumption basket than US$1 in the United States.

The WDI calculations are done using a standardized methodology across countries and over time. While the poverty lines are referred to as “US$1-a-day” and “US$2-a-day,” they are actually US$1.075 and US$2.15 in 1993. To convert these values into the corresponding poverty lines in current local currency in a particular year, they are converted to local currency at 1993 PPP exchange rates and then adjusted to price index for local currency. Local currency price indices are those developed by the IMF and can be found in the WDI database.

For example, for Nicaragua, WDI uses a PPP exchange rate of 6.24 córdobas/US dollar.

and the Nicaragua price index according to the IMF database shows values of 49.4090 in 1993 and 139.298 in 2005.

Accordingly, the $1-a-day line in 1993 córdobas is calculated as follows:

$1.75 * 6.24 córdobas/US dollar = 10.92 córdobas1993 in PPP terms

To convert this to 2005 córdobas, the figure is multiplied by the ratio of the price index in 2005 to the price index in 1993:

10.92 córdobas1993 * (139.298 córdobas2005 / 49.40905 córdobas1993)

=30.78655 córdobas2005/day

The $2-a-day poverty line is double the value of the $1-a-day line. To convert these to a yearly basis, they should be multiplied by 365.

Based on these calculations, the $1-a-day poverty line in 2005 was approximately equal to the official national poverty line. Consequently, the WDI $1-a-day headcount and the official moderate poverty headcount are both approximately 46 percent.

The choice of 1993 PPP exchange rate is clearly crucial to the poverty line calculations. The value of 6.24 used by the WDI for Nicaragua comes from version 5.7 of the Penn World Tables. This value is approximately equal to the official exchange rate during the period. More recent versions of the Penn World Tables show very different values. Version 6.2 has a value of 1.51 for the 1993 PPP.7 If such a value was used for the $1-a-day and $2-a-day calculations, the associated poverty rates would be dramatically lower.

7 The Penn World Tables are available at http://pwt.econ.upenn.edu/php_site/pwt_index.php

21

Annex 3. Decomposition of Gini Coefficient by Income Source in 1998, 2001, 2005

Table A.3.1. Decomposition of Gini Coefficient by Income Source in 1998, 2001, 2005

1998

% of individuals

living in hholds that

received some

income of this type

Share of this type

of income in total national income

Gini coefficient

for this income

Gini correlation of income of this type with total income

Relative concentration coefficient of this income source in

overall inequality

Contribution of income source

to overall inequality (absolute

value)

Contribution of income source to

overall inequality

(%)

Income Source Pk Sk Gk Rk gk Sk*Gk*Rk ShareAg wages 27% 7% 0.88 0.23 0.38 0.01 3%Non-ag wages 59% 38% 0.75 0.79 1.11 0.22 42%Non-ag self-employment 38% 22% 0.87 0.75 1.22 0.15 27%Ag self-employment 49% 8% 0.86 0.29 0.46 0.02 4%Imputed value of housing 93% 12% 0.67 0.77 0.97 0.06 12%Education transfers 16% 0% 0.92 -0.06 -0.10 0.00 0%Gifts of food 30% 1% 0.90 0.13 0.22 0.00 0%Remittances 19% 6% 0.94 0.65 1.13 0.04 7%Institutional donations 1% 0% 1.00 0.20 0.37 0.00 0%Returns to capital 3% 2% 1.00 0.94 1.75 0.02 3%Pensions 4% 1% 0.98 0.60 1.11 0.01 1%Other 12% 1% 0.96 0.26 0.47 0.00 0%Undefined sources 4% 1% 0.98 0.40 0.74 0.00 1%Total Income 100% 0.54 0.54 99%

2001Income Source Pk Sk Gk Rk gk Sk*Gk*Rk ShareAg wages 25% 6% 0.88 0.11 0.17 0.01 1%Non-ag wages 61% 38% 0.74 0.77 1.04 0.21 39%Non-ag self-employment 42% 29% 0.88 0.81 1.29 0.20 37%Ag self-employment 46% 10% 0.89 0.44 0.72 0.04 7%Imputed value of housing 94% 11% 0.65 0.77 0.91 0.05 10%Education transfers 11% 0% 0.94 -0.16 -0.27 0.00 0%Gifts of food 27% 1% 0.91 0.15 0.26 0.00 0%Remittances 13% 3% 0.94 0.46 0.78 0.01 2%Institutional donations 1% 0% 1.00 0.09 0.17 0.00 0%Returns to capital 3% 1% 0.99 0.85 1.53 0.01 2%Pensions 6% 2% 0.97 0.60 1.06 0.01 2%Other 7% 1% 0.98 0.53 0.95 0.00 1%Undefined sources 0% 0% 1.00 0.15 0.26 0.00 0%Total Income 100% 0.55 0.55 100%

2005Income Source Pk Sk Gk Rk gk Sk*Gk*Rk ShareAg wages 26% 6% 0.89 0.17 0.30 0.01 2%Non-ag wages 59% 35% 0.73 0.75 1.08 0.19 38%Non-ag self-employment 44% 20% 0.83 0.72 1.17 0.12 24%Ag self-employment 43% 15% 0.91 0.55 0.98 0.07 14%Imputed value of housing 95% 11% 0.65 0.77 0.98 0.06 11%Education transfers 46% 1% 0.72 -0.28 -0.40 0.00 0%Gifts of food 36% 1% 0.88 0.12 0.22 0.00 0%Remittances 22% 6% 0.90 0.57 1.01 0.03 6%Institutional donations 1% 0% 1.00 0.19 0.38 0.00 0%Returns to capital 3% 1% 0.99 0.84 1.64 0.01 2%Pensions 7% 2% 0.97 0.65 1.25 0.01 2%Other 5% 1% 0.99 0.68 1.32 0.01 1%Total Income 100% 0.51 0.51 100%

Source: Own analysis of household survey data.

22

2. VOICES OF NICARAGUA:

A QUALITATIVE AND QUANTITATIVE APPROACH TO VIEWING POVERTY IN NICARAGUA8

By Ximena del Carpio and Vanessa Castro*

“It is difficult for the poor to improve; he either stays poor or gets poorer because he has too many children. The one that is born without opportunities stays like that, the one that is born with opportunities is always going to have them, money brings money and poverty brings poverty. That is how it is…”

Individual Interview in Quilali, Nueva Segovia 2007.9

Introduction

This paper explores some of the key dimensions of well-being and addresses the difficulty inherent in researching poverty by using the voices of the people who live in it every day. Given the richness of quantitative data available in Nicaragua, much of the research on poverty in the last years has measured socio-economic outcomes quantitatively; the objective of this exercise is not to switch to a purely qualitative method but rather to enhance our understanding of poverty and poverty determinants through qualitative techniques (Bourguignon 2003). The mixed methods approach seeks to lessen the limitations from quantitative techniques in understanding poverty measurement, mainly the identification and referencing problems, by contextualizing the information (Ravallion 2003) into the social and cultural systems of the country and the realities lived by poor people that may have influenced the lack of change in poverty headcount rates in the last eight years. The simultaneous and sequential mixing of methods seeks to attenuate the disjuncture that exists among stake holders involved in the poverty dialogue in Nicaragua, yield insight that neither method alone can produce and lead to policy recommendations that are more operational (Rao and Woolcock 2004).

The extent of convergence between the two methods for this analysis is very strong10; it integrates statistical principles from the design and sampling stage and throughout the entire analytical process (Mani 2001). Sampling was based on the existing quantitative panel of approximately 2,400 households collected by the National Institute of Statistics (INEC), at three different points in time, for the last eight years11. The selection of the communities for the qualitative study were selected based on various

8 All data and figures used in the analysis of this background paper are based on the MECOVI Panel for 1998, 2001, 2005 (Approximately 2,400 households). These calculations do not incorporate the data of household incorporated into the MECOVI sample in 2001 and 2005. It is important to note that as shown in the poverty profile of the Nicaragua Poverty Assessment 2007, panel households have fared, on average, better in terms of economic measures than the cross section for the years of 2001 and 2005. * The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), and Aline Coudouel (Senior Economist, LCSHS) for their valuable comments and suggestions. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent. 9 Statements made by people during the field work are incorporated in the text throughout the document; they can be identified as they are in quote marks. These are sometimes paraphrased for clarity and translation. Single words with difficult translation are also italicized in the document. 10 See Castro, Del Carpio, Premand and Vakis (2007) “Do Voices Echo Quantitative data? A Q2 Study of Well-being Dynamics in Nicaragua”. 11 The tracking was done for the dwellings not the household members which adds another layer of complication given the potential variation of the households composition that may occur over time.

23

economic experiences using expenditure data to classify poor and non-poor communities (using the national poverty line) and upward and downward trajectories in the last eight years. The final sample selection of villages and urban locations nationwide was drawn after a systematic stratification of the full panel into economic trajectories (see Appendix II for municipal locations, Map). Since the focus of the study is on communities that on average hovered around the poverty line (above or beyond), communities with very high mean consumption levels in 1998 were excluded from the original sample selection. The final sample contains 18 communities in 16 municipalities which account for approximately 150 un-weighted survey participant households, 8 households per community on average, and hundreds of non-survey participant households, located in all communities studied in all regions in Nicaragua (Central, Pacific, Atlantic south and north, and Managua)12. It is important to note that the quantitative panel data was a representative sample in 1998 however attrition rates throughout subsequent waves (2001 and 2005) has been sufficiently high for the quantitative panel to lose national representativeness.

The entire study is part of a broader report on poverty and downward mobility that studies Nicaragua as one of three case studies (India and Ethiopia are the other case studies) on poverty and downward mobility around the world. The full set of investigative tools applied in the study reflect the philosophical underpinnings that Q-squared analysis requires; the instruments used in the field range from commonly applied qualitative tools to newly created instruments used for the first time. The tools are: community profile, social mapping, wealth ranking, semi-experimental exercise on needs, perceptions and use of programs and services, youth aspirations, citizenship and security, life histories and social assets exercise13. Each one of these instruments has a specific focus and procedure—focal, semi-structured, fully-structured, individual, and dynamic— of information gathering. The procedures were set ex-ante to obtain the best results and facilitators went through intensive training to learn the process and apply it homogeneously throughout the regions. Because of time constraints and specific focus of the full Poverty Report, this paper does not make use of all the instruments and modules applied during the two month field work. Moreover, at the time of the completion of this paper not all communities had been analyzed which implies that the general patterns derived do not reflect the experiences of all 18 communities. It is also important to note that this paper does not attempt to evaluate social programs, policies or institutions; it simply offers a view of Nicaragua from the perspective of the people and presents patterns that emerged from people’s economics with poverty (Narayan 2000) in the sampled villages.

On the terminology used throughout this paper, all estimates using the full panel (all 2,400 households) will be referred to as panel throughout the text and the word data, panel data or quantitative data will refer to quantitative data only. All qualitative estimates or figures using qualitative communities, which encompass only 150 households and 18 communities, will be referred to as qualitative sample in the text and figures; the information obtained will be referred to as qualitative data or qualitative information. The paper is structured as follows. Part II provides an in depth look at what poverty is and how it has evolved over time and space in Nicaragua. This part has five sub-sections that range from definitions of poverty to methods of looking at poverty beyond poverty lines to inequality. Part III presents findings related to inequality of opportunity and focuses on three key services: water, education and health as well as present an evaluation of the role of leadership as an enabler or hinderer to accessing opportunities. Part IV is a brief conclusion of the findings presented through the paper.

12 17 communities were selected using the stratified random sample selection method however the 18th community was purposely selected because it is a Mayagna community in the municipality of Rosita in RAAN. The Mayagna community in RAAN is NOT included in this version of the report because it was the last community to be available. 13 See Castro and Del Carpio (2006) “Voices of Nicaragua: Qualitative Research Manual”, for all details on the instruments, methods of recording qualitative data and overall process and procedures for qualitative field work. Final draft

24

Poverty Over Time

Given the multi-dimensionality of poverty, development specialists in the past years have requested an enrichment of conventional quantitative methods by adding participatory approaches to the analysis of poverty. In this section the analysis focuses on the evolution of poverty over time and space from a participatory perspective. The findings from this analysis can serve as the first step toward the identification of relevant development obstacles, identified by the people in Nicaragua, which may assist in contextualizing policy for the country in the near future. The section is organized as follows. Part A provides a definition of poverty as related by the people and recorded by facilitators. Part B will present a brief analysis on the importance of relative versus absolute poverty. Part C develops alternative methods of looking at poverty, beyond consumption based estimates; this part includes quantitative as well as qualitative methods to compare and contrast measurement outcomes. Part D presents a set of causes for mobility identified through various qualitative instruments. Part E closes the poverty section by looking at inequality across time and space

Defining Poverty

A variety of participatory techniques, nine in total—social mapping, individual interviews, focal groups, youth groups, leader exercises and interviews etc.— were used to investigate poverty, mobility and inequality of opportunities in Nicaragua. The analysis of the qualitative information collected provides valuable insight on what people perceive as the causes and consequences of poverty as well as identify opportunities and enablers for positive change. The information collected provides a rich picture of the variety of experiences of poverty and mobility in the country; although not statistically representative, the national coverage and social disaggregation (age groups, social status, and gender) of the study permits for findings to be illustrative of the rest of the country.

Among the communities visited the majority cited a rich set of causes and consequences of poverty as people see them, in one category or another; unfortunately many of them overlap which makes them difficult to disentangle from each other. Causes and consequences cited are: poor quality housing, ill health, insecure food sources, low levels of education, poor judgment, mismanagement of resources, no land or low quality land, no livestock, no job opportunities, no skills, discrimination and war. The social consequences of poverty cited by some or all are: disempowerment, lack of pride and honor, lack of dignity (particularly as related to jobs and food), exclusion from the community, no freedom, and vulnerability to all shocks. In terms of opportunities and enablers out of poverty people identified: education, good jobs outside of agriculture, livestock ownership, migration/remittances, good administrative skills, vocational training and technical assistance. From a community stand point people noted that good leadership and ability to gestionar or submit requests was important to their progress because it resulted in improved access to necessary infrastructure (water, schools, health, roads, and social programs). The analysis in this paper focuses on some of the major categories mentioned by linking quantitative and qualitative data whenever possible in order to strategize next steps and inform policy (see Appendix I, Tables 1a-d). It is important to note that the inability to disentangle causes from consequences has a strong limiting implication in policy making and future development program design.

Poverty is not having a place to live, renting substandard land and having to pray for the weather to cooperate every day. In a rural community in Waslala people identify, through a wealth ranking exercise, the rich and the poor in the community by the assets they hold; they state that poverty to them means having to squatter and rent agricultural land. People also commonly say they pray to keep from starving every day. Some identify small agricultural producers as being mostly poor and livestock owners as being rich. Agricultural production of basic grains in one visited community is reserved for the poor because land quality has been deteriorating over time and people can no longer sell any of their production; large land owners are now cattle owners or large agricultural producers. People state that a

25

large part of the community still lives off the land but are unable to save any food for later times in the year and sometimes cannot even produce enough for their own consumption in the season. A small group of families in the community however, have been able to improve their economic well-being substantially over the past years by buying large amounts of cattle; according to the people these families hold at least 200 heads of cattle. People also point to the fact that large cattle owners are buying the land from people leaving the community and from the poor at a very low cost to use for their cattle; the data shows that households who own large amounts of cattle (10 head or more) also have large land holdings (100 manzanas or more) in this community (see Appendix, Table 1a for other material dimensions of poverty).

Exogenous shocks pose high barriers to accruing assets and keep the poor from progressing. 2.1 illustrates an increase in durable asset holdings of the poor and non-poor, as classified by using consumption, for all panel households living in the central region separated into urban and rural, over the last eight years. In terms of productive assets, a general finding derived from the field work regarding the determinants of poverty and lack of progress in reducing poverty is that unforeseen expenses related to health and death (shocks) tends to force people to deplete assets, productive ones such as land, cattle and equipment and non-productive ones such as houses, televisions and radios. The elderly seem to be in a particularly difficult situation because they cannot find work, have high health expenses which force them to sell assets they have had for years in order to survive (see Appendix I, Table 1a for other material dimensions of poverty).

Figure 2.1. Log of durable asset values in the Central Region (Urban-Rural) as classified by being poor or not-poor in 1998

0

1

2

3

4

5

6

7

8

9

10

Not-Poor Poor Not-Poor Poor Not-Poor Poor

Central Rural Central Urbano

Source: Own calculations using EMNV panel Note: Durable assets includes 25 items ranging from all household electronics (radio, television, refrigerator, blender, rice maker) to productive durables (sewing machines, computers, typewriters) and transport durables (bicycles, cars, boats, motorcycles).

Historically disadvantaged regions, stemming from ethnic differences, social exclusion or customary practices, are often characterized by having strong group-based inequities that become reflected in regional inequalities14. According to a UNDP regional study of the Atlantic region, the Coast continues to be perceived by some institutions in the central government as a natural reserve subject to extractive practices and not as a society in its own social, political and economic context. Changes in the autonomy law in 2003 however have improved relations between indigenous groups and the central government through recognition of their rights and approval of their autonomy. The autonomous governments have more power to approve projects related to the exploration and mining of the natural

14 World Development Report 2006

1998 2001 2005

26

resources in the region; unfortunately, lack of transparency within the autonomous region, low levels of administrative training and democratic participation and high incidence of corruptive practices have led to the proliferation of contracts of convenience and a transference of possession toward the highest bidders (UNDP 2005). In RAAN people mention that they law forbidding the cutting of wood for sale works against the poor and favors the rich. One person mentioned that “people are not allowed to cut wood from their own community to sell unless they pay a fee; however the fee is not a problem for the richer people who can then commercialize the wood but its an inhibitor for the poor to engage in commerce”. The diversity and lack of clarity that exists in land rights, laws and tenure around the country are also perceived as deterrents for progress by the people. In the agricultural region of the coast, indigenous groups (Miskitos, Mayagnas, Creoles, Garifunas and others) have had to fight against external forces (foreign companies and national traders) as well as internal forces (rent-seeking locals and politicians) contributing to the deterioration of their resources; the illegal traffic of wood, over-fishing and drugs are three key examples. In RAAN one informant said that “they have abundant natural resources but many people abuse them. People use nets to fish and take even those little ones from the lake and other species, killing their natural habitat. Moreover, there are people in the community with motorized cutting equipment for wood cutting who cut a tree, do not pay the fees and only use parts of it and leaves the rest to waste”.

Access to formal employment seems to be limited by geography, skills and social connections. From the communities evaluated in the qualitative analysis there seems to be a pattern emerging in terms of formal employment; people in the urban area tend to have wider access to formal low skilled jobs located in nearby communities or urban centers. People in an urban community in a municipality in Managua were able to count and report the status on 10 people in the community who are formally employed in a local cement factory. One person said “those people have social security benefits and have a fixed income whereas the rest of the people in the community work informally as drivers, carpenters, builders, welders (all men) and making tortillas, bread, washing, cleaning and domestic duties (for women)”. People who gained employment in the cement factory or maquilas usually have someone already working who helped them gain employment; the majority of the people however cannot access these jobs because they either lack the skills necessary, a connection or both. In RAAS, the ability to speak English is a necessary skill to work as an embarcado (in cruise ship); people with low levels of education, ability to speak English and an initial fee for paperwork (passport, medical exams etc.) can access these jobs. In Managua, a high school degree is required to work in the maquila factories; age (under 30) is also a factor that some of the youth mention is a requirement. People in some rural communities exhibited frustration toward the lack of employment outside of agriculture; many are hopeful about the prospect of finding a job, particularly the youth who often aspire to work in an activity different than that of their father (agriculture). There is wide heterogeneity in economic functions and opportunities related to agriculture; disparities within communities and between them as well as gender divisions. Livestock and commerce of livestock products—milk, cuajada (special cheese), meat—are generally reserved for the more affluent members of the community.

Relative Poverty

Discussion of poverty and how to break poverty cycles often focus on people’s absolute levels of living standards and conditions as measured by their consumption or income and their ability to command goods and services (Ravallion and Lokshin 2005). Figure 2.2 illustrates the positive relation that exists between consumption and inequality (annual percentage change from 1998 to 2005), as measured by the standard inequality measure GINI, using the qualitative sample municipalities. This upward sloping curve indicates that as consumption increases inequality also rises; however the relationship is not as straightforward as it seems particularly when perceptions of the people are incorporated into the mix and relative positioning in the distribution affect their views.

27

Relative position in the welfare distribution affects people’s perception of poverty and inequality. Individual relations to others in their community and between groups within and between communities appear to influence how people respond to poverty and inequality questions; in other words, relative income affects well-being (Easterlin 1995). People place value on intangible elements such as status, access to networks and power etc. to their own measure of wealth and poverty (Van Campenhout 2006). Prior research on subjective well-being finds that households, both poor and rich, compare their well-being and overall position in the distribution to their neighbors (Frey & Stutzer 2002). One paper finds that households living far away from markets care more about their relative position (Fafchamps and Shilpi 2006) which implies expanding the analysis framework to account for human experiences. Moreover, In a group exercise in a rural community in the Central region people identified the primary school teacher as acomodada because she had a regular salary, in the same municipality in an urban location, during the same exercise, teachers were classified as poor because they sacrifice much and do not earn commensurate to their education and effort according to the people.

Figure 2.2. Positive Relationship between Consumption and Inequality (1998-2005 annual % change)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-40 -30 -20 -10 0 10 20 30 40

%Change in Real Consumption Per Capita (1998-2005)

%C

han

ge

in In

equ

alit

y (G

INI 1

998-

2005

)

Source: Own analysis using EMNV panel Note: Markers are the average for all households in the 16 qualitative sample municipalities. Square marker represents all households in the panel and the round marker represents the aggregate of all households in the 16 municipalities. 1998 weights applied.

Moving beyond the use of poverty lines to measure poverty: relative vs. absolute poverty as perceived by the people. Poverty indicators used in conventional poverty work are not always able to reflect on the complexity of community dynamics and settings. In a rural community in the municipality of Quilali most households are considered poor or extremely poor by using the quantitative data (headcount using the panel has gone from 93% to 100%); consumption has gone up slightly by 6% while inequality has gone down by 20%. People’s welfare interdependence and aversion toward inequality appears to shape the discussion in a wealth ranking exercise and moves their focus away from how the community has fared economically as a whole toward individual achievement and overall position of each household (and often person) in the wealth ladder. During the exercise people categorize the community into poor, very poor, moderate and rich. Contrary to other more wealthy communities in the sample where having a car, a profession or a stable job matter, in this community having food daily makes a household not poor, and owning arable land (1 manzana or more, the average for the community is 2 manzanas in 2005 versus 4 in 1998) makes a person fall somewhere between moderate and rich15. The

15 Having one manzana of land or more is at times considered to be a criterion for poverty program targeting. Recently in Nicaragua a proposed program “Hambre Cero” is planning to target households with 1 manzana as its rural anti-poverty strategy, to receive a productive program.

28

main indicator however is the house; owning a home versus renting or squatting and occasionally the type of house is identified in this community as a measure for economic welfare. A household with a bigger and nicer dwelling is considered not poor, regardless of their consumption ability. In a rural community in Waspam, people noted that the abundance of fish and other food resources enable even the poorest people to eat; one person said “food is not the problem in the community, it is the lack of housing and good land that make people vulnerable.”

Even the rich in rural areas consider themselves poor when they compare themselves to others in the urban center. A person’s status among peers and surroundings affect how he or she perceives well-being (Narayan and Petesch 2001). The third wealthiest household in a rural community, in terms of consumption and food expenditure, places itself among the poor in the municipality. The main reason is that people have various classifications of poverty and they emphasize it differently depending on certain characteristics they value more. In this community, the richer households had a common notable difference; they had inherited assets, mainly land and cattle, and had access to credit. When compared to other communities, this one is poor because the quality of the land is low and there is no access to electricity and transport. The observed behavior in this community is consistent with the reference point hypothesis which states that people judge their well-being relative to others like them (Layard 2002).

Identifying the Poor

It is often argued that quantitative poverty measures miss some important poverty dimensions such as community dynamics, leadership, social networks, risk etc. To further illustrate how people’s perception compare against quantitative poverty estimates this paper applies a variant of a technique used by Nguyen and Rama (2007) to make a comparison between consumption based estimates, proxy estimates and subjective measures obtained through the wealth ranking exercise for four communities sampled in the Central region, one urban and three rural (see Table 2.1). The analysis includes one more method, self-assessment for a community in the Atlantic region for illustrative purposes. The first two methods are standard poverty measures used for identifying poor households using quantitative methods; the proxy differs in that it predicts household economic well-being by using observable characteristics correlated with poverty16.

The self assessment method relies on the households own assessment of its economic conditions, vulnerability to shocks, social status; the facilitator guides the conversation and the analyst makes a decision on the final ranking based on the households present and past experience. It is considered the most subjective of all the methods presented here because households use their own benchmark. The wealth ranking method involves the collective perceptions on the status of a subset of 30 households picked at random from a comprehensive list of 50-100 households in the community. This exercise was conducted in each community (two times), once with common people and the other with leaders and educated persons living in the community. When overlaps occurred and differences of opinion about the economic wellbeing of a family arose, documented discussion (to include all reasons given by informants) were evaluated and a decision taken from the analysis. In all communities people defined at least four categories (very poor, poor, middle income, rich, very rich) from which poor and non-poor for the comparison analysis are derived. It is important to note that the qualitative based measures are based on a number of assumptions and arbitrary decisions; alternative ways of proceeding may have been possible for coding even if it impossible to replicate the assumptions used by Nguyen and Rama exactly.

People classify themselves to be less poor, on average, than what quantitative measures indicate. The estimated poverty rates for the four locations in the Central region obtained from each method

16 See Nguyen and Rama (2007) for the calculation functions of the proxy means estimates.

29

calculated differ from each other substantially. The estimated poverty rates range from 52% when applying the wealth ranking measure to 85% when utilizing the proxy means; the consumption based method falls in between the others at 64% (see Table 2.2). In Quilali the expenditure based method and the proxy both estimate 100% poverty rates (8 households in the quantitative sample versus 32 households in the qualitative) however community members identified high levels of inequality in the community during the ranking exercise, to include coffee exporters, large land owners and cattle traders in the community.

Table 2.1. Alternative Poverty (Targeting) Methods

Can be used at the level of Label Assess household poverty status based on

Community Household

Data sources

Household expenditure

Actual household consumption per capita, as measured in the LSMS expenditure module and the poverty line for the country

Yes Yes, but respecting confidentiality

LSMS-Panel

Proxy means Predicted household consumption per capita, with the prediction relying on observable correlates and the poverty line for the country


LSMS-Panel/Census

Self assessment Self declared (and assisted by facilitator) poverty status, according to an extensive life history account and a long guided conversation


Life history exercises, household level

Wealth ranking Poverty classification by local households with the assistance of facilitators with local knowledge

Yes Yes, but matching data for the specific households required

Wealth ranking exercise

Note: this is an abbreviated version extracted from Nguyen and Rama (2007) Note2: Self-Assessment calculations are not included for the Central Region analysis only for RAAN in this paper

Quality of services and availability of basic infrastructure affect people’s perception of well-being. The correlations coefficients across methods computed show there is a 36% correlation at the community level between the proxy means and the consumption measure, and a 24% correlation between consumption and the qualitative measure. At the municipality level, the correlation between the proxy method and the consumption based methods significantly increases to 45%. A closer look at the actual categories identified by participants (very poor, poor, not poor, better off and wealthier) in Quilali shows that 13% of the people are considered rich and only 6% are destitute. In the urban location in the municipality of Jinotega, people identified that 14% of the households ranked were destitute, handicapped and living in makeshift houses with no possibility of improvement. For this location, the proxy method finds that 63% are poor and 25% are extremely poor, the qualitative method finds that 43% are poor and 14% are extremely poor; both methods estimate significantly higher rates than the consumption method of 25% for poor and 0% below the extreme poverty line. One possible explanation is that the proxy puts emphasis on assets, quality of services and infrastructure (house quality, roads and distance to services) which differs from what a consumption aggregate includes. Like the proxy, people during a semi-

30

experimental exercise with leaders and common people, stated that the quality of the road in the community made it difficult for some people to obtain appropriate transport and services such as water and electricity were not adequate for most families.

Table 2.2. Poverty Rates by Applying 3 Methods (Central Region)

MunicipalityPoor Not-poor Poor Not-poor Poor Not-poor

Quilalí (RU) 100% 0% 100% 0% 58% 42%Jinotega (UR) 25% 75% 63% 38% 43% 57%Matagalpa (RU) 63% 38% 88% 13% 65% 35%Ciudad Dario (RU) 67% 33% 89% 11% 44% 56%Total 64% 36% 85% 15% 52% 48%

Consumption Proxy means Qualitative (WR)

Source: Authors calculations using EMNV and qualitative transcripts Note: The name of the communities visited is not listed to respect anonymity; municipality calculations represent 1 community within each municipality Note 2: There are approximately 8 households per community in the panel and proxy, 25+ in the wealth ranking

Difficult to measure household characteristics such as happiness and hope appear to influence poverty perceptions and bias subjective estimates of poverty down.. The comparison of averages of poverty obtained by all four methods considered for the RAAN community exercise are very different even if they all conclude that this rural community is very poor (see Table 2.3). The method whose outcome is closest to the wealth ranking benchmark is the household self assessment of poverty exercise. The results suggest that the two subjective measures, which take into account unobservable characteristics in the household and community such as happiness, hope and historical events, yield substantially lower poverty estimates than the quantitatively based methods. It is worth noting however, that from the 12 households used for the calculation of the self-assessment method, 6 were panel household and all 6 considered themselves poor even if the gradients of poverty they placed themselves in were very diverse. It is difficult to say which methods is more accurate to the reality and whether the differences stem mostly from the difference in sample size (from 8 for quantitative methods to 26 households with the wealth ranking method) and the level of aggregation.

Table 2.3. Poverty Rates by Applying 4 Methods (1 Rural Community in RAAN added)

MunicipalityPoor Not-poor Poor Not-poor Poor Not-poor Poor Not-poor

Waspam (RU) 100% 0% 88% 13% 73% 27% 64% 36%

Quilalí (RU) 100% 0% 100% 0% 58% 42% na naJinotega (UR) 25% 75% 63% 38% 43% 57% na naMatagalpa (RU) 63% 38% 88% 13% 65% 35% na naCiudad Dario (RU) 67% 33% 89% 11% 44% 56% na naTotal 71% 29% 85% 15% 56% 44% 64% 36%

Qualitative (SA)Consumption Proxy means Qualitative (WR)

Source: Authors calculations using EMNV and qualitative transcripts Note: There are approximately 8 households per community in the panel and proxy, 12 in the self-assessment and 25+ in the wealth ranking.

Causes of Economic Mobility: Entrants, Leavers and Stayers

People throughout Nicaragua face highly uneven playing fields in their capacities to acquire necessary endowments and aspirations to improve their well-being. The government can improve the playing field by broadening opportunities through various supply side programs in the areas of health, education and

31

risk management. People on the other hand can contribute their part, and this can be enhanced through demand driven approaches, that encourage them to invest in themselves and modify their behavior to improve their life-outcomes. Educational outcomes will not improve unless schools are there and people are willing to attend them. Technical programs will not succeed in the long-term if people do not seize the opportunity to learn from them and implement their lessons effectively. The intergenerational transmission of poverty can be halted if both, supply and demand side interventions and actions work together. People throughout the qualitative work recognized the importance of doing their part; this section presents the main drivers and interrupters of progress and factors prolonging poverty, as perceived by the people in the study.

Educational attainment and availability of good jobs appear to be equally important as ability, community cohesion and empowerment. From the communities visited for the qualitative work, there is a significant variation in individual experiences of poverty movements, the factors for this heterogeneity in outcomes vary from asset depletion resulting from negative climate shocks (such as drought in the central region and floods in the Atlantic) to positive unexpected changes (like credit and income generation programs, inheritance and free drugs for trafficking). Among the significant correlates of consumption presented in the profile section (Table 2.6, section poverty profile), educational attainment, the labor sector in which household participates in and services available are the most commonly cited factors either positively or negatively influencing socio-economic change. Other factors not presented in the correlates and often cited by people, such as intelligence/ability, community cohesion and tensions, leadership, empowerment and social networks, indicate that income or consumption are insufficient measures for well or ill-being.

People able to seize opportunities experienced the most progress in the last years. For a subset of households in Waspam (RAAN) experiencing a positive trajectory in the quantitative 3 period panel, measured by an increase in their real consumption during the eight year period, the qualitative analysis shows they had at some point experienced a unique positive shock (the fall of an airplane in 2003 full of drugs). The drugs, through sale or trade, may or may not have been the catalysts for their economic improvement but most of the people interviewed recognize it as an important turning point for the community and some even say that for some it was like, as a person put it, “winning the lottery”. In other communities where households experienced more standard positive shocks such as inheritance, remittances or technical/credit programs, these were able to harness the opportunity by investing the gains in new businesses or purchasing productive assets such as land and cattle. In one case in the southern Atlantic (RAAS) a man, identified by other as a person with opportunities and later interviewed, mentioned that many years before he had the opportunity to buy a small old boat with savings, which he fixed up and transported people for a fee and slowly he was able to buy another boat to transport commercial goods. Now the man has multiple boats and transports both, goods and people, from the rural to the urban area of El Rama. “A man stated that the intelligence of some people to manage and administer what they have contributed to their success; even if they had a cow or land inherited initially they make it grow and augment their wealth in a continuous manner”.

Bad judgment and vices contribute to the perpetuation of poverty. People throughout the communities visited repeatedly cite poor judgment, commonly stated as having mala cabeza or bad judgment and weak administrative skills (particularly as related to natural resource management) as a key contributor to the persistence of poverty. People who have had positive shocks sometimes sell their house, land or cattle, according to informants, because “they don’t know any better”; in other words, according to people not having a job is not the only reason to blame for a decrease in economic welfare. Other examples revolve around depleting assets, such as land and cattle too quickly and without substantiated reasons. There are various cases where people came across free fertile land, with formal title, and sold it to buy consumption goods such as medicine, food and in some cases alcohol. In a semi-rural community in Managua a man stated that some people are poor because they stop fighting. He said

32

that “because of poor administration from the part of government officials people lost their jobs and had to borrow to survive; men and women however, have to look for a balance and identify their next move and most importantly leave laziness behind”. People do not always blame others or the system for their failure to seize opportunities. A woman in RAAN said that “when the airplane full of drugs crashed some people use the resources obtained from the sale of the drugs to improve their households but other like herself did not know how to use the money productively so she spent it all and is now the same or worse than before”.

Remittances and wise use of resources are cited as drivers of economic improvement. In the urban area of el Viejo, during a wealth ranking exercise, people were able to distinguish between the poor and the destitute in a very clear manner. The poor, one person said and others agreed, “are those who have to look for food and work every day and the destitute are the ones that cannot seek work, because of their age, bad health or having too many children and no help”. The Barrio (neighborhood) has a very diverse group of families, a person stated that “some are very poor and have always been poor and others are wealthy and lucky”. There are people who have improved their well-being in the last ten years because they have a family member who has emigrated abroad, to Costa Rica, Spain or the United States, and send money back to their families. Others who emigrated and later returned with savings were able to use the money to buy land and livestock; one man in the Central region stated that “after the divorce from his wife, who owned the local pulperia or corner store, he went to Costa Rica to work and was able to save enough money to educate his children and buy assets upon his return and now lives off the livestock and land he purchased”. In the Atlantic coast people often cite migration to Costa Rica for agriculture work for men and domestic labor for women as common outlets for improving their economic well-being. In the recent years, there has been a surge in a different type of migration in the Atlantic coast, embarcados or cruise ship workers but this activity is mostly male dominated which makes mothers be fathers as well. Cruise ship workers send remittances amounting to approximately $500 every 15 days (compared to $200 every 15 days for fishermen); in the last years these families have become the ones with the best houses and the highest purchasing power; they are followed only by teachers, formal sector workers, public employees and small business owners.

Migration and remittance stories however, are not always positive. The negative side of high volumes of remittance income and temporariness of the cruise ship work in RAAS is that it has led to a surge in drug use in the community. A statement made by one person in the community was that “Young people who have access to resources but have no jobs turn to drugs and drug dealing; crime and murders have surged as a result of this”. In one rural community in RAAN, people who immigrated to the community because of the abundance of cheap land faced harsh economic conditions due to the low quality of the land purchased, plagues and bad weather which forced them to lose their assets and are now nearly destitute. Recent waves of emigration in one rural community in the same region influenced the change in asset holdings; people sell assets and accrue debt to leave and only in some cases they are able to repay their debt and use remittances to buy back their assets. One woman in Managua stated that “families of people living in Costa Rica are still poor because the members who left do not come back and their family does not improve because the migrants forget that they exist and do not send anything.”

Social programs and benefits positively affect the socio economic dynamic of beneficiaries. Other people have been able to use their pension to open small businesses and have made wise investments with their time and money; women are known for making more money a day than men in this community (C$100-300/day women vs. C$50/day for men) through the sale of cosmetics and other goods. The quantitative data however, shows that female headed households in the community had an average consumption level 20% higher than male headed households in 1998 but the pattern has reversed in 2005, and women are now making 33% less. In a rural community in RAAN, a man mentioned that technical assistance programs for the landed have been very good for diversifying crops in the community—through an organic seed program and work tool loans—and helping them make links to outside markets;

33

unfortunately some people chose not to participate expecting that the new government would come with another, less work intensive program. People throughout the field work mentioned vocational programs (mostly related to agriculture) in a very positive light; in one example it was said to have inspired the inception of a cooperative which is now benefiting the whole community. A man stated that “he finally appreciates that the presence of programs such as road repairs, cattle reproduction, cooperative capacity building among others have improved the lives of some members in the community in a manner that is irreversible; the knowledge they have can be applied always and no one can take it away.”

Property titles are required to access electricity services. Having a house is important but having property rights to a house or land can also influence economic well-being, objectively and subjectively. People throughout the qualitative field work shared various opinions about not having formal title to their property; some felt threatened by the possibility of their property being taken away and even cried during the interview while others felt secure and showed no concern about not having a formal title. In one example in urban Bluefields, where the data shows that all panel households and 82% of households in the municipality have access to electricity, the community leader stated that “there is a whole sector of the community where poor people living around the bay could not access electricity because the electric company required property titles to install the service and none of them had titles”. This problem is common place throughout other parts of the country; in a rural community in Managua a group of low quality houses in a sector of the community were all connected illegally to the public electric source through makeshift wires. The houses were all located in a property owned by a large cement company and could not formalize their ownership which meant that the electric company could not provide the service to them, even if they wanted to pay for it. In terms of credit, having a title is very beneficial, one person in an urban community of the central region said that “he knows what it implies to have title to a property, it means having access to loans something that could be a great help for those who have titles and a great hindrance for those who don’t”. Figures 2.3 and 2.4 show that overall rates of access to electricity have improved over the last 8 years for all panel households but the bottom 40 percent of the population improved at lower rates; in other words expansion favored the wealthier households.

Figure 2.3. Access to Electricity in the 1st and 2nd Quintiles

0

0.2

0.4

0.6

0.8

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

1998

2005

CEN-RU ATL-RU PAC-RU CEN-UR ATL-UR PAC-UR

Source: Own analysis using EMNV panel

34

Figure 2.4. Access to Electricity in the 4th and 5th Quintiles

0

0.2

0.4

0.6

0.8

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

1998

2005

CEN-RU ATL-RU PAC-RU CEN-UR ATL-UR PAC-UR

Source: Own analysis using EMNV panel

The church serves as a source for social interaction and leadership. Whether it is beliefs, culture or the networking effect (Mc Cleary and Barro 2006), religion in Nicaragua plays a very strong role in people’s daily life and acts as a principal source for gathering and organization. No claim is made on whether it is beliefs or belonging to the church and without further empirical work on this is not possible to answer this puzzle however there seems to be a clear relationship between religiosity and well-being. In some communities religion acts as a catalyst for social capital; religious activities are the main source of interaction among women, particularly young girls. The church is perceived as an institution of influence and it provides indirect avenues for economic interaction through communal activities. A negative aspect with respect to religious affiliation found, not commonly, is the feeling of exclusion from social activities; some people stated that they needed to contribute to the church or be active members of a particular denomination to be included in community activities organized by the church. In a semi-rural community in the Pacific where the quantitative data shows 0% religious participation, the Jehovah’s Witness and Catholic churches are considered focal points of the community, they provide food for children on Sunday, a venue for social activities for the youth and women and a center for coordination of social programs. Some people characterized the followers of these churches as “pudientes” (economically enabled), for example “they are the only ones that have telephone connections in the community” stated an informant. In 2005, 7.4% of the panel household participated in some church association; regionally, Managua leads the country in church membership with 13% and the rural Atlantic region follows closely (11%).

It is unclear from qualitative data if there is a direct effect on measurable economic well-being however there are various examples of church participation—direct or indirect—in economic activities. Small credit (loans), tools being shared by church members, jobs being created and assets distributed to affiliated followers are just some examples mentioned in the qualitative data. One woman received a wagon-tricycle as a gift from the church, she uses it to share the gospel with others in the community and to distribute goods; being handicapped inhibited her from working and the tricycle changed her life. In an urban location in Managua people referred to the church as the only organization that cared about peoples’ well-being; one person mentioned that “the church brings doctors, nurses, provides appointments and medicines for everyone that attends, regardless of their religious beliefs”. One man said that the pastor was the only person who “mediated conflicts among community members and was respected enough to see the conflict through until it was solved”. It is important to note that in some communities people report spending several hours a day in church related activities (spreading the word, arranging events, attending mass/ceremony etc.). This may reflect on Azzi and Ehrenberg (1975) where they do a time-cost trade-off analysis and posit that people who invest the most time in religious based activities (i.e. attendance) are those with low value of time—non-labor participating women and retirees are two groups mentioned.

35

Box 2.1. A Q-Squared View of Economic Trajectories from Two typical Nicaraguan Families

These are two stories of two Nicaraguans who began their lives under similar poverty conditions in the same community in the late 60’s but have had very dissimilar economic trajectories, one richer and the other one poorer, resulting from the experiences they had and choices they made throughout their lives. At one point in 2001 the two families experienced transitory income shocks that placed both families in the same income group, however despite the drastic change the gap in consumption levels between the two families has not narrowed at all.

Rosario is a woman entrepreneur (family 1, Graph 1 shows that consumption expenditure has improved by more than 50% in 8 years, from C$6,000 in 1998 to C$9300 in 2005) who is considered economically comfortable, by others in the community, because she owns a pulperia (corner store) and engages in commercial activity. This woman entrepreneur was born to a poor family, has 1 year of secondary school completed and grew up more or less under similar conditions of others in the community. Rosario has three children, one from her first marriage when she became a widow, and is remarried to a nurse who works in the health center. She moved to the community after her first husband died and began selling lollipops and basic grains using the savings she had from the sale of a small piece of land she inherited from her father earlier in her life. Five years ago, Rosario had a difficult birth that required a blood transfusion and an extended hospital stay. Due to the difficulties experienced during birth, Rosario could not work for an extended period which affected her ability to produce and the overall income of the family (see the income dip in 2001, graph 1). According to Rosario, the family did not face major financial hardship because they had health coverage, through her husbands work, and were able to keep their consumption levels more or less stable during this period as well as keep their asset base untouched. Rosario has access to a credit program in her community and uses it to purchase inventory for her store (most recently she purchased school supplies to sell) regularly. Two years ago, she partnered with her mother in buying a windmill to grind corn and make tortillas and bread and sell it in the store. According to Rosario, her experience with credit has been good because she has always been careful in how she uses it and conscious of making her payments. She has seen people around her lose their assets because they did not pay back their credit and misused it in vices and non-productive activities. Rosario believes that her economic situation is superior to that of her parents because she is able to send her children to school and not require their help to sell goods like her parents used to with her and her sister. Her outlook of the future, in terms of economic wellbeing, is bright and her expectations for her children are very high.

Graph 1. Two Economic Trajectories

Source: Own author’s calculations using EMNV household information

The second family is that of Marvin, a janitor in the local cement factory who has had a semi-regular job in the formal sector for the last five years (See spike in income, Graph 1) but has no access to health insurance. Marvin was born to a more or less poor family, engaged in agricultural activities. He attended school until 4th grade (he did not like school and dropped out for several years) and returned to school later in his life, at age 17, to finish the 6th grade. Marvin is married and has 4 children, 3 in school and 1 at home; he is the sole provider of the family and feels lucky for having a non-agricultural job as a main source of income and some land he can cultivate year round. Marvin, like Rosario, inherited a piece of land from his mother and received a house from his father as inheritance; he uses this land to farm consumption goods such as corn and to hedge in times of unemployment. He has no formal documentation for his land or house and is unable to sell it or use it as collateral for a loan even if he feels he is not at risk of expropriation. Marvin uses his farm products as savings and sells them in times of sickness to cover health expenses. He believes his economic situation is equal to that of his parents, not better or worse, and his children face a similarly unstable present as he did growing up, to include food scarcity in times of bad winters and no formal income earning opportunities. The only difference between him and his children is the hope he has for them in the future; unlike his parents, Marvin pushes his children to study and stay in school so they can continue learning and some day work in a company or factory using their education.

36

Inequality

As established in the poverty profile, Nicaragua has experienced welfare improvements for the average poor even though poverty headcount has barely changed since 1998. In terms of inequality, the country has seen a substantial decline at the urban, rural and overall national level; however it is unclear from the quantitative data what the drivers of this decline are and whether it is due to a structural change or a loss of wealthier participating households from the sample. A close look at the panel sample reveals that households, from 1998 to 2005, experienced a decline in the GINI coefficient, from .42 in 1998 to .40 in 2005 (see Figure 2.5); a noticeable decline in the urban areas between 1998 and 2001. The communities visited for the qualitative study, randomly selected from a stratified sample17, also reflect an overall similar declining pattern between the periods, with a noticeable decline in the rural area between 1998 and 2001, from .40 to .30 respectively (Figure 2.6). It is important to note that qualitative findings are only generalizeable to the area studied (community and possibly comarca or district), despite careful sampling, the use of larger samples (18 communities nationwide) and a highly standardized methodology.

Figure 2.5. Inequality 1998-2005 Panel Households

0.42 0.41

0.35

0.400.37

0.34

0.40 0.38

0.33

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45


Gin

i co

effi

cien

t

1998 2001 2005

Source: Own analysis of EMNV panel, using all households in the panel for 1998-2001-2005

Figure 2.6. Inequality 1998-2005 of Communities Visited for Qualitative Work

0.400.38

0.40

0.340.37

0.30

0.350.37

0.32

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45


Gin

i co

effi

cien

t

1998 2001 2005

Source: Own analysis of EMNV panel Note: This figure is for illustration only given that the sample used is approximately150 households from the qualitative communities.

17 The qualitative sample accounted for in this calculation is 150 households, from which approximately 50 households are located in the urban area and 100 in the rural area. Given the size of the sample, this data is used for illustration only and to get insight into the communities visited for the qualitative work; in other words it is not accurate to make general statements from a statistical stand point.

37

The inverse relationship between changes in consumption and changes in poverty headcount observed with survey data is true on average but requires further analysis. Figure 2.7 illustrates through the experiences of households in the 16 municipalities visited for the qualitative work that the inverse relationship between growth and poverty, as shown for the entire panel in the poverty profile section in the full poverty report (Figure on GDP per capita and Inequality), still holds true with the small sub-sample. This figure also shows that there is a large heterogeneity in change of the poverty headcount and consumption which can be attributed to the diversity that exists between and within the regions and the divergence between rural and urban locations. The following two communities located in one municipality and in the same sampling segment (segmento compacto or UPM) for the quantitative data help illustrate the heterogeneity that exists within a close geographical location and the persistence of spatial inequality.

How heterogeneous can two communities in a municipality in Bluefields be? In the municipality of Bluefields for example, the two communities visited have very distinct characteristics; one rural and one urban, poverty headcount (66% and 11%), maximum years of education (4 and 10.6 years) and female headed households (16% and 66%) all for 2005, rural and urban respectively. The average consumption in the rural area has gone from C$5300 to C$3200 per capita, a decrease of 39% whereas in the urban community the average consumption for the community has gone from C$11000 to C$8200 per capita, a 25% decrease; an average decrease of 32% for the municipality in the figure. According to the community profile obtained through the qualitative work, these two communities differ in many more ways than what is apparent through the quantitative indicators. Community 1: The rural community in Bluefields, accessible only by boat (only 2 boats exist in the community), emerged 25 years ago as a result of forced migration during the war. It is conformed by less than 20 families (100 people) who own all the land under a cooperative set-up and are ethnically the same (mestizo) as they originated from the same location in the Pacific before the war. They have no access to electricity or potable water, only one contaminated well exists. The school offers pre-school through 6th grade with one teacher for all grades, and the health center is two hours away by boat. The economic activities revolve around the production of coconut for commerce and basic grains for auto-consumption; no commercial activities beyond coconut are available. Community 2: The urban community on the other hand, was established in 1960 and is conformed of five sectors housing over 2,000 dwellings and 160,000 inhabitants. One man described the ethnic composition of this urban community as a true gallo pinto (Mestizo, Creoles, Sumos, Mayagna, Miskitos, Ramaqui); and the languages vary from Spanish to English to indigenous tongues. In terms of economic activities, men work in fishing, agriculture (locally and abroad), cruise ships (temporary migrants), commerce, airport, and a small percentage in professional jobs. Women work mostly in domestic jobs locally or in Costa Rica, tourism, trade jobs (sewing, cooking etc.) and take care of small businesses. The youth of both genders engage in most of the activities mentioned above; however drug trafficking for this age group has surged in the last decade. Most of the people have electricity and only some families have private wells; the poor have no access to a stable source of potable water in the community. Primary education is available locally but health services and secondary education are available in a different sector nearby; a portion of the population in this community use private education and health services.

Interregional disparities in economic wellbeing, as measured by education, persist despite general economic improvements. Despite the average increase in years of education and value of durable asset holdings in all regions in the country (except Atlantic rural for asset values where the median in 2005 is lower than the median in 2001) regional disparities, within and between regions, are still very apparent (see Figure 2.8 and Figure 2.9). In education, as measured by maximum years of education of any household member, the average improvement is noticeable for all communities visited for the qualitative work (represented by their regions). Figure 2.8 shows that all regions appear above the 45 degree line signifying an improvement from 1998; however the disparity within regions and between regions, as

38

presented in the non-intersecting ovals in the figure (rural locations encased in the oval with the dotted pattern and urban with the solid pattern above) still persists over time.

Figure 2.7. Inverse Relationship between Poverty and Consumption (1998-2005 annual % change)

-60

-50

-40

-30

-20

-10

0

10

20

30

40

-40 -30 -20 -10 0 10 20 30 40

%Change in Real Consumption Per Capita (1998-2005)

%C

han

ge

in H

ead

cou

nt

(P0

1998

-20

05)

Source: Own analysis of EMNV panel Note: Markers are the average for all households in the 16 qualitative sample municipalities. Square marker represents all households in the panel and the round marker represents the aggregate of all households in the 16 municipalities. 1998 weights applied.

Figure 2.8. A Persistent Regional Disparity in Maximum Education (1998-2005)

ATL-RU

ATL-UR

CENT-UR

PAC-URMAN-RU

MAN-UR

CENT-RU

PAC-RU

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.001998

2005

Source: Own analysis of EMNV panel

The urban-rural divide in education becomes more evident during the winter months as the distance to schools gets more difficult. Educational services differ noticeably in the urban and rural areas; in urban areas or locations near the city center access to primary schools and usually secondary schools is generally available. According to some rural communities, the roads become more ruralizadas and less appropriate for access during the winter. This problem not only affects students but also teachers who travel from outside the community. The qualitative work evidences that the supply of educational services is vastly reduced in the rural areas, particularly remote sectors within rural communities. Costs to study secondary for example are higher for families living in remote areas because of transport from the sector in the community to the sector or urban area where secondary education is offered. In a rural community in San Juan del Sur people stated that during the winter months children cannot cross the river because the waters are too high and teachers cannot reach the school for the same reason.

39

Figure 2.9. A Persistent Regional Disparity in Durable Asset Values (1998, 2001, 2005)

Source: Own calculations using EMNV panel for the whole panel in 1998-2001-2005 respectively Note: This figure represents the log of durable assets; the vertical lines show the lowest median and the highest median in real terms

Chronic regional underperformance leads to persistent disparities in asset value holdings over time. In terms of assets, as measured by durable asset values (appliances, cars, electronics etc.), in 2005, a household in the 75th percentile in the rural area of the Atlantic had an equal or lower asset value than a household at the 25th percentile of the urban area in the Atlantic region and well below the 25th percentile of households in Managua (see Figure 2.9). The dispersion within region and among regions is also very striking. People throughout RAAN and the Central region cite the war as a turning point for them in terms of asset depletion (durables and productive) and change, for the worse, in their economic well-being. Many life histories reflect that people, who grew up in decent living conditions to include access to education, had to leave key productive assets behind, namely land and cattle, during the war and return to find nothing. One person shared through his life history that “during the war between 1980 and 1987 the contras came at night and took my three brothers and killed them; my mother died because of the sadness and a few years later my father passed as well. Because of all the deaths I had to sell all the land and cattle for expenses and the sadness was so ingrained that I was never able to improve my life”.

40

Box 2.2. On-going Security Concerns and War Memories Negatively Affect Nicaraguans

Security concerns range from petty theft (mostly related to livestock) to gang violence to natural resource conflicts. In RAAN, community members reported that outsiders come into the community to steal wood and the locals are afraid of them because they threaten to burn the wood if they are not allowed to cut it when they want. In some cases the municipality intervenes but the majority of the cases are handled by local leaders. In a rural community in the Pacific region, a whole community is dominated by one big family known as the anti-socials; members of this family threaten outsider from coming into the community while terrorizing the locals via threats (to include gun use and poisoning of the waters) in some occasions. One man said, “They stand on the corners and night, drunk and on drugs. Sincerely I am very afraid at night because I have been robbed by them before and there are two gangs in the community. However we are lucky because they attack outsiders more than insiders”. The police avoid coming into the community to deal with crime reported because of the strength of the gang and service providers, such as water and electricity, refuse to come into the community to install new services. One man mentioned that when he applied for a job and gave his address he was turned away by the employer because his address was identified as being part of that community and he was thought of being associated.

Remnants from the war still permeate the memories of people. The large scale displacements during the war re-shaped the social dynamics of many communities and interrupted the strengthening of existing social networks; people in a rural community in RAAS agreed that some of the leadership and organizational voids that exist today evidence the effects of the war on the people. In one community in RAAN community members consider that the divisions that currently exist in the community are a direct result from the war. The ex-combatants have auto designated themselves as leaders however only part of the community supports them and the other part supports the elected leaders. One informant stated that her fear of being killed is still very present because her son deserted during the war. Another woman who fought being displaced stated that “in the time of war the Sandinistas wanted us to leave the community, but we did not leave because we felt that if we were going to die it was better to die here; my son however was taken when he was 12 by the military service but he was so small that they left him on the road and he escaped. When he was 16 the contras captured him and I never knew anything about him again”.

INEQUALITY OF OPPORTUNITIES

The World Development Report on Equity and Development (2006) found that in many developing countries, the actions of the state in providing services magnify inequalities at birth. In Nicaragua the provision of human services such as health, education and even water appear to be unevenly distributed among the population; we see an urban rural divide as well as regional disparities. As stated in the WDR, equity and efficiency are complementary because human capital requires adequate levels of investment in order to obtain higher rates of return. The qualitative work in Nicaragua indicates that even though households fall into various categories in terms of physical, economic, human, and social dimensions (see Annex), policies and development efforts thus far have not been able to level the playing field to compensate for disparities in initial endowment. Based on the definition of equity it is apparent, based on the q-squared analysis so far, that people’s outcomes in life are not, on average, commensurate to their talents, efforts and aspirations; instead, they reflect, almost solely, the conditions they were born in (family circumstances, race, gender, place of birth etc.). In terms of inequality itself, impact and effectiveness of service provision, social and productive programs in Nicaragua over the last ten years seems to have mixed outcomes; some with strong measurable impacts and others with no impact at all.

The poor have limited access to services and generally cannot afford to use them. Tables 2.1a-d show that people measure poverty along similar dimensions economists do however they define each dimension differently and characterize them according to each economic sub-grouping (the extreme poor

41

or destitute, poor, non-poor and well-off). This section presents examples of some key services: water, education and health to illustrate why people perceive that services are limited and often reserved for those with better opportunities. The role of leadership is also evaluated in this section as a factor that directly limits or enables access to opportunities.

Water

Limited water access affect people’s quality of life from an instrumental and intrinsic stand point. Many of the communities visited mentioned having an inadequate source of water. The quantitative data shows that in 2005 people’s access to piped water (by gravity, pump or private well) in the lowest wealth quintile in the rural areas of the country (except Managua were rural and urban are not dissagregated) range from 10% to 25% while those in the highest quintile ranged from 20% to 55% (see Figure 2.10). The averages change upward drastically in the urban area (see Figure 2.11) however people throughout the field visits mentioned that even though they have a water pipe going into their house in the urban area, the pump in the community does not work or works occasionally. Availability of water varies from holes on the ground or vertientes to in-house piped chlorinated water; these however often dry up or contain contaminated water. Intrinsically this is objectionable because people feel it’s a human right to have access to water and it should be provided to them. Instrumentally, water is a major contributor to productivity and a powerful instrument necessary for economic prosperity.

Productiveness and safety are negatively affected by the scarcity of water. People engaging in agriculture, for example, need large amounts of water and property values go up and down in some locations, contingent on water access. Moreover, people spend large parts of their day gathering water and in some cases children miss school to fulfill this basic requirement. One man in the Central region stated that water is an essential service for the quality of life; the community has relied on water holes to drink, we have to walk 10 to 30 minutes depending on which source we choose but most of them are undrinkable. In the case of community wells, when people do not have to pay water tends to be rationed which at times causes conflict among neighbors. Children are often tasked with gathering water; people, mostly children and women, have to walk long distances to the river to gather water, the characteristics of the ground and the geography of the community force them to go up and down to get water. A leader in the Atlantic region stated that “during the summer time men have to gather water because the heat dries up shallower wells and deeper, more dangerous, wells are the only ones that still hold water ; women and children can and sometimes try to get water from these wells but given the difficulty of the task they put themselves at risk”.

Quality of water in both, urban and rural, is unreliable and often unsafe. What the people say is that even when a well is available the water is often contaminated. In one community people mentioned that the Red Cross had installed several wells in the community several years ago but the lack of maintenance led to the contamination, with parasites, of the wells; people want to maintain them but do not know how to do it. In the case of rivers, in San Juan del Sur in the Pacific region one group mentioned that the water has human waste and the water cannot be consumed. In terms of quality of water, community members have to monitor each other from washing clothes, bathing and dumping human waste in the same water source used for drinking. The construction of latrines in small plots of land near a water source causes major difficulties in the prevention of contamination. In an urban community in RAAS people complained of not having a community well; however one person stated that “some people have private wells and are kind enough not to charge others to use it”. In other cases people opt for connecting a hose to a natural hole in the ground to bring it closer to the house; in one community in the Pacific region the community had a water pump but when it burnt the mayor decided not to turn it back on because some people were not paying for the service and were obtaining the water illegally.

42

Figure 2.10. Access to Private Water Source—Rural Regions (Except Managua)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1st Q 2nd Q 3rd Q 4th Q 5th Q

% a

cces

s to

wat

er s

ou

rce

Atlántico Rural Central Rural managua Pacíf ico Rural


Figure 2.11. Access to Private Water Source—Urban Regions (Except Managua)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1st Q 2nd Q 3rd Q 4th Q 5th Q

% a

cces

s to

wat

er s

ou

rce

Atlántico Urbano Central Urbano managua Pacíf ico Urbano


Education18

Does having a school in the municipality mean having access? All groups have access to primary education in the qualitative sample (most up to 4th grade; 5th and 6th are not generally available within the community) however only the non-poor and better off can afford to pay for school fees, transport and materials. It’s important to note that access is often interrupted by weather and geographic conditions; in some cases communities have come together to remedy the situation by building bridges for the children to cross the river. Another important point in terms of quality of educational services is the inadequacy of infrastructure relative to population size; poor people as well better off people complain that the children have to attend school in a makeshift location because the actual school cannot fit all the children. Another notable difference between the groups is the perception of quality of services; the destitute and poor rarely cite lack of quality of doctors or teacher as a concern, but rather costs and distance are the biggest problem for them in using these services. The non-poor on the other hand focus their voice on quality and rarely complain about discrimination; some seem keenly aware of the low educational level of

18 Parts of this section were contributions by Vanessa Castro

43

teachers, overcrowding (multi-grade classrooms) and irregular schedule of health providers.

Girls are not sent to school and young children delay their entrance to primary school because of distance. Evidence from qualitative work finds that unequal access and quality of education is a problem in need of addressing. Distances and rains affect people living in sectors far from schools. Families in poverty tend to be the ones that send their children to school less, to stay home and work. People generally agree that the poor in rural areas are the most disadvantaged. The distance from some of the sectors of the communities visited to the closest school deters young children and young girls from attending. Parents tend not to send their daughters to school in a rural community in RAAN because it is risky for girls to travel to school because of the distance. In the central region people mentioned that young children in the community are unable to enter school at the appropriate age because they are too young to walk the distances required. In various communities segments of children do not attend school because parents cannot afford the costs to get the ready for school; in one community in the Pacific region interviewed parents reported that they had to spend 1,000 Cordobas in shoes, uniforms, notebooks which makes basic education, and most notably secondary, inaccessible to the poorest.

Children in single female headed and/or socially excluded households are the most affected in terms of education. Given that the only income for survival comes from the woman in single female headed households their children are less likely to be sent to school; according to people interviewed these children are the ones dropping out of school most often in their communities. In some cases it was observed that parents send only some of the children to school because they cannot afford the load of sending all of them at the same time. Some times social exclusion from poverty also leads to the decision of not sending children to school; in one community in the Pacific region, parents were not interested in educating their children and decided not to send them to school because in the sector they live people view them as the “anti-socials”.

Some educational services, beyond primary and secondary, are available however they are limited in scope. Pre-school programs are not generally available and the supply of all-grade secondary education is inexistent in rural areas. One of the rural communities in RAAN offers partial secondary education only; the last two years requires for students to travel far away and pay the fees for the student to stay in Waspam. Members of some communities reported having local access to adult education programs. In one of the communities visited, a foreign NGO offers an educational program for school drop outs (children out of the normal age range); the program consists of an accelerated primary degree and vocational skills. In a rural community in the Pacific region people mentioned a very structured literacy program imparted by the mayors office through a program “Yo Si Puedo”; through this program people not only learn to read and write but will eventually be able to take secondary level classes. In the central region a US department of labor funded program called “Yo Aprendo” offers vocational courses to school drop outs in subjects such as carpentry and cooking.

Demand of schooling, as reflected by overcrowding, exceeds supply of classrooms, teachers and educational materials. Inadequate supply of classrooms and teachers is a general problem. In various communities there are too many children for what the infrastructure can accommodate and one teacher to teach multi-grade or the whole primary; this mismatch of supply and demand is a problem because it lowers quality and is overall anti-pedagogical. In a semi-urban community in Managua a leader stated that the instituto or secondary school they have in the community is sufficient for the children from the community however the school is overcrowded because it is a feeder school for 13 other communities nearby. In a semi-rural community in Managua there are 170 students in 3 6th grade sections which results in approximately 55 students per teacher; well above the national average. In another community in the pacific region the qualitative work revealed that a primary school only had one teacher assigned for the entire school; while in other cases parents reported that 2 of the teachers had not completed secondary education and lacked the appropriate teaching credentials. A school in a rural community in the central

44

Atlantic region had no teachers assigned and children were unable to attend classes despite the infrastructure being in place.

Work deters children from participating in school. Desertion from school was reported to be high, up to 20% in one community, around the harvesting period. The combination of work and school work negatively affects attendance as well as focus on the part of secondary students. Many parents report assigning labor and work related tasks to their children when they arrive from school. Figure 2.12 shows the percentage of young adults, from 15 through 18 years of age, who contribute income to the household, for both the panel and the 18 qualitative communities sampled. This percentage is derived as the average number of teens working divided by the average number of teens in the age cohort. The comparison shows that the averages for the panel are different from the sub-sample, particularly for the Atlantic rural and the Pacific urban however the differences do not appear to be systematically biased. What the figure shows is that teens in the rural areas on average work more than in the urban areas, with no significant regional disparities except for teens in the qualitative sample in the Atlantic rural region and Managua. The implication of this finding is that given that the combination of work and school is reported to be difficult to handle it may be expected to see less school achievement from teens in the rural area.

Figure 2.12. Percentage of Youth (ages 15-18) Contributing to Household Income in 2005

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

ATL-RU ATL-UR CEN-RU CEN-UR MAN-BOTH PAC-RU PAC-UR

Panel Qual-Comm


The youth recognize that education is the path out of poverty and have ideas on how to improve their participation. In terms of hope, the youth expressed concern for their future and believe education is very important and a contributing factor out of poverty however they also expressed their concern for the lack of adequate employment. In one youth group children said that a proper source of employment is essential to them as a motivator to study and in order to demonstrate their capacities. They ask for vocational programs (carpentry, construction), weekend primary schools and literacy programs taught locally, high school classes taught in their local church so they don’t have to cross the river and improvement in transportation. In a rural community in Waslala, a group of young people highlighted a strong inhibitor to high school participation. Despite a new road program (quantitative data shows that at least half of the panel households in the community benefited from the new road, some from the income earned from working on the construction) and increased access to markets and overall communication improvement (public good) high school students feel left behind. None of the modes of transport are available during the schedule of high school students (6 am and late in the evening) and there are not enough students to make it worth while for transport providers to switch the schedule. A youth group in RAAS expressed that there are no alternatives for them in the community in terms of vocational education, professional, entertainment or social; only agricultural work is available. They want to have a better life than their parents and would like to have a profession but see no alternatives, no one has come to help them or worry about them.

45

Health

Costs, difficult access, discrimination and lack of medicines push people away from going to traditional health providers. In regards to health, some communities have health centers located in the community but a good part of them do not; people are forced to travel outside the community, sometimes walking for hours, incurring transport costs. One person in an indigenous community in RAAN stated that as a poor person she is discriminated against by health providers; she said that “the service in the health center is not equally administered to all people, they help the people with higher economic opportunities first and then the poor ones, they leaves us for last she said and they deny us the best medicines”. A common problem is that health staff is not always present at the center when people arrive; another problem is the shortage of free medicines. The basic basket of health products is composed of over 100 medicines and is regularly sent to health centers and hospitals, to be distributed free of charge. People count on these free medicines and expect to get them when they visit the health center; however a large part of the interviewees nationwide complained that these medicines are rarely available and all they receive are prescriptions from the health centers. Many people expressed that they walk long hours, pass rivers and difficult road conditions to get to the health center and end up walking away with a prescriptions they cannot afford. A community leader in RAAN stated that he “has no formal training in medicine and practices natural medicine on people in the community who, because of lack of transport, seek his help for pulling teeth to more complex obstetrical procedures. He states that he helps people with home remedies which he learned empirically by watching Pablo McDavis who taught him natural medicine and by reading documents and books (Wan in Lanka in Miskito, which means our awakening) about natural medicine”.

Leadership as an Enabler (or Hinderer) to Access Opportunities

Strong and organized leadership enable access to opportunities and services for the poor. According to people interviewed in Waslala, other communities nearby are better off now than ten years ago because they are more organized and have strong leadership; the leaders listen to the people and are able to make formal requests for programs in the mayors office. An example often cited by the people as an indicator of bad leadership is the fact that there is no school in the community. The children have to go to the next community to attend school by foot or bike and many do not go because it is not always safe. The quantitative data shows that the average number of children has gone up in this community since 1998 (from 1.3 to 2.1 in 2005) but enrollment in primary school has declined from 1.25 to .87 for the same age cohort. The municipal leadership is also weak, they raise the issue of roads and health during the campaign but have not supported any programs or projects to improve either one. In the case of health, wealthier people have access to private doctors but the majority of people travel for two hours to get health services as there is no health center in the community and they are assigned to the nearest hospital two hours away. In an urban community in Jinotega the leaders were accused of not submitting a proposal to improve the roads and currently the community is the only one in the vicinity that has no adoquinado or road brick project. Another very common leadership source is religion; people often rely on their church leaders for more than spiritual help. Various Christian denomination churches are actively involved in development projects (education, documentation etc) in the communities they inhabit; people often mention church and their source of community cohesion and their pastor as the safety net they go to in times of need.

Corruption and mis-targeting of social programs hinder existing opportunities from reaching the poor. People, in general, cite politics, corruption and theft as big problems in the region for accessing social programs and progressing. Some state that corruption is the reason why some people have been able to improve their economic standing so drastically in less than five years and now own mansions and have cars. In a rural community in RAAN people stated that the leaders have made the community poorer because they only distribute the donated goods that comes for the poor in the community among

46

themselves and the donors get upset and leave. Other reasons contributing to uneven growth within communities is the mis-targeting and mis-allocation of funds from large social programs. State institutions, NGOs and other social program providers fail the very people they are tasked to help by disempowering them through corruption and mistargeting; program designs could attenuate mis-targeting and social exclusion problems by incorporating people’s realities and experiences (Narayan 2000). A small group during a focal interview in the Atlantic region expressed that most social programs are badly targeted; for example providing food to people that do not need it and instead, sell it to the poor in the community for profit. Another example is a road program in the Waslala. The quantitative data shows that at least 50% of the panel households in the community interviewed benefited from a road program. Although the road program benefited most members in the community directly through employment and access (public good) and indirectly from increased transport, secondary students still have no motorized transport available because of divergent schedules. The school schedule (6 am and late in the evening) does not coincide with the transport schedule and there are not enough students to make it worth while for transport providers to switch.

Putting Priorities to the Test

Previous empirical work found evidence to support the argument that the identity of the decision maker influences the type of program or policy put forth rather than reflect on the wishes of their constituency (Chattopadhyay and Duflo 2004). In their research the authors find that women leaders invest in public goods closely linked to their own concerns; in the case of senators in the U.S., Levitt (1996) found that a senator’s ideology is the primary determinant of his voting behavior. The implication is that efficiency is not always a priority to leaders but rather personal preferences and unobserved utility; thereby knowing the identity of the leaders, and their personal situation, can provide good insight into policy decisions to be expected.

This section explores whether the findings of Chattopadhay and Duflo apply to local level leaders in Nicaragua by analyzing the similarities and differences identified by both, leaders and constituents, when given the chance to propose initiatives to improve the community’s well-being. The information is derived based on a semi-experimental game19conducted in 15 communities20 with leaders and their constituency, separately. The participants for this instrument are two distinct groups, the first a group of 4-5 leaders in the community and the second, a group of 4-5 regular people (constituency) living in the community. Participants are given a hypothetical amount of money totaling C$1,350,000 (approximately US$8,000) in two stages, in a structured manner, and asked to assign the funds to anything they believe would contribute to the development of the community. The recording methodology for the themes derived for this exercise is based on Hargreaves et al (2005) where they used wealth ranking instruments to identify the poor and derive categories to describe them. It is important to note that there are drawbacks based on comparability across heterogeneous locations and time using wealth ranking exercises to derive themes because households have idiosyncratic benchmarks and individual points of reference. However this part of the analysis relies solely on information derived from the semi-experimental game, not wealth ranking, to reduce some of the biases resulting from subjectiveness characterizing the wealth ranking exercise, increase comparability across groups and reflect directly the contrast of leaders and constituents propose.

19 The instrument is considered semi-experimental game because it is intended to be exact everywhere and hypothetical money is distributed to identify patterns of behavior across a group. 20 See Castro, Del Carpio, Premand and Vakis (2007) “Do Voices Echo Quantitative data? A Q2 Study of Well-being Dynamics in Nicaragua” for more details on the exact methods applied and the background work done.

47

Table 2.4. Priority Programs as Reported by Leaders and People

LEADERS PEOPLE BOTH

MAJOR THEMES % % %

Water project (drinking and septic) 13.86% 14.07% 13.96%

Construction and repair of street/roads 11.88% 13.03% 12.45%

Productive opportunities 12.87% 11.96% 12.43%

Health center/personnel/goods 10.89% 10.02% 10.47%

House building or improvements(poor or single moms) 6.93% 8.25% 7.58%

School (pre-school, primary and secondary) 7.92% 4.99% 6.49%

Electricity/solar power/street electricity 5.94% 4.12% 5.05%

Vocational school/training 5.94% 4.08% 5.03%

Recreational park/sports for youth 4.95% 4.12% 4.55%

Church 2.97% 1.03% 2.02%

Other 15.84% 24.31% 19.99% Source: Data derived from the outcomes obtained through Semi-Experimental exercise Note: Various Categories are aggregated into one for ease of presentation

Leaders and regular people agree on the top priority: Potable Water. This analysis took into account the project requests, as well as the discussion between participants, of 15 out of all 18 studied in the qualitative work. Drinking water alone was requested 19 times out of 198 total requests; regular people requested a water program (which means from a cured well to a pump) in every region except Atlantic rural (see Table 2.4 and Appendix III Table 1). Leaders in the Atlantic rural region however, made water a priority in their list and listed it more than once. This can be interpreted as an indicator that the inadequacy of water sources is a general problem with a strong likelihood of being true beyond the communities visited. The other water related project (aguas negra or sewages, latrines, septic tanks) directly related to hygiene and indirectly linked to the availability of clean drinking water was also among the top 10 priorities where leaders and common people coincided; neither Managua rural nor Managua urban mentioned it as a concern in either group.

The construction of roads and repair of existing roads is also listed as a top priority by leaders and people in general. Roads, whether adoquinado (road brick construction) or repair, are mentioned by the leaders in all communities visited as priorities for improving well-being. This theme includes roads inside the community (streets) or roads leading to the urban area; other communication programs mentioned as lacking and affecting progress are bridges and a pier in the Atlantic rural and infrastructure maintenance in the Atlantic urban region. Having water and roads as the two top priorities on the part of leaders and people can be perceived as a positive for Nicaragua. Some of the development literature finds that collective investments have more equitable benefits than individual investment and contribute more to building community capacity through cooperative consumption (Rao 2001).

Housing for the poor is a common theme mentioned as a development concern but does not rank high on the priority list of development projects. It is interesting to note that in most of the discussions during the social mapping and the wealth ranking exercises in most communities housing quality was mentioned as a priority—in terms of material, social and physical well-being—by leaders and people in general. However, during the semi-experimental exercise, leaders did not place it high in the priority list or assign large funds toward remedying this problem. The regular people mentioned throughout their discussion that inadequate housing for the destitute that live in plastic houses with no roof was a problem that mostly affected single mothers with small children who pass cold weather and endure rain in their makeshift homes. However, even though inadequate housing seems to be a concern nationwide in every community except rural Managua, it still ranks secondary to other projects.

48

Health and education, infrastructure and services, are at the top of the list for leader and people. Leaders and regular people mentioned that the lack of a pre-school, inadequate size primary, a local secondary school and a health center inhibited the development capacities and limited their progress. Both, the services (paying for doctors and nurses and buying books and hiring more teachers) and the infrastructure appear to be equally important; and priorities to both groups. Scholarships for secondary students and vocational training for carpentry, computers, sewing and other skills were also mentioned as projects with strong potential given the demands for their skills in the market. In terms of health, the chronically ill are identified as a vulnerable group in need of assistance; the short supply of free medicines for this segment of the population is a common concern among leaders and regular people in the Atlantic and Central regions.

Having limited productive opportunities is a resonant theme nationwide. The limited availability of productive tools, skills and resources leads both leaders and people to believe that programs such as a locally managed credit fund, agricultural inputs, livestock (cows, chickens and pigs), the creation of a local market and a distributing center are part of the answer to development in their communities. In table 4 it can be observed that both leaders and regular people (13% and 12% respectively) place productive opportunities among the highest categories of programs they perceive as important. The 24% allocated to the other category under the people column represents various poverty themes in the social dimension; for example food for the poor, help for the disabled, help for street children, single mothers, old folks home and assistance in financing holiday celebrations. Leaders tend to assign resources toward infrastructure and productive projects such as those mentioned in table 3, as well as the purchase of land for cultivation. 3% of the projects assigned by the leaders and not included in the table below assign funds to creating businesses in the community such as a bakery; regular people concentrate more of the resources in helping women and enhancing security in the community.

Recreational outlets for the community, such as sports complexes, parks and gathering centers (casa base) rank high for both regular people and leaders. In terms of other less obvious necessities that contribute to people’s perceptions are activities for the youth in the community to keep them from leaving and from engaging in crime. A recreational complex was unanimously proposed during this exercise as a potential investment because it would motivate the youth to stay away from delinquency and encourage cohesiveness among community members. In a rural community in the Atlantic north drugs have become a problem for the youth; a person stated that “they not only traffic it but also consume it”. Some people shared with the facilitators through a focal group that the youth are engaging in other criminal activities because they have no activities to distract them, no job opportunities and feel pressure to contribute to the income of the household. People of all ages mentioned parks and casa base (gathering house) as necessary for community dynamics and to deter the youth from leaving the community. In more than one occasion young people stated their desire to move out of the community because there are low or no opportunities to study, learn, and have fun or access the internet or a “cyber”.

The youth aspiration module reveals four priorities for young people nationwide: Jobs, education, recreational facilities and family. The findings related to the youth found through the semi-experimental exercise were limited because the participation of young people in that activity was minimal because the exercise was conducted while children and young adults were in school or working. However it is important to note that the findings from the semi-experimental resonate with findings in all community visits during the youth aspirations focus groups, for both males and females. In all communities, four areas of priority are mentioned: Access to job opportunities, ability to continue education, access to sports related infrastructure and having children and a spouse. Males tend to focus on jobs, sports activities (soccer and baseball) and vocational education. Women focus more on regular education with some mention of vocational training, having a husband and children, and having entertainment activities related to church as well as a gathering place to hang-out with other women their age. Most of the young people interviewed believe that education is the solution to escaping poverty; one

49

young lady in the Atlantic urban said “I want to study more to have more, it’s possible if I study, I know I can be better off”. Another important finding that is widely shared among both genders is that they want to have a better job than their parents because their life and their parent’s lives are too hard and they want a better life for their children. One woman in RAAN said “I don’t want to wash clothe and iron for a living like my mother, life is too tough that way”.

Conclusions

The information obtained through the use of participatory methods sheds light on differences in behavior across households, and regions, that were not obvious through panel data. Both methods, qualitative and quantitatively provide insight into what constitutes well-being. Quantitatively, poverty estimates find that poverty has not reduced in a significant manner in Nicaragua over the last eight years; qualitatively people perceive themselves as poor despite many changes experienced in the last decade. This paper does not seek to find whether poverty has gone up or down but rather whether the quantitative definition of poverty, to include its determinants, resemble the definition given by people and whether factors beyond survey data can explain mobility and identify inequalities that hinder progress. This analysis shows that people coincide in many dimensions by which they measure poverty with those used in quantitative analysis. However, people also include unobservable characteristics such as happiness, history and hope in their perception of well-being thereby making it difficult to measure accurately. Despite the large volume of data collected, the Q-squared approach revealed that there are five key messages that emerged through the analysis that not only reflect the findings of the quantitative data but also the voices of the people in Nicaragua.

First, infrastructure availability and quality affects people’s perception of well-being. This is evidenced by the importance placed on housing—owning versus renting or squatting—and through the placement of water, roads, electricity and recreational facilities at the top of the priority list. The inadequate quality of school infrastructure—typically too small to match the number of children attending or made of materials that are not suitable for a school in all weather conditions—reduces participation levels. The second finding, quality, accessibility and affordability of services influence people’s willingness to use them and overall human development outcomes. Water, health and education are three notable examples. People mentioned the existence of wells in the community and in-house piping but the majority noted that the water is inaccessible due to contamination, malfunctioning equipment or dryness of the wells. In the case of education, long distance to schools inhibit children from attending in bad weather conditions, rough terrains and places with security concerns; it appears that girls and young (entry level) children are the most affected. In terms of health, the inaccessibility of health personnel and free medicines deter people from utilizing health posts and centers. Long distances to hospitals and health clinics (mostly from rural areas), particularly in the Atlantic and Central regions, contribute to lower use of formal health care; this may explain regional disparities and lower health outcomes.

The third key finding is how crucial good leadership and organization are to accessing opportunities. This message is evident throughout the country; leadership either enables or hinders access to services and programs, depending on the strength, interest and influence of the leaders. Various communities visited have low organizational capacity and exhibit weaker social cohesion due to the lack of leadership. People cite the war—through massive displacements, loss of assets and lingering fear—as a reason for existing polarization and disorganization within communities and disparities in progress between regions. Through this analysis it is also observed that NGOs, particularly religious organizations, often fill the void in leadership and organization at the community level. Church leaders act as substitutes to formal leadership, and the institution of the church complements (or substitutes) other forms of institutionalized leadership such as elected or appointed political groups.

50

The fourth key message is the importance of good employment opportunities for all skill levels, locations and age groups. Job types differ greatly across regions, depending on the industries located in the region and the skills required by employers. Urban activities tend to revolve around low skilled production and formal and informal commerce. Low capacity agriculture is still common among rural inhabitants however the negative effects of weather shocks appear to limit productivity among small producers and hinder growth opportunities. Young people express a desire to do something different than their parents, in both rural and urban, and want professions such as teachers, engineers and various medical professions; migration is reported as a very desirable option. People want good jobs, defined as not being exploitative and having benefits. In RAAN natural resource based activities and commerce of primary products are typical. In RAAS, however activities veer toward services in the formal sector and drug trade in the informal sector. Both regions benefit from their natural resource endowments and proximity to the ocean. The fifth finding is the qualitative affirmation of the relationship between higher educational attainment and poverty alleviation. Informants measure others’ economic success by various measures and educational attainment is among the top. The youth view education as the path out of poverty and often refer to it as the reason why some have been able to do better, or in the absence of it as the reason why some are poor. Vocational education and skill transfer programs are popular requests among young adults and people with low or no formal skills.

51

REFERENCES

Azzi, C. and R. Ehrenberg (1975). “Household Allocation of Time and Church Attendance”. Journal of Political Economy, Number 83 (February).

Bourguignon, F. (2003). “Qualitative and Quantitative Approaches to Poverty Analysis: Two Pictures of the Same Mountain?” Pages 68-72 in R. Kanbur, editor. Q-Squared: Combining Qualitative and Quantitative Methods in Poverty Appraisal.

Castro, V., X. Del Carpio, P. Premand and R. Vakis (2007). “Do Voices Echo Quantitative Data? A Q2 Study of Well-being Dynamics in Nicaragua” Draft in progress

Castro, V. and X. Del Carpio (2006). “Voices of Nicaragua: Qualitative Research Manual”. Final draft

Chattopadhyay, R and E. Duflo (2004). “Women as Policy Makers: Evidence from a Randomized Policy Experiment in India”. Econometrica, Vol. 72, No. 5 (Septemeber)

Easterlin, R. (1995). “Will Raising the Incomes of All Increase the Happiness of All?”. Journal of Economic Behavior and Organization 27 (1)

Fafchamps, M. and F. Shilpi (2006). “Subjective Welfare, Isolation, and Relative Consumption”. Mimeo

Frey, B. and A. Stutzer (2002). “What Can Economists Learn from Happiness Research?”. Journal of Economic Literature XL

Hargreaves, J.R., L.A. Morison, J.S.S. Gear, J.D.H. Porter, M.B. Makhubele, J.C. Kim, J. Busza, C. Watts, P.M. Pronyk (2005). “Hearing the Voices of the Poor”: Assigning Poverty Lines on the Basis of Local Perceptions of Poverty: A Quantitative Analysis of Qualitative Data from Participatory Wealth Ranking in Rural South Africa’, Q-Squared Working Paper no. 4, Toronto: Centre for International Studies, University of Toronto.

Layard, R. (2002). “Rethinking Public Economics: Implications of Rivalry and Habit.” mimeo

Levitt, S. D. (1996). “How Do Senators Vote? Disentangling the Role of Voter Preferences, Party Affiliation and Senator Ideology”. American Economic Review, 86

Lokshin, M. and M. Ravallion (2002). “Self-Rated Economic Welfare in Russia”. European Economic Review, 46(8) (Septemeber)

Mani, D. (2001). Data Analysis and Interpretation Integrating Quantitative and Qualitative Data. Seminar on Vulnerability and Assessment. United Nations Center for Regional Development. www.uncrd.org.jp/hs/doc/01a_mani.pdf

McCleary, R. and R. Barro (2006). “Religion and Economy”. Journal of Economic Perspectives, Vol. 20, Number 2 (Spring).

Narayan, D., R. Patel, K. Schafft, A. Rademacher and S. Koch-Schulte (2000). Voices of the Poor: Can Any-one Hear Us?, New York: Oxford University Press.

Nguyen, N. and M. Rama (2007). “A Comparison of Quantitative and Qualitative Poverty Targeting Methods in Vietnam”. Q-Squared Working Paper No. 32

52

Rao, V. and M. Woolcock (2004). Integrating Qualitative and Quantitative Approaches in Program Evaluation. Pages 165-190 (Chapter 8) in F. Bourguignon and L.A. da Silva, editors. The Impact of Economic Policies on Poverty and Income Distribution: Evaluation Techniques and Tools. World bank & Oxford University Press, NY

Ravallion, M. (2003). “Qualitative and Quantitative Approaches to Poverty Analysis: Two Pictures of the Same Mountain?” Pages 68-72 in R. Kanbur, editor. Q-Squared: Combining Qualitative and Quantitative Methods in Poverty Appraisal.

Ravallion, M. and M. Lokshin (2005). “Who Cares about Relative Deprivation?”. World Bank Policy Research Working Paper 3782, (December)

United Nations Development Program (2005). Informe de Desarrollo Humano 2005. Las Regiones Autonomas de la Costa Caribe: Nicaragua Asume su Diversidad?

Van Campenhout, B. (2006). “Locally Adapted poverty Indicators Derived from Participatory Wealth Rankings: A Case of Four Villages in Rural Tanzania”. Journal of African Economies (December)

World Development Report 2006 (2007). Equity and Development. World Bank. A co-publication of the World Bank and Oxford University Press

53

Ann

ex A

.1

Tab

le A

.1a.

Def

inin

g P

over

ty th

roug

h th

e V

oice

s of

Nic

arag

uans

(Mat

eria

l Wel

l-be

ing)

Des

titu

teP

oo

rN

on

-po

or

Wel

l-o

ffH

ou

sin

gO

ften

a da

y-to

-day

unc

erta

inty

. P

eopl

e w

ho m

ostly

squ

atte

r, le

gally

or

ille

gally

and

dep

end

on o

ther

s ki

ndne

ss fo

r ho

usin

g. O

ften

loca

ted

in th

e le

ast d

esira

ble

area

in th

e co

mm

unity

. In

the

urba

n ar

ea th

ese

peop

le a

re c

onsi

dere

d ho

mel

ess

and

have

dif

Live

in u

nsta

ble

hous

es o

ften

mad

e of

lo

w q

ualit

y m

ater

ials

(di

rt, c

ardb

oard

, pl

astic

, jun

k an

d re

mna

nts)

. S

mal

ler

in

size

(on

e ro

om).

Do

not n

orm

ally

hav

e fo

rmal

title

. Lo

cate

d in

a r

emot

e pa

rt o

f th

e co

mm

unity

or

far

from

prim

ary

serv

ices

Hou

sing

mad

e of

per

man

ent

mat

eria

ls, f

loor

s m

ade

of c

emen

t (n

on-d

irt)

with

mor

e th

an o

ne r

oom

. F

orm

al ti

tle is

not

unc

omm

on.

Loca

tion

varie

s, o

ften

clos

e to

ne

wly

inst

alle

d se

rvic

es

Hou

sing

mad

e of

per

man

ent

mat

eria

ls, f

inis

hed

floor

s, w

ith m

ore

than

one

roo

m. H

as b

een

impr

oved

or

rece

ntly

bui

lt. F

orm

al ti

tle o

f pro

pert

y an

d of

ten

clos

e to

the

vent

a a

nd

serv

ices

in th

e co

mm

unity

Lan

d (

rura

l an

d s

emi-

rura

l)T

hey

lack

acc

ess

to c

ultiv

able

land

an

d ca

nnot

aff

ord

inpu

ts fo

r fa

rmin

g.

Whe

n th

ey fa

rm, t

hey

have

to r

ent

land

by

seas

on p

ay in

kin

d at

the

end

at a

ver

y hi

gh c

ost.

O

ccas

iona

lly h

ave

acce

ss to

a s

olar

(k

itche

n ga

rden

)

The

y ha

ve a

cces

s to

low

qua

lity

culti

vabl

e (r

entin

g/sh

arin

g) a

nd o

ften

cann

ot a

ffor

d in

puts

for

farm

ing.

The

y pa

y fo

r la

nd a

nd in

puts

to s

omeo

ne

else

at a

ver

y hi

gh c

ost.

Do

not h

ire

moz

os,

wor

k w

ith o

ther

fam

ily

mem

bers

. P

lant

bas

ic c

rops

and

eat

m

ost o

Hav

e ac

cess

to la

nd, h

ave

reso

urce

s to

buy

inpu

ts o

r ac

cess

to

cre

dit t

o bu

y th

em.

Wor

k al

ongs

ide

hire

d la

bor

to p

lant

and

ha

rves

t. O

ften

farm

bas

ic c

rops

in

larg

er q

uant

ities

, vul

nera

ble

to

wea

ther

sho

cks.

Hav

e a

sola

r an

d us

e it

for

hous

ehol

d co

nsum

p

Hav

e ac

cess

to la

nd, h

ave

reso

urce

s to

buy

inpu

ts a

nd a

cces

s to

cre

dit

whe

n ne

eded

. H

ire o

utsi

de h

elp

to

plan

t and

har

vest

. O

ften

farm

a

dive

rsity

of p

rodu

cts

but d

o no

t de

pend

sol

ely

on th

eir

land

for

inco

me.

Som

e do

not

farm

for

com

mer

ce a

nd u

se th

e

Cat

tle/

anim

als

Don

't ha

ve a

nyR

arel

y ha

ve c

attle

, onl

y ch

icke

ns a

nd

pigs

(<2

)H

ave

cattl

e an

d va

rious

pat

io

anim

als

(<10

)H

ave

cattl

e (2

0+)

Pro

du

ctiv

e A

sset

s (o

ther

)D

on't

have

any

Hav

e ac

cess

to fa

rm to

ols

but d

o no

t ow

n th

em, a

cces

s to

bes

tias

(mul

e or

ho

rse)

. U

se a

sset

s to

hed

ge a

gain

st

any

econ

omic

dow

ntur

n in

duce

d by

ill

ness

, wea

ther

sho

cks

and

hung

er

Ow

n ba

sic

farm

ing

tool

s, h

ave

acce

ss to

a m

otor

ized

ass

et (

car

or

boat

), o

wn

bicy

cle

and/

or b

estia

s (h

orse

or

mul

e).

Use

ass

ets

as

savi

ngs

and

to p

ay fo

r dr

astic

ec

onom

ic c

hang

es

Ow

n to

ols

need

ed fo

r ca

ttle

and/

or

farm

ing,

ow

n a

mot

oriz

ed a

sset

and

re

nt it

to o

ther

s

Material

Dim

ensi

on

sQ

ual

itat

ive

Mea

sure

s

54

Tab

le A

.1b.

Def

inin

g P

over

ty th

roug

h th

e V

oice

s of

Nic

arag

uans

(Phy

sica

l Wel

l-be

ing)

Des

titu

teP

oo

rN

on

-po

or

Wel

l-o

ffO

uts

ide

inco

me

op

po

rtu

nit

ies

Prin

cipa

lly u

nem

ploy

ed, l

ives

off

ch

arity

from

fam

ily o

r co

mm

unity

m

embe

rs, d

onat

ions

from

pro

gram

s.

Whe

neve

r po

ssib

le w

ork

in

agric

ultu

re.

Wan

ts to

mig

rate

for

wor

k bu

t can

rar

ely

affo

rd to

go

and

com

e ba

ck.

Reg

ular

ly u

nem

ploy

ed fr

om th

e fo

rmal

se

ctor

, liv

es o

ff a

gric

ultu

re (

moz

o) o

r in

form

al s

pora

dic

jobs

(w

ashi

ng

clot

hes,

sel

ling

woo

d).

Mig

rate

s te

mpo

raril

y to

wor

k in

agr

icul

ture

(i.e

. co

ffee

in N

icar

agua

and

cot

ton

in e

l sa

lvad

or o

r cr

ops

in C

osta

Ric

a).

Nor

mal

ly in

and

out

of t

he fo

rmal

se

ctor

. E

ngag

es in

sm

all

com

mer

ce (

cuaj

ada,

milk

etc

) or

ha

s a

prof

essi

on a

s a

teac

her,

m

unic

ipal

ity o

r he

alth

cen

ter.

Ofte

n ha

s a

fam

ily m

embe

r w

ho r

emits

re

gula

rly.

In th

e A

tlant

ic th

e m

ost

com

mon

act

ivity

is e

mba

rc

Eith

er p

artic

ipat

es in

a p

rofe

ssio

n in

th

e fo

rmal

sec

tor

or is

eng

aged

fully

in

com

mer

ce.

Ow

ns v

enta

s an

d em

ploy

s pe

ople

in th

e co

mm

unity

. O

ften

has

mig

rant

/rem

itter

s w

ho li

ve

outs

ide

the

coun

try.

Inv

ests

reg

ular

ly

and

trav

els

to th

e m

unic

ipal

hea

dqua

r

Sel

f in

com

e o

pp

ort

un

itie

sN

one

Sm

all a

gric

ultu

re (

sola

r), f

ishi

ng, w

ood

cutti

ng a

nd g

old

digg

ing

in th

e A

tlant

ic.

Sel

l pro

duct

ion

ahea

d to

inve

stor

s fr

om o

utsi

de to

fina

nce

inpu

ts

Has

a s

tabl

e ag

ricul

tura

l inc

ome

from

ow

n la

nd o

r ow

ns a

sm

all

pulp

eria

or

mis

cella

neou

s st

ore.

F

ishi

ng a

nd w

ood

cutti

ng fo

r co

mm

erce

in th

e lo

cal w

ater

s an

d fo

rest

s. S

ell p

rodu

ctio

n to

inve

stor

s fr

om th

e ou

tsid

e at

the

end

of th

e ha

rves

t at a

sta

ndar

d

Ow

ns la

nd, c

attle

and

com

mer

cial

ac

tiviti

es (

mun

icip

al m

arke

t and

w

hole

sale

sal

es).

Sel

ls o

r bu

y pr

oduc

tion

in a

larg

er s

cale

. O

wns

ac

opio

for

fish

and

shrim

p or

cof

fee.

Wat

erN

o ac

cess

to w

ater

sou

rces

oth

er

than

the

river

, "oj

os d

e ag

ua/v

ertie

ntes

" or

com

mun

al w

ells

. O

ften

cons

ume

cont

amin

ated

w

ater

s to

sur

vive

. C

hild

ren

gath

er

wat

er n

orm

ally

Hap

haza

rd s

ourc

es o

f wat

er a

cces

s,

river

s, c

omm

unal

wel

l, "o

jos

de

agua

/ver

tient

es"

or c

omm

unal

wel

ls.

Doe

s no

t nor

mal

ly u

se c

hlor

ine

and

cons

umes

con

tam

inat

ed w

ater

. C

hild

ren

gath

er w

ater

nor

mal

ly

Con

sum

es w

ell w

ater

, ow

n or

co

mm

unal

, ofte

n ha

s pi

ped

wat

er

that

is p

ushe

d by

gra

vity

and

oc

casi

onal

ly b

y eq

uipm

ent.

Use

s ch

lorin

e or

boi

ls w

ater

reg

ular

ly.

Buy

s w

ater

from

oth

er s

ourc

es

whe

n no

t ava

ilabl

e lo

cally

Nor

mal

ly p

iped

wat

er in

urb

an a

reas

or

gra

vity

in r

ural

are

as.

Buy

s w

ater

fr

om tr

uck

whe

n pi

ping

not

ava

ilabl

e lo

cally

. T

ends

to o

wn

irrig

atio

n sy

stem

s fo

r ag

ricul

ture

.

Fo

od

(ac

cess

)Ir

regu

lar

(0 to

2 ti

mes

per

day

), lo

w

nutr

ition

al v

alue

(to

rtill

a w

ith s

alt)

Irre

gula

r in

term

s of

tim

es p

er d

ay

(fro

m 1

to 3

). L

ow n

utrit

iona

l qua

lity

and

low

var

iety

. Li

mite

d ac

cess

to

loca

l sto

re fo

r ba

sic

food

s

Reg

ular

ly a

vaila

ble,

of a

dequ

ate

nutr

ition

al v

alue

. A

cces

s to

cre

dit

in th

e lo

cal s

tore

whe

n tim

es a

re

toug

h

Eat

s as

nee

ded,

var

iety

ava

ilabl

e an

d ac

cess

to o

utsi

de fo

ods

Physical

Qu

alit

ativ

e M

easu

res

Dim

ensi

on

s

55

Tab

le A

.1c.

Def

inin

g P

over

ty th

roug

h th

e V

oice

s of

Nic

arag

uans

(Hum

an W

ell-

bein

g)

Des

titu

teP

oo

rN

on

-po

or

Wel

l-o

ffE

du

cati

on

(ac

cess

)A

cces

s to

prim

ary

scho

ol b

ut n

o m

oney

to p

ay fo

r sc

hool

sup

plie

s an

d fe

es.

No

acce

ss to

sec

onda

ry

scho

ol

Acc

ess

to p

rimar

y sc

hool

but

lim

ited

mon

ey to

pay

for

scho

ol s

uppl

ies

and

fees

and

will

pul

l kid

s to

wor

k w

hen

need

ed.

No

acce

ss to

sec

onda

ry

scho

ol b

ecau

se o

f dis

tanc

e an

d fe

es

Acc

ess

to p

rimar

y sc

hool

, can

us

ually

cov

er th

e fe

es a

nd c

hoos

e pr

ivat

e sc

hool

whe

n av

aila

ble.

P

aren

ts c

ompl

ain

abou

t the

qua

lity

of te

ache

rs a

nd s

ervi

ce.

Acc

ess

to

seco

ndar

y an

d vo

catio

nal s

choo

l in

term

s of

fund

s bu

t lim

ited

reso

urce

s m

ake

it di

ffic

ul

Acc

ess

to a

ll sc

hool

ing

type

s an

d un

iver

sity

for

child

ren.

In

rura

l are

as

child

ren

tend

to v

alue

voc

atio

nal

trad

es o

ver

univ

ersi

ties

and

aspi

re to

be

eng

inee

rs a

nd te

ache

rs.

Ed

uca

tio

n (

par

tici

pat

ion

)C

hild

ren

do n

ot n

orm

ally

go

to

scho

ol u

nles

s th

ey a

re p

art o

f a

prog

ram

that

cov

ers

expe

nses

or

ther

e is

a fo

od p

rogr

am a

ttach

ed to

it.

No

part

icip

atio

n in

sec

onda

ry

scho

ol o

r as

pira

tion

for

it.

Chi

ldre

n go

to s

choo

l irr

egul

arly

, are

of

ten

behi

nd b

y a

few

yea

rs a

nd a

re

pulle

d of

f dur

ing

plan

ting

and

harv

est

seas

on in

the

rura

l are

a an

d to

wor

k or

lo

ok fo

r fo

od in

the

urba

n ar

ea.

Ext

rem

ely

rare

par

ticip

atio

n in

se

cond

ary

scho

ol b

ut s

tron

g ho

pe fo

Chi

ldre

n go

to s

choo

l reg

ular

ly a

nd

are

norm

ally

at g

rade

leve

l.

Wea

ther

con

ditio

ns a

ffec

t pa

rtic

ipat

ion

of s

econ

dary

sch

ool

stud

ents

and

wor

k be

com

es a

n al

tern

ativ

e or

cho

ice

for

stud

ents

af

ter

6th

grad

e. O

ften

atte

ndin

g w

eeke

nd s

econ

dary

or

voca

tiona

Par

ticip

ate

in a

ll sc

hool

ing

leve

ls

unle

ss k

id p

ulls

out

due

to d

islik

e fo

r ed

ucat

ion

Hea

lth

(ac

cess

)A

cces

s to

the

loca

l hea

lth c

ente

r bu

t fe

els

disc

rimin

ated

and

ofte

n no

t tr

eate

d in

a d

igni

fied

man

ner.

Sel

f-di

agno

ses.

No

acce

ss to

priv

ate

doct

ors

or th

e ho

spita

l due

to

tran

spor

t cos

ts.

Acc

ess

to th

e lo

cal h

ealth

cen

ter

but

com

plai

ns o

f lac

k of

med

icin

es a

nd

occa

sion

ally

of t

he q

ualit

y of

ser

vice

pr

ovid

ed.

No

acce

ss to

priv

ate

doct

ors

and

limite

d ac

cess

to h

ospi

tals

due

to

tran

spor

t.

Acc

ess

to h

ealth

ser

vice

s bu

t co

mpl

ains

of q

ualit

y an

d av

aila

bilit

y of

med

icin

es a

nd d

octo

rs.

Pre

fers

pr

ivat

e pr

ovid

ers

but u

ses

them

oc

casi

onal

ly.

Acc

ess

to h

ealth

ser

vice

s bu

t co

mpl

ains

of q

ualit

y an

d av

aila

bilit

y of

m

edic

ines

and

doc

tors

. P

refe

rs

priv

ate

prov

ider

s lo

cate

d in

the

mun

icip

al c

ente

r an

d us

es th

em.

Hea

lth

(af

ford

abili

ty)

Can

not a

ffor

d th

e co

sts

and

self-

med

icat

es m

ost o

f the

tim

e. A

ttend

s on

ly if

a s

ocia

l pro

gram

is a

ttach

ed

to th

e ce

nter

and

pro

vide

s th

em w

ith

food

Can

not u

sual

ly a

ffor

d th

e co

sts,

sel

ls

asse

ts to

pay

for

them

or

relie

s on

so

cial

pro

gram

s to

fund

par

ticip

atio

n in

va

ccin

atio

n. T

ends

to g

o fo

r th

e ch

ildre

n an

d to

com

ply

for

a fo

od

prog

ram

or

othe

r so

cial

pro

gram

. U

ses

cura

nder

os a

nd n

atur

al r

emed

ies.

Can

aff

ord

to g

o to

the

doct

or a

nd

pay

for

med

icin

es b

ut o

ften

sells

as

sets

to b

uy th

em w

hen

the

prob

lem

is s

erio

us.

Occ

asio

nally

us

es lo

cal c

uran

dero

s bu

t rel

ies

on

stan

dard

med

icin

e.

Can

aff

ord

to g

o to

the

doct

or a

nd p

ay

for

med

icin

es b

ut o

ften

sells

ass

ets

to

buy

them

whe

n th

e pr

oble

m is

se

rious

. O

ccas

iona

lly u

ses

loca

l cu

rand

eros

but

rel

ies

on s

tand

ard

med

icin

e.

Ele

ctri

city

Non

e, w

ood

is to

o ex

pens

ive

Rar

ely

have

ele

ctric

ity, u

ses

woo

d or

ca

ndle

in r

ural

are

as o

r ill

egal

ene

rgy

in

the

urba

n ar

eas

Has

ele

ctric

ity w

hen

avai

labl

e or

us

es fu

el o

r w

ood

to c

ook

whe

n in

ru

ral a

reas

. U

ses

batte

ry fo

r ra

dios

an

d te

levi

sion

and

occ

asio

nally

so

lar

ener

gy

Has

ele

ctric

ity w

hen

avai

labl

e or

use

s ba

ttery

, fue

l or

woo

d to

coo

k w

hen

in

rura

l are

as.

Sol

ar e

nerg

y is

not

un

com

mon

in r

ural

are

as

Human

Qu

alit

ativ

e M

easu

res

Dim

ensi

on

s

56

Tab

le A

.1d.

Def

inin

g P

over

ty th

roug

h th

e V

oice

s of

Nic

arag

uans

(Soc

ial W

ell-

bein

g)

Des

titu

teP

oo

rN

on

-po

or

Wel

l-o

ffP

rod

uct

ive

or

inco

me

pro

gra

ms

Not

a ta

rget

hou

seho

ld fo

r pr

oduc

tive

prog

ram

sT

arge

ted

by s

ome

prod

uctiv

e pr

ogra

ms

but d

o no

t alw

ays

part

icip

ate

beca

use

lack

of l

and

owne

rshi

p, ti

me,

inab

ility

to

com

mit

or c

ircum

stan

ces

Nor

mal

ly b

elon

gs to

ass

ocia

tions

en

gage

d in

pro

duct

ive

ende

avor

s an

d te

nds

to b

e th

e pr

omot

er o

f th

ese

prog

ram

s or

cou

nter

part

s lo

cally

. S

ome

peop

le in

this

gro

up

are

pens

ion

rece

pien

ts

Not

targ

eted

by

prog

ram

s an

d pr

efer

s no

t to

be p

art o

f one

.

Fo

od

pro

gra

ms

Pot

entia

lly ta

rget

ed if

it r

equi

res

no

cond

ition

ality

but

ofte

n m

isse

d be

caus

e la

cks

docu

men

tatio

n or

st

able

add

ress

Tar

gette

d fo

r m

ost f

ood

for

wor

k pr

ogra

ms

and

othe

r fo

od p

rogr

ams

Tar

gette

d by

food

pro

gram

s if

hous

ehol

d is

loca

ted

in a

rel

ativ

ely

poor

com

mun

ity.

Ofte

n or

gani

zes

and

lead

s pr

ogra

m a

nd fo

od

dist

ribut

ion.

Vol

unte

ers

to h

elp

with

ta

rget

ting

and

stor

ing

the

food

Not

targ

eted

by

prog

ram

s bu

t oc

casi

onal

ly in

clud

ed.

Sel

ls, g

ives

aw

ay fo

od o

r us

es it

for

anim

al

feed

ing.

Co

mm

un

ity

coh

esio

nN

ot p

art o

f com

mun

ity a

ctiv

ities

, onl

y el

derly

peo

ple,

wid

ows

and

som

e si

ngle

mot

hers

in th

is c

ateg

ory

are

reco

gniz

ed a

s m

embe

rs o

f the

co

mm

unity

and

rec

eive

hel

p fr

om

neig

hbor

s

Par

t of t

he c

omm

unity

but

exc

lude

d fr

om le

ader

ship

rol

es n

orm

ally

. R

ecei

ve s

ome

help

from

nei

ghbo

rs a

nd

get i

nfor

mal

loan

s fr

om p

eopl

e w

hen

they

are

est

ablis

hed

in th

e co

mm

unity

Ofte

n in

volv

ed in

lead

ersh

ip r

oles

or

rega

rded

as

lead

ers

by o

ther

s.

Fee

l th

e po

or a

re o

ften

at fa

ult f

or

not u

sing

thei

r re

sour

ces

wis

ely

"mal

a ca

beza

" an

d ev

en e

ngag

e in

vi

ces

Ofte

n di

scon

nect

ed fr

om th

e co

mm

unity

in te

rms

of s

ocia

l act

iviti

es

and

pref

ers

to lo

ok o

utw

ard

for

soci

al

inte

ract

ion.

The

y in

tera

ct w

ith

com

mun

ity le

ader

s m

ostly

.

Em

po

wer

men

t F

eel v

oice

less

and

hea

vily

de

pend

ent o

n lu

ck a

nd k

indn

ess

of

othe

rs.

Sin

gle

wom

en w

ith c

hild

ren

and

wid

ows

feel

par

ticul

arly

di

sem

pow

ered

and

exp

ress

littl

e ho

pe in

the

syst

em.

Oth

ers

refe

r to

pe

ople

in th

is c

ateg

ory

as p

ower

less

an

d ex

clud

ed.

The

you

t

Exp

ress

dis

cont

ent w

ith th

e sy

stem

(p

oliti

cal a

nd s

ocia

l nor

ms)

. O

ften

feel

fr

ustr

ated

by

how

they

are

trea

ted

by

othe

rs (

empl

oyer

s, n

eigh

bors

, loc

al

polit

icia

ns a

nd s

ervi

ce p

rovi

ders

).

The

y re

ly o

n re

ligio

n an

d ex

pres

s hi

gher

leve

ls o

f hop

e th

an th

e de

Ten

d to

be

lead

er ty

pes,

eng

aged

an

d so

mew

hat p

ower

ful a

t the

loca

l le

vel.

Ofte

n ex

pres

s fr

ustr

atio

n w

ith o

utsi

de p

ower

s an

d fe

el

forg

otte

n by

hig

her

leve

l pol

itici

ans

and

soci

al p

rogr

ams.

Rel

igio

n is

im

port

ant t

o th

is g

roup

and

re

cogn

ize

impo

rtan

ce o

f

Ten

d to

be

natu

ral l

eade

rs in

the

com

mun

ity b

ut o

ften

choo

se to

be

less

atta

ched

to lo

cal g

estio

nes

. S

omet

imes

thei

r re

latio

nshi

p to

oth

ers

in th

e co

mm

unity

is n

ot v

ery

good

be

caus

e th

ey fe

el e

nvie

d an

d di

ffer

ent.

Dep

endi

ng o

n ho

w w

ealth

w

as a

cqui

red

(l

Social

Qu

alit

ativ

e M

easu

res

Dim

ensi

on

s

57

Ann

ex I

I. M

ap o

f Qua

litat

ive

Mun

icip

alit

ies

LE

GE

ND

MU

NIC

IPA

LIT

IES

RE

GIO

N1

SA

N J

UA

N D

EL

SU

R (

RU

)P

AC

IFIC

2E

L V

IEJO

(U

R)

PA

CIF

IC3

TE

LIC

A (

RU

)P

AC

IFIC

4LA

CO

NC

EP

CIO

N (

RU

)P

AC

IFIC

5S

AN

RA

FA

EL

DE

L S

UR

MA

NA

GU

A6

MA

NA

GU

A (

UR

)M

AN

AG

UA

7JI

NO

TE

GA

(U

R)

CE

NT

RA

L8

QU

ILA

LI (

RU

)C

EN

TR

AL

9M

AT

AG

ALP

A (

RU

)C

EN

TR

AL

10C

IUD

AD

DA

RIO

(R

U)

CE

NT

RA

L11

BLU

FIE

LDS

(U

R)

RA

AS

12S

IUN

A (

UR

)R

AA

N13

NU

EV

A G

UIN

EA

(U

R)

RA

AN

14W

AS

PA

M (

RU

)R

AA

N15

EL

RA

MA

(R

U)

RA

AS

16W

AS

LALA

(R

U)

RA

AS

17R

OS

ITA

(R

U-M

AY

AG

NA

)R

AA

N

Not

e: S

an R

afae

l del

Sur

and

Blu

efie

lds

have

2 c

omm

uniti

es e

ach

58

Ann

ex I

II. T

hem

es I

dent

ifie

d by

the

Peo

ple

Dis

aggr

egat

ed b

y L

ocat

ion

198

AT

L-R

UA

TL-

UR

CE

N-R

UC

EN

-UR

PA

C-R

UP

AC

-UR

MA

N-R

UM

AN

-UR

%A

TL-

RU

AT

L-U

RC

EN

-RU

CE

N-U

RP

AC

-RU

PA

C-U

RM

AN

-RU

MA

N-U

R%

TO

TA

L A

LL%

BO

TH

Wat

er p

roje

ct (

drin

king

)3

21

12

8.91

%2

21

11

12

10.3

1%19

9.60

%

Hou

se b

uild

ing

or im

prov

emen

ts (

poor

or

sing

le m

oms)

12

21

16.

93%

11

21

11

18.

25%

157.

58%

Con

stru

ctio

n S

tree

t/roa

ds2

11

11

11

18.

91%

11

12

11

7.22

%16

8.08

%

Hea

lth c

ente

r/pe

rson

nel

21

11

26.

93%

11

21

16.

19%

136.

57%

Wat

er p

roje

ct (

sept

ic o

r la

trin

es)

11

11

14.

95%

22

15.

15%

105.

05%

Ele

ctric

ity/s

olar

pow

er/s

tree

t ele

ctric

ity1

21

11

5.94

%2

11

4.12

%10

5.05

%

Rec

reat

iona

l par

k/sp

orts

for

yout

h/ca

sa c

omm

unal

21

11

4.95

%1

11

14.

12%

94.

55%

Rep

air

stre

et1

11

2.97

%1

21

4.12

%7

3.54

%

Inpu

ts fo

r ag

ricul

ture

/mai

nten

ance

of f

ield

s/ag

equ

ipm

ent

11

1.98

%1

21

4.12

%6

3.03

%

Hel

p fo

r m

edic

ines

for

chro

nica

lly il

l1

11

13.

96%

11

13.

09%

73.

54%

Agr

icul

ture

for

fam

ilies

/sol

ar1

11

2.97

%1

11

3.09

%6

3.03

%

Cow

s1

11

2.97

%1

11

3.09

%6

3.03

%

Cre

dit/r

evol

ving

fund

12

2.97

%2

13.

09%

63.

03%

Kitc

hen

for

child

ren

for

child

ren

10.

99%

11

13.

09%

42.

02%

Foo

d fo

r pe

ople

10.

99%

11

13.

09%

42.

02%

Prim

ary

scho

ol in

vest

men

t and

pre

-prim

ary

2

23.

96%

22.

06%

63.

03%

Oth

er p

atio

ani

mal

s1

11

2.97

%1

12.

06%

52.

53%

Littl

e m

arke

t/dis

trib

utio

n ce

nter

/pro

duct

div

ersi

ficat

ion

11

1.98

%1

12.

06%

42.

02%

Pla

ntin

g tr

ees

for

refo

rest

atio

n1

0.99

%1

12.

06%

31.

52%

Voc

atio

nal s

choo

l/tra

inin

g1

0.99

%

11

2.06

%3

1.52

%

Con

stru

ctio

n of

sec

onda

ry s

choo

l 1

11

13.

96%

11.

03%

52.

53%

Chu

rch

11

12.

97%

11.

03%

42.

02%

Mac

hine

ry fo

r se

win

g ta

ller/

beau

ty1

11.

98%

11.

03%

31.

52%

Sch

olar

ship

s fo

r yo

uth

11

1.98

%1

1.03

%3

1.52

%

Car

pent

ry to

ols

for

a ta

ller

10.

99%

11.

03%

21.

01%

Con

stru

ctio

n of

brid

ge1

0.99

%1

1.03

%2

1.01

%

Pur

chas

e of

pig

s0.

00%

11.

03%

10.

51%

Cle

an la

nd/tr

ees/

herb

s/pl

ant

0.00

%1

1.03

%1

0.51

%

Wom

en's

pro

ject

0.00

%1

1.03

%1

0.51

%

Tra

sh/o

vera

ll hy

gien

e 0.

00%

11.

03%

10.

51%

Mai

nten

ance

of p

roje

cts

0.00

%1

1.03

%1

0.51

%

Sec

urity

0.

00%

11.

03%

10.

51%

Hel

p fo

r di

sabl

ed0.

00%

11.

03%

10.

51%

Hel

p fo

r st

reet

chi

ldre

n0.

00%

11.

03%

10.

51%

Hel

p fo

r fu

nera

ls0.

00%

11.

03%

10.

51%

Hel

p fo

r ce

lebr

atio

ns0.

00%

11.

03%

10.

51%

Old

folk

s ho

me

0.00

%1

1.03

%1

0.51

%

Veh

icle

to d

rive

out t

he s

ick

11

1.98

%0.

00%

21.

01%

Bus

ines

s2

1.98

%

0.00

%2

1.01

%

Mat

eria

ls fo

r sc

hool

libr

ary

10.

99%

0.00

%1

0.51

%

Tru

ck fo

r di

strib

utio

n1

0.99

%0.

00%

10.

51%

Ric

e pr

oces

sor

10.

99%

0.00

%1

0.51

%

Con

stru

ctio

n of

a b

aker

y fo

r w

omen

10.

99%

0.00

%1

0.51

%

Con

stru

ctio

n of

a p

ier

10.

99%

0.00

%1

0.51

%

LEA

DE

RS

RE

GU

LAR

PE

OP

LE

TH

EM

ES

/LO

CA

TIO

N

60

3. THE ROLE OF LABOR MARKETS FOR SHARED GROWTH IN NICARAGUA

By Catalina Gutiérrez and Marco Ranzani*

Summary

During the past decade Nicaragua has been consolidating a stable macroeconomic environment and has registered modest growth rates

In recent years Nicaragua has experienced modest growth rates, averaging 3.8 percent between 1998 and 2005, and 1.7 percent per year between 2001 and 2005. The country has consolidated its structural adjustment programs and completed the requirements for benefiting from the HIPC initiative, freeing the country from a debt service burden that amounted to 250 percent of GDP in 2001. This growth has been closely tied to investment and exports.

Despite the consolidating macroeconomic environment, macroeconomic uncertainty and instability are still reported by firms as one of the main investment climate concerns; and lack of credit is a major constraint for private investment growth.

Nicaragua is witnessing a change in its demographic structure which may provide a window of opportunity for poverty reduction

During this period the country has also seen an increase in the working age population as a fraction of the total population, and a growth in employment at a rate of 3.9 percent per year. This decrease in the number of dependents per working age person, generated by the increasing labor force, presents an important opportunity for poverty reduction, as each working member now has a smaller number of dependents to support. Moreover, this new labor force has managed to find employment: the share of the working age population employed increased from 62 percent to 64 percent.

Despite the increase in employment, the new employment opportunities were low paying, in particular for the poor

* The authors are with the World Bank. This work was prepared as a contribution to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Pierella Paci (Lead Economist, PRMPR), Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), Jose Ramon Laguna (Head Research Assistant for the Poverty Assessment), Norman Hicks (Consultant), Gabriel Demombynes (Economist, LCSPP), Diego Angel-Urdinola (Economist, LCSPP), and Ximena Del Carpio (Consultant) for their valuable comments and suggestions. We also wish to thank the participants at the Poverty Workshop in Managua in March 2007 for helping us improve our understanding of the Nicaraguan socioeconomic context, and methodological and data issues. We are particularly grateful to the Minister of Labor of Nicaragua, Ms. Janeth Chavéz Gómez; the Central Bank of Nicaragua (BCN) General Manager, Mr. José de Jesus Rojas, the BCN Director of the Research Department, Mr. Mario Alemán, and, amongst his staff, Hiparco Loaisiga, Ligia Miranda, Miguel Aguilar and Lisbeth Laguna; the Institute of Statistics and Census (INEC) Poverty Specialist, Mr. Juan Rocha; and, Fundación Internacional para el Desafío Económico Global (FIDEG) Director, Mr. Alejandro Martinéz Cuenca. Finally, special thanks go to Nydia Betanco at the Nicaragua World Bank Country Office for all her help and support while on mission in Nicaragua. The views expressed here are those of the author and need not reflect those of the World Bank, its Executive Directors, or the countries they represent.

61

However, despite the higher employment and the lower dependency rate, headcount poverty did not change. An important fraction of the employment generated appeared to be in “bad jobs.” Most of the employment seems to have been absorbed by the manufacturing and agricultural sectors.

Agriculture offers the lowest returns among economic activities and has historically concentrated the largest number of poor. Moreover, Value Added per worker in agriculture has decreased. These characteristics make employment growth in agriculture an unlikely driver of significant poverty reduction. However, it is still unclear as to what extent the employment growth in agriculture that has been captured through the household surveys adequately reflects the reality in this sector. It is possible that the 2001 household survey underestimates rural population, so that, when compared to the 2005 survey, it appears as if the rural population increased, which is at odds with the census trends. This dubious rise in rural population may be behind at least part of the agricultural employment growth.

The manufacturing sector contributed to a significant share of employment generation, but an important fraction of employment generated in this sector was not tied to better incomes. First, 45 percent of the employment generated in this sector was in family enterprises, which is, and has been, associated with low income generation. Second, a large share of the new jobs was concentrated in the food and beverage sector, which registered a decrease in wages. The clothing sector was the only sector that contributed to the generation of “good jobs.” This was probably maquila employment. However, the apparent requirement that those employed in maquilas have completed secondary education may have limited the poor from benefiting from growth in this sector.

In addition, wages in manufacturing seemed to have decreased. This might be a response to the rising labor force and the decrease in productivity in the sector.

The very poor saw important increases in their labor income that were due to higher prices of goods produced by the agricultural poor; this growth, although important, was not enough to bring them out of poverty

When decomposing growth in household per capita income into its components, we find that the poorest 20 percent benefited from an important increase in the share of working age persons within the household and in the participation rates. The rise in the share of working age population as a fraction of total members in the household explains 37 percent of per capita household income growth in the lowest quintile. On the other hand, income per self-employed in agriculture contributed with 44 percent of the change in household per capita labor income. This means that the poorest basically gained because of better earnings from self-employment in agriculture, as well as from an important demographic transition.

The rise in agricultural income from self-employment can be attributed to better producer prices of coffee, meat, maize and beans, which are all produced by small farmers. Both yield per hectare and cultivated area remained almost constant. Despite increases in the production of sensitive agricultural products (due mainly to an expansion of harvested areas), the output growth was not sufficient to compensate the inflow of labor into agriculture, so that Value Added per worker decreased. This decrease in (constant) Value Added per worker was offset by higher producer prices. Had it not been for price increases, rural poverty would have most likely increased. In addition, the relative productivity of Nicaragua with respect to other countries in the world, and in particular with respect to its main trading partners, remains astonishingly low, especially for rice and milk.

62

There is some evidence that agricultural wages increased as well, despite the decreases in (constant) Value Added per worker. Increases in the relative prices of export goods may be behind this behavior.

Constraints to the generation of better jobs seem to lie outside the realm of labor market regulation, and there is some evidence of agricultural/non-agricultural sector segmentation

Labor regulation does not seem to be a hindering factor for formal employment generation. The labor regulation does not appear particularly high, neither from the perspective of employers nor compared to other trading partners and neighbors.

Despite this, there is some evidence of segmentation between the agricultural and non agricultural sectors. There is an important earnings premium in non agricultural earnings. An important part of this premium is the results of a selection process where more educated individuals opt out or can access non agricultural jobs. However, even after selection effects are taken into account, there is an earning premium for working outside of agriculture, with returns to education, location, and gender explaining most of the differential for otherwise equivalent workers. Qualitative results suggest some barriers to moving outside agriculture.

An important earning premium of being a wage worker outside of agriculture, when compared to self employment outside of agriculture was also found. Part is explained by selection of more educated and male workers into wage employment, and another part explained by differences in returns to education for otherwise equivalent workers. Thus the earning differential between self employment and wage employment in non agricultural jobs can be potentially explained by segmentation. Although among the self employed, who are mostly informal, the majority did not choose self-employment because of lack of wage employment but rather because of schedule flexibility, for an important fraction of the unskilled (26 percent) self-employment was a response to lack of wage employment. Qualitative results also support some segmentation between wage employment and self-employment.

Geographic barriers to mobility and low levels of education constrain the rural poor from moving to better jobs

There is evidence of segmentation between agricultural and non-agricultural jobs. There is an important earnings premium outside of agriculture, which is mostly explainable by differences in returns to individual characteristics (controlling for selection). The most important factor determining whether workers have a non-agricultural job is having primary and secondary education and being older and more experienced. Being a male and living in the Pacific region makes it more likely for a worker to end up in agriculture.

Although education is an important determinant of being employed in a high earning sector or occupational category, there is no evidence of skill mismatch within occupational categories or sectors

Low levels of education among the poor seem to be restricting their access to the most dynamic sector in the economy—the manufacturing maquila sector—as employment in firms in this sector requires a completed secondary education. In addition to being important to finding employment outside of agriculture, education brings significant returns even within occupational categories and within sectors.

63

Despite the fact that education does affect earning levels, there appears to be no evidence of skills mismatch. Firms do not report skills to be a constraint for business functioning or growth. In addition, the evidence suggests that, while the demand for skills may be rising, the supply of skills is rising more than proportionally. This rise in the availability of skills is likely to reduce wages unless there is a substantial boost in production and in the demand for labor.

Exploring policy options

If growth is to translate into poverty reduction, increasing the level of education of the labor force should be at the forefront of the policy agenda, in particular in the rural sector. Despite the fact that this may exert an important downward pressure on wages, it increases the returns of the wage employment and employers in agriculture as well as the likelihood of being employed outside of agriculture. Both higher returns in agriculture and a moving out of agricultural employment are key elements for poverty reduction.

To prevent educational expansion from resulting in lower wages, the demand for skills must keep pace with the supply of skills. Labor regulation does not appear to pose a constraint to labor demand, and labor is cheap relative to capital. Thus, increasing the demand for wage employment is likely to be achieved only if the most binding constraints to growth are addressed: namely, macroeconomic uncertainty and lack of credit. Fostering investment in unskilled intensive sectors, in areas outside of Managua (such as in the tourism sector) is a policy that merits careful consideration. In addition, given the small domestic market in Nicaragua, promoting export in order to increase labor demand and reduce the downward pressure on wages will be imperative.

The very low levels of productivity found in the agricultural sector, together with some indirect evidence of low mobility between urban and rural areas, suggests that rising productivity in agriculture should also be at the forefront of policy initiatives. Observed income rises among the poor seem to be tied to rises in the prices of agricultural products and foreign remittances. This behavior increases the vulnerability to foreign shocks. Without targeted investments in agricultural productivity and agricultural exports, decreasing rural poverty in the short and medium runs seems implausible.

On the labor market front, the most promising route to fostering the creation of poverty reducing jobs is to identify and address the barriers to moving out of agriculture. Infrastructure and transport costs, land titling, information problems and education merit further study. Finally, it might be worth exploring whether a simpler minimum wage structure than the current multi-sector scheme might lead to fewer distortions in the labor market that may be particularly binding for the unskilled.

Introduction

Why do we care about employment, earnings and labor markets in the search for growth?

The degree to which growth is able to translate into poverty reduction depends on how its benefits are distributed among different segments of society. There is little doubt that growth—measured by changes in average income—contributes significantly to poverty reduction.21 However, it is also clear that countries differ in the degree to which income growth spells have translated into

21 Kraay (2006) finds that in the short and medium terms income growth accounts for 70 percent of the variation in headcount poverty, and in the long run, it accounts for as much as 97 percent.

64

poverty reduction; and, although differences in the responsiveness of poverty to income growth account for a small fraction of the overall differences in poverty changes across countries, from the point of view of an individual country these differences may have significant implications for poverty reduction, especially in the short term.22

There is a general consensus that the availability of employment opportunities and their characteristics constitute an essential transmission channel from growth to poverty reduction and, in this way, play a key role in poverty’s response to growth. For one thing, the poor derive most of their income from work, either as self-employed or as employees, so that what happens to their income and employment status seems tautologically relevant. In addition, the ease with which the poor may take up the opportunities afforded by growth may depend crucially on (i) the structure of employment, (ii) the returns to labor and their distribution, and (iii) the existence of imperfections and frictions in the labor markets. For example, one may be inclined to believe that when the poor face flexible labor markets and low barriers to mobility across labor market segments, geographic regions or sectors of production they are in a better position to take the opportunities generated by growth, by “moving” more easily to the growing sectors. Similarly, the effectiveness of growth in reducing poverty may also depend on whether growth is unskilled labor-intensive and whether the poor have or can easily acquire the skills required by the growing sectors. Moreover, there is some evidence of strong links between labor market regulations, such as minimum wages, and the incidence of poverty in developing countries.

The concern that employment, returns to labor and imperfections/rigidities in the labor markets play a crucial role in the poverty impact of growth has been reflected in the emphasis in the policy debate on the idea that “jobless” growth has been responsible for the disappointing results seen by some countries in the effectiveness of growth in reducing poverty. As a result, debates addressing how to foster employment-intensive growth have followed.23 However, it is also often recognized that poverty is less an outcome of open unemployment than of adequate levels of income, and, as such, emphasis should be placed not on increasing employment levels but on increasing the productivity of the working poor.24 The debate has also been concerned with whether policy interventions should concentrate on increasing earnings in the sectors where the poor are found (such as agriculture), or whether they should be targeted to sectors where the poor are not found, so that more of the poor can be drawn into the higher-earning sectors (Fields 2006). To date, there is very little evidence to illuminate the debate. Moreover, the questions are hard to address, because there is lack of clarity on how to achieve the alternative objectives and because it is inherently difficult to identify the costs and benefits of the possible policy alternatives.

Objectives, Scope and Structure

The objective of this paper is to shed light on some of the issues discussed above in the case of Nicaragua, and to provide some policy guidelines for the fight against poverty. In particular, we

22 See, for example, Bourguignon (2002), Kakwani, Khandker and Son (2006), Lucas and Timmer (2005) and Ravallion (2004), for evidence on heterogeneity in the poverty impact of growth. See Ravallion (2004) for a discussion of the relevance of this heterogeneity from the perspective of a country: a 1 percent increase in income levels could result in a poverty reduction of as much as 4.3 percent or as little as 0.6 percent. 23 One of the core elements of the global employment agenda “Macroeconomic policies for growth and employment” calls for addressing four key questions, one of which is “How can the employment intensity of growth be increased?” ILO (2003). 24 ILO (2003).

65

hope to be able to identify the growing sectors, as well as the constraints faced by the poor in benefiting from this growth.

This paper is part of a series of studies conducted within the PREMPR, to foster our understanding of the role of employment earnings and labor markets in shared growth. In addition, this paper on Nicaragua is intended to feed as a background document for the Nicaragua Poverty Assessment 2007.

The paper is structured in four sections. Section 1 briefly describes the evolution of the Nicaraguan economy, in terms of its macro indicators as well as of employment and poverty. Section 2 analyzes the profile of growth as well as the way in which it helps explain the observed behavior of poverty, using aggregate data from National Accounts and employment from household surveys. It describes growth and employment by the sector of economic activity and its employment productivity profile. It goes more deeply into the evolution of the manufacturing sector and the maquila production. Section 3 looks at the income profile of the population, using household surveys. Section 4 provides a brief statement on policy implications and further research.

Definitions of terms used throughout the paper are presented in Box 1 bellow. Workers have been classified into 4 occupational categories: wage and wage and salaried workers, individual self employed workers, family enterprise workers and employers. We believe these are qualitative distinct types of labor, across which segmentation or barriers to mobility may exist. In particular we have opted to divide the non-wage workers into the above mentioned categories for several reasons: i) employers (those who employ paid labor) receive substantially higher income than other non wage workers and are better educated. They often have assets which other non wage workers do not, ii) returns to labor, for family enterprise workers and self-employed not working with other members of the family, need different methodologies of calculation. While the income reported by the self employed working alone is the return for labor for his/her individual work, reported income for self employed workers working with other unpaid family members is the income earned by all the family members, and a methodology has to be devised to assign a proportion of household income to each member of the family and, iii) individual self employed are more prevalent in urban areas, while the family household enterprise workers are more prevalent in rural agricultural work.

Box 1: Definitions

Employment Labor market The place where labor services are bought, sold, and exchanged. The

labor market comprises wage and salaried workers and their employers, but also non-wage family enterprise workers and the self-employed, who make up the largest share of workers in Madagascar.

Labor force The sum of the working age employed and unemployed. Employed An individual who performed market activities for at least one hour in

the week prior to the survey, or who has a permanent job. Unemployed A working age individual who is not employed but is actively looking

for work. Inactive A person who is neither employed nor actively looking for work. Wage worker A worker who has declared being salaried for his/her work. It

includes those self reported as jornaleros and peones, which work for a daily or per job rate in manual agricultural labor, often only during the harvest season.

66

Self-employed A self-declared self-employed person, living in a household in which there are no other self-employed or unpaid family workers.

Household enterprise worker, family enterprise worker

A self-declared self-employed person living in a household with other self-employed or unpaid family workers.

Formal employment

Employment for which social security contributions are paid by workers and firms

Working age population The population between 15 and 64 years of age. Child labor A child between 6 and 14 years old, who performed market activities

for at least one hour in the week prior to the survey, or who has a permanent job.

Maquila employment The maquila sector comprises all production units located in the ‘Special Export Processing Zones’ which are clearly defined zones, often within a wired complex. Production is undertaken with mostly imported materials using local labor and all output is destined for export markets.

Earnings Earnings, labor income All cash payments, payments in kind, and benefits received in

exchange for labor services in wage and salaried employment, self-employment and other forms of labor exchange. “Earnings” and “labor income” are used interchangeably, although the latter is more often used when referring to the labor income of a household rather than of an individual. Depending on the context, earnings include only primary job earnings (e.g., when comparing earnings in the different sectors) or the sum of earnings in all reported jobs.

Wage earnings Total cash and in-kind earnings as declared in the survey. Earnings of the self-employed and employers

For non-agricultural work: it is calculated as declared in the survey For agricultural work: it is calculated as net profits using the survey’s agricultural enterprise module.

Household enterprise earnings

For non-agricultural work: earnings for each individual are calculated as a proportion of the sum of earnings declared in the survey of all the workers employed in the household enterprise. In 2001 each worker is assigned a portion of earnings proportional to reported hours of work. In 2005 total enterprise income is divided equally among total number of adult workers. For agricultural work: Earnings are derived from the survey’s agricultural enterprise module and divided by the number of adult household members reported as working in the enterprise.

Low earner An employed individual whose earnings are below the national poverty line.

Country Context

Macroeconomic Context

Over the past 12 years, Nicaragua has witnessed a very significant transformation: from a nation torn by war, political instability and natural disasters, with its economy plunged into chaos, it has re-emerged as an inclusive democracy where the foundations for economic growth and sustainable development are being laid. Notwithstanding this

67

progress, Nicaragua still remains among the poorest countries in the Western Hemisphere. It is classified as a lower middle income economy with a per capita Gross National of Income of US$1,000 in 2005, which is a third of the average value for the Latina American and the Caribbean Region and half the average off all lower middle income countries. It has a population of 5.1 million, with a life expectancy at birth of 70 years.

Consolidating a stable macroeconomic environment

During the past years Nicaragua has experienced modest growth rates, averaging 3.8 percent between 1998 and 2005. The country has consolidated its structural adjustment programs and completed the requirements for benefiting from the HIPC initiative, thereby freeing the country from a debt service that amounted to 9.5 percent of GDP in 2001.

Between 1998 and 2001, GDP per capita grew at an average rate of 3.8 percent, and then decelerated, averaging a per capita growth rate of 1.7 percent between 2001 and 2005. This growth has been closely tied to investment and exports (see Figure 3.1). Investment has been fueled by foreign assistance. In 1998, after Hurricane Mitch struck the country, massive reconstruction efforts were undertaken. The country received US$250 million in emergency assistance, and a further US$1.4 billion was pledged by the international community. Until 2001, recovering from the aftermath of the hurricane was a prime policy objective which, together with important flows of foreign assistance, led to an increase in public investment of 27 percent in 1999.

The last 10 years have also seen a consolidation in the IMF-led stabilization policies adopted in the early 1990s, which were concentrated in controlling hyperinflation, reducing the fiscal deficit and privatizing public utility companies. A second wave of reforms was initiated in 2002 with the signature of the Poverty Reduction and Growth Facility (PRGF) with the IMF. Its aim was to achieve fiscal sustainability through the broadening of the tax base, the elimination of tax exemptions, improved revenue collection, more effective budgeting and the improvement of the financial position of the Central Bank. The government also sought access to a HIPC initiative to gain foreign debt relief. In 2004, Nicaragua reached the completion point under HIPC, and bilateral and multilateral debt relief was granted for debt incurred prior to 2005. On the international front Nicaragua has signed several trade and integration agreements with its Central American partners, and trade with Honduras, El Salvador and Guatemala is gaining in importance, although the United States remains the main trading partner.

68

Figure 3.1: Investment, Exports and Growth, 1995-2005

-

1,000.0

2,000.0

3,000.0

4,000.0

5,000.0

6,000.0

7,000.0

8,000.0

9,000.0

10,000.0

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 p/ 2005 e/0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

InvestmentExportsGDP growth

Inve

stm

ent a

nd e

xpor

ts

Mill

ions

of 1

994

$C

GD

P G

rowth

-

500.0

1,000.0

1,500.0

2,000.0

2,500.0

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 p/ 2005 e/0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

Public InvestmentGDP growth

Inve

stm

ent M

illio

ns o

f 199

4 $C

GD

P G

rowth

Source: Own calculations with data from BCN.

Some important developments in an otherwise unchanging economic structure

There have been no major changes in the sectoral structure of production and in the urban/rural composition of the population. However, Nicaragua has experienced an important demographic transition as the share of working age population (15 - 64 years) increased faster than other age ranges, reducing the dependency ratio.25 In addition, maquila and financial intermediation have experienced important developments.

Population growth has slowed down, and Nicaragua has started to see a change in its demographic structure. Between 1995 and 2005, the population grew at a rate of 1.6 percent annually, a number well below the projected growth rate of 2.04 percent. The ratio of working age population (15 - 64 years) to total population increased from 53 percent in 1998 to 55 percent in 2001 and 58 percent in 2005, significantly reducing the dependency ratio. Despite this overall demographic change, there was little gain in the share of urban population, which increased its share in total population by 1 percentage point in the last 10 years (See table 3.1)

The sectoral structure of GDP remained relatively constant during these 10 years, with the secondary sector gaining only a 1 percentage point share during the whole period. Although there were no major changes in the structure of production, within the secondary and tertiary sectors there were some important developments, namely, the growth of the maquila sector and an important surge in financial intermediation.

Financial intermediation has grown at an average annual rate of 9 percent. This increase in intermediation is an important development, as Nicaragua has the smallest banking system in Central America and as it is the main source of credit for the private sector. Still, financial intermediation is weak and accounts for only 3.6 percent of GDP.

25 The dependency ratio is the ratio of total population to working age population, and it indicates, on average, how many persons a working adult has to support.

69

Table 3.1: Main Macroeconomic Indicators, 1998-2005

1998 1999 2000 2001 2002 2003 2004 2005

GDP real growth (%) 3.7 7 4.1 3 0.8 2.5 5.1 4

Real GDP per capita growth (%) 2 5.3 2.4 1.3 -0.9 0.8 3.4 2.2

Share of Value Added in

primary sector (% ) 21.3 20.8 22.3 22.1 21.8 21.6 21.3 21.2

Share of Value Added in

secondary sector (%) 26.7 27.5 27.1 27.5 27.1 26.8 27.6 27.8

Private consumption per capita

real growth % 3 4.1 3.5 3.1 2.7 0.1 1.9 1.8

Gross fixed investment real

growth % 4.3 27.1 -16.8 -8.4 -7.1 -1 4.2 10.1

Consumer price inflation (year

to year % change) 13.04 11.22 11.55 7.36 3.99 5.15 8.44 9.42

Real effective exchange rate

2000=100 98.9 96.9 100 100.9 96.9 91.2 89 88.7

Urban population as a share of

total population 54.9 55 55.2 55.3 55.5 55.6 55.8 55.9

Total Population (thousands)

4,579

4,655

4,733

4,812

4,892

4,974

5,057

5,142

Sources: INEC, BCN, and World Bank.

Growth in the maquila sector has had important implications in terms of the availability of foreign reserves and employment. The maquila sector, which started in the early 1990s with the development of the first public Free Trade Zone, has experienced an amazing dynamism. Between 2001 and 2005, the share of maquila exports in total exports jumped from 32 percent to 50 percent, and the sector generated a little over 53,000 new jobs during these four years. The value of transformation services in the maquila, a measure of the value of domestic inputs used in the process, reached 5.5 percent of total Value Added in 2005.26

26 It corresponds to the difference between the value of imported raw materials and value of final exports and corresponds mostly with the cost of labor and utilities.

70

Labor Market Context

The labor market structure in Nicaragua is typical of countries with a similar level of development

The labor market profile of Nicaragua is very similar to that of low income and low middle income countries, which is characterized by low unemployment rates27 (as defined by the ILO), low formality and waged employment rates, high shares of population working in agriculture, and relatively high child labor.

This structure of employment is mainly a reflection of the stage of industrialization of these countries. Low and middle income countries still have a large agricultural sector in which productivity is generally low and workers are mostly self-employed. Most of the population has very low incomes, so that they cannot afford to be unemployed. Instead, an important fraction of the working age population is self-employed in informal activities, many in agriculture. As industrialization progresses, the share of employment in the modern sectors, mainly manufacturing and services, rises. Industrialization spreads predominantly in urban areas, which leads to a process of urbanization, as rural workers leave low productivity jobs in agriculture in search of higher paying jobs out of agriculture.

As urbanization progresses, urban self-employment in low productivity jobs (in many cases informal) increases. The reason behind this increase is still a matter of debate and may depend on the particulars of the labor market structure and regulation. In many cases workers who are searching or “queuing” for good jobs are still too poor to afford to be unemployed and must engage in self-employment “survival activities” while they seek a job. In other cases, monopsonistic behavior by firms leads to very low wages for the unskilled, so that many low skill migrant workers find self-employment as attractive as wage employment. With urbanization, unemployment begins to be noticeable, as higher incomes resulting from higher productivity permit the luxury of shopping for good jobs.

The development of a modern sector also comes with a rise in formalization and in the share of waged and salaried employment, as modern firms grow and demand labor. The growth of the modern sector is usually also accompanied by a rise in agricultural productivity, although the links and causalities for this are less clear. In many cases purposeful investment in agriculture frees rural labor, as more productive techniques mean that fewer workers are needed to exploit the available land. This free urban labor migrates to urban markets, providing new labor that feeds the process of urbanization and industrialization. In other cases, migration to non-agricultural jobs with higher productivity and higher pay allows households to generate savings that can translate into new investments that raise agricultural productivity.

As industrialization progresses child labor may decrease, as higher incomes mean lower opportunity costs of sending children to school. Additionally, as the demand for skill

27 ILO defines as unemployed, those not employed but who had actually looked for a job in the past week.

71

increases, so does its returns, and the benefits of acquiring an education become more evident. Thus it is costlier not to send children to school.

In Nicaragua, unemployment as defined by the ILO is low, slightly less than 4 percent (seeTable 3.2). Most of the employed work in the informal sector (82 percent), wage employment accounts for half of the employed, agriculture absorbs a high share of employment (37 percent) and child labor is relatively high (9 percent). Agriculture is still a sector with low returns and productivity has been declining, suggesting that employment in this sector still acts as a “last resort” option for the working population. In the discussion that follows, the labor market structure is described in more detail.

Unemployment is not a major problem among the poor and is only a minor problem for the non-poor

Table 3..2 presents the main indicators of the labor market. We find that unemployment rates according to the ILO definition are very low and that they remained almost constant during the period under analysis. The broad unemployment rate, which also includes discouraged workers, is slightly higher but its still low compared to other countries (less than 10 percent). The working age population, defined as those between 15 and 64, as a proportion of the total population increased 5 percent (or 3 percentage points). And the number of employed as a fraction of the total working age population also rose slightly. Child labor saw a small increase, from 8.7 percent to 9.2 percent. If we look at the poverty rate among unemployed workers, it stands out that it is half the overall poverty rate, which suggests that unemployment is not strongly correlated with poverty.

The table also shows the number of workers affiliated to social security, which is often a measure for formalization. In Nicaragua only some 19 percent of the labor force has social security, and this ratio decreased slightly in 2005. Finally, the table shows the number of workers holding more than one job concurrently. It has been pointed out that in many cases workers cannot generate enough income from their main job and must find additional work to complement their income, so that the share of workers holding more than two jobs concurrently is often used as a measure of the (poor) quality of the jobs. The share of workers holding two or more jobs is less than 10 percent, a figure below that of low income countries.

Agricultural jobs offer the lowest returns. Outside of agriculture, the self-employed do not earn less per hour worked than the waged employed, but they appear to earn less annually owing to shorter spells of work during the year

Table 3.3 shows the median annual labor income and median earning rates for the different employment categories. Earnings are lower for all agricultural categories, and among agriculture the lowest income is obtained by household enterprise workers and the individually self-employed. It is also obvious that wages decreased for non-agricultural employment while they increased for agricultural workers. The self-employed in non-agricultural work have similar earning rates to the wage employed, which suggests that wage employment is not necessarily a better earning option. However, yearly earnings among the self-employed are lower, which suggests that they are employed for shorter periods or work fewer hours.

72

Table 3.4 presents a description of the employment status of the population 6 years of age and above. The first column (the stub) lists the tiers, meaning the group and subgroup of labor force categories. The second column shows the number of persons under each tier for 2001. The third column shows the hierarchical rates, meaning the percentage of people in the subcategory (or tier) for 2001. The fourth and fifth columns show the equivalent numbers and rates for 2005. The last column gives the percent change.

Table 3..2: Main Indicators of the Labor Market, 2001 and 2005

Level 2001 Level 2005 % change

Unemployment rate* 3.47 3.39 (2.44)

Broad unemployment rate** 8.07 6.87 (14.84)

Employment-to-working-age-population ratio 62.15 62.78 1.01

Working age population as a fraction of total population 55.30 58.22 5.27

Child labor rate 8.87 9.28 4.69

Share of long-term unemployed*** 0.04 1.50 3,913.41

Poverty rate among unemployed workers (national poverty line,

poor) 22.14 28.52 28.80

Poverty rate among unemployed workers (national poverty line,

extremely poor) 6.96 6.69 (3.83)

Poverty rate among unemployed workers (international poverty

line 1$/day) 24.76 28.52 15.19

Share of workers holding 2 or more jobs concurrently**** 9.03 8.41 (6.92)

Share of workers affiliated to social security***** 19.70 18.05 (8.40)

* ILO unemployment definition is those not employed but that actively searched for a job in the past week. ** Broad unemployment rate also includes discouraged workers. *** Ratio of long-term unemployed over total active labor force=(employed + unemployed), the questions are not strictly comparable: in 2001, how long have you been unemployed? In 2005, how long have you been searching for a job? **** Defined as holding two jobs in the past week. ***** Affiliated to Social Security in main occupation.

Source: Own calculations with data from EMNV 2001 and 2005.

The first three tiers (child population, population 65 and above, and working age population) illustrate the basic population structure of those 6 years and older. It is very evident that the share of the population between 6 and 14 increased its participation among the group. The population 65 and older and the population of working age reduced their shares. As noted before, the share of the working age population as a fraction of the total population actually increased (in other words, the dependency ratio decreased). This means that the main reason for the decrease in the dependency ratio is a reduction in the share of those 65 and older and those 6 and younger. As will be seen in the next section, this has important implications for the

73

evolution of the labor market in the coming decade, as the cohort between 6 and 14, which represents the largest fraction of the population, will enter the labor market in the coming years.

Table 3.3: Earnings and Income by Employment Category, 2001 and 2005

Level 2001 Level 2005 % change

Non-

Agri-

culture

Agri-

culture

Non-

Agri-

culture

Agri-

culture

Non-

Agri-

culture

Agri-

culture

Wage and salaried workers

Median annual labor income 21,064.36 10,532.18 20,844.00 11,700.00 (1.05) 11.09

Median hourly earnings rate 8.32 4.22 8.52 5.06 2.41 20.01

Low earnings rate 20.57 37.76 17.66 24.86 (14.17) (34.16)

Individual self-employed

workers

Median annual labor income 13,374.20 6,423.42 12,000.00 6,319.25 (10.28) (1.62)

Median hourly earnings rate 11.20 3.88 6.22 - (44.43) -

Low earnings rate 28.32 55.96 34.13 51.59 20.50 (7.81)

Employers

Median annual labor income 46,809.70 9,298.00 45,000.00 31,751.19 (3.87) 241.48



Household enterprise

workers

Median annual labor income 10,532.18 5,190.74 16,053.45 5,891.44 52.42 13.50



Notes: Median annual labor income refers to all the occupations and includes monetary, non-monetary and in-kind earnings. The median hourly earnings rate is calculated from the main occupation only, except for the agricultural self-employed, agricultural employers and agricultural family enterprises, for which it is calculated as profits per hour worked using the agricultural enterprise module. In 2005, the agricultural enterprise module does not report hours worked, and therefore the hourly earning rate cannot be calculated for the self-employed in agriculture.


74

The table further disaggregates the working age population (15 to 64) into active and inactive. The rate of inactivity has remained almost constant at 35 percent. The inactive include the discouraged workers and the seasonally inactive. As has been mentioned, the proportion of discouraged workers as a fraction of the active population has decreased. But this is also true for the seasonally inactive (from 5 percent to 1.6 percent). Among the active population, 96 percent are employed with very little change between the years.

The employed population (tier 1.3.2.2) is disaggregated into different employment categories. The bulk of the non-agricultural population is employed as wage and salary workers (43 percent in 2001). In the agricultural sector employment is evenly distributed among the individually self-employed in agriculture (11 percent in 2001), the employed in agricultural family enterprises (10 percent in 2001) and the wage and salary workers (11 percent in 2001).28 There has been little change in this structure. Under each employment category, the share of low earners is shown. These are the workers who earn incomes below the poverty line. The highest low earnings rates can be found among those individually self-employed in agriculture (55 percent have low earnings) and those that work in household family enterprises in agriculture (40 percent have low earnings). Employment in all agricultural categories has increased. As will be discussed further, it is unclear how much of this increase might be due to errors in the urban/rural weights of the 2001 survey.

Finally, Table 3.5 shows the sector of employment and level of education of the employed population. The tertiary sector absorbs most of the employed population, namely, more than two-thirds of the employed. This share decreased slightly between 2001 and 2005, owing to the increase in employment in the primary and secondary sectors. Finally the low level of education of the labor force stands out, as nearly 40 percent of the employed have an incomplete primary education or below, and only 10 percent have a completed secondary education.

In summary, between 2001 and 2005 Nicaragua’s labor markets saw either no change or very subtle changes in this labor market profile. Perhaps the important events have been the increase in the share of the working age population as a fraction of the total population and the increase in employment in the agricultural sector. In general, as industrialization progresses the share of the population in the rural sector tends to decrease. In very few cases increases in the rural population are seen as response to urban crisis. However, this has not been the case in Nicaragua, and thus the reason for this increase is yet to be determined. One possible explanation is that the population weights used in the 2001 was not in accordance with the census behavior of the population (see Box 3.1). However, it is not clear to what extent this may be affecting the results.

28 The individually self-employed are the self-employed who do not work with other family members. The employers are those who are self-employed but have paid workers. The household enterprise workers are the self-employed who work with unpaid family members or unpaid helpers.

75

Tab

le 3

.4: H

iera

rchi

cal D

escr

ipti

on o

f the

Pop

ulat

ion

6 Y

ears

and

Abo

ve, 2

001

and

2005

2

00

1 l

evel

(in

mil

lio

ns)

Hie

rarc

hic

al

rate

s

20

05

lev

el (

in

mil

lio

ns)

Hie

rarc

hic

al

rate

s%

ch

an

ge

1. T

ota

l po

pula

tio

n 6

yea

rs a

nd

ab

ove

4,1

54,9

83

10

0.0

0

4

,48

6,7

08

100

.00

7.9

8

1

.1 C

hil

d p

op

ula

tio

n (

6-1

4 y

ears

of

age)

1

,03

1,1

58

2

4.8

2

1

,18

5,5

78

26

.42

1

4.9

8

1.1

.1 C

hil

d l

abo

rers

9

1,4

43

8

.87

11

0,0

71

9

.28

2

0.3

7

1

.2 P

op

ula

tio

n 6

5+

yea

rs o

f ag

e 2

57

,45

9

6

.20

26

5,7

42

5

.92

3.2

2

1.2

.1

9

2,8

16

3

6.0

5

9

7,4

48

36

.67

4.9

9

1

.3 W

ork

ing a

ge

po

pula

tio

n (

15

-64

yea

rs o

f ag

e)2

,86

6,3

65

6

8.9

9

3

,03

5,3

87

67

.65

5.9

0

1.3

.1 I

nac

tivea

1

,02

0,7

27

3

5.6

1

1

,06

2,8

46

35

.02

4.1

3

1

.3.1

.1 D

isco

ura

ged

9

2,2

06

9

.03

7

3,7

23

6

.94

(2

0.0

5)

1

.3.1

.2 T

emp

ora

rily

Inac

tive

48

,292

4

.73

1

6,7

98

1

.58

(6

5.2

2)

1.3

.2 A

ctiv

e

1,8

45,6

38

6

4.3

9

1

,97

2,5

40

64

.98

6

.88

1.3

.2.1

Unem

plo

yed

6

4,0

99

3

.47

6

6,8

37

3

.39

4.2

7

1

.3.2

.2 E

mp

loyed

b

1,7

81,5

38

9

6.5

3

1

,90

5,7

03

96

.61

6.9

7

1

.3.2

.2.1

Wag

e an

d s

alar

ied

agri

cult

ure

19

6,3

48

1

1.0

2

20

9,3

95

.

10

.99

6.6

4

Wit

h l

ow

ear

nin

gs

72

,922

3

7.1

4

5

2,0

64

.

24

.86

(2

8.6

0)

1

.3.2

.2.2

W

age

and

sa

lari

ed

no

n-

77

1,3

18

4

3.3

0

78

5,0

37

41

.19

1.7

8

Wit

h l

ow

ear

nin

gs

14

3,9

63

1

8.6

6

13

8,6

19

17

.66

(3.7

1)

1

.3.2

.2.3

In

div

idual

se

lf-e

mp

loyed

5

9,8

54

3

.36

6

7,0

48

3

.52

1

2.0

2

Wit

h l

ow

ear

nin

gs

32

,950

5

5.0

5

3

3,4

40

49

.88

1.4

9

1

.3.2

.2.4

In

div

idual

se

lf-e

mp

loyed

no

n-

20

0,8

74

1

1.2

8

25

3,1

25

13

.28

2

6.0

1

Wit

h l

ow

ear

nin

gs

58

,200

2

8.9

7

8

6,3

86

34

.13

4

8.4

3

1

.3.2

.2.5

Em

plo

yer

s ag

ricu

lture

3

5,7

07

2

.00

2

3,9

57

1

.26

(

32

.91

)

W

ith l

ow

ear

nin

gs

13

,592

3

8.0

7

1

,31

5

5

.49

(

90

.32

)

1

.3.2

.2.6

Em

plo

yer

s no

n-a

gri

cult

ure

3

6,7

64

2

.06

5

7,4

93

3

.02

5

6.3

8

Wit

h l

ow

ear

nin

gs

2,2

88

6

.23

3

,92

5

6

.83

7

1.5

1

1

.3.2

.2.7

In

H

ouse

ho

ld

ente

rpri

ses

19

3,3

77

1

0.8

5

30

0,6

39

15

.78

5

5.4

7

Wit

h l

ow

ear

nin

gs

79

,136

4

0.9

2

9

4,2

46

31

.35

1

9.0

9

1

.3.2

.2.8

In H

ouse

ho

ld e

nte

rpri

ses

no

n-a

g c

12

2,0

07

6

.85

18

3,9

49

9

.65

5

0.7

7

Wit

h l

ow

ear

nin

gs

32

,071

2

6.2

9

10

4,7

63

56

.95

2

26

.66

1

.3.2

.2.9

Oth

ers

16

5,2

86

9

.28

2

5,0

55

1

.31

(1

69.2

0)

a T

he

surv

ey d

oes

not

allo

w f

or

a d

isti

nct

ion

bet

wee

n d

isco

ura

ged

work

ers

wh

o a

re w

illi

ng t

o w

ork

and

th

ose

wh

o a

re n

ot.

Tem

pora

rily

inac

tive

wo

rker

s ar

e th

ose

wai

tin

g t

o s

tart

a j

ob

or

wai

tin

g

b E

mp

loyed

work

ers

are

sub

-cla

ssif

ied

acc

ord

ing t

o t

he

mai

n o

ccu

pat

ion

. T

he

cate

gory

does

not

sum

to t

ota

l em

plo

yed

bec

ause

oth

er e

mp

loyed

are

cla

ssif

ied

as

mem

ber

s of

coop

erat

ive

c In

clu

des

un

pai

d f

amil

y m

emb

ers.

76

Table 3.5: Other Characteristics of the Employed, 2001 and 2005

Share of total

employment 2001

Share of total

employment 2005

Sector of activity (primary occupation)

Primary 18.45 20.60

Secondary 11.90 12.56

Tertiary 69.64 66.84

Formal schooling attainment

No-school 20.30 17.63

Incomplete primary 26.89 24.13

Primary 13.72 14.79

Incomplete secondary 21.45 22.50

Secondary 8.23 10.17

Tertiary 9.41 10.78


Box 3.1: Urban/Rural Population: Possible Data Problems

Labor regulation in Nicaragua does not seem to pose a constraint to investment

The largest share of non-labor costs corresponds to social security contributions, which amount to 15 percent of the wage. Workers contribute with 6.25 percent of their wage for social security. Workers are entitled to one month of paid vacations and an annual bonus that is equivalent to one month of work. They are also entitled to seniority bonuses. In addition to these costs, employers have to pay 2 percent of the total payroll for INATEC,

The table below shows the population calculated from the census and the surveys. According to the surveys,

urban population increased substantially from 1998 to 2001, from 54 percent to 58 percent, and decreased

afterwards between 2001 and 2005, from 58 percent to 55 percent. It is hard to estimate whether the behavior in

the surveys is actually true. It is surprising that urbanization increased substantially and then reversed in such a

short time. The available census information suggests that there was an increase of 1 percentage point between

1995 and 2005, but there are no data points in between to illustrate the inter-census behavior. Moreover, the 2001

population estimations used in the 2001 survey overestimate the population growth. The present report corrects

the weights for this overestimation, but makes no adjustments for regional or urban/rural composition, as there

were no data available to do so. It is unlikely that the population overestimation was uniform across regions or

urban/rural populations.

77

the technological training institute. Moreover, there are minimum wages by sectors and there is strong support for unionization. Firms are allowed to hire temporary workers and can extend this type of contract indefinitely. The working week consists of six days and the working week can be extended up to 50 hours. Termination of the employment contract is authorized with no third party involvement, and workers are entitled to severance pay upon termination, which varies with tenure.

Table 3.6: Labor Market Flexibility. Comparative Performance

Region or Economy

Difficulty of

Hiring Index

Rigidity of

Hours Index

Difficulty of

Firing Index

Rigidity of

Employment

Index

United States 0 0 0 0

Jamaica 11 0 0 4

Dominica 11 20 20 17

Chile 33 20 20 24

El Salvador 33 40 0 24

Nicaragua 11 60 0 24

Colombia 22 40 20 27

Uruguay 33 60 0 31

Latin America & Caribbean 34 34.8 26.5 31.7

Costa Rica 56 40 0 32

OECD 27 45.2 27.4 33.3

Guatemala 61 40 0 34

South Asia 41.8 25 37.5 34.8

Middle East & North Africa 29.7 44.7 32.9 35.8

Honduras 67 40 0 36

Mexico 33 40 40 38

Brazil 67 60 0 42

Dominican Republic 56 40 30 42

Sub-Saharan Africa 44.3 52 44.9 47.1

Ecuador 44 60 50 51

Bolivia 61 60 100 74

Venezuela 67 60 100 76

Source: World Bank, Investment Climate surveys.

78

Investment climate surveys collect information among firms of constraints to growth and business activities. The information collected includes the level of non-wage labor costs and the perception among firms of the rigidity of labor regulation. Using this information, the Enterprise Survey Unit at the IFC constructs relative hiring and firing rigidity indexes. Table 3.6 compares the results for Nicaragua with other countries in the region (and elsewhere) and its main trading partners (shown in grey). As can be seen, Nicaragua does not appear particularly rigid when compared to other countries in the region. In fact, it appears to be one of the least rigid economies, ranking only below Jamaica and Dominican Republic and having an overall performance equal to Chile and El Salvador. It is relatively low compared to the United States, one of its main trading partners but also the most flexible economy in the world.

Minimum wages are set by the Minimum Wage Commission in which representatives of the unions, the government and the private sector negotiate their level. Minimum wages are differentiated according to sector of economic activity in an attempt to “take into account the level of education of the labor force in each sector” (see Table 3.7).

Table 3.7: Minimum Wage and Lowest Wage Paid as a Proportion of Minimum Wage, 2001 and

2005

Minimum wage Lowest paid wage as a

proportion of minimumSector

2001 2005

Agriculture 542 736 1.23 1.17 Fishing - Mining 942 1,377 2.12 1.53 Manufacturing 664 988 1.58 1.24 Electricity, Gas and Water 887 1,242 1.63 1.73 Construction 1,001 1,410 1.71 1.32 Commerce 1,292 1,752 1.01 0.98 Transport and Communications 1,001 1,410 1.39 1.22 Financial Intermediation 1,001 1,410 1.18 1.26 Services 1,110 1,752 0.88 0.70 Municipal and Central Gov. 778 1,066 0.79 0.90

Note: Monthly average minimum wage calculated as weighted average of ongoing minimum wages during the year. Source: Own calculations based on MITRAB and BCN.

Assessing whether the minimum wage is binding in Nicaragua is a hard task. The fact that there are 12 different minimums implies that one would have to estimate separately for each sector of economic activity whether the wage is binding or not, and this would reduce the sample size and thus the reliability of any estimate. We opt for analyzing minimum wages in the four largest sectors in terms of employment: agriculture, manufacturing, commerce and community services.

A first step in analyzing minimum wages is to plot kernel density estimates of wage earnings and explore whether the distribution of earnings displays a kink at the minimum wage.

79

Fig

ure

3.2:

Dis

trib

utio

n of

Wag

es b

y Se

ctor

and

For

mal

ity,

200

1

0.2.4.6.81

-4-2

02

4lo

g ho

uly

earn

ings

rat

e

For

mal

Info

rmal

Agr

icul

ture

0.2.4.6.81

-20

24

log

houl

y ea

rnin

gs r

ate

For

mal

Info

rmal

Man

ufac

turin

g

0.2.4.6.8

-20

24

6lo

g ho

uly

earn

ings

rat

e

For

mal

Info

rmal

Com

mer

ce

0.2.4.6.81

-4-2

02

46

log

houl

y ea

rnin

gs r

ate

For

mal

Info

rmal

Ser

vice

s

Sour

ce:

Ow

n ca

lcul

atio

ns w

ith d

ata

from

EM

NV

200

1 an

d 20

05 a

nd N

icar

agua

Cen

tral

Ban

k.

80

Figure 3.2 illustrates the results for the log of hourly wage earnings for 2001, for both formal and informal wages. The vertical line is the corresponding log of the hourly minimum wage for the sector. In the case of the manufacturing sector there are three different minimum wages. The lowest (C$670) corresponds to the manufacturing non maquila sector. The middle wage (C$895) corresponds to the maquila sector and the highest (C$1,010) corresponds to utilities (electricity, gas and water).

As expected the distribution of formal wages is to the right of the distribution of informal wages. From the figure, there is some evidence of a kink around the minimum wage in formal agriculture, while in the informal agricultural sector minimum wages do not seem to have any effect. The maquila minimum wage is binding in manufacturing, but the non-maquila minimum wage seems to be setting the standard for minimum pay in the informal sector, although not in the form of a king, but rather by affecting the mode. The effects of minimum wages on commerce are unclear: there seems to be a slight kink for the formal sector, but results are sensitive to the assumption about hours worked29. Again there does not seem to be any effect on the informal sector. Minimum wages in the community services sector have no effect on either formal or informal wages; although both distributions show a kink, it is located at a higher level than the minimum wage. In any case, the distribution of wages does not show important distortions around the minimum wage when compared to other Latin American Countries. Moreover, there is no evidence of important effects of minimum wages on the informal sector. In many Latin American Countries minimum wages have been shown to leak to informal markets, suggesting that both segments are more integrated than previously thought. This does not seem to be the case for Nicaragua. This opens the possibility that minimum wages in Nicaragua are acting as a barrier to formal job creation, and may contributing to an informal sector that does not comply with minimum wage regulation. The magnitude and importance of this effect is merits further study.

It is unclear whether the current structure of minimum wages provides much benefit over a unique minimum wage. If the idea of sectoral minimum wages is to “take into account the different average skill levels of the labor force in each sector” it might be better to set a minimum wage by level of education (for the low skilled) rather than by sector. The current structure of the minimum wage might be introducing unnecessary distortions into the labor market, and might be segmenting the market according to skills. This might explain the behavior of maquila factories, which face a higher minimum wage than overall manufacturing and, as a response, may restrict employment to those with a secondary or higher education. If we assume that more productive firms have larger profits and a higher share of skills (as is often the case), the current minimum wage setting mechanism is acting more as a central collective bargaining mechanism to distribute profits between low skilled workers and firms, rather than as than as a mechanism for setting the lowest paid wage. But even if this is this were the objective of having a differential minimum wage by sectors, it is unclear what are the advantages of this centralized bargaining system over a decentralized (firm level) bargaining system.

When firms make hiring decision they compare the marginal cost of labor –i.e. the minimum wage- with the marginal product –i.e. the value of output produced by one

29 For this exercise a working week of 48 hours was assumed.

81

additional worker-. More productive firms are usually more competitive, account for larger shares of employment and grow faster. In ‘competitive’ labor markets, if firms differ in productivity and minimum wages are higher for the most productive firms, then low skill workers (for which minimum wages are binding) will be rationed out of the most dynamic sectors of the economy. Instead a minimum wage by skill level, will mean that more productive firms will have an advantage with respect to low productive firms when hiring low skilled workers: relative to the marginal cost (i.e. the minimum wage), the marginal benefit of having an unskilled worker is larger. Therefore, high productivty firms might be more inclined to increase their unskilled labor intensity technology while workers will be equally off (for a given average skill level) in any sector or firm. If on the other hand labor markets are characterized by frictions and wages are either bargained or the results of firms competing for labor (as in wage posting models), higher productivity firms will pay higher wages to equivalent workers (i.e. for the same level of education), distributing in this way profits between workers and capital, regardless of ‘the average level of education’ in the firm or sector, but proportionally to productivity in the firm. Under this setting it is unclear what is gained by introducing a centralized bargaining system that sets minimum wages according to the average level of education of workers in the sector. The efficiency of the outcome will depend on the bargaining power of both parties (firms and workers), but not on the average level of education of the sector or firm. Therefore there would seem no clear reason to use the average level of education in a sector as a guide to set minimum wages.

Understanding the employment effects of minimum wages and the impact of its sectoral structure on employment and the relative demand for unskilled labor is beyond the scope of this paper. But it is an area that merits further research.

Investment climate constraints to increasing employment lie outside of the labor market

Nicaragua conducted an investment climate assessment for 2003. Although an investment climate assessment is outside the scope of this paper, the survey can be used to pinpoint the main bottlenecks that are present and that may be hampering growth and employment generation.

Table 3.8 shows the percent of firms responding that a particular constraint was “severely” hampering business functioning and growth. Labor regulation and the skills of the labor force are among the least problematic constraints, while macroeconomic stability and uncertainty and credit issues are severely constraining business functioning and growth.

82

Table 3.8: Issues Affecting the Investment Climate

Ranking Investment climate Issue

Percent of firms answering

that it is a very severe

problem

1 Corruption 38.3

2 Cost of credit 34.2

3 Macroeconomic and regulatory uncertainty 31.4

4 Access to credit 28.7

5 Macroeconomic stability 27.0

6 Non-competitive practice 26.8

7 Availability of credit 26.7

8 Efficiency of justice administration and conflict resolution 19.7

9 Crime and Violence 18.8

10 Transport 17.3

11 Electricity 17.3

12 Taxes 14.6

13 Red tape on taxes 8.4

14 Property rights 5.5

15 Skills of the labor force 5.5

16 Access to land 5.1

17 Import taxes regulation 4.9

18 Telecommunications 4.7

19 Permits and operating licenses 4.4

20 Labor regulation 3.1

21 Trade regulation 3.0

Source: Own calculations with data from World Bank Enterprises Surveys.

Output, employment and Poverty

This section describes the labor and productivity profile of growth and links it to poverty reduction. It also takes a closer look at the manufacturing sector and the maquila sector.

83

A first section describes the main trends in output, poverty and employment and a second section decomposes growth into sectoral employment and productivity changes, while a final section takes a closer look at manufacturing.

Main Trends in Output, Employment and Poverty

Value added grew at an annual average rate of 4.2 percent between 1998 and 2005. Between 1998 and 2001 growth reached 5.42 percent. Growth decelerated dramatically between 2001 and 2005, reaching an average annual growth rate of 3.24 percent (Table 3.9). Agriculture, construction and services suffered the largest growth losses. Only transport and the financial sector kept their growth pace, but these sectors are small in terms of employment and output. Furthermore, the share of the poor employed in these two sectors is less than 4 percent. Despite this strong deceleration of economic activity, the manufacturing sector managed to grow at an average annual rate of 4.4 percent. As will be discussed below, this has important implications for poverty reduction. As has been mentioned, this growth was fueled in part by the reconstruction efforts after Hurricane Mitch struck the country in 1998. These reconstruction efforts meant that the construction sector grew at an average of 11 percent per year, although most of this growth was concentrated in the year after the hurricane, in which construction grew 36 percent. Manufacturing, agriculture and services also registered growth rates above the average. Growth in these sectors has important implications for both employment and poverty, as 60 percent of total employment is concentrated in these sectors and 76 percent of the poor earn their livelihood in these three sectors.

The working age population is growing, but skills remain stagnant

As has been mentioned, Nicaragua began to see an important change in the population structure, with the working age population (between 15 and 64 years of age) increasing its share in the total population. The working age population grew at an average annual rate of 2.7 percent per year, compared to a 1.7 percent average annual population growth. Figure 3.3 illustrates the significant change observed in population structure between 2001 and 2005.

This population change presents both challenges and opportunities for poverty reduction. One the one hand, a larger fraction of the population will have to find jobs. Between 2001 and 2005 the economy had an inflow of around 350,000 new workers. Had these new workers not been able to find jobs, poverty would have increased. On the other hand, each working adult now has to support a lower number of dependents, which provides an opportunity for poverty reduction if these new working adults are able to find sufficiently well-paid jobs.

84

Table 3.9: Sectoral Growth, 1998-2005

Average annual

growth 1998-2001

Average Annual

growth 2001-2005

Average annual

growth 1998-2005

Agriculture 6.84 2.37 4.26

Mining and Utilities 5.23 3.00 3.95

Manufacturing 5.71 4.42 4.97

Construction 11.30 1.11 5.36

Commerce, Restaurants and Hotels 4.17 3.66 3.88

Transport and Communications 4.29 4.67 4.51

Services 6.22 3.07 4.41

Government 1.55 1.67 1.62

Financial Sector 8.51 9.76 9.22

Total 5.42 3.24 4.17

Source: Own calculations based on data from BCN and EMNV.

Furthermore, the cohort aged 10 to 15, which will have completed the transition to the working age within the next five years, will imply an additional 590,000 workers in the labor market.30 Thus the opportunities and challenges offered by this population transition will continue to be present in the next decade.

Figure 3.3: Change in Population Structure, 2001-05

2001

00-04

5-10

10-15

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85 +

8 6 4 2 0 2 4 6 8

00-04

5-10

10-15

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85 +

MaleFemale

2005

8 6 4 2 0 2 4 6 8

Source: Own calculations based on Nicaragua Census data.

Unfortunately, the level of education of this new labor force has not shown much improvement. Although higher primary completion rates were observed in 2005 compared to 2001, the share of the working age population with an incomplete secondary education or less decreased only 3 percentage points (from 68 percent of the employed 30 This increase is net of those aged 60 to 64 who will be exiting the labor market.

85

working age population to 65 percent). This means that each year the share of the employed population with a less than complete secondary education decreased by only 1 percent or, equivalently, the share of the employed working age population with a completed secondary education or above increased by 1 percent annually. In other words, at this rate it would take 23 years to reach a stage at which at least 50 percent of the working age population had the level of complete secondary education or above.

Nicaragua has one of the lowest education levels in Latin America and Central America. It ranks only above Guatemala in terms of the education level of its urban and its rural populations (Table 3.10). If the population transition is to lead to poverty reduction, two policies will need to be at the front of the national agenda: increasing good employment opportunities and accelerating educational achievement.

Table 3.10: Average Level of Education of Population 25 to 64

Country Year Urban Rural

Guatemala 2004 6.5 2.4

Nicaragua 2001 6.9 3.1

Honduras 2003 7.5 3.5

Brazil 2005 7.8 3.8

El Salvador 2004 8.6 3.8

Bolivia 2004 8.9 4.9

Venezuela (national total) 2005 8.9 …

Dominican Rep. 2005 9.1 6.2

México 2005 9.6 6.0

Costa Rica 2005 9.6 6.8

Colombia 2005 9.7 …

Uruguay 2005 9.9 …

Ecuador 2005 10.4 5.6

Peru 2003 10.6 5.3

Panama 2005 11.1 7.0

Sources: Nicaragua’s own estimations based on 2005 survey. Other data: CEPAL.

The growing labor force was absorbed disproportionately by the manufacturing and agricultural sectors

All sectors, with the exception of mining and utilities, and construction, experienced positive employment growth. The average annual total employment growth was 4 percent. Moreover, the growth in employment was greater than the growth in the labor force (3 percent).

86

The growing labor force was absorbed by the agricultural, manufacturing and commerce sectors. These sectors accounted for around 67 percent of total employment, and they all experienced average annual growth rates above 2.5 percent, thus accounting for 84 percent of total employment growth (see Table 3.11). On the other hand, community services, which is the other important sector in terms of its employment size, was stagnant, growing at an annual average rate of 1 percent. This meant that its contribution to total employment generation was a 5 percent. Figure 3.4 illustrates the sectoral shares of employment for 2001 and 2005.

Although commerce absorbed an important fraction of the new labor force, its growth rate was lower than aggregate employment growth, thus losing its participation in total employment. The third column in see Table 3.11shows the change in the employment share of each sector. The gain of 2.5 percentage points in manufacturing employment and of 1.1 percentage points in agricultural employment stands out. A further look at these sectors will be taken later in this section.

Figure 3.4: Share of Employment by Sectors, 2001 and 2005

2001

Agriculture, 31.71

Manufacturing, 11.99Construction, 5.27

Commerce, 22.87

Transport, 3.93

Financial Services, 2.70

Government Services, 2.97

Community Services, 17.26

Mining and Utilities, 1.30

2005

Agriculture, 32.82

Manufacturing, 14.51

Construction, 4.51Commerce, 21.79

Transport, 3.70

Financial Services, 3.05

Government Services, 3.15

Community Services, 15.48

Mining and Utilities, 0.98


But headcount poverty has remained stagnant

Despite the increase in the working age population and in the share of working age population employed, headcount poverty did not change. The number of poor according to the national poverty line stayed at 46 percent, and extreme poverty stayed at 15 percent. Poverty in the urban sector (29 percent) is substantially lower than poverty in the rural sector (68 percent).The incidence of poverty (the poverty gap) decreased a little less than 1 percentage point.

87

Table 3. 11: Evolution of Employment by Sectors, 2001 and 2005

Average annual

employment growth

(%)

Share of total

employment generation

(%)

Change in the

share of total labor

force (percentage

points)

Agriculture 4.80 39.5 1.11

Mining and Utilities -3.25 -1.0 -0.32

Manufacturing 8.99 29.8 2.52

Construction -0.04 -0.1 -0.76

Commerce 2.65 15.2 -1.09

Transport 2.35 2.3 -0.23

Financial Services 7.18 5.2 0.36

Government Services 5.39 4.2 0.17

Community Services 1.13 4.8 -1.77

Total employment 3.90 100.0 0.02

Labor force 2.98

Source: Own calculations based on data from EMNV.

Tabel 3.12 shows the poverty rates of the working age population by area of residence and employment status. The poverty rate increased from 41 to 42 percent between 2001 and 2005. It is clear that poverty rates decreased among the employed and increased among the unemployed and inactive. The increase in poverty among the rural unemployed was particularly strong, but the unemployed in the urban sector make up less than 1 percent of the total working age population, so that this increase does not affect the overall poverty rate in any significant way. Thus, while the poverty rate among the rural employed decreased it was more than compensated by an increase in poverty among the rural inactive. As has been mentioned, the inactive include discouraged workers and the seasonally unemployed, among others.Table 3.13 shows the evolution of employment by sector and poverty level. The table clearly shows that the poor are overly represented in agriculture, and this share may have increased from 2001 to 2005.31 It also calls attention to the increase in the share of the poor employed in manufacturing from 8.8 percent to 11 percent, while they are losing their share in community services and commerce.

31 The Household Survey for 2001 shows a rural share of the population that is inconsistent with the census. According to the Household Survey (EMNV 2001), the share of rural population increased between 2001 and 2005. The census shows the opposite. Apparently the survey of 2001 underestimates the rural population. If this is the case, the increase in the share of employed in agriculture might be due exclusively to the under-representation of rural households in the survey of 2001.

88

Table 3.12: Poverty Rates of the Working Age Population by Employment Status, 2001-2005

2001 2005

Rural 63 62 Employed

Urban 25 25

Rural 48 61 Unemployed

Urban 18 22

Rural 66 69 Inactive

Urban 27 28

Rural 64 65

Urban 26 26 Total working age

Total 41 42

Rural 64 68

Urban 29 29National poverty level

Total 46 46


Table 3.13: Employment by Sector and Poverty Level, Shares of Total Employment, 2001 and 2005

Poor Non Poor Total

2001 2005 2001 2005 2001 2005

Agriculture 53.57 55.66 16.98 17.25 31.71 32.82

Mining and Utilities 0.95 0.76 1.54 1.13 1.30 0.98

Manufacturing 8.85 11.04 14.10 16.89 11.99 14.51

Construction 5.09 3.83 5.39 4.98 5.27 4.51

Commerce 12.99 11.85 29.53 28.56 22.87 21.79

Transport 1.86 1.90 5.33 4.93 3.93 3.70

Financial Services 1.15 1.03 3.74 4.44 2.70 3.05

Gvt. Services 1.24 1.40 4.14 4.34 2.97 3.15

Community Services 14.30 12.54 19.25 17.49 17.26 15.48

Total 100.00 100.00 100.00 100.00 100.00 100.00


89

Decomposing per Capita Income Growth

The aim of this section is to show how growth is linked to changes in employment, productivity (output per worker) and population structure at the aggregate level and by sector. The main idea is to profile growth in per capita Value Added, to see whether growth has been accompanied by productivity or employment increases and if so in which sectors.

The change in per capita Value Added between 2001 and 2005 is decomposed into: (i) changes in the demographic composition of the population, (ii) changes in productivity, and (iii) changes in the share of working age population employed. The decomposition is performed at the aggregate level and by sectors. Per capita Value Added can change from one year to another if any of these components changes. For example, if there is an exogenous increase in productivity (Value Added per worker) for the same number of workers and for a constant population structure, the higher productivity per worker will imply more Value Added per person. Equally, there might be a change in the structure of the population so that each working person has fewer dependents; if productivity and employment do not change, then Value Added will increase as more workers are producing for the same total population. In reality, however, many factors are changing at the same time, so that it is difficult to disentangle what has happened to each component of per capita Value Added for a given observed growth. There are several techniques for decomposing and attributing to each component a share of total observed growth. The result described used Shapley decompositions which are described in Annex A in more detail.

Table 3.14 shows the change in Value Added per capita and in its main components. Per capita Value Added saw a growth of 7.14 percent for the period, while employment grew 16.54 percent, the population share of the working age population grew 3.62 percent and Value Added per worker (productivity) decreased almost 2 percent. This means that the new labor force was absorbed by employment, but at a lower productivity level (a lower level of output per worker).

Growth was employment intensive, but overall productivity decreased.

Figure 3.5 illustrates the results for the decomposition at the aggregate level. It shows that 74 percent of the change in per capita Value Added can be linked to changes in the structure of the population. In other words, had everything else stayed the same, the sole change in the number of dependents per working age person would have generated a growth equivalent to 74 percent of the actual observed growth (i.e., a total growth for the period of 5.3 percent). Changes in employment were also important, accounting for some 51 percent of observed growth. This means that if productivity had stayed the same and the number of dependents per working age member had also remained constant, the higher rate of employment would have generated a growth of 3.6 percent. Unfortunately, changes in productivity acted in the opposite direction. Had productivity not changed, observed growth would have been 9 percent, but decreases in productivity meant that growth was 1.6 percentage points lower.

90

Table 3.14: Percent Change in Selected Variables, 2001-05

% change

Average

Annual

growth %

Value added 14.47 3.44

Value added per capita 7.14 1.74

Population 6.85 1.67

Population of working age 12.47 2.98

Employment 16.54 3.90

Employment/pop. working age 3.62 0.89

Value added per worker -1.78 -0.45


Figure 3.5: Aggregate Employment and Productivity Profile of Growth 2001-2005


The key question, then, is why did productivity decrease? There are many reasons why output per worker might have decreased: Workers might have moved to a sector where marginal productivity is lower, TFP might have decreased or, the capital labor ratio was reduced as a result of the large inflow of workers into the economy. To explore these possible alternatives we further decompose changes in aggregate output per worker into change due to: i) intersectoral labor shifts i.e. movements of worker between sectors of different productivity levels- ii) changes in the capital labor ratio and iii) changes in Total Factor Productivity (TFP).

-25.99 51.64

74.35

-40.00 -20.00 0.00 20.00 40.00 60.00 80.00

Changes in output per worker

Changes in share of working age population

employed

Changes in inverse of dependency ratio

Percent contribution to total change

91

Figure 3.6 shows the result of the decomposition of changes in output per worker for the aggregate economy. Total output per worker decreased 1.78 percent. Of this decrease intersectoral employment shifts exerted a positive effect on output per worker (half a percentage point) or 6.72% of total productivty growth. The capital labor ratio also increased, contributing with 1.68 percentage points. But TFP suffered an important reduction which explains 3.66 points of the decrease. From this data we can clearly say that TFP changes are responsible for the decrease in output per worker.

For this decomposition TFP was calculated as a residual (see Appendix A). This means that it is capturing all factors other than capital and intersectoral shifts. For example, the average skill of the labor force which comprises both experience and education. It will also capture changes in the structure of employment by employment categories (rather than by sectors). For example if an important increase of total employment was concentrated among family enterprises or the self employed, which have lower than average productivty, then TFP will decreases. As pointed out in the previous section (see Table 3.5) the average years of education of the labor force increased, so the explanation for TFP decreases do not lie on the level of education (at least if quality did not change). In the previous section (Table 3.3 and Table 3.43.4) we found that employment had increased disproportionately among household enterprise workers and individual self employed, which are the categories with lowest earnings. If they are also those with the categories with the lowest productivty (as is most likely the case) then this increase in the share of employed in low earning categories might account for part of the TFP decrease. New entrants to the labor market also have less experience and as such may have lower productivities.

In addition, in a segmented labor market, where marginal products differ between sectors, it is possible that as workers move from low marginal product sectors to high marginal product sectors, the average product of labor falls in the sector where employment rises as marginal decreasing returns to labor set in. Unfortunately we have no data to decompose changes in output per worker by sector, to see whether capital labor ratios decreased in the expanding sectors or whether TFP changes explain these decreases. We can however, look at changes in overall productivity and employment by sectors, as well as intersectional shifts, to understand further the aggregate behavior.

Figure 3.6: Decomposition of Changes in Output per Worker. 2001-2005

Total percentage Growth of Output per Worker = -1.78%

1.68

0.43

-3.88

-5.00 -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00

Capital labor ratio

Total factor productivty

Intersectoral employment shifts

Percentage points

92

Decomposing intersectional shifts

It is possible to understand further how changes in the share of employment in the different sectors help explain the overall contribution of intersectoral shifts to per capita growth. An important literature has found that structural change, which is movements of labor force shares form low productivty sectors to high productivty sectors, is an important factor behind growth. Increases in the share of employment in sectors with above average productivty will increase overall productivty and contribute positively to the intersectoral shift term. On the contrary, movements out of sectors with above average productivty will have the opposite effect. By the same token, increases in the share employment in sectors with bellow average productivty should reduce growth, while reduction in their share should contribute positively to growth.

Table3.15 shows the results of decomposing intersectoral shifts using the above intuition (see Annex A for details and formulas). The results show that the increasing shares of employment in manufacturing and government explain most of the positive effect of intersectoral shifts. While movements into agriculture and out of mining and utilities exerted a negative effect on overall per capita growth.

Table3.15: Decomposition of Intersectoral Shifts

Direction of

Employment Share shift

Contribution to

Intersectoral Shifts

(%)

Sectoral contributions

Agriculture Movements into -84.72

Mining and utilities Movements out off -161.67

Manufacturing Movements into 279.11

Construction Movements out off 6.39

Commerce Movements out off 46.14

Transport Movements out off -48.24

Government Movements into 49.52

Other Movements out off 13.48

Total Contribution of

intersectoral shifts 100.00

In other words, had employment growth been proportionally distributed among all the sectors, per capita Value Added growth would have been 6 percent lower (i.e. the contribution of intersectoral shifts to total per capita growth). But because employment growth was disproportionately concentrated in manufacturing, a sector with high productivty, it spurred growth.

93

Decreases in productivity were not economy-wide, while increases in employment were concentrated in manufacturing and agriculture

Table 3.16 shows changes in total Value Added by sector as well as changes in the share of each sector in total Value Added. All sectors experienced positive growth, and overall employment growth was 14.5 percent for the whole period. Manufacturing, commerce, transport and “other” saw a Value Added growth that was above average, thus gaining share in total Value Added. The sector referred to as “other” groups community and enterprise services as well as financial services, but this last has a very small share of the total. Agriculture reduced its share. Overall changes in shares were relatively small: manufacturing gained a 1 percentage point share while agriculture lost a 1 percentage point share.

Table 3.16: Sectoral Growth, 2001-05

Share of total Value Added

Total Value

Added growth

2001-2005 2001 2005 % change

Agriculture 9.80 22.1 21.2 -4.08

Manufacturing 18.91 19.0 19.8 3.87

Mining and utilities 12.55 3.6 3.5 -1.69

Construction 4.52 4.9 4.5 -8.69

Commerce 15.45 18.2 18.3 0.85

Transport 20.03 7.1 7.4 4.85

Government 6.86 7.0 6.6 -6.65

Other 18.39 18.1 18.7 3.42

Total 14.47 100.0 100.0 -

Source: Own calculations based on data from BCN.

Table 3.17 shows changes in productivity and employment shares by sectors. All sectors experienced positive employment growth, but growth was disproportionately concentrated in manufacturing and agriculture. Value added per worker decreased in both manufacturing and agriculture, which were the sectors that experienced the highest increases in employment. It is also worth noting that all sectors that saw an increase in employment also saw a decrease in productivity, while sectors that experienced an increase in productivity had a decrease in their share of employment. There may be several explanations for this. One possible explanation is that changes in employment mainly capture new entrants to the labor market and these new entrants have lower productivity than more experienced workers. The sectors that have a stronger increase in employment growth absorb most of this new labor force, and thus have a stronger negative effect on average productivity. Alternatively as explained above, the inflow of

94

workers into these sectors implied a lower capital labor ratio, as decreasing marginal returns to labor set in, and as such a decrease in average productivty.

Decreases in productivity as well as increases in employment were concentrated in agriculture and manufacturing. Increases in the relative size of the manufacturing sector (in terms of employment) account for an important share of growth

Finally, Table 3.18 illustrates the contribution of each sector to total per capita Value Added growth. Sectoral contributions are decomposed into: i) contribution of changes in output per worker (first column), ii) contribution of the sector to employment rate growth (second column), and iii) contributions of the sector to the intersectoral shift component (see Annex A for details). Overall, manufacturing, commerce, transport and other services contributed positively to growth, while agriculture, mining and utilities, construction and government had a negative contribution.

Table 3.17: Employment Shares and Productivity, by Sectors of Economic Activity, 2001- 2005

Output per worker (1994 $C) Employment/pop. of working age

2001 2005 % change 2001 2005

Absolute

Change

Agriculture 10,973

9,988

(8.97) 19.21 20.60 1.39

Manufacturing 25,032

21,097

(15.72) 7.26 9.11 1.85

Mining and utilities 43,097

55,364 28.47 0.79 0.62 -0.17

Construction 14,701

15,393 4.70 3.19 2.83 -0.36

Commerce 12,505

13,005 4.00 13.86 13.68 -0.18

Transport 28,418

31,084 9.38 2.38 2.32 -0.06

Government 37,223

32,239

(13.39) 1.80 1.98 0.17

Other 14,308

15,643 9.33 12.09 11.64 -0.45

Total 15,757

15,477

(1.78) 60.59 62.78 2.19


95

Despite the enormous employment growth in manufacturing, the decrease in output per worker was so large that it more than offset the employment growth. However, shifts of labor into manufacturing (i.e. the relative size of the manufacturing sector) and away from other sectors of lower productivty, more than compensates this effect , so that in the aggregate manufacturing contributed with 12 percent of total per capita growth. Other services accounted for a non-negligible 14 percent of observed growth of output per capita; and commerce and transport contributed with 9 and 5 percent of the growth, respectively. In all three sectors the effect was mostly due to increases in output per worker.

Agriculture, on the other hand contributed negatively to per capita growth, via to different effects, first it saw a decrease in output per worker and second it increased its share of total employment. Given that it has productivty bellow average this shift toward agriculture reduced growth.

These results suggest that had productivity in agriculture and manufacturing not decreased, then Value Added per worker would have been 5 percentage points higher. The next section tries to provide an understanding of what was happening in manufacturing. A closer look at the agricultural sector will be undertaken in a later section..

Despite decreasing productivity in agriculture, agricultural wages increased while manufacturing wages saw a decrease

Table 3.19 presents the median wages by sector of economic activity, calculated from the household surveys. The table shows that, in real terms, wages in agriculture increased 17 percent while wages in manufacturing decreased 2 percent between 2001 and 2005. As will be seen, lower Value Added per worker in agriculture was compensated by higher producer prices, so that this may have limited the falls in productivity from translating into lower wages. In the case of the manufacturing sector, wages did decrease, but much less than productivity, so that productivity falls were not totally passed on to wages.

96

Table 3.18: Total Sectoral Contribution to Growth, 2001-05

Contribution of

Changes in Output

per Worker (%)

Contribution to

Total

Employment

Rate Changes

(%)

Contributions

of Intersectoral

Shifts (%)

Total (%)

Sectoral contributions

Agriculture -29.53 32.78 -5.30 -2.05

Mining and utilities 13.03 -4.11 -10.12 -1.20

Manufacturing -48.47 43.48 17.48 12.49

Construction 3.14 -8.44 0.40 -4.90

Commerce 10.39 -4.23 2.89 9.05

Transport 9.46 -1.37 -3.02 5.08

Government -14.19 4.11 3.10 -6.97

Other 23.89 -10.59 0.84 14.14

Subtotals -32.26 51.64 6.26 25.64

Demographic

component - -

74.36

Total 100.00

Total % change in output per capita 2001-2005 7.14


Table 3.19: Wages by Sector of Economic Activity, 2001 and 2005

Median wage $C 2001

2001 2005

Real growth

(%)

Agriculture 6,840 8,018 17.2

Mining and Utilities 20,000 23,495 17.5

Manufacturing 12,600 12,364 -1.9

Construction 10,080 10,000 -0.8

Commerce 13,329 13,364 0.3

97

Transport 18,900 18,327 -3.0

Financial Services 18,175 21,238 16.9

Gvt Services 23,665 25,833 9.2

Community Services 12,179 14,118 15.9


A Closer Look at the Manufacturing Sector

Employment generated in manufacturing was concentrated in low return employment categories and in sectors with no overall wage increases

Table3.20 shows employment generation by sub-sector. When analyzing which sectors absorbed most of the employment generation, we find that 66 percent of the employment growth was in the food and beverage sector and the clothing sector. The tobacco sector contributed to an additional 8 percent of employment generation. The last row of the table shows employment generation by maquila, which contributed an amazing 32 percent of employment generation between 2001 and 2005, most of which was clothing (see Box 2.1)

Table 3.213.21 shows the median wages for these sectors. The two sectors that generated most of the employment growth in manufacturing (food and clothing) saw differing behavior. In the food sector, median wages decreased, while wages in the clothing sector increased.

Table3.20: Employment Generation by Sub-sector, 2001 and 2005

Total employment Employment Employment Share of total

2001 2005 2001-2005 2001-2005 2001-2005

Food and beverage 69,754 100,398 30,644 43.93 38.54

Tobacco 1,118 7,394 6,276 561.35 7.89

Textiles 4,305 6,472 2,167 50.34 2.73

Clothing 62,811 84,989 22,178 35.31 27.89

Wood products 8,766 10,571 1,805 20.59 2.27

Paper and prints 2,455 5,210 2,755 112.25 3.47

Petroleum 469 622 153 32.52 0.19

Chemicals 1,752 2,926 1,174 67.02 1.48

Plastic and rubber 1,441 1,362 (79) -5.46 -0.10

Other non-metallic 10,917 13,461 2,544 23.31 3.20

Metal and metal products 14,597 16,460 1,864 12.77 2.34

Machinery and equipment 1,945 3,695 1,750 89.99 2.20

98

Transport equipment 443 634 190 42.96 0.24

Other 15,271 21,355 6,084 39.84 7.65

Total Manufacturing 196,043 275,549 79,506 40.56 100.00

Maquila employment 35,565 61,000 25,435 71.52 31.99


Another way of looking at the types of jobs generated is to see which type of employment showed more growth. As was discussed, in Nicaragua the lowest income is observed in those working in household family enterprises, followed by the individual self-employed, while the highest income corresponds to employers, followed by wage and salaried workers. Table 3.22 shows that 32 percent of the employment generated in manufacturing was concentrated in family enterprise workers, 17 percent was in the individual self-employed, and 55 percent was in the wage and salaried categories. This means that 48 percent of the jobs created in manufacturing were of low income generation.

Table 3.21: Wages in the Manufacturing Sector, 2001 and 2005

Median Income $C 2001

2001 2005

Real

growth

(%)

Food and beverage 13,800 12,364 -10.4

Tobacco 16,960 9,848 -41.9

Textiles 9,960 8,379 -15.9

Clothing 11,495 13,424 16.8

Wood products 11,700 11,455 -2.1

Paper and prints 19,840 14,773 -25.5

Chemicals 30,805 20,606 -33.1

Plastic and rubber 16,350 18,038 10.3

Other non-metallic mineral products 10,032 11,882 18.4

Metal and metal products 11,970 15,273 27.6

Machinery and equipment 7,036 16,743 138.0

Other 4,944 17,438 252.7


99

The maquila sector may have counteracted the negative forces in income generation

It is worth noting that the maquila sector may have played an important role in counteracting the negative effect of the growing number of employed in family enterprises and the decreasing wages in the food sector. No data are available on wages for the maquila sector, but most of this is concentrated in the clothing sector, which saw an overall wage increase of 17 percent. If the maquila sector participated in this wage increase, then it may have had important positive effects on income generation, since this sector contributed to 32 percent of total employment (see Box 3.3)

Table 3.22: Employment Generation in Manufacturing by Type of Employment, 2001 and 2005

Total employment

Employment

growth

(number)

Employment

growth (%)

Share of total

employment

generation

2001 2005 2001-2005 2001-2005 2001-2005

Wage and salaried

workers 127,748

171,355 43,607 34.14 54.85

Individual self-

employed 33,153

46,837 13,683 41.27 17.21

Employers 12,443

10,850

(1,592) -12.80 -2.00

Family enterprise

workers 22,277

47,526 25,249 113.34 31.76

Other 421 -

(421) - -0.53

Total 196,043

276,569 80,526 41.08 101.28


The main picture that emerges from this section is as follows: growth was mainly concentrated in the manufacturing sector. Employment growth was concentrated in both manufacturing and agriculture. Unfortunately, this growth had a limited impact on the income opportunities of the poor. On the one hand, despite the fact that wages in agriculture increased, returns in this sector still offer the lowest income generation, so that growing employment in this sector is not likely to reduce poverty. On the other hand, employment generated in manufacturing was divided evenly between family enterprise employment and wage employment. Family enterprise employment has a very low income generation potential. Wage employment has a better potential, but its benefits to the poor seem to have been limited by two main factors: (i) wages decreased in the food sector, which saw most of the employment growth in manufacturing; and (ii) clothing,

100

The first maquila factories started in 1990, with the establishment of the first publicly owned Export

Processing Zone (EPZ), “Las Mercedes.” In 1994 the EPZ law was reformed to allow private ownership by

both foreign and domestic investors and an expansion of EPZ to other regions in the country, with the

particular aim of providing employment opportunities for poorer areas. Currently there is the zone of Las

Mercedes. The rest are distributed in 30 parques industriales (industrial parks), which are private EPZs. The

main investors are from Taiwan, the United States, Korea, Nicaragua, Italy, Honduras, Belize and Mexico.

Initially, EPZ firms were not allowed to use domestic raw materials, which limited their spillover effects to

those of employment generation, but with the signing of the CAFTA agreement these restrictions were

removed. EPZs have a 100 percent exemption on corporate income tax for the first 15 years of operation.

They are exempt from capital gains on real estate, all corporate taxes, excise, sales and municipal taxes as

well as import duties on machinery, inputs and equipment. Currently there are 84 firms, of which the large

majority are in the clothing sector.

In 2006 there were 68,300 employees in the EPZ, which corresponds to just over 3 percent of total

employment,18 percent of formal employment and 64 percent of formal employment in manufacturing. This

employment is mostly female (90 percent). EPZ firms have to comply with all of the labor regulations. There

is a special minimum wage for the EPZs, which is above the manufacturing EPZ wage. The rationale for this

is that labor force quality in the EPZ is higher than that in the average manufacturing labor force. Apparently,

to be a worker in the maquila sector employees have to have completed secondary education, which

substantially reduces access to this employment for the very poor. The fact that the minimum wage is higher

for the maquila sector may contribute to this selection.

For 2005 Maquila exports represented 50 percent of total exports, and the value of transformations services

(which corresponds to the sum of wages, utilities and services paid) was equivalent to 24 percent of Value

Added in manufacturing. The graph below illustrates the evolution of employment and output in the maquila

sector.

which was the other important sector in terms of job creation, offered limited employment access for the poor because of its skill requirements (secondary or above).

Box 3.2: Evolution of the Maquila Sector and Its Importance in the Employment Growth in

Manufacturing

101

EMPLOYMENT AND LABOR INCOME PROFILE OF THE POPULATION

Some background knowledge of how labor income and its components affect household poverty is useful to an understanding of what the priority policies should be. A labor profile of the population should inform policymakers as to how households are distributed among sectors, what their status in employment is and what the determinants of per capita household labor income are. This can be done by dividing the population into the poor and the non-poor – defined according to national and international poverty lines – or by using income quintiles. Another method for an understanding of how labor markets have affected household welfare is to disentangle the sources of labor income growth that are responsible for the observed changes in total labor income.

This section is structured as follows: the first section sketches a labor profile of the population, while the second focuses on the decomposition of household labor income growth through the use of the panel component of the 2001 and 2005 surveys. A final section is concerned with the agriculture sector.

Income and Employment Profile

Inactivity among the poor is more significant than unemployment

.1 Table3.23 and

Table 3.24 show the employment status of the working age population by quintile and poverty level. There has been an increase among the poor in total working age population, which means that there are fewer dependent people within a household and there are potentially better employment and income opportunities for the household as a whole. Among poor households, not all the members of working age looked for a job (the inactive members increased from 39 percent in 2001 to 40 percent in 2005), but most of those who sought a job actually found one: the proportion of unemployed remained almost constant, while the number of employed increased by almost 1 percentage point.

Table3.23: Employment Status of the Working Age Population by Quintile, 2001 and 2005

Q1 Q2 Q3 Q4 Q5 Total

200

1

200

5

200

1

200

5

200

1

200

5

200

1

200

5

200

1

200

5

200

1

200

5

Employed 9.3 10.4 10.4 11.6 12.4 12.2 13.6 13.4 16.4 15.2 62.2 62.8

Unemployed 0.2 0.2 0.2 0.3 0.5 0.4 0.5 0.7 0.8 0.6 2.2 2.2

Inactive 6.1 6.7 6.5 6.9 6.6 7.4 8.2 6.6 8.2 7.4 35.6 35.0

Total 15.6 17.3 17.1 18.7 19.6 20.1 22.3 20.7 25.5 23.2 100 100


102

Table 3.24: Employment Status of the Working Age Population by Poverty Level, 2001 and 2005

Poor Non Poor Total

2001 2005 2001 2005 2001 2005

Employed 59.7 60.6 61.2 64.4 60.6 62.8

Unemployed 1.3 1.5 3.0 2.7 2.3 2.2

Inactive 38.9 37.9 35.8 32.9 37.1 35.0

Total 100.0 100.0 100.0 100.0 100.0 100.0


The 2001 Household Survey asks the reasons for inactivity. Discouraged workers represent 11 percent among the poor and 8 percent among the non-poor. Those who are temporarily inactive (who have occasional jobs, are waiting for the harvest season, or are waiting to start a new job) correspond to 3 percent of the inactive among the poor and 2 percent among the non-poor. The largest shares of the inactive are homemakers and students: 52 percent and 34 percent are homemakers among the poor and the non-poor, respectively; 14 percent and 32 percent are studying (among the poor and non-poor, respectively).

The poor do not benefit from formal employment

Table 3.25 and Table 3.26 describe the structure of employment by quintile and poverty level. The fraction employed in each category is shown as a proportion of employed individuals. As discussed above, employment in the formal sector is very small; only 17 percent of the employed have formal jobs, most of which are held by the non-poor (14 percent). This close relationship between poverty and formal employment does not have an immediate interpretation. Either poverty is a consequence of lack of formal employment or being poor hampers access to formal employment. Alternatively both informal employment and poverty may be a consequence of lack of education and skills. It would be important to look further into this relationship, to assess the importance of generating formal employment vis-à-vis removing barriers to employment mobility among the poor, including access to education.

The number of waged workers employed in the informal sector decreased by 7.8 percent. What is more important is that the decrease was even greater among the poor than among the non-poor (-13 percent compared to -3.78 percent). The individual self-employed with no paid employees and also employers with paid employees decreased their share among the poor, while the number of family enterprises – often associated with low income generation – rose by around 1 percentage point for both the poor and the non-poor. The number of poor employed in the public sector remained almost constant.

10

3

Tab

le 3

.25:

Em

ploy

men

t Cat

egor

ies

by Q

uint

ile, 2

001

and

2005

Q

1 Q

2 Q

3 Q

4 Q

5 T

otal

2001

20

05

2001

20

05

2001

20

05

2001

20

05

2001

20

05

2001

20

05

Wag

ed e

mpl

oyed

pri

vate

form

al s

ecto

r 0.

5 0.

6 1.

4 1.

5 2.

8 2.

7 4.

6 5.

0 7.

6 7.

7 16

.9

17.4

W

aged

em

ploy

ed p

riva

te in

form

al s

ecto

r 6.

4 5.

4 6.

6 6.

2 8.

2 7.

2 6.

7 6.

9 6.

5 6.

3 34

.5

32.0

W

aged

em

ploy

ed p

ublic

sec

tor

0.1

0.2

0.3

0.2

0.4

0.3

0.9

0.9

1.5

1.7

3.1

3.4

Self

-em

ploy

ed w

ith

no p

aid

empl

oyee

s 2.

5 2.

2 3.

2 3.

3 3.

8 3.

6 4.

2 4.

4 5.

1 5.

0 18

.7

18.5

E

mpl

oyer

s w

ith

paid

em

ploy

ees

0.3

0.1

0.8

0.4

0.8

0.6

1.3

1.3

2.4

2.6

5.7

5.1

Fam

ily e

nter

pris

e w

orke

rs

5.1

6.0

4.5

4.7

4.0

4.9

4.4

3.8

3.2

4.3

21.2

23

.6

Tot

al

14.9

14

.5

16.7

16

.2

20.0

19

.4

22.0

22

.3

26.3

27

.7

100.

0 10

0.0

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

data

fro

m E

MN

V.

Tab

le 3

.26:

Em

ploy

men

t Cat

egor

ies

by P

over

ty L

evel

, 200

1 an

d 20

05

P

oor

Non

-poo

r E

xtre

mel

y P

oor

Tot

al

20

01

2005

20

01

2005

20

01

2005

20

01

2005

W

aged

em

ploy

ed p

riva

te fo

rmal

sec

tor

2.5

2.7

14.4

14

.7

0.2

0.4

16.9

17

.4

Wag

ed e

mpl

oyed

pri

vate

info

rmal

sec

tor

15.3

13

.5

19.2

18

.5

4.8

4.1

34.5

32

.0

Wag

ed e

mpl

oyed

pub

lic s

ecto

r 0.

4 0.

5 2.

6 2.

9 0.

1 0.

2 3.

1 3.

4 Se

lf-e

mpl

oyed

wit

h no

pai

d em

ploy

ees

6.9

6.4

11.8

12

.1

1.9

1.4

18.7

18

.5

Em

ploy

ers

wit

h pa

id e

mpl

oyee

s 1.

4 0.

6 4.

2 4.

4 0.

2 0.

0 5.

7 5.

1 F

amily

ent

erpr

ise

wor

kers

10

.7

12.1

10

.5

11.5

3.

9 4.

3 21

.2

23.6

T

otal

37

.3

35.9

62

.7

64.1

11

.1

10.4

10

0.0

100.

0

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

data

fro

m E

MN

V.

10

4

Tab

le 3

.27:

Str

uctu

re o

f Inc

ome

by Q

uint

ile, 2

001

and

2005

Q

1 Q

2 Q

3 Q

4 Q

5

20

01

2005

20

01

2005

20

01

2005

20

01

2005

20

01

2005

% in

com

e fr

om w

aged

em

ploy

men

t agr

. 29

.2

26.0

15

.1

14.5

12

.7

9.1

5.9

5.5

3.0

3.3

% in

com

e fr

om w

aged

em

ploy

men

t non

-

agr.

18

.1

15.0

29

.1

26.7

35

.4

32.3

37

.0

36.7

40

.0

34.1

% in

com

e fr

om s

elf-

empl

oym

ent a

gri.

22.3

27

.6

17.3

21

.4

13.7

14

.3

8.5

8.6

3.6

5.6

% i

ncom

e fr

om s

elf-

empl

oym

ent

non-

agri

. 8.

8 7.

2 16

.4

12.5

17

.0

16.7

21

.9

21.4

25

.2

22.1

% in

com

e fr

om p

ublic

tran

sfer

s 1.

9 6.

8 0.

8 5.

4 1.

6 4.

4 2.

6 3.

1 2.

3 3.

0

%

inco

me

from

fa

mily

re

mitt

ance

s

(int

erna

l)

3.6

3.3

4.0

3.3

2.1

3.5

4.3

3.9

2.6

3.4

%

inco

me

from

fa

mily

re

mitt

ance

s

(ext

erna

l)

0.9

1.4

1.3

2.6

2.6

4.2

2.6

4.9

4.4

8.8

% in

com

e fr

om o

ther

non

-lab

or s

ourc

es

15.3

12

.7

15.9

13

.7

15.0

15

.6

17.1

15

.9

18.9

19

.7

Tot

al

100.

0 10

0.0

100.

0 10

0.0

100.

0 10

0.0

100.

0 10

0.0

100.

0 10

0.0

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

data

fro

m E

MN

V.

10

5

Tab

le3.

28:

Stru

ctur

e of

Inc

ome

by P

over

ty L

evel

, 200

1 an

d 20

05

P

oor

Non

-poo

r E

xt P

oor

20

01

2005

20

01

2005

20

01

2005

% in

com

e fr

om w

aged

em

ploy

men

t agr

. 20

.7

18.6

5.

9 5.

0 30

.8

28.2

% in

com

e fr

om w

aged

em

ploy

men

t non

-agr

. 24

.6

22.2

38

.6

35.1

16

.4

14.0

% in

com

e fr

om s

elf-

empl

oym

ent a

gri.

19.0

23

.2

7.3

8.2

22.2

27

.4

% in

com

e fr

om s

elf-

empl

oym

ent n

on-a

gri.

13.8

10

.5

22.1

21

.1

8.7

6.4

% in

com

e fr

om p

ublic

tran

sfer

s 1.

4 6.

0 2.

3 3.

2 1.

8 6.

6

% in

com

e fr

om fa

mily

rem

itta

nces

(int

erna

l)

3.6

3.3

3.1

3.6

4.0

3.4

% in

com

e fr

om fa

mily

rem

itta

nces

(ext

erna

l)

1.4

2.4

3.4

6.4

0.9

1.2

% in

com

e fr

om o

ther

non

-lab

or s

ourc

es

15.6

13

.7

17.4

17

.5

15.2

12

.6

Tot

al

100.

0 10

0.0

100.

0 10

0.0

100.

0 10

0.0

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

data

fro

m E

MN

V.

106

Public transfers increased for the poor while remittances increased for the entire population but less so for the poor

Table 3.27 and Table 3.28 show the earnings profile of the population by quintile and by poverty level. There are no significant differences in the income structure of the population. For all levels of income, wage employment is the main source of income, accounting for a little over 45 percent of total income. The main difference is that poor wage earners receive their earnings from agriculture. This share decreased slightly for all income levels as remittances and public transfers increased. Remittances for the non-poor showed the largest increase, while public transfers showed the greatest increase for the poor, and among them for the extremely poor. It is worth noting that in 2001 public transfers were regressive in the sense that the poor received fewer transfers as a proportion of their total income. In 2005 this situation changed and public transfers became progressive.

Decomposing Changes in Labor Income

A traditional way to understand how labor markets have affected welfare is to disentangle the sources of household per capita labor income that are responsible for observed growth or decreases in average household income.32 Per capita household labor income – that is, the total income that the household earns from labor divided by the number of members in the household – can change for several reasons: (i) because income per employed member increases, (ii) because unemployment decreases, (iii) because the number of members that actively participate in the labor market rises, or (iv) because the dependency rate decreases.

In this section we attempt to understand what the main source of per capita labor income growth was between 2001 and 2005.33 For this purpose we use the panel component of the survey. For each household we decompose how much of the change in labor income was attributed to changes in each of the components mentioned above. In addition, we differentiate between four types of employment: (i) waged work in agriculture, (ii) waged work in non-agriculture, (iii) self-employment in agriculture, and (iv) self-employment in non-agriculture. This means that, in addition to studying the share of income growth that was due to increases in employment, we can further discuss whether this employment growth took place in any of the employment categories mentioned above. In the same way we can determine whether increases in earnings (income per employed member) were due to increases in earnings in agriculture/non-agriculture or to waged/self employment.

We take a sample of 1,250 households whose members are classified according to their occupation (waged and salaried workers versus the self-employed) and their sector of employment (agriculture and non-agriculture).34 We decompose labor income growth into

32 See Kakwani, Neri and Son (2006) for an application of this decomposition to the analysis of pro-poor rates of growth. 33 See Annex B for the methodology. 34 Some methodological clarifications are important. First, we select people of working age by dropping child laborers and the working elderly. This might imply overestimating productivity and underestimating employment shares and thus their contribution to total labor income growth. This is not a serious issue as they represent less than 2 percent and

107

four main terms: (i) income in “sector” j to total employment, (ii) the employment rate (which is equal to one minus the unemployment rate), (iii) the activity rate, and (iv) the share of working age people within a household. Furthermore, we disaggregate the first component into four sub-terms which represent productivity gains and employment shares in each “sector” of employment.

Table 3.28 presents the labor profile of the population by poverty level. The table shows the average labor income per employed person by employment category, the share of employed that work in each category, the employment rates (number of employed as a fraction of the active population), the participation rates (the active among the working age population), and the ratio between working age members and total members. All values are in 2001 córdobas.

Several features are worth noting. First, wage employment in non-agriculture is the highest earnings option in agriculture for both the poor and the non-poor. It should also be noted that the non-poor have a lower earnings rate than the poor in agriculture. This is most likely due to the fact that the non-poor work fewer hours in agricultural activities, and not because they earn less per hour. As discussed earlier, the poor are mostly found in agricultural self-employment: around 40 percent of the members of a poor household are employed in this category. Employment rates are very high among both the poor and the non-poor, being slightly higher for the poor, which merely reflects the fact that they cannot afford to be unemployed. Conversely, participation rates are slightly higher for the non-poor, and, as expected dependency rates are substantially higher for the poor: while for 2005 the poor showed 59 percent of their members of working age, among the non-poor this ratio was 67 percent.

There have been some important changes in the labor profile for the years analyzed. First, dependency rates among the poor decreased substantially, even more than for the non-poor. Second, there was an important increase in income per worker in self-employment for both agriculture and non-agriculture. On the other hand, the share of the employed in each employment category remained almost constant, with a slight increase in waged agricultural employment.

7 percent, respectively, of the total working population (average over the two years of the survey). Second, we are forced to drop a little over one-third of the sample in both 2001 and 2005 since there is no correspondence between workers and reported income within households: there are some households reporting income in a sector where there is no one employed, and vice versa. It might be an issue of misreported income for some of the households. Additionally, in many cases it seems that many people reporting earnings from agricultural self-employment are actually receiving rents from farms and are not directly employed in agriculture. As we believe this is not labor income but rents, we drop these households from the sample. Finally, we need to drop a small number of households that report a jump in the employment rate from 0 to 1 as they have an analogous increase in income and therefore they have a growth rate of income that goes to infinity. The selection is completely neutral across quintiles since we drop proportionally more poor than rich households as they are more likely to suffer from misreported income in agricultural business.

108

Table 3.28: Labor Profile by Poverty Level, 2001 and 2005

Poor Non-poor

2001 2005 2001 2005

Average labor income per worker employed in

waged work agriculture (annual 2001 C$) 2,104 2,220 1,079 1,264

Average labor income per worker employed in

waged work non-agriculture (annual 2001 C$) 6,235 6,227 21,476 18,219

Average labor income per worker self employed

in agriculture (annual 2001 C$) 5,898 6,032 2,942 3,279

Average labor income per worker self employed

in non agriculture (annual 2001 C$) 3,133 3,840 12,552 12,152

Share of waged employed in agriculture 11.2 12.4 2.5 2.9

Share of waged employed in non-agriculture 31.9 31.7 57.8 54.8

Share of self employed in agriculture 41.0 39.8 9.2 9.9

Share of self employed in non agriculture 15.9 16.0 30.1 31.7

Employment rate 98.4 98.0 96.5 96.1

Activity rate 63.1 65.0 66.5 68.9

Share of working age members within a HH 52.9 59.0 64.2 67.1

Average per capita labor income 3,588 4,522 9,989 10,085


Table 3.8 shows the same labor profile as that in Table 3.7 by quintile, which permits a clearer understanding of labor profiles and their changes among the poor and non-poor households. Two important phenomena stand out. First, for the poorest 20 percent, income from non-agricultural waged employment is not the highest earnings option, while for all other quintiles it is. For the poorest 20 percent, income from self-employment is the best earnings option. Two different effects might be responsible for this: (i) the poor might work fewer hours as waged employees in non-agriculture, and (ii) the poor might earn less per hour worked. For the poorest 20 percent the highest earnings option is self-employment in agriculture. The second phenomenon is the increase in income for the very poor households.

Table3.30 shows the average change in per capita household income by quintile. Between 2001 and 2005 the poor benefited more from economic growth as their labor income grew substantially more than that for the other groups. For the poorest quintile the annual per capita labor income growth rate was 14 percent. It was around 5 percent in the second quintile and it became negative in the last quintile (-1 percent). It is interesting to note that agriculture was the sector in which the poor, both the waged and the salaried

109

workers and the self-employed, saw their income decreasing, while the income of the poor working in other sectors showed a substantial increase. Despite the important growth in the per capita income of the lowest quintile, it was not sufficient to bring them above the poverty line. In 2001 the poorest 20 percent had an average per capita income of C$ 2,609; the 14 percent increase still left it well below the C$ 5,241 of the poverty line. Figure 3.7 illustrates this growth.

11

0

Tab

le 3

.29:

Lab

or P

rofi

le o

f the

Pop

ulat

ion

by Q

uint

ile, 2

001

and

2005

Q

1 Q

2 Q

3 Q

4 Q

5

20

01

2005

20

01

2005

20

01

2005

20

01

2005

20

01

2005

Ave

rage

lab

or i

ncom

e pe

r w

orke

r in

wag

ed a

gric

ultu

ral

empl

oym

ent (

annu

al 2

001

C$)

3,

057

2,74

2 1,

716

1,77

3 1,

243

1,90

4 45

8 97

0 1,

474

1,24

2

Ave

rage

la

bor

inco

me

per

wor

ker

in

wag

ed

non-

agri

cultu

ral e

mpl

oym

ent (

annu

al 2

001

C$)

3,

841

3,63

3 7,

864

7,32

8 9,

303

10,5

51

15,4

95

14,4

82

33,6

83

24,9

55

Ave

rage

lab

or i

ncom

e pe

r w

orke

r in

agr

icul

tura

l se

lf-

empl

oym

ent (

annu

al 2

001

C$)

5,

175

6,87

8 6,

623

6,10

7 4,

321

4,62

1 2,

792

3,85

9 2,

579

1,95

5

Ave

rage

lab

or i

ncom

e pe

r w

orke

r in

non

-agr

icul

tura

l

self

-em

ploy

men

t (an

nual

200

1 C

$)

1,51

1 1,

536

3,56

4 3,

929

5,78

9 7,

043

10,2

21

11,4

59

18,7

23

15,8

09

Em

ploy

men

t sha

re in

wag

ed a

gric

ultu

ral e

mpl

oym

ent

15.6

18

.1

9.8

9.1

4.8

6.7

1.4

2.8

2.2

1.4

Em

ploy

men

t sh

are

in

wag

ed

non-

agri

cultu

ral

empl

oym

ent

20.6

22

.5

39.0

33

.6

49.4

47

.6

56.8

53

.6

60.6

59

.2

Em

ploy

men

t sha

re in

agr

icul

tura

l sel

f-em

ploy

men

t 53

.3

49.4

36

.7

38.6

20

.9

22.2

8.

6 9.

1 3.

5 4.

2

Em

ploy

men

t sha

re in

non

-agr

icul

tura

l sel

f-em

ploy

men

t 10

.6

9.8

14.5

18

.7

24.4

23

.5

33.0

34

.4

33.2

33

.5

Em

ploy

men

t rat

e 98

.4

97.8

98

.2

98.1

97

.2

96.8

96

.5

97.3

96

.5

95.0

Act

ivity

rat

e 63

.9

64.6

61

.2

65.5

65

.6

65.8

65

.9

69.6

67

.7

69.7

Shar

e of

wor

king

age

mem

bers

with

in a

hou

seho

ld

49.0

55

.9

55.2

59

.5

55.9

62

.8

62.4

65

.7

70.7

70

.8

Ave

rage

per

cap

ita la

bor

inco

me

2,59

8 3,

459

4,02

2 5,

041

4,70

5 5,

977

6,97

8 8,

201

15,9

06

13,6

96

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

data

fro

m E

MN

V.

111

Table3.30: Per Capita Household Income Changes, by Quintile, 2001-05

Quintile Annual growth rate

of per capita

household income

Y/N

Level of per capita

household income

$C 2001 Y/N

1 14.22 2,608.9

2 5.36 3,844.4

3 8.71 4,862.36

4 6.98 7,015.593

5 -1.00 14,897.57

2005 Poverty line in $C 2001: 5,241


Figure 3.7: Growth in Average per Capita Income by Quintile in 2001 (panel)

-2.00

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

1 2 3 4 5


Finally, the decomposition results are shown in Table 3.31The table shows the contribution of each component to the observed change in per capita labor income by quintile. The two most important factors in raising the income of the poorest 20 percent of the population were (i) the observed increase in income per employed worker in agricultural self-employment (44 percent of the total increase in per capita household income) and (ii) the important increase in the share of working age people within a household (38 percent of the observed change in per capita household income). Participation rates also made an important contribution (11 percent).

For the second quintile the main source of income growth came from lower dependency rates and higher participation rates. The larger fraction of employed household members in non-agricultural waged jobs also contributed to the higher labor income.

Decreases in the number of dependents per working age person were seen in all but the richest 20 percent, and they were an important opportunity for poverty reduction. In all

112

but the richest 20 percent, participation rates also increased, and in all but the middle quintile employment rates increased, contributing positively to poverty reduction.35

It should be noted that increases in agricultural wages for the panel sample seem to be relatively small compared to the wage increases seen for the whole sample. It is unclear from this exercise whether agricultural wages may have played a more important role in reducing the incidence of poverty.

Table 3.31: Shapley Decomposition of per Capita Labor Income, by Quintile

Q1 Q2 Q3 Q4 Q1

Income per waged worker in

agriculture -0.9 -17.1 -1.55 1.29 -3.3

Share of employed in waged agriculture 5.15 -2.44 -1.17 2.76 -9.8

Income per waged worker in non-

agriculture 3.45 19.13 33.21 -3.69 289.27

Share employed in waged non-

agriculture 2.7

-

32.69 3.02 -8.71 77.7

Income per self-employed worker in

agriculture 44.64 -5.14

-

14.72 -3.39 55.61

Share of self-employed agriculture -14 3.38 -7.47 -4.9 -20.84

Income per worker self-employed in

non-agriculture 10.29

-

20.86 3.28 30.82 136.31

Share of self-employed in non-

agriculture -0.96 14.07 6.76 1.05 -31.38

Employment rate 0.53 4.5 -3.74 5.46 4.0

Participation rate 11.14 63.99 16.4 42.49 -342.18

Inverse of dependency 37.94 73.17 65.96 36.83 -55.39

Total 100 100 100 100 100


35 We should be careful in making an interpretation given the small number of households in each quintile. For these estimations we excluded the 2.5 percent in the tails of changes in total labor income, and results change when these outliers are included. Additionally, the sample selected shows no increase in wages except for the third and fourth quintiles, and agricultural wage increases are lower than for the whole sample.

113

Agriculture: What Happened?

Household survey data show a sizable increase in real per capita labor income for the self-employed in agriculture, which points to a need to understand where this gain came from: was it driven by relative prices, quantities or productivity?

.2 By using data from the FAO and the Central Bank of Nicaragua, we can disentangle the effect of each component of the increase in agricultural production which represents 26 percent of total Nicaraguan production. We are interested in goods produced by the poor (precisely, by the self-employed among the poor), even if none of them could be considered as contributing to a sufficient proportion of the GDP or generating a large number of jobs. In addition, we analyze the behavior of export products which might have affected agricultural wages. The products considered are beans, coffee, meat, milk, rice and corn. The first two items are export goods: they represent 8.5 percent and 35.5 percent, respectively, of agricultural GDP (in 2001). The rest are considered “sensitive” goods, according to the definition used by the Nicaraguan Agriculture and CAFTA Report (2004): they “…are those with high tariff protection, are economically vulnerable and possess significant socio-economic importance.” This means that they are produced by small and medium scale farmers (see Table 3.32). Beef production represents the largest share of GDP (4 percent in 2001), and the production of white corn generates the largest number of jobs (175,000), or 9 percent of total employment. Figure 3.8 to Figure 3.11 look at productivity (yields), area harvested and relative producer prices.

Table 3.32: Number of Farms, by “Sensitive” Product, according to Farm Size, 2001

Product Small Medium Large Total

Rice 6,714 4,873 5,742 17,329

Corn 58,378 53,087 29,919 141,384

Milk 64,855 26,391 5,718 96,964

Beef 64,855.00 26,391.00 5718 96,964.00

Source: Nicaraguan Agriculture, and CAFTA, 2004.

Absolute productivity remained constant for most products while productivity decreased relative to trading partners

Figure 3.8 shows the evolution of absolute productivity measured as yield per hectare. Productivity remained constant for the period under analysis for beans, coffee and milk, while rice and maize saw increases in productivity of around 13 percent.

114

Since the United States and some Central American countries (Costa Rica, El Salvador and Honduras) are the main trade partners of Nicaragua, we compare relative productivity (as a ratio to U.S. productivity) across these countries. For all products analyzed (except for dry beans) Nicaragua seems to be the least efficient country among the four countries in Figure 3.3. In some cases, relative productivity, measured as yield per hectare of cultivated land with respect to U.S. productivity, decreased over the 15 year period analyzed as well as over the years of the surveys (2001-05). Relative productivity actually decreased for three products out of a total of four (Figure 3.9), the exception being coffee. Then again, the relative gains in productivity were modest.

Despite these low levels of productivity and the decreases in relative productivity, the observed gains in absolute productivity for maize and rice may have helped the small farmers of these products.

The cultivated area increased, which pulled production up

Between 2001 and 2005, the area harvested increased for three of the products analyzed (milk, maize and beans), while for the others it remained relatively constant (Figure 3.10). Given that productivity remained almost constant, this was the main source of the production increases seen for all of the goods considered ( Figure 3.11) In the case of maize, the increase in output was close to 30 percent, while for the other products it was less than 15 percent. In any case, and despite the important increases in output, it is unlikely that aggregate output growth was above the observed employment growth of 21 percent for the whole period, which would explain the decrease in Value Added per worker reported in the second section of this paper.

Figure 3.8: Productivity of “Sensitive” Products, by Yield per Hectare, 1990-2005

Source: Own calculations with FAO data.

Yield per hectare (Kg/Ha)

0

500

1000

1500

2000

2500

3000

3500

4000

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Years

Kilo

s p

er h

ecta

r

Beans (incl. cow peas), dry

Coffee, green

Maize

Milk, whole, fresh

Rice, paddy

115

Figure3.9: Nicaragua Relative Productivity by Product, 1990-2005

Source: Own calculations with data from FA.

Figure 3.10: Area Harvested for “Sensitive” Products, 1990-2005


Yield per hectare (relative to US) - Milk

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salvador

Honduras

Yield per hectare (relative to US) - Dry Beans

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salavador

Honduras

Yield per hectare (relative to US) - Coffee

0

0.5

1

1.5

2

2.5

3

3.5

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salavador

Honduras

Yield per hectare (relative to US) - Rice

0

0.2

0.4

0.6

0.8

1

1.2

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salvador

Honduras

Yield per hectare (relative to US) - Milk

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salvador

Honduras

Yield per hectare (relative to US) - Dry Beans

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salavador

Honduras

Yield per hectare (relative to US) - Coffee

0

0.5

1

1.5

2

2.5

3

3.5

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salavador

Honduras

Yield per hectare (relative to US) - Rice

0

0.2

0.4

0.6

0.8

1

1.2

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

Kg

/Ha

Nicaragua

Costa Rica

El Salvador

Honduras

Area harvested (1000 Ha)

0

100

200

300

400

500

600

700

800

900

1000

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Years

Th

ou

san

ds

hec

tars


Coffee, green

Maize

Milk, whole, fresh

Rice, paddy

116

Figure 3.11: Production Volume for “Sensitive” Products, 1990-2005


As regards prices, we first present the pattern of producer price indexes for three baskets of goods as they are computed by the Central Bank: the three aggregates are cereals, export goods and meat. We observe a significant price increase for all the products over the survey years (Figure 3.12 to Figure 3.15). If we look at the producer prices of each single good, we observe an increase for all the goods considered except for milk (according to the data from the Central Bank of Nicaragua).

This suggests that the terms of trade improved for agricultural producers, as the producer prices increased more than the overall consumer price index (CPI). The case of export goods (Figure 3.12) deserves special attention: after 1999, the basis year, the producer price index dropped dramatically. This may be attributed to the 2000 crisis in coffee prices in the world market that affected the price of green coffee, which is the producer price of coffee but not the price of coffee in grains.

Thus, it seems that the gains made by the self-employed in agriculture between 2001 and 2005 are due to the evolution of the terms of trade (i.e. relative prices). Increases in the area harvested were important but were probably not sufficient to keep Value Added per worker from falling in response to the apparent inflow of workers to agriculture. Rises in agricultural production may also explain the increases in agricultural wages seen for the overall sample.

The decrease observed in productivity (for two out of a total of four goods) relative to the United States poses a considerable challenge for Nicaragua with respect to other Central American trade partners: Nicaragua needs investments in order to recover productivity and to fill the gap with the main trade partners. This is particularly important, because income for rural households appears to be tied to price variations, which increases the vulnerability of this population to price shocks.

Production (1000 tons)

0

100

200

300

400

500

600

700

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Years

Th

ou

san

ds

ton

nes Beans (incl. cow peas), dry

Bovine meat

Coffee, green

Maize

Milk, whole, fresh

Rice, paddy

117

Figure 3.12: Relative Prices for Export Goods, 1999-2006

.


Figure 3.13: Relative Prices of Trade for Meat, 1999-2006


Figure 3.14: Relative Prices for Cereals, 1999-2006


Cereals (Goods Price Index/CPI)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1999 2000 2001 2002 2003 2004 2005 2006

Years

Producer Price Index/CPI

Meat (Goods Price Index/CPI)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1999 2000 2001 2002 2003 2004 2005 2006

Years

Producer Price Index/CPI

Yield per hectare (Kg/Ha)

0

500

1000

1500

2000

2500

3000

3500

4000

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Years

Kilo

s p

er h

ecta

r


Coffee, green

Maize

Milk, whole, fresh

Rice, paddy

118

Figure 3.15: Relative Prices for Sensitive Products, 2001-06


Policy Implications and Further Research

Despite modest growth and important employment growth, Nicaragua saw no major decrease in poverty. Two main factors explain this outcome: (i) an important fraction of new jobs were generated in agriculture, which offers the lowest returns among economic activities; (ii) jobs generated outside of agriculture with good earnings were either accessible only to the most educated sector (the maquila sector) or experienced a decrease in wages (industrial food and beverage sector).

The depth of poverty, as measured by the poverty gap, saw significant reductions. This was due to the following: (i) increases in the relative prices of products produced by the agricultural poor; and (ii) an increase in the amount of remittances. However, the increase in income was not sufficient to bring the poorest out of poverty.

The analysis presented here has implications for action on five policy fronts: (i) skills, (ii) productivity, (iii) employment generation, (iv) geographical mobility, and (v) minimum wage regulation.

Skill levels in Nicaragua are substantially lower than in neighboring countries. Despite the important progress on this front, substantial efforts are still needed. Increasing skills in Nicaragua would involve both benefits and risks, and any educational policy must try to magnify the former while minimizing the later. Currently, the supply of skills seems to be growing at a higher rate than the demand, and, as a consequence, wages for the skilled population might drop. Such a drop in the returns to schooling could serve as a disincentive to acquiring education and might reduce employment growth in the waged sectors of the economy, mainly manufacturing and services. It is thus imperative that

Producer Prices/CPI (Basis Year 2001)

0

50

100

150

200

250

2001 2002 2003 2004 2005

Year

café pergamino

arroz en granza

frijol

carne de res (en pie)

leche cruda

maiz

119

policies to increase growth in these sectors (and with this to increase the demand for labor) are undertaken. Currently, constraints to growth appear to lie outside of the labor market, and involve macroeconomic uncertainty and lack of affordable credit. An increase in the demand for waged work and in particular of higher skills is unlikely to be seen unless these constraints are addressed. However, higher skills are important determinants of earnings among the urban self-employed and family enterprises as well as urban informal wage workers. It is also important in increasing earnings in wage agricultural work and among agricultural employers. Targeting the expansion of education to the rural sector appears to have important potential as a poverty reducing strategy. Education is a key determinant in accessing better earning opportunities and in moving out of agriculture.

Productivity in Nicaragua declined during the period studied, with important decreases within agriculture. Nicaragua has the lowest levels of agricultural productivty among its neighbors and trading partners. This factor, together with some indirect evidence of low mobility between urban and rural areas, suggests that raising productivity in agriculture should also be in the forefront of policy initiatives. Without targeted investments in agricultural productivity and agricultural exports, decreasing rural poverty in the short and medium runs seems implausible.

Employment generation should be targeted towards the formal secondary and tertiary unskilled intensive sectors. Exploring targeted interventions to foster growth in such sectors as tourism, with training programs for the unskilled specifically designed for the industry, seems a policy worth exploring. Nicaragua recently conducted a tourism investment climate survey which may provide initial input for the design of this policy. Regulation does not seem to pose a constraint for job creation and growth. The most urgent policies in this area would seem to be to address macroeconomic uncertainty and credit constraints.

The study pointed towards possible geographical barriers to mobility between the urban and the rural sectors. It is as yet unclear as to how important these barriers are, and what its main determinants are. Further study on this issue may yield promising policy implications. Becoming an employer is also linked with availability of non-labor income, which lead to the natural question of whether alleviating credit constraints might help more family enterprise workers or self employed become employers.

It is unclear to what extent the current minimum wage structure in Nicaragua provides any benefits or manages to take into account the skills of the labor force. More productive sectors have higher minimum wages and thus may be constraining the poorest from accessing jobs in precisely the sectors that offer the highest earnings potential. Workers with productivity below the minimum wage will be rationed out of formal employment. The higher the minimum wage is, the less is the access the unskilled have to these sectors. Results of this study suggest that minimum wages may be binding for the maquila manufacturing sector and probably for commerce. Further study in this area may clarify the extent to which the current sectoral structure of minimum wages is beneficial to the poor or is actually constraining them from accessing jobs in the most productive sectors of the economy.

120

REFERENCES

Arias, Omar; Blom, Andreas; Bosch, Mariano; Cunningham, Wendy; Fiszbein, Ariel; López Acevedo, Gladys; Maloney, William; Saavedra, Jaime; Sánchez-Páramo, Carolina; Santamaría, Mauricio; and Siga, Lucas (2005). “Pending Issues in Protection, Productivity Growth, and Poverty Reduction.” Policy Research Working Paper 3799, World Bank,Washington D.C.

Besley, Timothy, and Burgess, Robin (2004). “Can Labor Regulation Hinder Economic Performance? Evidence from India,” Quarterly Journal of Economics, 119 (1): 91-132.

Bibi, Sami (2005). “When Is Economic Growth Pro-Poor? Evidence from Tunisia,” Paper Presented at the Economic Research Forum, 12th Annual Conference, Egypt.

Borjas, G. J. 1996. Labor Economics. New York: McGraw-Hill.

Bosworth, D., P. Dawkins, T. Stromback. 1996. The Economics of the Labor Market. Essex, England: Addison Wesley Longman, Ltd.

Bourguignon, François (2002). “The Growth Elasticity of Poverty Reduction: Explaining Heterogeneity across Countries and Time Periods,” DELTA Working Papers No. 2002-03.

Bourguignon F., Fournier M. Gurgand M. “selection bias corrections based on the multinomial logit model: monte carlo simulations” Journal of EconomicSurveys”, Vol21 No.1, 2007.

Bourguignon, François, and Fields, Gary S. (1997). “Discontinuous Losses from Poverty, Generalized Pα Measures, and Optimal Transfers to the Poor,” Journal of Public Economics, Vol. 63(2): 155-175.

Calderón, César, and Chong, Alberto (2005). “Are Labor Market Regulations an Obstacle for Long Run Growth?” in Labor Markets and Institutions, Ed. Restrepo, J.E., and Tokman, A. Banco Central de Chile, Santiago, Chile, 2005.

Chen, Shaohua, and Ravallion, Martin (2004). “How Have the World’s Poor Fared since the Early 1980’s,” World Bank Research Observer, No. 19: 141-169.

Contreras, Dante (2001). “Economic Growth and Poverty Reduction by Region: Chile 1990-1996,” Development Poverty Review, 19(3): 291-302.

Cukierman, Alex; Rama, Martin; and van Ours, Jan (2001). “Long-run Growth, the Minimum Wage and Other Labor Market Institutions. Preliminary notes,” mimeo, 2005.

Datt, Gaurav, and Ravallion, Martin (1993). “Growth and Redistribution Components of Changes in Poverty Measures: A decomposition with Applications to Brazil and India in the 1980s,” Journal of Development Economics 38: 275-295.

Datt, Gaurav, and Ravallion, Martin (1998). “Farm Productivity and Rural Poverty in India,” Journal of Development Studies, No 24(4): 62-85.

121

Del Carpio, Ximena (2006) “Voices of Nicaragua. “A Qualitative and Quantitative Approach to Viewing Poverty in Nicaragua” Mimeo, World Bank.

Dickens, W. T., and K. Lang. (1985). “Testing Dual Labor Market Theory: Reconsideration of the Evidence.” Working Paper Series No. 1670, 1–27. NBER, Cambridge, MA.

Dollar, David, and Kraay, Aart (2002). “Growth Is Good for the Poor,” Journal of Economic Growth, 7(3): 195-225.

Essama-Nssah, B. (2005). “A Unified Framework for Pro-Poor Growth Analysis,” Economics Letters, 89: 216-221.

Esfahani, H. S., and D. Salehi-Isfahani. 1989. “Effort Observability and Worker

Productivity: Towards an Explanation of Economic Dualism.” The Economic Journal 99: 818–36.

Fields, G. (2006). “Employment in Low-Income Countries beyond Labor Market Segmentation?” Document prepared for the World Bank conference “Rethinking the Role of Jobs for Shared Growth,” Washington D.C., June 19, 2006.

Fortreza, Alvaro, and Rama, Martin (2001). “Labor Market ‘Rigidity’ and the Success of Economic Reforms across More Than 100 Countries,” Policy Research Working Paper 2521. World Bank, Washington D.C.

Freeman, Richard, and Oostendorp, Remco H. (2000). “Occupational Wages around the World (OWW) Database,” www.nber.org/oww.

Goodman, Leo A. (1960). “On the Exact Variance of Products,” Journal of the American Statistical Association, 55(292): 708-713.

Huassman, Ricardo; Rodrik, Dani; and Velasco Andrés (2005). “Growth Diagnostics,” Unpublished; John F. Kennedy School of Government, Harvard University, Boston MA.

http://ksghome.harvard.edu/~drodrik/barcelonafinalmarch2005.pdf.

ILO (2003). “Review of the Core Elements of the Global Employment Agenda,” Committee on Employment and Social Policy; Geneva, March 2003.

Islam, Rizwanul (2004). “The Nexus of Economic Growth, Employment and Poverty Reduction: An empirical Analysis,” Issues in Employment and Poverty Discussion Paper No. 14. ILO, Geneva.

Kakwani, Nanak, and Pernia, Ernesto M. (2000). “What Is Pro-Poor Growth?” Asian Development Review, No. 18(1): 1-16.

Kakwani, Nanak; Khandker, Shahid; and Son, Hyun H. (2006). “Pro-Poor Growth: Concepts and Measurement with Country Case Studies,” UNDP International Poverty Center, Working Paper No. 1, Brazil.

122

Kakwani, Nanak; Neri, Marcelo; and Son, Hyun H. (2006). “Linkages between Pro-Poor Growth, Social Programmes and Labour Market: The Recent Brazilian Experience,” UNDP International Poverty Center, Working Paper No. 26, Brazil.

Kapsos, Steven (2004). “Estimating the Requirements for Reducing Working Poverty: Can the World Halve the Working Poverty by 2015?” Employment Strategy Papers 14, Employment Strategy Department, ILO, Geneva.

Katz, L. F. and K. M. Murphy (1991). "Changes in relative wages, 1963- 1987: supply and demand factors." National Bureau of Economic Research. Working Paper Series No. 3927: [1]-38

Katz, L. F., and L. H. Summers. (1988) “Can Inter-Industry Wage Differentials Justify Strategic Trade Policy?” Working Paper Series No. 2739. NBER, Cambridge, MA.

Kraay, Aart (2006). “When Is Growth Pro-Poor? Evidence from a Panel of Countries,” Journal of Development Economics, No. 80(1): 198-227.

Landman, Oliver (2004). “Employment, Productivity and Output Growth,” Employment Strategy Papers No. 2004/17, Employment Trends Department, ILO, Geneva.

Layard, P. R. G., S. J. Nickell, R. Jackson (1991). Unemployment: Macroeconomic Performance and the Labour Market. Oxford, England, New York: Oxford University Press.

Loayza, Norman, and Raddatz, Claudio (2006). “The Composition of Growth Matters for Poverty Alleviation,” mimeo, June 2006.

López, J. Humberto (2004a). “Pro-Poor Growth: A Review of What We Know (and What We Don’t),” Paper prepared for the Operationalizong Pro-Poor Growth Program, mimeo, World Bank, Washington D.C.

López, J. Humberto (2004b). “Pro-Growth Poor-Poor: Is There a Trade-off?” World Bank Policy Research Paper No. 3378, World Bank, Washington D.C.

López, J. Humberto, and Servén, Luis (2006). “A Normal Relationship? Poverty, Growth and Inequality,” Policy Research Working Paper No. 3814, World Bank, Washington D.C.

Lucas, Sarah, and Timmer, Peter (2005). “Connecting the Poor to Economic Growth: Eight Key Questions,” Working Paper Center for Global Development, Washington D.C.

Lustig, Nora C., and McLeod, Darryl (1997). “Minimum Wages and Poverty in Developing Countries: Some Empirical Evidence,” in Labor Markets in Latin America, Edwards and Lustig, Eds. Brookings Institution Press, Washington D.C.

Maloney, William F., 2004. "Informality Revisited," World Development, Elsevier, vol. 32(7), pages 1159-1178, July.

Menezes-Filho, Naercio, and Vasconcellos, Ligia (2004). “Operationalizing Pro-Poor Growth: A Country Study for Brazil,” Background Paper in the Operationalizing Pro-Poor Growth Program Series by Agence Française pour le Développement, Bundesministerium für

123

Wirtschaftlieche Zusammenarbeit, U.K, Department for International Development and The World Bank, Washington D.C.

Neuman, S. and Oaxaca, R (2004), “Wage Decompositions with Selectivity-Corrected Wage Equations: A Methodological Note. Journal of Economic Inequality (2) 3-10.

Osmani, S.R. (2005). “The Employment Nexus between Growth and Poverty: An Asian Perspective,” Sida Studies No. 15, Sweden.

Prasada Rao, D.S; Colleli, Timothy J.; and Alauddin, Mohammad (2004). “Agricultural Productivity Growth and Poverty in Developing Countries, 1970-2000,” Employment Strategy Papers No. 2004/9, Employment Trends Department, ILO, Geneva.

Rama, Martin, and Artecona, Raquel (2002). “A Database for Labor Market Indicators Across Countries,” mimeo, World Bank, June 2002, Washington D.C.

Ravallion, Martin (2004). “Pro-Poor Growth: A Premier,” Policy Research Working Paper No. 3242, World Bank, Washington D.C.

Ravallion, Martin (2005). “Inequality Is Bad for the Poor,” Policy Research Working Paper No. 3677, World Bank, Washington D.C.

Ravallion, Martin, and Chen, Shaohua (2003). “Measuring Pro-Poor Growth,” Economics Letters, 78: 93-99.

Ravallion, Martin, and Datt, Gaurav (2002). “Why Has Economic Growth Been More Pro-Poor in Some States of India Than Others?” Journal of Development Economics, 381-400.

Satchi, Mathan, and Temple, Jonathan (2006). “Growth and Labor Markets in Developing Countries,” Discussion Paper No. 06/581, Department of Economics, University of Bristol, U.K.

Shorrocks, Anthony F. (1999). “Decomposition Procedures for Distributional Analysis: A Unified Framework Based on the Shapley Value,” mimeo, University of Essex.

Sundaram K. and Suresh K. Tendulkar, (2002). The Working Poor in India: Employment-Poverty Linkages and Employment Poverty Options, IEPDP 4, ILO, Geneva.

Temple, Jonathan, and Woessmann, Ludger (2006). “Dualism and Cross-Country Regressions,” Journal of Economic Growth, September, 11(3), 187-228

Timmer, Peter C. (2005). “Agriculture and Pro-Poor Growth: An Asian Perspective,” Working Paper No. 63, Center for Global Development, Washington D.C.

World Bank (2005). “Pro-Poor Growth in the 1990’s: Lessons and Insights from 14 Countries,” Agence Française pour le Développement, Bundesministerium für Wirtschaftlieche Zusammenarbeit, U.K, Department for International Development and The World Bank, Washington D.C.

124

Annex A: DECOMPOSITION OF PER CAPITA VALUE ADDED GROWTH Step1: decompose aggregate growth:

A simple way of understanding how growth has translated into increases in productivity and

employment at the aggregate level and by sectors (or regions), is to perform a simple

decomposition of growth in per capita GDP. To do so, note that per capita GDP, Y/N=y can be

expressed as: Y Y E A

N E A N=

Equation 1

Or:

* *y e aω=

where Yi is total Value Added E is total employment, A is the total population of working age

and N is total population. In this way Y/E=ω it total output per worker, E/A is the share of

working age population (i.e. the labor force) employed and A/N is the labor force as a fraction of

total population.

Thus change sin per capita Value Added can be decomposed into changes in output per worker,

changes in employment rates and changes in the size of the labor force. Using shapely

decompositions this will be equal to: 1 1 0 0 1 0 0 1 1 1 0 0 1 0 0 1

1 1 0 0 1 0 0 1

3 6 3 6

3 6

t t t t t t t t t t t t t t t t

t t t t t t t t

e a e a e a e a a a a ay e

e e e ea

ω ω ω ωω

ω ω ω ω

= = = = = = = = = = = = = = = =

= = = = = = = =

+ + + +⎡ ⎤ ⎡ ⎤Δ = Δ + + Δ + +⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦

+ +⎡ ⎤Δ +⎢ ⎥⎣ ⎦

The first term in the summation will be the contribution of changes in output per worker, the

second term the contribution of changes in the employment rate and the third term the

contribution to changes in the demographic component.

With this information we can present aggregate growth in terms of each of these components: 1 1 0 0 1 0 0 1 /

3 6t t t t t t t te a e a e a e a

yω ω = = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥⎣ ⎦

will be the fraction of growth that can be

linked to changes in output per worker, 1 1 0 0 1 0 0 1 /3 6

t t t t t t t ta a a ae e y

ω ω ω ω= = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥⎣ ⎦

will be the fraction of growth that can be linked to changes in the employment rate, and 1 1 0 0 1 0 0 1 /

3 6t t t t t t t te e e e

a a yω ω ω ω= = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥⎣ ⎦

will be the fraction of growth that can be

linked to changes in the share of total population that is of working age; where the bar denotes the

fraction of growth explained by the component. In this way percentage growth between two

periods can be expressed as: y y y y

e ay y y y

ωΔ Δ Δ Δ= + +

Once we have decomposed aggregate employment growth we can go further and understand i)

the role played by different sectors in changes in employment and ii) the role of capital, Total

Factor Productivty and intersectoral shifts in explaining changes in output per worker, both at the

aggregate level and by sectors. This amounts to doing a step wise decomposition: first

decomposing aggregate growth into employment and productivity changes and the decomposing

employment and productivty changes by sectors.

125

Step 2: Understanding which sectors contributed most to employment generation.

To understand which sectors contributed to most of the employment generation we can further

decompose employment growth ( eΔ ) by sectors. The easiest is of course to express the total

growth in employment as the sum of employment generation in each sector.

1

s

ii

e e=

Δ = Δ∑

Where ii

Ee

AΔ = Δ is just the change in employment in sector i as a share of total working age

population. Let /i

eie e e≡ Δ Δ , denote the fraction of the aggregate employment rate change that

can be linked to changes in employment in sector i. The supra-index e will make explicit that it is

the contribution to employment growth (as opposed to total per capita growth).

Step 3: Decompose changes in output per worker by sectors and in between and within components

We can further decompose output per worker into sectoral employment shifts and changes in

output per worker by sectors by noting that:

i i

S i

Y EY

E E E=∑

Or equivalently:

1

S

i ii

sω ω=

=∑

where Yi is Value Added of sector i=1…S, Ei is employment in sector i, and E is total

employment. This means that ii

i

Y

Eω = will correspond output per worker in sector i, i

i

Es

E= is

the share of sector i in total employment. This equation juts states that changes in output per

worker are the weighted sum of changes in output per worker in all sectors, where the weights are

simply the employment share of each sector.

Using the shapely approach, changes in aggregate output per worker can be decomposed as:

1, 0 1, 1 2, 0 2, 1 , 0 , 1 , 0 , 11 2

1

* * ... * *2 2 2 2

w B

St t t t i t i t i t i t

i ii

s s s s s ss

ω ω

ω ωω ω ω ω= = = = = = = =

=

Δ Δ

+ + + +⎛ ⎞ ⎛ ⎞ ⎛ ⎞ ⎛ ⎞Δ = Δ + Δ + + Δ + Δ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟

⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠∑

14444444444444244444444444443 14444244443

Each terms , 0 , 1*2

i t i ti

s sω = =+⎛ ⎞

Δ ⎜ ⎟⎝ ⎠

are the change in output per worker due to changes in output per

worker in sector s. The last term in the equation ΔωB is the change in output per worker due to

due to intersectoral employment changes (i.e. between sectors). That is employment movements

from low productivty sectors to high productivity sectors should increase total output per worker,

and the flows from high productivty sectors to low productivity sectors should reduce aggregate

output per worker. If this last term is negative the reallocation of employment by sectors was

detrimental to overall productivty growth. Finally, the term Δωw corresponds to total changes in

output per worker net of relocation effects (or within component).

126

We can then denote the fraction of aggregate output per worker growth that can be linked to

growth in output per worker in sector i as , 0 , 1* /2

i t i ti i

s sωω ω ω= =+⎛ ⎞≡ Δ Δ⎜ ⎟

⎝ ⎠, where again the bar

denotes the fact that we are referring to contributions, and the supra-index denotes the fact that it

is a contribution to aggregate output per worker growth ω, rather than a contribution to output per

capita growth y.

Similarly we can define the contribution of within sector productivity growth as /w wωω ω ω≡ Δ Δ

and the contribution of intersectoral shifts as /B Bωω ω ω≡ Δ Δ

Step 4: understanding the sources of changes in output per worker (net of intersectoral shifts) at the aggregate level and by sectors.

The terms ω and ωi, will capture changes in output per worker, but its interpretation is not so

straight forward. Increases in output per worker can come from three different sources: i)

increases in capital labor ratio ii) increases in Total Factor Productivty (TFP) and iii) relocation of

jobs from bad jobs sectors (low productivty) to good jobs sector (high productivity). To see the

first two points, note that under constant returns to scale, if Yt=Φtf(Et,Kt) where Kt is the capital

stock and Φt a technological, then output per worker Yt/Et= Φtf(1,Kt/Et). Therefore it will capture

changes in capital labor ratio and in TFP growth. Note that it may also capture cyclical behavior

of output: firms operating in economic downturns may have underutilized capital, when the

demand rises again; it will be reflected as rise in output per worker. The third point is simply the

result of worker moving from a low productivty sector (or firm) to a high productivty sector (or

firm), so that in the aggregate average output per worker will rise. From step 3 we found that it is

possible to isolate the effect of intersectoral shifts: wωΔ is just changes in output per worker net

of intersectoral shifts.

If data on capital stock is available then we can assume a particular functional form for the

production function and separate the contribution of higher capital labor ratios and the rest. For

example if we are willing to assume that the production function is Cobb-Douglas then: 1

Y K

E E

α−⎛ ⎞= Φ ⎜ ⎟⎝ ⎠

In competitive markets 1-α is the share of payments to capital in total Value Added. It is usually

available from national accounts data or if there are enough time series then it can be estimated

by taking logs and estimating:

ln ln (1 ) lnY K

tE E

α μ⎛ ⎞= Φ + − + +⎜ ⎟⎝ ⎠

Where t is an (optional) time trend capturing technological change and μ is a residual. Once we

have a value of α we can proceed to decompose changes in output per worker net of intersectoral

shifts, into changes in Total factor Productivity and changes in the capital labor ration.

Once we have an estimate of α, we can calculate total factor productivty as a residual: In the first

period it will be: (1 )

00 0

/ tt t

Y KTFP

E E

α−

== =

⎛ ⎞ ⎛ ⎞ =⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠

.

127

In the second period we need however to take into account that part of the change in output per

worker was due to relocation shifts so that: (1 )

11 1

/B tt t

Y KTFP

E E

α

ω−

== =

⎡ ⎤⎛ ⎞ ⎛ ⎞− Δ =⎜ ⎟ ⎜ ⎟⎢ ⎥⎝ ⎠ ⎝ ⎠⎣ ⎦

The term in square brackets I just output per worker in period two net of relocation effects. In this

way we are able to see whether changes in output per worker net of relocation effects, where due

to increases in capital per worker or in total factor productivity: 1 1

1 0 1 0 1( ) ( )2 2

t t t tw

TFP TFP k kk TFP

α ααω

− −− = = = =+ +Δ = Δ + Δ

Where k is simply the capital-labor ratio. The first tem in the right hand side is the contribution of

changes in the capital labor ratio to growth in output per worker net of relocation effects, and the

second term is the contribution of changes in TFP.

This means that changes in total output per worker can be expressed as the sum of changes in

TFP, changes in the capita labor ratio and intersectoral shifts: 1 1

1 0 1 0 1( ) ( )2 2

w

t t t tB

TFP TFP k kk TFP

α αα

ω

ω ω− −

− = = = =

Δ

+ +Δ = Δ + Δ + Δ144444444424444444443

As before let 1 0 1( )/

2t tTFP TFP

k kω α ω− = =+≡ Δ Δ denote the share of output per worker that can

be linked to changes in the capital labor ratio, 1 1

0 1( )/

2t tk k

TFP TFPα αω

ω− −

= =+≡ Δ Δ denote the

share of growth in output per worker that can be linked to TFP changes /B Bωω ω ω≡ Δ Δ denote

the share of changes in output per worker that can be attributed to intersectoral employment

shifts.

Step 5: Understanding the role of each sector on intersectoral shifts.

.3 It is possible to understand further how changes in the share of employment in the different sectors help explain the overall contribution of intersectoral shifts to per capita growth. An important literature has found that structural change, which is movements of labor force shares form low productivty sectors to high productivty sectors, is an important factor behind growth. Increases in the share of employment in sectors with above average productivty will increase overall productivty and contribute positively to the intersectoral shift term. On the contrary, movements out of sectors with above average productivty will have the opposite effect. By the same token, increases in the share employment in sectors with bellow average productivty should reduce growth, while reduction in their share should contribute positively to growth.

.4 Using the above intuition we can rewrite the intersectional shift as:

.5 , 0 , 1 0 1

1 2 2

Si t i t t t

B ii

sω ω ω ωω = = = =

=

+⎛ ⎞+Δ = Δ −⎜ ⎟⎝ ⎠

∑

The term in parenthesis is the difference between a sector i’s productivty (averaged between the

two periods) , 0 , 1

2i t i tω ω= =+

and the average (over the two periods) productivity of all the

128

economy (note there is no sectoral sub-index) 0 1

2t tω ω= =+

. Therefore, the contribution of sector i

to the intersectoral shifts term will be:

, 0 , 1 0 1

2 2i t i t t t

isω ω ω ω= = = =+⎛ ⎞+Δ −⎜ ⎟

⎝ ⎠

Thus if sector i has productivty bellow the average productivty, and increases its share si, its

contribution will be positive, that is outflows from this low productivty sector have contributed to

increase output per worker. If on the other hand, if the sector sees an increase in its share, these

inflows into this low productivity sector will decrease output per worker and thus have a negative

effect on the intersectoral shift term. The magnitude of the effect will be proportional to: i) the

difference I the sector’s productivity with respect to the average and ii) the magnitude of the

employment shift.

As before we can denote the share of intersectoral shift that is explained by sector i as:

, 0 , 1 0 1 /2 2

B i t i t t ti i Bs sω ω ω ω ω ω= = = =+⎛ ⎞+= Δ − Δ⎜ ⎟

⎝ ⎠

Step 6: putting everything together

Once the above steps are completed the percent contribution of each factor to total changes in

GDP per capita can be obtained as follows:

Contribution of Formula Comments

1. Demographic

shifts

1 1 0 0 1 0 0 1 /3 6

t t t t t t t te e e ea a y

ω ω ω ω= = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥⎣ ⎦

As in step 1

2. Contribution of

aggregate changes

in output per

worker

1 1 0 0 1 0 0 1 /3 6

t t t t t t t te a e a e a e ayω ω = = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥

⎣ ⎦

As in step 1

3. Contribution of

changes in the

employment rate

1 1 0 0 1 0 0 1 /3 6

t t t t t t t ta a a ae e y

ω ω ω ω= = = = = = = =+ +⎡ ⎤≡ Δ + Δ⎢ ⎥⎣ ⎦

As in step 1

4. Contribution of

increases in

sectoral

employment

[ ]*

/ *i

ei

i

e e e

e e e

=

= Δ Δ

Is calculated as the

contribution of

changes in

employment in sector

i to total employment

rate changes (step 2),

times the contribution

of employment rate

changes to changes in

129

total GDP per capita

(step 1)

5. Contribution of

changes sin output

per worker within

sectors

*

, 0 , 1

1

* / *2

w w

Si t i t

ii

s s

ωω ω ω

ω ω ω= =

=

=

⎡ ⎤⎛ + ⎞⎛ ⎞= Δ Δ⎢ ⎥⎜ ⎟⎜ ⎟

⎢ ⎥⎝ ⎠⎝ ⎠⎣ ⎦∑

It’s the contribution

of within changes in

output per worker to

total changes in

output per worker

(step 3) times the

contribution of

aggregate output per

worker to GDP per

capita (step 1)

6. Contribution of

intersectoral

employment shifts

*

, 0 , 1

1

* / *2

B B

Si t i t

ii

s

ωω ω ωω ω

ω ω= =

=

=

⎡ + ⎤⎛ ⎞= Δ Δ⎢ ⎥⎜ ⎟

⎝ ⎠⎣ ⎦∑

It’s the contribution

of between changes

in output per worker

to total changes in

output per worker

(step 3) times the

contribution of

aggregate output per

worker to GDP per

capita (step 1)

7. Within changes

in output per

worker in sector i

, 0 , 1

*

* / *2

i i

i t i ti

s s

ωω ω ω

ω ω ω= =

=

⎛ + ⎞⎛ ⎞= Δ Δ⎜ ⎟⎜ ⎟

⎝ ⎠⎝ ⎠

It is the contribution

of sector i, to within

changes to total

changes in output per

worker (step 3) times

the contribution of

output per worker to

changes in per capita

GDP (step1)

8. Contribution of

shifts in the share

of employment

witnessed by

sector i

, 0 , 1 0 1

*

/ *2 2

Bi i B

i t i t t ti B B

s s

s

ω ωω ω ω ω ω ω= = = =

=

⎡ + ⎤⎛ ⎞+= Δ − Δ⎢ ⎥⎜ ⎟⎝ ⎠⎣ ⎦


of sector i , to the

between component

of changes in output

per worker (step5)

times the contribution

of the between

130

employment shifts

component to total

GDP per capita

(calculated as above

in Numeral 6)

9. Contribution of

TFP (net of

intersectoral shifts)

1 10 1

*

( )/ *

2

w

t tw

TFP TFP

k kTFP

ω

α α

ω

ω ω− −

= =

=

⎡ ⎤+= Δ Δ⎢ ⎥⎣ ⎦


of TFP growth to


worker net of

intersectoral shifts

(step 4) times the

contribution of within


worker to total GDP

(calculate above

Numeral 5)

10. Contribution of

capital labor ratio

1 0 1

*

( )/ *

2

w

wt t

k k

TFP TFPk

ω

α

ω

ω ω− = =

=+⎡ ⎤= Δ Δ⎢ ⎥

⎣ ⎦


of changes in the

capital labor ratio to


worker net of

intersectoral shifts

(step 4) times the

contribution of within


worker to total GDP

(calculate above

Numeral 5)

131

Annex B: DECOMPOSITION OF LABOR INCOME GROWTH The labor income profile is best described at the household level. A simple and useful

characterization of households in terms of labor indicators can be obtained by noting that the

average labor income of household j can be written as (borrowing from Kakwani, Neri and Son,

2006):

j

j

j

j

j

j

j

j

j

Lj

j

Lj

N

A

A

L

L

E

E

H

H

I

N

I=

Equation 2

where LjI is the total labor income of household j, Hj is the total hours worked by working age

members of the household, j, Ej is the total number of employed in the household, Lj the number

of participants in the labor market, and Aj the number of working age members. In this way

ϖ =IL/H corresponds to average earnings per hour worked, h=H/E corresponds to average hours

worked, E/L is the employment rate, l=L/A is the participation rate, and a=A/N is the ratio of

working age members to total household members, or the dependency rate. For simplicity let the

above equation be rewritten as:

jjjjjLj aluh )1( −= ϖι

Equation 3

where (1-uj) corresponds to the employment rate of household j, which can be rewritten as 1 minus the household’s unemployment rate uj. Note that ϖ (omega bar) is different to ω (simple omega), which refers to output per worker in the previous sections.

In many contexts there is an important fraction of child laborers and elderly workers, and

calculating earnings per hour worked by the employed of working age is overestimating real

household “productivity.” In these cases, it might be better to abstract from the structure of the

household according to working age (Aj in Equation 2) and calculate dependency rates as the

number of participating individuals over the working total household members (Aj/Lj), and define

Ej as the number of working individuals irrespective of whether they are of working age or not;

and hours worked Hj, as total hours worked for all employed individuals irrespective of age.

By averaging each of the components of the household’s per capita labor income over sub-

groups of population we can obtain a full profile of labor market characteristics. For example, if

we divide households by quintile of income, it will describe the average labor market

characteristics of each quintile. Let Ω denote the subset of households belonging to a particular

quintile. It is possible to compare deciles by average dependency rates ∑Ω∈Ω j

jaN

1, average

participation rates ∑Ω∈Ω j

jlN

1, average hours worked ∑

Ω∈Ω jjh

N

1, incidence of unemployment

∑Ω∈Ω j

juN

1 and earnings per hour worked ∑

Ω∈Ω jjN

ϖ1.

Analyzing the sources of changes in labor incomes

132

A traditional way to understand how labor markets have affected welfare is to disentangle the

sources of labor income growth that are responsible for observed changes in total labor income.36

From Equation 3 the average per capita labor income of the subset Ω of households (whether

poor or non-poor households, or households falling within an income range or with particular

demographic characteristics), will then be:

⎟⎟⎠

⎞⎜⎜⎝

⎛++−++= ∑ ∑∑∑∑∑

Ω∈ Ω∈Ω∈Ω∈Ω∈ΩΩ∈Ω j jj

jj

jj

jjj

j

Lj aluh

NNlnln)1ln(lnln

1ln

1 ϖι

Equation 4

It is thus possible to decompose the change in the average per capita household labor income of

group Ω into changes in its different components: changes in average log earnings/per hour

worked, changes in average of log hours-worked, changes in average log unemployment rates,

etc. In particular:

∑ ∑∑∑∑∑Ω∈ Ω∈ΩΩ∈ΩΩ∈ΩΩ∈ΩΩΩ∈Ω

Δ+Δ+−Δ+Δ+Δ=Δj j

jj

jj

jj

jjj

Lj a

Nl

Nu

Nh

NNNln

1ln

1)1ln(

1ln

1ln

1ln

1 ϖι

Equation 5

In this way we can easily see whether growth in the average labor income of the poor (or any

group Ω) was due to changes in employment rates, participation rates, hours of work or earnings

per hour worked.

We can go a step further and decompose average earnings per hour into earnings per hour from

self employment (πj) and earnings per hour from waged employment (wj): jjjwj

Lj hwh πϖ π+= ,

with wjh corresponding to the share of waged employment in total hours worked and π

jh the

share of self employment. In this case, however, log-linearization of Equation 3 is no longer

possible and we would have to perform Shapley decompositions to analyze income changes.

Comparing in this way changes in average incomes of the poor and their components with

changes in average incomes of the non-poor can shed some light on what the channels are

through which a growth process is affecting the income of the poor. In many cases, however,

there might be considerable heterogeneity among employment sectors. In many cases it is useful

to perform these decomposition dissagregating households according to other characteristics: for

example, dividing households depending on their main occupation (e.g., differentiating between

rural farmers, rural non-farm workers, sector of occupation of household head, etc.).

36 See Kakwani, Neri and Son (2006) for an application of this decomposition to the analysis if pro-poor rates of growth.

13

3

Ann

ex C

: es

tim

atio

n re

sult

s (0

)

Tab

le A

C 1

: M

ean

and

Stan

dard

Dev

iati

on b

y E

mpl

oym

ent C

ateg

ory

Var

iabl

e W

aged

wor

kers

E

mpl

oyer

s Se

lf-e

mpl

oyed

H

H e

nter

pris

es

A

gric

ultu

re N

on-a

gric

ultu

reA

gric

ultu

reN

on-a

gric

ultu

reA

gric

ultu

reN

on-a

gric

ultu

reA

gric

ultu

reN

on-a

gric

ultu

re

Ear

ning

s 6.

073

13.2

4029

.293

48.9

8316

.367

15.7

878.

071

15.0

39

Age

31

.244

32

.716

42.3

2042

.627

36.3

6739

.438

44.7

0844

.896

Year

s of

Edu

catio

n 3.

121

8.40

73.

339

8.02

72.

654

5.37

51.

729

4.03

9

Gen

der

0.91

6 0.

537

0.94

90.

781

0.88

50.

393

0.93

00.

696

No.

of c

hild

ren

< 6

yea

rs

1.37

0 0.

878

0.97

60.

740

1.32

30.

904

1.31

90.

927

No.

of c

hild

ren

ages

7-1

5 1.

502

1.29

61.

506

1.11

81.

281

1.34

62.

276

1.55

6

Num

ber

of a

dults

3.

326

3.54

93.

509

3.29

92.

879

3.14

63.

589

3.32

6

Num

ber

of e

lder

ly

0.11

6 0.

200

0.12

70.

183

0.16

90.

133

0.07

20.

109

Non

-Lab

or I

ncom

e 3.

210

10.5

295.

857

15.0

763.

653

8.36

44.

481

7.59

7

Man

agua

0.

082

0.43

10.

000

0.31

20.

046

0.38

30.

007

0.23

4

Pac

ific

0.32

9 0.

303

0.27

00.

361

0.20

80.

348

0.22

70.

360

Cen

tral

0.

468

0.19

70.

580

0.24

70.

520

0.21

10.

509

0.28

8

Atla

ntic

0.

121

0.07

00.

150

0.08

00.

226

0.05

80.

257

0.11

7

Sou

rce:

Ow

n ca

lcul

atio

ns b

ased

on

EM

NV

.

13

4

Tab

le A

C 2

: Mea

n an

d St

anda

rd D

evia

tion

, by

Sect

or o

f Eco

nom

ic A

ctiv

ity

and

For

mal

ity

Lev

el

Var

iabl

e P

rim

ary

Seco

ndar

y T

erti

ary

Form

al

Info

rmal

Fo

rmal

In

form

al

Ear

ning

s 11

.492

15.8

9910

.776

21

.421

15.6

97

Age

37

.235

32.3

5034

.513

35

.486

37.0

11

Year

s of

edu

catio

n 2.

672

8.24

95.

631

10.6

496.

268

Gen

der

0.91

60.

609

0.71

0 0.

517

0.45

6

No.

of c

hild

ren

< 6

yea

rs1.

299

0.83

31.

025

0.76

00.

893

No.

of c

hild

ren

ages

7-1

51.

660

1.03

61.

452

1.17

41.

376

Num

ber

of a

dults

3.

318

3.41

93.

516

3.47

03.

358

Num

ber

of e

lder

ly

0.11

80.

128

0.17

6 0.

211

0.16

8

Non

-lab

or I

ncom

e 3.

969

10.8

706.

461

14.6

649.

107

Man

agua

0.

045

0.57

10.

274

0.48

50.

374

Pac

ific

0.26

80.

254

0.39

6 0.

275

0.32

9

Cen

tral

0.

504

0.15

00.

248

0.17

00.

224

Atla

ntic

0.

184

0.02

50.

082

0.07

00.

074

Sour

ce:

Ow

n ca

lcul

atio

ns b

ased

on

EM

NV

.

135

Table AC 3: Earnings Equations by Employment Category, 2001

Employment category

Variable Waged workers Employers Self-employed HH enterprises

Agriculture Non-agriculture Agriculture Non-agriculture Agriculture Non-agriculture Agriculture Non-agriculture

Age .0234*** .0261*** -0.0034 -0.0126 0.0043 -0.0015 -.0344*** 0.0065

Years of Education .0667** .1279*** .1374* 0.0324 .1144* .0731*** 0.0692 0.0221

Gender 0.053 .2296*** -1.319 -.8265** -0.3815 .4915*** -.8999*** .4435***

Pacific -.5601* -.1525* -1.383 -0.5319 -1.36* 0.0715 0.3933 -.6348**

Central -0.4818 -.1692* -1.057 -.6464** -0.4303 0.1019 0.3589 -.477*

Atlantic -0.0964 -0.0079 0 -0.0989 0.4632 .4823** 0.7036 -0.3808

-0.0317

0.2883

-1.996

-1.72**

0.3646

-.6512**

-.8981***

0.2035

Constant 0.8791 -0.1851 7.793 7.476*** 1.151 2.306*** 4.293** 1.381

No Observations 409 1939 159 234 283 873 356 245

R-squared 0.1471 0.2465 0.1379 0.3532 0.1685 0.0539 0.0965 0.0908

Note: * significant at 10%; ** significant at 5%; *** significant at 1%; lambdas standard errors are bootstrapped standard errors. Source:

Own calculations based on EMNV.

136

Table AC 4: Earnings Equations by Sector of Employment, 2001

Sector of economic activity

Variable Primary Secondary Tertiary

Formal Informal Formal Informal

Age 0.0045 0.014 .0354*** .027*** .0253***

Years of Education -0.0356 .1181*** .0649*** .177*** .0715***

Gender .8978* 0.2768 0.1972 .2212** 0.1668

Pacific 0.0546 -.3224* -.3854*** -.2501** -0.0094

Central 0.8212 -0.1309 -.405*** -.3088* -0.063

Atlantic 1.399** -0.5493 -0.1137 -0.0783 0.2034

1.109*

1.066*

-0.0587

0.7258

0.1745

Constant -1.017 -1.189 0.4249 -1.142 0.3396

No Observations 1210 219 569 693 1812

R-squared 0.0814 0.2304 0.159 0.2732 0.1097

* significant at 10%; ** significant at 5%; *** significant at 1%; lambdas

standard errors are bootstrapped se.

Note: *significant at 10%; ** significant at 5%; *** significant at 1%; lambdas standard errors are bootstrapped standard errors. Source: Own calculations based on EMNV.

…

137

4. MIGRATION, OPPORTUNITIES AND POVERTY REDUCTION IN NICARAGUA

By Catalina Herrera y Edmundo Murrugarra*

RECENT ECONOMIC TRENDS AND THE MIGRATION TO COSTA RICA

The recent slowed down growth observed in Nicaragua between 2001 and 2005 was preceded by natural disasters, and associated to reduced investments, lagging agricultural sector and adverse international prices. GDP growth in Nicaragua was about 4.4 percent during 1997-2001 when the country suffered the effects of Hurricane Mitch (1998) and a number of droughts in the following years. During the period 2001-2005, GDP growth was reduced to 3 percent mainly due to the poor performance in 2002 and 2003 (0.8 and 2.5 percent respectively). This slowdown in economic growth is clearly observed in the agricultural sector where one-fifth of Nicaragua’s GDP is produced and 40 percent of the labor force is employed. Agricultural growth rates were halved from 5 percent to less than 2.4 percent between 1997-2001 and 2001-2005. On the expenditure side, both internal and external factors are associated to this slowdown. First, there were steady declines in public and private investment since 1999-2000: total investment in 2003 was only 70 percent that of 1999.

On the external front, the coffee price crisis in 2002 stopped the annual 10 percent increases in export values since 1998, resulting in a small contraction of total export values of 3.5 percent. The impacts of the coffee price crisis were larger in Nicaragua than in other coffee producer countries in Central America, given the concentration of export agriculture on this grain and the large fraction of people involved in it. Costa Rica, another important coffee producer, had a more diversified agricultural production for exports including bananas, flowers and melons (Siegel et al, PRWPS- World Bank).

Central America, and Nicaragua in particular, is a region prone to natural disasters adding an important vulnerability component to the economic context. The Central America region has the second largest number of deaths per population associated to natural disasters after Africa, and Nicaragua is the country that shows the largest losses as percent of GDP due to natural disasters (Baez and Santos, 2006). The role of natural disasters is shaping migration flows and has been recognized before. In 1998, Costa Rica granted legal status to 152,000 immigrants after Hurricane Mitch hit Nicaragua (Bail, 2007).

* The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We wish to thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), Diego Ángel-Urdinola (Economist, LCSPP) and David McKenzie (Staff, Nicaraguan National Institute of Statistics and Census, INEC, now Nicaraguan National Institute of Information for Development, INIDE), for their valuable comments and suggestions. We also wish to thank the participants at the Poverty Workshop in Managua in March 2007 for helping us improve our understanding of the Nicaraguan socioeconomic context, and methodological and data issues. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent.

138

Source: Baez and Santos (2006)

During the recent years, migration and remittances had an increasing role in the Nicaraguan social and economic dynamics. Around 10 percent of the population is abroad -- mostly in the U.S. and Costa Rica -- and official remittances have increased 90 percent during the last ten years (Figure 4.1) reaching $600 US millions in 2005,37 representing 40 percent of total exports, 2.6 times of the foreign direct investment and 12 percent of GDP (IMF, 2006).

Figure 4.1. Remittances flows to Nicaragua 1995-2005

600

75

50

150

250

350

450

550

650

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005e

US

Mill

ions

Source: World Bank , DECPG

The main destinations for Nicaraguan migrants are Costa Rica and the U.S. but the composition is changing over time. The early outmigration was predominantly to the U.S., mostly due to the economic and political crisis during the eighties. About 60 percent of the total Nicaraguan migrants between 1979 and 1990 went to the U.S. according to information on household migrants in the 2005 Nicaraguan Household Survey.38 These earlier outflows are corroborated with evidence from the 2000 US Census.

37 Remittances are defined as workers' remittances, compensation of employees, and migrant transfers. 38 The information comes from the module 6 of LSMS 2005 where a migrant is defined as a former household member that lives in other country recently or many years ago. In the case that an entire household migrated, the LSMS does not measure these people as migrants, so the total number of migrants could differ from other sources like the Nicaraguan Census. This measurement issue is discussed for the Mexican case by McKenzie (2006).

139

According to this data, 51 percent of Nicaraguan migrants into the U.S. had arrived 25 years ago.39 After the Mitch hurricane in 1998 and the coffee price crisis in 2002, this emigration pattern changed, showing a drastic shift towards Costa Rica and away from the United States. In the nineties, 50 percent of Nicaraguan emigrants went to Costa Rica mainly for better opportunities of employment, mostly in the agricultural and household services sectors. Moreover, this trend has been sustained with the rapid increase of migration during the last years. Between 2001 and 2005, according with the LSMS data, the percentage of households with a migrant abroad augmented from 11 to 14 percent, with 60 percent of migrants going to Costa Rica compared to only 30 percent to the U.S, and the rest to other neighboring countries (Figure 4.2).40 The survey evidence is corroborated by the recent 2005 Census of Nicaragua. The Census data shows the recent acceleration of outflows since 48 percent of reported migrants left the country between 2000 and 2005, compared to only 22 percent between 1995 and 1999. It also shows the composition change between the U.S. and Costa Rica: only 30 percent of reported migrants went to Costa Rica between 1985 and 1995, but it was more than 55 percent between 1995 and 2005 (INEC, 2007).

Figure 4.2 – Nicaraguan Migrants to Costa Rica and USA

Graph 2. Nicaraguan Migrants to Costa Rica and USA

0

5000

10000

15000

20000

25000

30000

35000

1980

1982

1984

1986

198

8

1990

1992

199

4

1996

199

8

2000

2002

2004

Num

ber

of

mig

ran

ts

USA Costa Rica

Source: LSMS (2005)

Nicaraguan migrants to Costa Rica represent 6 percent of the Costa Rica population. According to the Costa Rican Census in 2000 there were more than 200,000 Nicaraguans in living in Costa Rica, but other estimates based on Costa Rican birth records and fertility patterns suggest about 283,000.41 While this shows the magnitude and relevance of the migration phenomenon for both countries, the impact could be still underestimated because the Census does not capture seasonal Nicaraguan migration associated to the peak harvest times in Costa Rica. Some estimates of this type of migrants indicate that this number could be around 100.000, but still is unknown.42 The number of Nicaraguan household members that were

39 This data was extracted from the 1% Sample of the 2000 U.S. Census. A migrant is defined as a person who was born abroad USA. Accordingly, there were 234,328 Nicaraguan migrants in USA in 2000, which represent 4.5% of the actual Nicaraguan population. 40 Guatemala, El Salvador, Honduras, Mexico, and Panama 41 Rosero-Bixby et al (200?). 42 There is no information in LSMS 2005 about the seasonal migration. However, some approximation to measure this type of migrants is to calculate the number of the household members that were absent less than nine months. According to that, this group of migrants represents 7% of the population; so far there is no more information in the LSLMS survey that could confirm that these people could be considered as seasonal migrants to Costa Rica.

140

absent from their households during the year is 60 percent higher than that of permanent migrants, suggesting that an important proportion of households is involved in seasonal migration.43 Corroborative evidence of seasonal migration is also found in a survey of more than 4,000 households in the North of Nicaragua where half of the households are involved in seasonal migration, mostly for only 3 months (Macours and Vakis, 2007).

Underreporting of migrants is still an issue since there is no official register of the undocumented Nicaraguan migrants into Costa Rica. However, it has been estimated that irregular migration could be of equal size to that of the migrant population captured by the Costa Rican Census in 2000 (CEPAL, 2006). This large number is also corroborated by the administrative data of Nicaraguans deported from the destination countries. Between 2001 and 2005 there were 69,111 migrants deported from Costa Rica which represented 95 percent of the total Nicaraguans deported back during this period.44

The substantial migration propensity to Costa Rica is also reflected in the internal migrant flows. According to the Nicaraguan Census 2005, 13 percent of the population lives in a department different from their birth place. The main attractors -- measured as those with the largest fraction of residents not born in the place of census -- are the departments of Rio San Juan, RAAS and Managua. Additionally, the most important outflows are from Matagalpa to RAAN, from Chontales to Rio San Juan, and from RAAS to Rio San Juan. This pattern has been almost stable compared with the one observed in the Census 1995 but this stability could simply hide faster rates of transition into Costa Rica rather than staying in Rio San Juan (See tables A1 and A2). In any case, Rio San Juan, despite its small size, is an attractor of internal inflows and could be associated to higher migration rates to Costa Rica, compared with the ones observed in Chontales and RAAS, given its location on the Costa Rica border.45

Migration selectivity and economic opportunities

The recent political and economic factors had a differential impact on potential migrants, thus creating important differences in the characteristics of migrants to both destinations, Costa Rica and the U.S. Migration destination choices are determined by the different net benefits associated to each destination, and by the differential ability of households (and individuals) to manage such financial and household constraints. Migration to the U.S. involves a larger financial cost of travel, additional cultural thresholds– such as language – and social networks to enhance employment opportunities among other factors. Migration to Costa Rica, on the other hand, is cheaper by proximity, language is less problematic and temporary or circular migration may reduce the burden of the integration process. In sum, migration choices could reflect different socioeconomic backgrounds that are related to uneven distribution of social and economic opportunities such as access to education, social networks, and infrastructure. This section examines how these differences in socioeconomic characteristics actually shaped migrants pools.

The combined effect of the individual, geographical and social network factors described above defined a migration flow to the U.S. that is distinctively different from that to Costa Rica. The recent shift of outflows to Costa Rica away from United States has consolidated even further a differentiated migration profile between these destination countries: Nicaraguans migrants to Costa Rica are poorer, less educated, younger and more likely to come from rural areas than the ones that go the USA (See table 4.1).

43 The survey, however, does not include information on the purpose of these absences or the destination of those short duration trips that could range between one and 11 months. In fact, members that are absent for more than 9 months are not included in the counting of the household membership. 44 Direccion General de Migracion y Extranjeria de Nicaragua. 45 The linkages between internal and external migration need to be further examined to better clarify the role of geographic migration paths compared to economic and social factors in determining migration destinations.

141

Education differences

Nicaraguan migrants’ average of years of education is higher than the national average (table 4.1) but those who migrated to Costa Rica have on average primary education while the ones moving to United States have in average three years in the secondary school. The lower educational levels of the Nicaraguan migrants in Costa Rica correspond to the reported activities they perform in the destination country. Migrants are concentrated in the northern provinces of Costa Rica – where agricultural work is the main activity – and in the capital, San Jose, where construction and housekeeping services are the core occupations. Evidence from Costa Rica shows also that Nicaraguan migrants receive the lower wages and occupational status in those sectors (Costa Rican Poverty Assessment 2005).

Table 4.1 Total

Migrants Costa Rica USA

Other countries

National

Actual age (mean) 32.5 29.4 36.9 31.3

Female Migrant (%) 48.6 51.0 44.1 52.8

Average years of education 7.7 6.4 9.4 7.9 5.9 b

Departure year 2,000 2,001 1,997 2,001

Poverty 2005 (%) aExtreme Poor 4.6 8.2 0.1 4.3 14.8Poor (including extreme) 22.2 34.3 5.3 25.3 46.1

Urban residence (%) 75.1 61.7 91.8 77.4 55.8

Sending remittances (%) 59.2 52.9 71.2 46.9

Region of Origin(%) Managua 26.7 13.7 44.8 22.7 25.0Pacifico 39.5 50.2 22.3 49.5 29.0Central 25.2 23.6 28.2 22.5 32.0Atlantico 8.7 12.5 4.7 5.4 14.0total 100.0 100.0 100.0 100.0 100.0

(a) The poverty levels are based on the migrant's household poverty levels in Nicaragua (b) This average was calculated over the individuals who are older than 15 years

Source: LMS, 2005. Notes: Average years of education at the national level correspnds to those of 15 or more years.

The Nicaraguan migrants in the United States are more educated but, according to the US Census of 2000, recent waves of migrants are also becoming less and less educated. Comparing the wave of migrants that arrived to the U.S. 20 years ago with the one that arrived recently (0-5 years), the percentage of migrants with none education increased from 4% to 10% within these groups, while the percentage of migrants with tertiary decreased from 40% to 24%.46

46 The analyses in this paper do not distinguish migrants by date of arrival to the US given the limited size of samples.

142

Figure 4.3. United States: Education level of Nicaraguan Immigrants

(% of migrants by level of education and date of arrival)

0

51015

202530

35

404550

0-5 years 6-10 years 11-15 years 16-20 years 21+ years T ota l

% o

f M

igra

nts

none pr ima_secun incomplsec completsec te rtiary

Source: US Census – IPUMS 1%

Geographic location of sending households

Nicaraguan outmigration is predominantly an urban phenomenon but very different between U.S. and Costa Rica. Three out of four Nicaraguan migrants come from an urban household, which is different from the characterization found in other Centro American countries where most of the migrants have rural residence (Cepal, 2006). However, almost 40 percent of the Nicaraguans that migrate to Costa Rica come from rural areas in comparison with only 8 percent that go to United States. This higher participation of rural households in migration to Costa Rica is close to the national rural population of 44 percent. If seasonal migration to Costa Rica associated with the peak harvest times were better captured in census and survey instruments, this percentage would be even higher. By geographical location, most of the migrants to United States are from Managua (45 percent) while the ones that go to Costa Rica come from the Pacific Region. The Pacific region that only accounts for 30 percent of the population, represents about half of the migration to Costa Rica (Table 1).

Socioeconomic distribution

Distinctive patterns of migration are observed across the income distribution, both in terms of propensity to migrate and choice of destination. The distribution of the total migrants by quintiles of consumption reflects a large selectivity process in migration. Graph 4 shows the proportion of all migrants in each income quintile. Since incomes may be affected by migration and remittances patterns, the distribution of migrants by household assets quintiles is also shown.47 Overall, migrants from households in better off quintiles (4 and five) outnumber those coming from poorer groups; in fact, two of three migrants come from the upper quintiles. The same pattern is found if the migrant population is stratified by households’ asset index.

47 Assets owned in a household reflect long term welfare status and, thus, are less likely to be affected by migrants’ departures. In the empirical analysis of conditional distributions, migration decisions are conditioned on households assets owned before the migrant’s departure. The asset index was constructed by PCA [spell out] method. It includes variables about the dwelling conditions (owning, material of roof, floor, access to water etc) and property of different housing assets (TV, radio, microwave, motorcycle, bicycle etc).

143

Figure 4.4. Nicaragua 2005: Distribution of Migrants by Quintiles

(percent of total migrants in each quintile)

0

10

20

30

40

50

1 2 3 4 5

% o

f M

igra

nts

Consumpt ion Asset index

Source: LSMS: 2005

The selectivity pattern is also observed in destination choices. More than a third of migrants to Costa Rica are coming from households under the poverty line, but this percentage decreases to only 5 percent for those that went to the U.S. (Figure 4.5)48 Moreover, of the total migrants from households in extreme poverty, for every migrant going to the U.S. there are more than ten going to other destinations including at least eight to Costa Rica. Migrants coming from non poor households are more evenly distributed between Costa Rica and U.S. (42% and 46%, respectively).

Figure 4.5. Nicaragua: Poverty and Migrant Destination

0

20

40

60

80

100

Poor ex treme or not NO Poor

% M

igra

nts

Cost a Rica USA Other

Source: LMS, 2005

The migration selectivity is corroborated when examining destination choices by income (or asset) quintiles. The poorer have a major incidence to migrate the South to South than to the North. The percentage of migrants to Costa Rica in each (income or asset) quintile is larger than the one that goes to the U.S. except for the highest quintile. Noticeably, the proportion of migrants to Costa Rica decreases as

48 About 8% of migrants to Costa Rica come from households under the extreme poverty line, while this number is negligible among migrants to the U.S.

144

household per-capita consumption (or asset ownership) increases. While more than 80 percent of migrants from the poorest quintile go to the South, only one fifth of those from the richest quintile go to Costa Rica (Figure 4.6).

Figure 4.6. Migrant Destination in each quantile ( Quantiles by Consumption)

0

20

40

60

80

100

1 2 3 4 5

% o

f M

igra

nts

Costa Rica USA

Source: LSMS, 2005

In sum, since migration is a household response to economic prospects (and uncertainties) the composition of migrants and their choice of destination seem to reflect differentiated opportunities to manage socioeconomic risks in Nicaragua. The next section examines these linkages in a regression framework to identify key drivers of the differentiated migration pools.

Determinants of Migration Destination

In this section, the differentiated patterns of migration destination are examined together with individual, household and community characteristics. The decision to migrate and the choice of destination are modeled jointly in a multinomial choice setting where individuals can choose between staying (not moving), migrating to Costa Rica, migrating to the U.S., and to other countries.49 The analyses used the information from the 2005 household survey. Modeling migration decisions requires some knowledge of households’ conditions at the time of departure, but since a number of migrants left since 1980, the analysis in this paper focuses on migration decisions during the 2002-2005 period. Notice this recent time span also corresponds to the increased flows to Costa Rica, where the poor are more likely to go. The analysis used information at the individual level for those individuals in Nicaragua and those leaving between the age range of 15 to 49. Information at the individual level included educational attainment, age, and gender. Household variables include demographic characteristics (size and composition), heads’ characteristics such as education, gender, and labor force participation, and the economic sector where most incomes are drawn from. Community characteristics include its urban/rural location, department, and the existing migration network in the community (measured as the relative importance of migrant to each destination in the corresponding Municipio). This section discusses the main findings of the analyses while the complete results of the multinomial choice model under different specifications can be found in Annex 1.

Households’ reliance on agriculture increases the odds of migration. The main economic activity of the household affects the destination choice. The probability of migrating to Costa Rica (compared to not moving) declines almost three times if the head household works in the financial sector and up to four

49 The exclusion of migration to other countries as a choice does not affect the results of the analysis.

145

times if the household is in urban areas. Working in almost any other sector rather than agriculture reduced the likely hoods of migrating to Costa Rica, consistent with the evidence from the Costa Rican census of 2000 that shows that most of the Nicaraguan migrants (32%) are employed in the agricultural sector.50 The lagging growth in agriculture during the last years discussed earlier, is a push factor to Costa Rican destinations.

Geographic proximity to Managua and to the Costa Rican border define migration patterns. The households’ residence in specific departamentos defines the destination of migrants. Individuals in households living in departments close to Managua like Jinotega, Madriz, Esteli -- are more likely migrate to the U.S. than those that live far away from the capital. On one hand, the probability of migrating to the South is increased for households that live in Rio San Juan or the Region Autonoma del Atlantico Sur (RAAS); one of the poorest and most affected by the coffee price crisis (Ref). On the other hand, the propensity of migrating to the USA is reduced drastically for individuals in the Pacific Region, particularly in departments with higher historical migration rates to Costa Rica like Carazo, Chontales, Rio San Juan and Rivas.

Access to infrastructure and services creates differentiation within departamentos. Besides the role of departamentos in shaping the migration patterns to both U.S. and Costa Rica, there are important differences within departamentos. These differences are determined by households’ access to roads and other social infrastructure since these distinguishing those that have access to migration routes. Households with lower access to social infrastructure, measured by the distance to the nearest health and education services, have lower propensities to migrate to Costa Rica.51 This result suggests that while households in, say, Rio San Juan or RAAS are more likely to migrate to Costa Rica (compared to not moving or going to the U.S.) only those households close to social infrastructure and routes will be able to do so. This is important to emphasize since the data shows that migration – even to Costa Rica - may not be an option for isolated or households without other financial or social assets.

Individual and households’ human capital distinguishes migrants by destination. The levels of education seem to affect the type of migration choice, reflecting the labor markets in the destination countries. An individual with primary schooling increases his or her chances to by 50% to migrate to Costa Rica, comparing with not moving. Notice that achieving primary education already positions adult individuals in the Nicaraguan average of 6 years of education. These separating effects of education are confirmed in other studies (Vargas and Barquero, 2005) based on a 2002 survey in six Nicaragua communities under the Latin American Migration Project (LAMP). LAMP data confirms that increasing human capital reduces migration to Costa Rica compared to the U.S.

The education results have important differences across genders, reflecting labor demand in countries of destination. Among females, attaining primary schooling augments their chances of migrating to the South by three-fold but for males this variable is not relevant. This evidence may also match the patterns of labor market insertion in Costa Rica, where females are predominantly absorbed in household services where some basic education may be required, while males are absorbed into agriculture and, sometimes, construction.

Besides the role of each individual’s education in migration choices, the household’s level of education is also important. Individuals in households where the head has a technical or tertiary education have migration odds to Costa Rica that are lower 77% and 65%, respectively. Unsurprisingly, these migration-

50 CEPAL (2005) 51 This distance was measured by a compound index of the distance from the household to the closest school and medical center. The index was constructed by principal component method where the variables were the distance in meters and time from the household to the closest medical center and school

146

reducing effects are larger in the urban areas where the odds of migrating to the South are lower by 78% and 88% for the corresponding education levels. Moreover, male´ s probability of migrating to Costa Rica is reduced twice if the head household has a technical education.

Social networks play an important role in flows to Costa Rica. The existing social networks of migrants in Costa Rica are playing an important role in the recent migration waves. Municipal migration networks for each destination are measured by the ratio of migrants to each destination that left Nicaragua before 2001 and the population at the municipal level. These migration network variables only play a role in increasing migration for those choices to the Costa Rica, not to the U.S. This shows both the different type of migration and labor markets insertion between the two destinations. In the North, education and direct family connections are critical for insertion in the economic life, while in the South, weaker social capital linkages are exploited to migrate and obtain a job. These weaker linkages utilize community or municipal contacts, very much beyond family connections, to assess the prospects of jobs and, in fact, get the contact with future employers in Costa Rica. Employers, in turn, rely on these broader networks to access new labor given the recommendations of current workers (Borge, 2005). The differentiated effect of social networks – that seem more important for the less educated -- is similar to the finding by MacKenzie and Rapoport (2006) in the Mexican case where the role of networks decreases with education level of the community. Overall, a one percentage point increase in the historical migration rate to Costa Rica represents an 12 times increase in the probability of migrating to Costa Rica.

The results of this exercise show that the migration choices during 2002 and 2005, when migration to Costa Rica was accelerated, were shaped by differences in households assets and in their ability to exploit those opportunities. Human and capital assets played a role in selecting individuals into migration and specific destinations, while geographic location and access to social infrastructure also constrained those choices.

Effects of Migration and Remittances in Poverty reduction

Almost a third of the Nicaraguan population is affected by migration or remittances flows. About 14 percent of households reported having a migrant abroad52 and 21.5 percent receives external remittances. In many cases, households receiving remittances do not report a household migrant abroad since remittances may flow from extended family members or friends, as found in other countries. Overall 31.2 percent of households either has a migrant or receives remittances from abroad, according with the LSMS.

The incidence of remittances reflects the migration patterns described before, where the upper quintiles have a higher propensity to receive remittances than the worse off. As shown in the graph 7, in the fifth quintile between 30% and 40% of households receive remittances, but this percentage decreases to 14% and 8% in the poorest quintile.53

52 This number is 10% according with the Census 2005 and the difference is attributed to the different ways of measuring migration. 53 The overall propensity to receive remittances across quintiles is maintained whether per-capita consumption or an asset index is used.

147

Figure 4.7 – Nicaragua: Proportion of households receiving remittances

05

10

152025

30

3540

45

1 2 3 4 5Quintiles

% o

f H

ouse

hold

s

Consumption Asse t Index

Source: LSMS, 2005

The actual distribution of remittances across quintiles accentuates this pattern given the higher average remittances among the better off. Moreover, of the total volume of remittances reported in the LSMS, 68 percent is perceived by the forth and fifth quintiles.54 This reflects the annual average remittance received per household in each quintile. While in the lowest quintile a household receives on average $750 USD per year, in the richest quintile this amount is almost duplicated (Figure 4.8). Remittances, however, play a more relevant role for the poorest households. They represent 15 percent of the households’ consumption in the lowest quintile and just 8 percent among the better off (Figure 4.9). In the case of seasonal migration, the average incomes brought back home are around US$ 200 for the selected communities in the North, and represents close to 19 percent of the total incomes (Macours and Vakis, 2006).

Figure 4.8. Nicaragua: Average Remittances by Quintiles

0

300

600

900

1 200

1 500

1 800

1 2 3 4 5Quin tiles by Consumption

Am

ount

rem

ittan

ces

(US

D)

Source: LSMS 2005

54 Expanding remittances data from the Nicaraguan LSMS provides an estimate of US$ 233 millions, still below the levels shown in Graph 1.

148

Figure 4.9. Nicaragua: Remittances as a share of household Consumption

Nicaragua: Remittances as a share of Household Consumption

0.00

0.04

0.08

0.12

0.16

1 2 3 4 5

Quintiles by consumption

Rem

ittan

ces

% C

onsu

mpt

ion

Source: LSMS 2005

Without remittances, the poverty rate in Nicaragua would have been 4 percentage points higher. Under the assumption that remittances are mostly consumed, the national poverty headcount rate would be 4 percentage points higher if households had not had remittances (Table 4.4), an effect clearly observed in urban areas (where most migrants are coming from). The effects are very important around the extreme poverty line, since the extreme poverty rate would have been more than 19 percent (from 14 percent) if remittances were not received in nicaragua. This reflects that while remittances to the poor are smaller, they still represent an important share of their budget.

Table 4.4 – Poverty Headcounts with and withourt remittances, 2005

Total consumption

(a)

Consumption without

remmittances (b)

Difference (a-b)

St Error

Poverty rate National 46.0% 49.9% -3.9% 1.4%Urban 28.9% 34.3% -5.4% 1.8%Rural 67.7% 69.6% -1.9% 1.5%

Extreme Poverty National 14.8% 19.3% -4.5% 0.8%Urban 5.4% 10.9% -5.5% 0.8%Rural 26.6% 29.8% -3.2% 1.4%Source: LSMS 2005, World Bank estimates

These poverty reduction effects were also found controlling for other household characteristics. Modeling either consumption or poverty indicators shows similar results in terms of migration and the associated poverty reducing effects.55 A probit model was implemented to show the effect of having a

55 Decisions on migration, remittances, labor supply, expenditure allocation, school attendance, child labor and so on are usually made simultaneously. Hence characteristics which explain migration and remittances may also shape household expenditures patterns. Moreover many of the characteristics which influence these decisions are

149

migrant abroad in the household’s propensity of being poor. Controlling for demographic, socioeconomic and geographic characteristics of the household, this marginal effect could potentially reduce the probability of being poor by 7.5 percentage points (table 4.5). This result was also corroborated with models of per-capita consumption. Controlling by the same group of variables, having a migrant abroad could increase in 13 percent the per-capita consumption at the household level. Examining the differential effects of migration across the consumption distribution shows that, for households with identical observed characteristics, those with higher level of consumption will benefit from larger gains due to migration (Annex 3). This corroborates that very poor households have little gains from migration, both because of their lower probability to migrate (as discussed in section 3) but also because of the lower remittances, if they remit at all.

The poverty reduction effects are around 5 percentage points if the number of migrants is taken into account. About one third of Nicaraguan households with migrants have more than one migrant abroad (INEC, 2006). The analysis shows that the number of migrants is associated to 8 percent increases in consumption or 5 percent reduction in poverty.56 A detailed analysis distinguishing the number of migrants in each household shows that the associated gains in consumption are revealed for urban households with 2 or more migrants, while among rural households additional migrants beyond the first do contribute only marginally.

Table 4.5 Marginal Impact of Migration g p g

Consumption Poverty Reduction (percent change) (percentage points)

Household with migrantAgregate 12.9%*** 7.5%***Urban 11%*** 4.8%Rural 13.0% 8.9%***

Number of migrants Agregate 8.4%*** 5%***Urban 9%*** 3%**Rural 5%*** 4%***, **and ***indicate significance level at the 10%, 5% and 1%

But are these associated gains in consumption linked to migration destinations? To address this question, the analysis distinguished the migrants (and their number) by country of destination. As expected, the marginal effect of migrating to the U.S. is significant higher and positive than the gains from migrating to Costa Rica (Table 4.6). Migration to the US is associated with consumption gains between 20 and 30 percent that are reflected in reductions of poverty between 12 and 19 percent. These associated gains for migration to Costa Rica are smaller, but still show significant effects in rural areas where households with migrants have an associated consumption gain of 9 percent, and a resulting reduction in poverty of 12 percentage points.

unobservable. These issues make it difficult to establish causality and bias the typical reduced form regression framework. See Sasin and MacKenzie (2006). 56 The equation of consumption was also estimated with dummy variables for cero, one, two, three or more migrants at the household level.

150

Table 4.6. Marginal Effects of Migration on Consumption and Poverty Table 6. Marginal Effects of Migration on Consumption and Poverty

National Urban Rural Consumption model (% gains)Households has migrant USA 34%*** 28.3%*** 50%*** Costa Rica 0.7% -4.7% 9.1%**

Number of Migrants USA 20.9%*** 18.9%*** 22.2%*** Costa Rica 0.4% -1.8% 3.1%

Poverty model (point reduction)Households has migrant USA 18,8%** 12.3%** 29.51*** Costa Rica 0.8% -5.6% 12.3%*

Number of Migrants USA 16.4%*** 13.7%*** 12.4%** Costa Rica 1.1% -1.8% 3,1%

The gains from migration may come with a direct cost on the households wellbeing. Migration processes may affect households’ allocation of labor resources with different potential outcomes. Migration of adult males, for example, could represent an increased need for labor from other members (females), especially if households are involved in self-employment or other household-based productive activities.57 This could represent a decline in schooling amongst the elder children, or less time of other members spent with the children. The additional income can, on the other hand, offset some of these negative effects. Arends-Kuening and Duryea (2006) examined the effects of parental presence on adolescents schooling and work in several countries and found that in Nicaragua, adolescents between 14 and 16 years living in a single parent household would see their enrolment decline from 67 to 55 percent, accounting for other income losses. Cox-Edwards and Ureta (2003), on the other hand, found that migration and their associated remittances had a positive effect on school retention. In a survey of Nicaraguan villages in the North, Macours and Vakis (2007) found that children from seasonal migrant mothers seem to fare slightly worse development indicators (cognitive and health status) compared to those from non migrants. These poorer outcomes, however, are attributed to other social and economic characteristics of the household and communities, and not to migration itself. In fact, once accounting for other factors, children from seasonal migrants do show better cognitive achievement outcomes underscoring the important role of extended family networks in providing care for young children and that of empowered women due to migration.

Summary and Policy Issues

Migration and remittances have an increasing role in social and economic dynamics of Nicaragua. Almost one third of Nicaraguans have either a migrant or receive remittances from abroad, and remittances are reaching 12 percent of the GDP. After the Mitch Hurricane, an important increasing migration outflow to Costa Rica has defined a bipolar migration profile. Nicaraguan migrants to Costa Rica are poorer, less educated, younger and more likely to come from rural areas than the ones that go to USA. The scarce opportunities in rural areas, living in departments neighboring Costa Rica, and access to roads and other social infrastructure are facilitating factors in the migration flow to Costa Rica, the country attracting 60 percent of the emigrants between 2001 and 2005. Still, the real magnitude of the migration phenomenon in Nicaragua may be larger if seasonal and irregular migration were well accounted for.

57 Or more generally, when local labor markets are absent or missfunctioning.

151

The less skilled migration to Costa Rica reflects the labor demand in the South. The migration to Costa Rica is primarily of working age individuals who are inserted into agriculture, construction, and household services, sectors do require little human capital from migrants. This way, attaining basic levels of education (primary) represents an important factor in choosing Costa Rica as destination. While this paper has examined the role of permanent migrants, the seasonal migration to Costa Rica and other neighboring countries is as important and needs to be studied in detail. The fast increase in the migration to the South has been supported by extended social networks that reduce the transaction costs of migration both in Costa Rica (information, housing, job search) and in Nicaragua (child care).

The associated gains from migration and remittances are expected to between 4 to 5 percentage points, but more important for the extreme poor. The associated gains through remittances could have reduced extreme poverty from more than 19 percent to less than 15 percent. But the gains also vary depending on the migration destination. Most of the benefits are associated to migration to the U.S. while the benefits from migrating to Costa Rica seem more modest and concentrated in rural areas. Still, those smaller benefits play an important role in those sending households. The income gains may come with a cost that needs to be examined in detail. The combined effects of increased incomes, changes in households’ labor supply, and child care availability are still to be accounted for. Preliminary evidence suggest that while seasonal migration is having a positive impact on cognitive development, average development outcomes of children from migrants are still behind those from non migrants.

How can Nicaragua gain the most out of migration and remittances? The observed migration patterns, especially that to Costa Rica, reflects the existing development challenges in Nicaragua where the rural population increasingly finds less attractive to stay and are willing to send one or more migrants abroad. The more educated, urban population keeps migrating out at a sustained pace. This dual pattern of migration suggests a two-pronged strategy in incorporating migration into the national development strategy.

The historical migration to the U.S., where the core of remittances come from, require policies to enhance the impact of those remittances on poverty reduction outcomes.

• Expanding financial coverage among the poor. Studies in Latin America and other regions showed that remittances can have the largest impact when financial sectors are well developed. This not only implies cheaper and easier transfers between countries, but the development of the financial sector in Nicaragua to deliver financial services to the poor (financial inclusion). Successful interventions in other countries like Ecuador have exploited the network of microfinance institutions to offer financial products that can have poverty reduction effects such as pre-paid medical care, or micro-credit, among others. These initiatives can leverage on the remittances flows to establish new uncovered markets. The existience and widespread use of microcredit institutions in Nicaragua provides the opportunity to develop other markets based on remittances flows.

• Return migration of the highly-skilled diaspora. Since migrants to the U.S. have a higher education compared to other groups, return or circular migration schemes could be encouraged to exploit the potential spillovers from the skilled diaspora in the U.S. Experiences in other countries like India, China or even Ireland show that diasporas, especially skilled ones, can play a key role in developing a critical mass of skilled labor that can complement investments of higher value added. The Intel engagement in Costa Rica for the last 10 years and the recent opening of the Dell call center in El Salvador underscores the need for human capital strategies where diasporas can play an important role in attracting investments.

152

The migration to Costa Rica case defines a different agenda given its younger, less educated, and, usually, irregular status.

• Bilateral agreements. Nicaraguan migration to Costa Rica is as important as the population in Rivas and Rio San Juan, and represent more than 6 percent of the population in Costa Rica. The recent increase in outflows provides an opportunity to establish a dialogue on bilateral agreements that provide temporary migrants a safe and well protected migration experience and ensures the return of migrants (Borge, 2006). These bilateral agreements are necessary to ensure the consistency of domestic labor market policies between two countries and identify the role of migration as a necessary adjustment mechanism.58 Spain is one of the countries with an active bilateral agreement policy with sending countries like Morocco, Colombia, Ecuador, or Romania, where the objective is to manage the market for immigrant labor, and ensure basic labor conditions. Better migration conditions are expected to provide higher benefits from the migration experience.

• Coordination of domestic social policies and migration processes. Well designed programs to benefit the poor may see their objective offset by household responses to policies. For example, it has been well documented that public transfers could displace between 30 and 40 percent of private transfers. If that is the case and remittances play an important role in the incomes of the poor, social transfers need to be designed to minimize these displacement effects and seek complementarities between public and private resources. Similarly, the design of early childhood education interventions needs to account for the absence of one parent and the heavy reliance on social networks for childrens’ care. The challenge in delicate coordination is in creating complementarities without generating additional incentives for migration.

• Exploit mechanisms of regional integration to treat migration issues. Experiences from other regions, such as the European Union, shows that regional integration processes have been the opportunity to raise and treat migration issues in a coherent fashion. Moreover, other policies – such as regional development investments in the EU – have played a role in keeping migration low even in the absence of mobility regulations. The lack of synchronization of monetary or exchange rate policies has been mentioned as an impediment for a monetary union In Central America. The efforts to raise migration as a common task like the Plan de Integracion Migratoria Centroamericana require further coordination and alignment of other policies. Otherwise, regional labor markets may rapidly adjust responding to misalignments in other markets (e.g. exchange rate).

58 In fact, consistency in other policy areas such as trade, monetary or exchange rates issues are equally important to ensure that migration incentives are not created

153

REFERENCES

Adams, R. (2004) “Remittances and Poverty in Guatemala”. DECRG, World Bank.

Baez, J. and I. Santos (2006) “Children’s Vulnerability to Shocks: Hurrican Mitch in Nicaragua as a Natural Experiment,” mimeo.

Bail, R. (2007) “Nicaragua exports its poor,” in Le Monde Diplomatique, January, 2007.

Banerjee, A., E. Duflo (2007) The Economic Lives of the Poor In The Journal of Economic Perspectives, Volume 21, Number 1, Winter 2007, pp. 141-167(27)

Baumistier, E. (2006) “Migracion Internacional y Desarrollo en Nicaragua”. Serie Población y Desarrollo no 67. CEPAL

Borge, D. (2005) “La fuerza de los vínculos débiles en la inserción laboral de los migrantes nicaragüenses,” in Población y Salud en Mesoamérica, Vol 3(1), Julio–diciembre. Centro Centroamericano de Población

Borge, D. (2006) “Migración y Políticas Públicas: elementos a considerar par ala administración de la migraciones entre Nicaragua y Costa Rica,” in Población y Salud en Mesoamérica, Vol. 3(2), Enero.

Gonzalez. M and Lizano. E (2006) “Bancarizacion de las remesas de inmigrantes nicarguenses en Costa Rica”. Series de trabajos ocasionales del FOMIN -BID

Filmer D. and Pritchett (1998) “Estimating Wealth Effects without Expenditure Data –or Tears : An application to educational Enrollments in States of India” World Bank.

International Monetary Fund (2006) Nicaragua Country Report No. 06/174.

Macours, K. and R. Vakis (2007) “Seasonal Migration and Early Childhood Development” World Bank.

Mackenzie, D. and Rapoport, H. (2006) “Self Selection patterns in Mexico-US Migration: The role of Migrant networks”. DECRG, World Bank.

Marquette,C (2006) “Nicaraguan Migrants and Poverty in Costa Rica”. Centro Centroamericano de Poblacion. Universidad de Costa Rica.

Sasin, M and Mackenzie. D (2007) “ Migration, Poverty and Human Capital” Migration Operational Vehicle. Operational Note 1 in www.worldbank/migration.

The Economist Intelligence Unit (2006) Country Profile- Nicaragua 2006

Vargas, J. and J. Barquero (2005) “Capital Humano y Social de los Nicaragüense con experiencia migratoria a Costa Rica y Estados Unidos,” in Estudios Migratorios Latinoamericanos, Vol. 19(56).

World Bank (2003). “Nicaraguan Poverty Assessment: Raising welfare and reducing vulnerability”. Central American Department- Latin American Region.

154

5. PROGRESS AND PROSPECTS FOR MDGS AND PRSP GOALS IN NICARAGUA

Salvador Leopoldo López G.*

“We will spare no effort to free our fellow men, women, and children from the abject and dehumanizing conditions of extreme poverty, to which more than a billion of them are currently subjected.”

United Nations Millennium Declaration (September, 2000).

Introduction

The Millennium Development Goals

The Millennium Development Goals (MDGs) grew out of the agreements and resolution of world conferences organized by the United Nations in the past decades. Brought together as a set of “International Development Goals” in 1996, they have been widely accepted as a framework for measuring development progress. The goals focus the efforts of the world community on achieving significant, measurable improvements in people’s lives. They establish yardsticks for measuring results, not just for the developing countries but for the rich countries that help to fund development programs and for the multilateral institutions that help countries implement them.

The MDGs were endorsed, in September 2000, by all 189 United Nations states. The means to achieve them were addressed at the “United Nations Conference on Financing for Development” held in Monterrey, Mexico, in March 2002.59

The Millennium Development Goals (MDGs) are comprised by a set of eight goals, eighteen targets, and forty eight indicators. The first seven goals include: erradicate extreme poverty and hunger, achieve universal primary education, promote gender equality and empower women, reduce child mortality, improve maternal health, combat HIV/AIDS, malaria and other diseases, and ensure environmental sustainability. The last goal - develop a global partnership for development – is about the means to achieve the first seven. They include eighteen targets that should be achieved by 2015. (Box 5.1)

Each of the goals is important by itself, however, they should be viewed together because they are mutually reinforcing. Better health care increases school enrollment and reduces poverty. Better education leads to better health. And increasing income gives people more resources to pursue better education and health care and a cleaner environment.

* The author is a Consultant for the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. I with to thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP) and Jose Ramon Laguna (Head Research Assistant for the Poverty Assessment) for their valuable comments and suggestions. The views expressed here are those of the author and need not reflect those of the World Bank, its Executive Directors, or the countries they represent. 59 Picciotto, Robert. World Bank Operations Evaluation Department. “Development Cooperation and Performance Evaluation: The Monterrey Challenge” June 2002.

155

Box 5.1. The Millenium Development Goals and targets

Goals Targets

Target 1. Halve, between 1990 and 2015, the proportion of people whose income is less than $1/ day.

Goal 1. Erradicate extreme poverty and hunger

Target 2. Halve, between 1990 and 2015, the proportion of people who suffer from hunger

Goal 2. Achieve universal primary education.

Target 3. Ensure that, by 2015, children everywhere, boys and girls alike, will be able to complete a full course of primary schooling.

Goal 3. Promote gender equality and empower women

Target 4. Eliminate gender disparity in primary and secondary education, preferably by 2015, and to all levels of education no later than 2015.

Goal 4. Reduce child mortality. Target 5. Reduce by two-thirds, between 1990 and 2015, the under-five mortality rate.

Goal 5. Improve maternal health. Target 6. Reduce by three-quarters, between 1990 and 2015, the maternal mortality ratio.

Target 7. Have halted by 2015 and begun to reverse the spread of HIV/AIDS

Goal 6. Combat HIV/AIDS, malaria and other diseases

Target 8. Have halted by 2015 and begun to reverse the incidence of malaria and other major diseases.

Target 9. Integrate the principles of sustainable development into country policies and programs and reverse the losses of environmental resources.

Target 10. Halve by 2015 the proportion of people without sustainable access to safe drinking water.

Goal 7. Ensure environmental sustainability

Target 11. By 2020 to have achieved a significant improvement in the lives of at least 1000 million slum dwellers.

Target 12. Develop further an open, rural-based, predictable, non-discriminatory trading and financial system.

Target 13. Address the special needs of the least-develop countries

Target 14. Address the special needs of landlocked countries and small island developing states.

Target 15. Deal comprehensively with the debt problems of developing countries through national and international measures in order to make debt sustainable in the long term.

Target 16. In cooperation with developing countries, develop and implement strategies for decent and productive work for youth.

Target 17. In cooperation with pharmaceutical companies, provide access to affordable essential drugs in developing countries.

Goal 8. Develop a Global Partnership for Development

Target 18. In cooperation with the private sector, make available the benefits of new technologies, especially information and communications.

Source: Picciotto, Robert. World Bank Operations Evaluation Department. “Development Cooperation and Performance Evaluation: The Monterrey Challenge” June 2002.

156

Monitoring the PRS and MDGs indicators

The process of monitoring the PRS (Poverty Reduction Strategy) have several advantages: i. it allows to track progress in achieving the poverty reduction goals, ii. It also help to validate the choices made in the first place, justifying government action to the public, iii. Reveal reasons for success or failure, allowing effective management of the strategy and improvements to be made; iv. mobilizes and sustains public support for the target; v. offers the opportunity for greater involvement of Civil Society in the process; vi. Acts as a means of accountability in the use of resources, generating transparency.60

Some General Concepts

Goals the objectives a country or a society want to achieve

Indicators the variables used to measure progress towards the goals

Targets the quantified level of the indicators that a country want to achieve in a given time frame

In general, indicators can be classified in two groups: intermediate indicators and final indicators. The first group can be sub-classified in input indicators and output indicators, and the latter in outcome and impact indicators.

Types of Indicators

Input Indicators Financial and Physical Indicators of Resources Used

Intermediate Indicators

Output indicators The intermediate goods and Services generated

Outcome indicators Access and Use of goods and Services and Satisfaction of Beneficiaries

Final Indicators

Impact Indicators Effect on Key Dimensions of Well-being (improvement in Living Standards)

The Nicaraguan PRSP. Long term goals and intermediate indicators.

In September 2001, the Nicaraguan Government presented its Poverty Reduction Strategy Paper (PRSP) to the Boards of the World Bank and IMF. The new government that took office in January 2002 has confirmed its commitment to the broad principles and priorities expressed in that PRSP.

The 2001 Nicaraguan PRSP rests on four pillars: i. broad-based growth with an emphasis on productive employment and rural development; ii. greater and better investment in human capital of the poor; iii. better protection of vulnerable population; and iv. strengthening of institution and good governance. It also includes three cross-cutting themes: i. Reducing environmental degradation and ecological vulnerability; ii. Increasing social equity; and iii. Promoting decentralization.61

In December 2005, the Nicaraguan government presented a new version of the PRSP, called the Nicaraguan Development Plan (NDP). The new version of the Nicaraguan PRSP includes goals, indicators and targets in the following areas: Poverty, Macroeconomic Performance, Economic Infrastructure, Regulatory Framework and Paperwork procedures, Property Rights, Access to financial

60 Pain, Chris, An introduction to the Challenges and Issues in the PRSP. June 2002. 61 World Bank. Country Assistance Strategy of the World Bank Group for the Republic of Nicaragua, November 2002

157

Services, Investment Attraction, Food Security, Sustainable Environmental Development, Education, Health, Social Protection, and Water and Sanitation.

The goals and targets that included in the Nicaraguan PRSP and that are part of the MDGs are the following:

Box 5.2. The Nicaraguan PRSP long term goals and targets

Goals Targets

Goal 1. Reduction of poverty Target 1. Halve, between 1995 and 2015, the proportion of people whose income is less than the extreme poverty line62. Extreme poverty would be 9.7 by 2015

Goal 2. Increase access to primary education.

Target 2. Ensure that, by 2015, all boys and girls alike, will be able to complete a full course of primary schooling.63 Net primary enrollment would be 100 % by 2015

Goal 3. Reduce infant and under-five mortality.

Target 3. Reduce by two-thirds, between 1994 and 2015, the under-five mortality rate and infant mortality. Under-five mortality would be 24 and infant mortality 20 by 2015.

Goal 4. Reduce maternal mortality rate.

Target 4. Reduce by three-quarters, between 1994 and 2015, the maternal mortality ratio. Maternal mortality target by 2015 is 22.

Goal 5. Reduce chronic malnutrition Target 7. Reduce chronic malnutrition to 7 % by 2015

Target 8. Increase to 90 % national water coverage by 2015 Goal 6. Increase access to water and sanitation

Target 9. Increase to 95 % national access to sanitation by 2015

Goal 7. Reduce Illiteracy Rate Target 10. Decrease illiteracy rate64 to 10 % by 2015 Source: IMF, Country Report No. 05/440, Nicaragua: Poverty Reduction Strategy Paper , December 2005

Achievement of 2005 PRSP Targets

“As long as you travel to a goal, you can hold on to a dream” Anthony de Mello65

From the ten 2005 PRSP targets (Table 5.1), four of them were achieved including: extreme poverty and net primary enrolment. Infant mortality and under-five mortality were on-track in 2001; there is no new data available for two indicators; their source data is the Demographic and Health Survey (DHS), which will be published at the end of 2007. PRSP targets that were not accomplished are: maternal mortality, access to reproductive health services, chronic malnutrition, access to drinking water and sanitation, and illiteracy.

62 The general poverty line in 1998 is US$403/year ($1.1 /day). The extreme poverty line is US$212/year ( 0.58 $/day), a peson consuming less that this amount per capita is considered extremely poor and she/he cannot meet her/his minimum daily caloric requirement even when the entire consumption is devoted to food. The calculation indicates that extreme poverty goal for year 2015 is 9.7 % 63 This target was increased from 90 % (2001 PRSP) to 100 % (2005 PRSP ) 64 Illiteracy rate is measure among persons with ages 10 years and over. 65 Vandemoortele, Jan. Are the MDGs feasible? UNDP, Bureau for Development Policy, June 2002

158

Table 5.1. Nicaragua: Progress toward meeting PRSP Goals and MDGs Data PRS-II Target

Source 1993 1998 2001 2005 Target 2005 2015Extreme Poverty (%) LSMS 19.4 17.3 15.1 14.9 16 Yes 10Net Primary Enrollment (%) LSMS 75.6 79.6 83 84.1 83.4 Yes 100Infant Mortality DHS … 40 31 … 32 Yes 20Under-five Mortality DHS … 50 40 … 37 Yes 24Chronic Malnutrition (%) LSMS 23.7 19.7 17.8 17 16 No 7

Maternal Mortality MINSA 98 106 115 95.7 b 93 No 22

Access to Reproductive Health MINSA … 21 a 24.5 12.9 24.8 No 100Access to Drinking Water (%) LSMS 68 71.7 70.3 71.5 75.4 No 90

Access to Sanitation (%) d LSMS 44.6 50.3 51.7 55.9 88 c No 95

Illiteracy (%) LSMS 21.5 18.8 18.7 18.4 16 c No 10

PRSP goals (MDGs) Actual Data On

Track?

Source: DHS, ENACAL, LSMS, MECD, MINSA, GON PRS-II (December 2005). (a) actual data for 1999, (b) actual data for 2006, (c) target is for 2004, (d) actual data for sanitation excludes untreated latrines

Regarding access to reproductive health, the indicator that is used is the share of women of childbearing age accessing reproductive health care services. The target to be achieved by 2005 was 24.8%. The figure from the Nicaraguan Ministry of Health indicates that the coverage is 12.9 %, which is far from the target, and also a considerable drop from the 2001 figure. The explanation from the Ministry of Health is that there was a change in the way the statistics was collected. They change it because there was duplication in the previous numbers, due to women that received the services several times in a year which distorted the results.

Prospects for MDGs and PRSP Goals.

Prospects of the Nicaraguan Economy

The 2001 Poverty Reduction Strategy Paper of the Nicaraguan Government was based on four main areas: broad-based economic growth, investment in human capital, protection of vulnerable groups, and good governance, and also three crosscutting themes: environmental vulnerability, social equity, and decentralization of decision-making and services provision.

In order to achieve a broad-based economic growth, the Nicaraguan government planned to encourage the effective development of highly competitive clusters – groups of related firms – to improve the perspectives for productivity and competitiveness, and produce and sell quality Nicaraguan goods and services in the regional and international markets.

In December 2005, the Nicaraguan government presented a new version of the PRSP, called the Nicaraguan Development Plan (NDP). The NDP is the culmination of a process of participatory planning and consultations from the bottom up that facilitated the incorporation of objectives, targets and indicators for economic growth and poverty reduction.66

“The objective of the NDP is sustained high levels of economic growth by supporting local development through the formation of human capital and social protection, democratic governance with wide citizen participation, equity, government transparency and accountability, and modernization of the State to take advantage of global commerce and regional free trade agreements.” 67

66 IMF, Country Report No. 05/440, Nicaragua: Poverty Reduction Strategy Paper, December 2005, p. 1 67 Idem

159

“The NDP seeks three general objectives: higher incomes and reduced poverty, higher and better employment, and increase investment and exports”. 68

The NPD includes prioritized core areas: regulatory framework, property rights, financial services, export promotion and attraction of investment, increase productivity through cluster promotional development, rural development and environmental sustainability. It also underlines investments in electric generation based on renewable resources. 69

Under the implementation of the NDP policies, the government expected to achieve a economic growth of 4.5 % during the period from 2006- 2010, and 5 % from 2010-2020 period. 70

There are several factors that are important to achieve the expected economic growth: (i) a responsible management of government expenditures by the new leftist government; (ii) renegotiation of internal debt, (iii) improving the management of public spending, (iv) improving the rule of law and (v) definition of clear, long term rules for potential foreign investors.

Forecasting Methodology

In order to determine the 2015 forecast figures for the indicators it was required to used different approaches: i. SIMSIP Goals regressions were recalculated using Stata. iii. Elasticities to growth were determined using POVCAL and used to predict future levels of poverty indicators, iv. Elasticities to growth for Nicaragua were calculated for the different indicators, and finally, v. SIMPSIP-Goals LAC elasticities to growth for some indicators were used. A more detailed description of the SIMSIP-Goals is presented in Box 5.3.

Box 5.3. SIMSIP-Goals: Assessing the Realism of Development Targets

SIMSIP (Simulations for Social Indicators and Poverty) is a set of user-friendly Excel based simulators that facilitate the analysis of issues related to social indicators and poverty. Many of the indicators correspond to the targets and areas of focus put forward in the Millenium Development Goals. The simulations/target for future levels can be based on either historical trends or model-based elasticities. For historical trends four different ways of fitting a historical trend line across the available data at the country level are considered for each indicator and each country. The best fit is selected. The second alternative is to rely on an econometric model yielding elasticities of the indicators to economic growth, population growth, urbanization, and time. These elasticities have been estimated with two different econometric models using world-wide panel data sets, and they are allowed to vary with a country’s level of economic development and urbanization.

68 Idem 69 Idem 70 Idem pag. 74

160

Box 5.4. POVCAL: a tool to obtain the elasticity of poverty to growth.

POVCAL assists with routine poverty assessment work by using sound and accurate methods for calculating poverty and inequality measures. It requires a basic PC and any of the various types of grouped income distribution data typically available, such as income shares of deciles of household ranked by per capita income.

You need to have your grouped distributional data and you will need to know the poverty line. The program estimates the Lorenz curve, Gini index, headcount index of poverty, poverty gap index, Foster-Greer-Thorbecke index, and the elasticities of these poverty measures with respect to the mean of the distribution, and the Gini index. It does all this for two alternative specifications of the Lorenz curve – the General Quadratic (Villasenor and Arnold) and the Beta model (Kakwani).71 From the different alternatives available these are the final forecasting methodologies that were used for each indicator:

Indicator Forecasting Methodology

Infant mortality

Under-five mortality

Net Primary Enrollment (%)

Elasticity to GDP growth using an econometric model for a sample of LAC countries.72

Extreme Poverty (%)

Illiteracy Rate (10 years and over)

Chronic Malnutrition

Access to Drinking Water

Access to Sanitation

Maternal Mortality

Access to reproductive health services

Elasticity to GDP growth based on Nicaragua data

Definition of likely performance evaluation in reaching targets for the MDGs73

The results are presented in terms of performance based on four categories: likely to achieve, may possibly achieve, unlikely to achieve, and very unlikely to achieve. The definitions vary for each target according to the table presented below.

The percentages are obtained by dividing the forecasted value in 2015 of an indicator by the base year (1990 or 1993/94 for the Nicaraguan case) and multiplying by 100, except for the case of net primary enrollment, where the objective is not specified in terms of progress in percentage terms. For water and sanitation the percentage is calculated dividing the forecasted value in 2015 by the 2015 Target. Thus a number close to 0 % indicates that the level of the indicator in 2015 is very small compared to 1990, while a number close to 100 % means that the value in 2015 is close to that of 1990.

71 Shohua Chen, Gaurav Datt and Martin Ravallion, POVCAL, A program for calculating poverty measures from grouped data. 72 Elasticities obtained from Wodon, Quentin. Poverty and Policy in Latin America and the Caribbean. World Bank Technical Paper 467. June 2000. 73 World Bank. Hicks and Wodon, Reaching the MDGs in Latin America: Preliminary Results. June 2002.

161

Target Likely Possible Unlikely Very unlikely

Reduce extreme poverty by 50 %

Reduce under 5 malnutrition by 50 %

Illiteracy rate (10 yrs or more)74

0 – 50 % 50 – 60 % 60 – 80 % > 80 %

Achieve Universal primary education

Universal drinking water access75

Sanitation76

Access to reproductive services

95 – 100 % 90 – 95 % 80 – 90 % < 80 %

Reduce infant mortality by 2/3

Reduce under 5 mortality by 2/3

Maternal mortality*

Chronic Malnutrition**

0 – 33 % 33 – 50 % 50 – 75 % > 75 %

Will MDGs be achieved by 2015?

Based on the current data and forecast, most of the long terms goals are unlikely or very unlikely to be achieved. It is possible, that the following goals will be achieved: extreme poverty, reduction of infant and under-five mortality. The indicators that seem unlikely to be achieved are net primary enrollment, chronic malnutrition, access to safe water and illiteracy rate; it is very unlikely that reductions in maternal mortality, access to reproductive health services and sanitation will be achieved.

Table 5.2. Nicaragua Achievement of MDGs and Medium and Long-Term PRSP Targets

PRSP goals (MDGs)PRS-I a

Base 1993

PRS-II b

Base 2001

Actual 2005

Target PRSP-II

2010

Forecast 2015 c

Target PRS-II 2015

Target 2015 will be

achieved?Extreme Poverty (%) 19.4 15.1 14.9 11.5 11.0 9.7 PossibleNet Primary Enrollment (%) … 82.6 84.1 90.5 87.0 100 UnlikelyInfant Mortality (per 1,000 live births) 58 31 … 27 24.1 20 PossibleUnder-Five Mortality (per 1,000 live births) 72 40 … 33 31.2 24 PossibleChronic Malnutrition (%) 19.9 17.8 17 12.8 11.7 7 UnlikelyMaternal Mortality (per 100,000 live births) 160 88.6 95.7 d 63 80.3 22 Very unlikelyAccess to Reproductive Health Services … 16.1 12.9 29 21.3 100 e Very unlikelyAccess to Water (%) … 75.8 71.5 83.5 76.4 90 UnlikelyAccess to Sanitation (%) ... 87.1 55.9 f 90 60.0 95 g Very unlikelyIlliteracy Rate (%) 19 18.7 18.4 15.6 15.3 10 UnlikelySource: PRS-I, LSMS 2005, PRS-I 1st and 2nd Progress Reports, PRSP-II, and own estimates. (a) MDGs base year is 1990, Nicaragua's PRS-I explains data was not always available, then closest year was used, for most cases 1993 or 1994, except malnutrition and illiteracy 1998; (b) PRS-II base year is 2001 for poverty, infant and child mortality, malnutrition and illiteracy, or 2004; (c) Estimated on the basis of SimSIP elasticities for Nicaragua and LAC, methodology cited in World Bank Technical Paper No.467; (d) 2006; (e) Target for 2010 is 29 from a 16.1 in 2004; (f) Actual 2005 excludes untreated latrines; (g) National target.

74 This indicator was added to the original table prepared by Hicks and Wodon 75 Idem 76 Idem

162

In the rest of this section, it will be described the data available and its comparison with the 2005 targets, the PRSP targets for 2010 and 2015, and the forecasting method and results used for each indicator.

PRSP Goal 1: Reduction of poverty

Target 1: Halve, between 1995 and 2015, the proportion of people whose income is less than the extreme poverty line

Indicator: Extreme poverty headcount index The goal for poverty reduction is possible to be achieved by 2015. The extreme poverty headcount ratio for 2005 is 14.9 percent; which is the proportion of extremely poor individual in the total population of Nicaragua77. The MDG target is to halve, between 1990 and 2015, the proportion of people with an income less than one dollar a day. Given the lack of data for 1990, the Government of Nicaragua (GON) selected a different base year. The PRSP target is to achieve a level of extreme poverty of 16 percent by 2005, 11.5 percent by 2010 and 9.7 percent by 2015.78

Extreme poverty elasticity to growth for Nicaragua for the period 1994 to 2005 is –1.06, which is lower than Latin American countries (-1.3 for LAC countries).79 Thus, forecasts for the future level of extreme poverty are at 13 percent by 2010 and 11 by 2015; assuming the Nicaragua economy will grow according to the medium term objectives established in the PRSP80 and the most recent forecast of population growth made by the Nicaraguan Institute of Statistics (INEC). 81

Extreme Poverty

5

10

15

20

1990 1995 2000 2005 2010 2015

Year

Hea

dco

un

t In

dex

Actual Data Forecast PRSP/ MDG Goals

Projected

Target

77 Living Standard Measurement Survey (LSMS), 2005 78 International Monetary Fund, IMF Country Report No. 05/440 “Nicaragua: Poverty Reduction Strategy Paper”, December 2005. 79 Wodon, Quentin and others. Poverty and Policy in Latin America and the Caribbean. World Bank Technical Paper 467. June 2000. 80 Medium-term economic growth objectives are 3.7 % for 2006, 4.3 % for 2007, 4.6% for 2008, 4.8 % for 2009, and 5 % for the period 2010-2015. 81 Population growth forecast is 2.63 % from 2001-05, 2.38 % from 2006-10 and 2.14 % from 2011 to 2015.

163

PRSP Goal 2: Increase access to primary education

Target 2: Ensure that, by 2015, all children, boys and girls alike, will be able to complete a full

Indicator: Net Primary Enrollment

The goal for increase access to a hundred percent primary education coverage is unlikely to be achieved by 2015; with a point estimate of 87 percent. Seven to twelve year-olds attending primary school are 84.1 percent; the 2005 target was 83.4 and it was achieved. Nicaragua has been approved to benefit from the Education for All (EFA-FTI) Fast Track Initiative. This program will allow expanding coverage. This program will allow expand coverage, to improve education quality and offering scholarships.82 Thus, it is likely that implementation of the EFA-FTI could support Nicaragua attaining 100 percent enrollment by 2015. Interventions to improve internal efficiency will be key.

Net Primary Enrollment

60.00

65.00

70.00

75.00

80.00

85.00

90.00

95.00

100.00

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

Year

NP

E

Historical Data Forecast MDG

Target

Projected

82 SETEP, Nicaraguan Government First Progress Report. November 2002. p. 19

164

PRSP Goal 3 : Reduce maternal mortality rate

Target 3: Reduce by three-quarters, between 1994 and 2015, the maternal mortality

Indicator: Maternal deaths per 100,000 live births The goal of maternal mortality is very unlikely to be achieved by 2015. The 2015 PRSP target is to reach 22 maternal deaths per 100,000 live births. Recent data from the Ministry of Health indicates mortality rate is currently at 95.7 (2006), however, it has to be taken into account that this figure is underestimated due to lack of reporting of all deaths.

Estimates for 2010 and 2015 figures are calculated based on the elasticity of maternal mortality to Nicaragua economic growth in the period from 1998 to 2006. Based on this model, the maternal mortality rate would be 89.1 deaths per 100,000 live births by 2010 and it would be 80.3 deaths per 100,000 live births by 2015. Interventions to improve skills of birth attendants will be critical as one-fourth of births are at home on average with more than half for the extreme poor.

Maternal Mortality

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

1985 1990 1995 2000 2005 2010 2015

Year

Dea

ths

per

100

,000

live

bir

ths

Historical Data Forecast PRSP/MDG Goals

Projected

Target

165

PRSP Goal 4: Reduce infant mortality rate

Target 3: Reduce by two-thirds, between 1994 and 2015, the infant mortality rate.

Indicators: Under-five mortality rate (deaths per 1000 live births)

Infant Mortality:

The goal for infant mortality is possible to be achieved by 2015. The Nicaraguan PRSP infant mortality target is a rate of 32 deaths per thousand live births by 2005. Most recent data indicates infant mortality is 31 deaths per thousand live births in 2001, which is very close to the PRSP target by 2005.

The PRSP targets for 2010 and 2015 are 27 and 20 death per thousand live births respectively. Based on the elasticity of the indicator to economic growth and the mid-term economic growth government objectives, infant mortality would be 27.2 deaths per thousand live births by 2010, and 24.1 deaths per thousand live births by 2015.

Interventions to reduce the incidence of diarrhea and acute respiratory diseases will be key, including increased access to safe water and safe sanitation.

Infant Mortality

0.00

10.00

20.00

30.00

40.00

50.00

60.00

1985 1990 1995 2000 2005 2010 2015

Year

Dea

ths

per

100

0 liv

e b

irth

s


ProjectedTarget

166

PRSP Goal 4: Reduce under-five mortality rate

Target 3: Reduce by two-thirds, between 1994 and 2015, the under-five mortality rate.

Indicators: Under-five mortality rate (deaths per 1000 live births)

Under Five-Mortality:

The goal for under-five mortality is possible to be achieved by 2015. The Nicaraguan PRSP defined a target by 2005 for under-five mortality rate of 37 per thousand live births and 24 by 2015. Most recent data (2001) indicates that under-five mortality rate is 40 per thousand live births, a rapid decline as compared to a level of 50 in 1998. Forecasts use the calculated elasticity of the indicator to economic growth. As in the case of extreme poverty, based on the economic growth objectives for the period 2006 – 2015, the under-five mortality rate would be 35.2 per thousand live births by 2010, and 31.2 per thousand live births by 2015. Interventions to reduce the incidence of diarrhea and acute respiratory diseases will be key, including increased access to safe water and safe sanitation.

Under-five Mortality

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

1990 1995 2000 2005 2010 2015

YEAR

deat

hs p

er th

ousa

nd li

ve b

irth

s

Historical Data Forecast PRSP Goals

Target

Projected

167

Goal 5: Reduce chronic malnutrition

Target 5: Reduce chronic malnutrition to 7 % by 2015

Indicators: Prevalence of chronic malnutrition among children under age five, measured by height to age.

The goal for chronic malnutrition is unlikely to be achieved by 2015. Forecasting chronic malnutrition makes use of the responsiveness (elasticity) of chronic malnutrition to Nicaragua’s economic growth. The elasticity is -1.29; therefore, if consumption level increases by 1 percent, chronic malnutrition could decrease by 1.29 percent. Based on projected economic growth, the projected level of chronic malnutrition by 2010 would be 14.4 percent and 11.7 percent by 2015. Interventions to reduce the incidence of diarrhea and acute respiratory diseases will be key, including increased access to safe water and safe sanitation.

Chronic Malnutrition

5.00

10.00

15.00

20.00

25.00

1990 1995 2000 2005 2010 2015

Year

Chr

onic

mal

nutr

ition

(hei

ght t

o ag

e)


Projected

Target

168

Goal 6: Increase access to water

Target 6: Increase to 90 % national water coverage by 2015

Indicator: % of population with access to drinking water The goal for access to safe water is unlikely to be achieved by 2015. In 2005, according to the LSMS survey 71.5 percent of the population has access to water, which is lower than the PRSP2 target of 75.4 percent. Institutional records from the Nicaraguan government indicate that the water coverage is 77.6 percent. Based on elasticity of the indicator to the Nicaraguan economic growth, water coverage would be 73.6 percent by 2010, and 76.4 percent by 2015.

Access to Drinking Water

40

50

60

70

80

90

100

1980 1985 1990 1995 2000 2005 2010 2015

Year

Per

cen

t C

ove

rag

e

Historical Data LSMS Forecast PRSP Target

Target

Projected

169

Goal 6: Increase access to sanitation

Target 6: Increase to 95 % national access to sanitation by 2015

Indicator: % of population with access to sanitation

Sanitation

The target for sanitation is very unlikely to be achieved by 2015. Sanitation coverage in Nicaragua is 55.9 percent, which is lower than the 2005 PRSP target of 88 percent. The target for the year 2010 is 90 percent and 95 percent by 2015.

Estimates using LSMS figures from 1993 to 2001 and economic growth indicate the responsiveness (elasticity) of sanitation to economic growth is 0.254. Applying the elasticity to the projected economic growth (mid-term economic objectives and forecast), then by 2010 the coverage of sanitation would be 57.7 percent, and by 2015 it would be 60.0 percent. Sanitation services, similarly to water, are not safe because they comprise more than one-half of all latrines being untreated or about one-third of all sanitation services. Interventions to increase access to “safe” sanitation will be critical, beyond increases in coverage of any sanitation services.

Access to Sanitation

50556065707580859095

100

1990 1995 2000 2005 2010 2015

Year

% C

ove

rag

e

Historical Data Forecast PRSP Target

Target

Projected

170

Goal 7: Reduce Illiteracy Rate

Target 7: Decrease illiteracy rate to 10 % by 2015

Indicator: % of illiteracy population older than 10 year old

The goal for illiteracy is unlikely to be achieved by 2015. Illiteracy rate is 18.4 percent in 2005,83 which is higher than the 2005 PRSP target at 16 percent. The PRSP target for 2010 is 15.6 percent and 10 percent by 2015. Projections are calculated based on the responsiveness (elasticity) of illiteracy to economic growth for Nicaragua, which is –0.66.84

According to theses projections, the illiteracy rate in 2010 would be 16.9 percent and 15.3 percent by 2015. Thus, forecasts deem the PRSP long-term target for illiteracy to be unlikely to be achieved. Improving internal efficiency of education and specific adult literacy interventions will be critical actions.

Illiteracy Rate

5.00

10.00

15.00

20.00

25.00

1990 1995 2000 2005 2010 2015

Year

illit

erac

y ra

te (

old

er t

han

10

year

s)


Projected

Target

83 LSMS, 2005 84 Wodon, Quentin. “Poverty and Policy in Latin America and the Caribbean”, World Bank Technical Paper no. 467, June 2000, p. 58.

171

Ann

ex I

Ann

ex I

SOC

IAL

IN

DIC

AT

OR

S F

OR

EC

AST

USI

NG

EL

AST

ICIT

IES

TO

GR

OW

TH

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Rea

l Eco

nom

ic G

row

th3.

00%

0.80

%2.

30%

5.10

%4.

00%

3.70

%4.

30%

4.60

%4.

80%

5.00

%5.

00%

5.00

%5.

00%

5.00

%5.

00%

Popu

latio

n G

row

th2.

05%

2.05

%2.

04%

2.03

%2.

02%

2.02

%2.

01%

2.00

%1.

97%

1.94

%1.

91%

1.89

%1.

86%

1.84

%1.

81%

Rea

l GD

P pe

r ca

pita

0.95

%-1

.25%

0.26

%3.

07%

1.98

%1.

68%

2.29

%2.

60%

2.83

%3.

06%

3.09

%3.

11%

3.14

%3.

16%

3.19

%

Pro

ject

ed L

evel

s of

Soc

ial I

ndic

ator

s(b

ased

on

elas

ticiti

es to

gro

wth

)

Soci

al I

ndic

ator

sE

last

iciti

es20

0120

0220

0320

0420

0520

0620

0720

0820

0920

1020

1120

1220

1320

1420

15E

xtre

me

Pove

rty

-1.0

615

.114

.914

.614

.313

.913

.513

.012

.612

.211

.811

.411

.0

Net

Pri

mar

y E

nrol

lmen

t (%

)0.

1283

.084

.184

.384

.584

.885

.085

.485

.786

.086

.386

.687

.0

Infa

nt M

orta

lity

(per

1,0

00 li

ve b

irth

s)-0

.78

31.0

31.3

31.2

30.5

30.0

29.6

29.1

28.5

27.9

27.2

26.6

25.9

25.3

24.7

24.1

Und

er-F

ive

Mor

talit

y (p

er 1

,000

live

bir

ths)

-0.7

640

.040

.440

.339

.438

.838

.337

.636

.936

.135

.234

.433

.632

.832

.031

.2

Chr

onic

Mal

nutr

ition

(%

)-1

.29

17.8

17.0

16.6

16.1

15.6

15.0

14.4

13.9

13.3

12.8

12.2

11.7

Mat

erna

l Mor

talit

y (p

er 1

00,0

00 li

ve b

irth

s)-0

.65

115.

095

.794

.392

.790

.989

.187

.385

.583

.882

.080

.3

Acc

ess

to W

ater

(%

)0.

2470

.371

.571

.872

.272

.673

.173

.674

.274

.775

.275

.876

.4

Acc

ess

to S

anita

tion

(%)

0.25

51.7

55.9

56.1

56.5

56.8

57.2

57.7

58.1

58.6

59.1

59.6

60.0

Illit

erac

y R

ate

(%)

-0.6

618

.718

.418

.217

.917

.617

.316

.916

.616

.315

.915

.615

.3

172

REFERENCES

International Monetary Fund, IMF Country Report No. 05/440 “Nicaragua: Poverty Reduction Strategy Paper”, December 2005.

López Gómez, Salvador Leopoldo. “Proyecciones de algunas metas e indicadores establecidos en la Cumbre del Milenio para Nicaragua” , Octubre 2002

Morales, Donald. “ El seguimiento a las metas de la cumber del milenio”, Agosto 2000

Nicaraguan Government. SETEC. “A strengthened Growth and Poverty Reduction Strategy”, July 2001.

Pain, Chris, An introduction to the Challenges and Issues in the PRSP. June 2002.

Reinikka, Ritva. The Millennium Development Goals and Policy Research in the Bank. “How to measure Impact” , June 2002.

Secretaria Técnica de la Presidencia (SETEP), “First Progress Report. Strengthened Growth and Poverty Reduction Strategy”, November 2002

Secretaria Técnica de la Presidencia (SETEP), “Third Progress Report. Strengthened Growth and Poverty Reduction Strategy “, November 2005.

Valadez, Joseph, Bamberger, Michael, “Monitoring and Evaluating Social Programs in Developing Countries”, A Handbook for Policymakers, Managers, and Researchers. Economic Development Institute of The World Bank, 1994.

Wodon, Quentin. Poverty and Policy in Latin America and the Caribbean. World Bank Technical Paper 467. June 2000.

World Bank . Nicaragua. Qualitative Poverty and Exclusion study. 2000.

World Bank. Country Assistance Strategy of the World Bank Group for the Republic of Nicaragua, November 2002

World Bank. Health, Nutrition and Population Development Goals. Measuring Progress Using the Poverty Reduction Strategy Framework. November 2001.

World Bank. Hicks and Wodon, Reaching the MDGs in Latin America: Preliminary Results. June 2002.

World Bank. Krishnan Ramadas, Bernadette Ryan, and Quentin Wodon. “SIMSIP Goals: Assessing the Realism of Development Targets”, February 2002.

World Bank. Nicaragua. Poverty Assessment Follow-up. Preliminary Draft. May 2, 2002

World Bank. Nicaragua. Poverty Assessment. Challenges and Opportunities for Poverty Reduction. February 21, 2001. Report No. 20488-NI

World Bank. Shohua Chen, Gaurav Datt and Martin Ravallion, “POVCAL, A program for calculating poverty measures from grouped data”

173

6. ACCESS TO AND QUALITY OF EDUCATION SERVICES IN NICARAGUA

Diego Angel-Urdinola and Jose Ramón Laguna*

Analysis of poverty in Nicaragua emphasizes the welfare gains from education: non-poor households have higher levels of educational attainment than poor ones (especially in post-primary education), and welfare gains have been associated with higher educational attainment. Despite rather equitable access to primary school, we find that there are substantial inequities in access and quality of preschool and post secondary education between richer and poorer households, between urban and rural areas, and between regions. Furthermore, Nicaragua still falls behind in the Latin America Contexts in primary and secondary education service delivery (both in relation to access and quality). This document is organized in three main sections: the first section quantifies returns to investments on education; the second section analyses constraints and inequities in access to school by socioeconomic level, area of residence, ethnic group and gender; and the third section examines inequalities in education quality. While it is important for Nicaragua to continue to invest in achieving universal primary education, the education system needs to pay attention to improving education quality and expanding access to pres-school and secondary education.

Introduction

Analysis of poverty in Nicaragua emphasizes the welfare gains from education: non-poor households have higher levels of educational attainment than poor ones (especially in post-primary education), and welfare gains have been associated with higher educational attainment85. Despite rather equitable access to primary school, we find that there are substantial inequities in access and quality of preschool and post secondary education between richer and poorer households, between urban and rural areas, and between regions.

Results in this chapter suggest that Education outcomes in Nicaragua have significant links with poverty. Investing in education is very profitable for individuals. Indeed, estimates indicate that a Nicaraguan is expected to earn 10 percent higher wages for each additional year of schooling attained. However, despite all the advantages that education has to offer, 72 percent of the population does not attain complete secondary education and therefore are expected to earn wages bellow the poverty line. International comparisons indicate that Gross primary (secondary) enrollment rates in Nicaragua are low (normal) for Latin American standards given its level of development. Late enrollment, high dropouts, and high repetition rates all together are preventing children, and especially those form poor families, to complete primary and secondary education. Young individuals who are poor, indigenous, and who live in households engaged in agriculture attain less than 5 years of education on average. Children living in the Caribbean Atlantic region display much lower enrollment rates than in other regions up to age 14.

Lack of access and affordability are the main reasons why children ages 7 to 12 are not enrolled in primary school. While lack of access to facilities/personnel constitute important reasons why poor children do no attend primary school (especially in the Central and Atlantic regions), lack of interest and family problems have risen in importance as factors explaining school non-attendance among urban children between ages 7 and 12. While work, lack of money, and lack interest are the main reasons for boys not to be enrolled in

* The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), Jaime Saavedra (Sector Manager, LCSPP), Alexandria Valerio (Task Manager Education Sector, LCSHE) and Vanessa Castro (Consultant) for their valuable comments and suggestions. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent. 85 Box 1 presents a brief description of the institutional framework of the education sector in Nicaragua.

174

secondary/post-secondary school; family problems, child care, and pregnancy are the main reasons for girls not to be enrolled.

Box 6.1. Institutional Framework

The Ministry of Education (Ministerio de Educación, MINED) is the public institution responsible for supplying general public education services in Nicaragua. Programs offered by the MINED include i) early education (preschool), ii) primary education, iii) secondary education, iv) adult education, v) teacher’s education (formación docente), and vi) especial education (for disabled children and children with especial needs).

The public education cycle in Nicaragua starts with 3 years of tuition-free preschool instruction for children between the ages of 3 and 6. Preschool education is not mandatory and children are not allowed to repeat years at this level. Parents are free to put their children in private preschools, generally paying tuitions out-of-pocket. In 2005 only 15.7 percent of all children enrolled in preschool did so in a private institution. The Primary education cycle, targeted to children between 7 and 12 years old, is free and mandatory. The primary cycle lasts 6 years and has four modalities: i) regular primary, ii) multigrado primary (children from different levels attend the same class and are taught by the same teacher), iii) primary for adults, and iv) bilingual intercultural education program (Programa Educativo Bilingüe Intercultural, PEBI).

Secondary education serves primarily the population between 13 and 17 years old that attained primary education. Secondary education lasts 5 years, is not mandatory, and has four modalities: i) daytime-secondary, ii) nighttime-secondary, and iii) distance secondary education (classes are conducted on Saturdays or Sundays weekdays), and iv) secondary education for adults.

The education system has a total of 10,721 schools; 85% of which are public and 15% private with and without a voucher. About 79% of all school infrastructure is located in rural areas (92% of which is owned by the government). In year 2005, the system served about 1,685,844 students in all modalities of basic, primary, and secondary education. Primary education accounts for 56.1% of all students; secondary for 24.6%, preschool for 12.7%, adult education for 5.5%, and all remaining modalities for 1.2% (see Figure 6.1)

Distribution of Children Enrolled by Type of Program, Year 2005

Primary55.9%

Secondary24.5%

Adult education5.5%

Techers' education0.3%

Other1.0%

Special education0.2%

Preschool12.6%

Source: Statistics Department, MINED (2005).

The analysis in this chapter suggests that relative to their income level, education is more expensive for the poor. While school fees represent a much higher share of overall expenditures on education for households

175

in the upper quintiles, non-fee related items (such as uniforms, school supplies, books, and transport) constitute a proportionally higher burden for the poor. Tuitions for tertiary education are found to be very expensive for the general income level of the population. Indeed, data suggest that university tuitions are not affordable by the poor and are hardly affordable by households in the middle class

In regards to education quality, this chapter finds that private preschools (generally associated with better learning indicators than public ones) are less accessible to the poor. Nicaragua is the Latin American country with the highest pupil-teacher ratio in the region; both in primary and secondary schools, and its teacher work force is one of the least qualified in the region. About 20 to 25 percent of all parents with children in the school system consider that their education is either regular or bad. Quality deficiencies are also reflected in the fact that less than 14 percent of all students in 3rd and 6th grade are found to be proficient in their curriculum.

This document is organized in three main sections. The first section quantifies returns to investments on education and estimates the expected years of study that are necessary to escape poverty in Nicaragua. The second section analyses inequities in school access by education level, socioeconomic group, area of residence, ethnicity, and gender. The section also presents and quantifies the main constraints (mainly related to access and affordability) households face to send their children to school. The third section examines inequalities in education quality: first, the section analyzes quality-outcomes by socio-economic group (such as repetition, attainment, and test scores); and second, it presents subjective user’s perceptions in relation to the quality of school service delivery. A brief set of conclusions and policy recommendations follows.

Returns to Education

Education outcomes in Nicaragua have significant links with poverty. Lack of qualified human capital decreases national competitiveness and limits the development of science and innovations that improve productivity. Lack of education constitutes one of the main determinants of poverty in Nicaragua. As indicated in Table 1, households having a head with technical and tertiary education consume on average 55 to 82 percent more respectively than otherwise similar households having a head with no education. But having a head with technical or tertiary education is a privilege of less than 10 percent of all the households in the population.

Table 6.1. Roughly 72 percent of all households in Nicaragua have a head with at most primary education.

% increase in expected Consumption vs. households

having a head with no education

Population Share

Head with no education/adult education 30.8% Head attained Primary 16.5% 41.7% Head attained Secondary 32.2% 17.8% Head attained Technical 54.7% 2.9% Head attained Tertiary 81.7% 6.8% Source: World Bank using the 2005 Nicaragua LSMS.

There is plenty of empirical evidence demonstrating the lineal relationship between the educational level of a country’s adult population and its wealth. Figure 6.1 shows that the poorest Latin American countries (Nicaragua and Honduras) are at the same time those countries displaying the lowest education levels among their adult population (5.6 and 5.4 years of education), whereas Argentina and Chile display the highest education rates along with the highest per capita incomes in the Latin American region.

176

Figure 6.1. Mean Years of Education in Nicaragua vs. LAC [period 1999-2004]

ARG

BOLBRA

CHL

ECU

SLV

GTM

MEX

NIC

PAN

PRYPER

URY

VENCOL

CRIDOM

LAC

HND

0

2

4

6

8

10

12

2,000 4,000 6,000 8,000 10,000 12,000

GDP per capita (average 99-04 in 2000 US$ constant PPP)

Yea

rs o

f E

du

cati

on

(25

-65

year

s o

ld)

Source: International Education Statistics (2007)

In Nicaragua, investing in education is very profitable for individuals. It is estimated that a Nicaraguan earns 10.3 percent higher wages for each additional year of schooling received86. Table 6.2 indicates that in the 1998-2005 period returns to education have been fairly stable and similar in magnitude across strata and gender groups (at approximately 9 to 10 percent per extra year of education). In addition, controlling for the selection bias arising from those unemployed or reporting zero labor income, our estimates suggest that these individuals would be willing to work lower than average salaries (especially males residing in urban areas). In contrast, we find a higher “reserve” salary among women and in rural areas (this could happen if these individuals have higher educational levels and thus can afford to stay unemployed a bit longer until they find better jobs). Thus, when entering the labor market, these individuals would receive higher returns due to their higher school attainment.

Table 6.2. Education returns by sex and area of residence (OLS and Heckman MVL)

OLS HECKMAN Education Categories 1998 * 2001* 2005 1998 * 2001 * 2005

National 10.3% 9.5% 10.3% 11.4% 11.2% 8.7% Male 10.9% 9.9% 11.0% 9.5% 9.5% 9.4% Female 10.2% 10.0% 9.8% 11.1% 12.2% 10.7% Urban 10.5% 10.2% 9.5% 11.1% 10.8% 9.2% Rural 6.5% 6.1% 7.5% 8.3% 7.7% 9.2%

Source: World Bank using the 2005 Nicaragua LSMS. * Laguna y Porta (2004)

There is evidence of decreasing returns to tertiary education and of greater returns to primary and secondary education between 1999 and 2005. This phenomenon may reflect a greater demand for semi-qualified labor, which could be attributed to an increasing demand for labor in the “maquilas”. In the same way, Figure 6.2 illustrates that private returns from primary, secondary and tertiary in year 2005 education were at 8.3, 9.7 and 16.5 percent,87 whereas social returns were at 7.3, 9.9 and 10.2 percent,

86 The methodology used to produce 2005 calculations follows Laguna (2003) and Laguna and Porta (2005). 87 Estimates take into account direct costs of education such as tuition fees, transportation, uniforms, books, and textbooks among others.

177

respectively. It should be noted that tertiary education’s lower social return may be explained by the government’s significant subsidy to public universities88.

Figure 6.2. Historically, tertiary education has generated the highest private returns, which is not the same in the case of social returns.

Private Returns

0%

4%

8%

12%

16%

20%

1998 2001 2005

Primaria Secundaria Terciaria

Social Return

0%

4%

8%

12%

16%

1998 2001 2005

Primaria Secundaria Terciaria

Source: World Bank using the 2005 Nicaragua LSMS.

Individuals with primary and incomplete secondary education (about 72 percent of the population) are expected to earn wages bellow the poverty line. Using a simple Mincer model, one can estimate the expected income of an individual based on his/her years of education and experience. Estimates for Nicaragua indicate that individuals with 1, 5, and 10 years of experience need to attain respectively 8, 9, and 10 years of education to earn wages that would be equivalent to the poverty line. This result implies that returns to experience among individuals with primary or incomplete secondary education are very low (5 years of experience achieve a wage gain equivalent to one extra year of education). As indicated by the shape of he curves in Figure 6.3, an extra year of education produces higher returns on wages once individuals have attained more than 11 years of education (which corresponds to the schooling necessary to complete secondary school). Returns to experience are also larger among individuals with post-secondary education. These findings are consistent with those in Laguna (2003), in the sense that the authors identify the need to invest more resources to develop human capital accumulation among the less favored segments of the population.

88 The social rate of return aggregates public and private subsidies to education, as well as any other positive externality that is not perceived by the individual. Given the difficulty to estimate economically positive externalities of education, the social rate of the return was calculated only accounting for public subsidies.

178

Figure 6.3. To earn wages above the poverty line an individual with no experience needs at least 8 years of education.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Years of schooling

Hou

rly w

age

rate

in C

ordo

bas

1 year 5 years 10 years (Exp) Poverty Line

Primary Secondary Tertiary


INEQUITIES IN ACCESS AND PERMANENCE IN SCHOOL 89

During the past decade, school attendance by children from the poorest households has increased significantly, among age groups corresponding to both primary and secondary education. Figure 6.4 shows that during the 1993-2005 period, the main increases in enrollment have benefited the poorest sectors of society, particularly the extremely poor. The percentage of children from the poorest quintile not attending school has diminished by 25.5 points among the 7 to 12 age group, and by 16.6 percentage points for those between ages 13 and 18. Despite efforts made to diminish the access gap among the poorest quintiles of income distribution, a large percentage of children still do not attend school.

Figure 6.4: One out of five children between ages 7 and 12 from the poorest quintile does not attend school, whereas half of the young people between ages 13 and 18 from the same quintile do not

attend school.

Percent 7-12 yrs old Not Attending School

0

10

20

30

40

50

1993 1998 2001 2005

All Extreme Poor Poor Non-poor

Percent 13-18 yrs old Not Attending School

0

20

40

60

80

1993 1998 2001 2005

All Extreme Poor Poor Non-poor

From 1998 to 2005, the poorest quintile reported a 10 point increase in the net primary school enrollment rate, whereas the main increases at the secondary level are found in the intermediate

89 See WorldBank (2005) and Porta and Laguna (2007) for Central American comparisons.

179

income quintile, with a rise of 15 points for the same period. Figure 6.5 illustrates that the highest rates of primary school enrollment are found among the intermediate income quintile. The apparent lower enrollment rate for the richest quintile can be explained by the fact that these families tend to enroll their children at a younger age than the official enrollment age. It should be noted that at the secondary level, school enrollment gaps are greater.

Figure 6.5. Changes observed in Primary and Secondary Net Enrollment Rates by quintiles in the 1998 - 2005 period.

Net Enrollment Rates for Secondary 13-18 years old

0.0

25.0

50.0

75.0

100.0

Poorest II III IV Richest

1998 2005

Net Enrollment Rates for Primary 7-12 years old

0.0

25.0

50.0

75.0

100.0


1998 2005

Source: World Bank using 1998 and 2005 LSMS data

Table 6.3. Youth from the poorest households and rural areas have the lowest literacy rates.

1993 1998 2001 2005 All 82.3 85.6 86.4 90.4

Extreme Poor 65.5 65.0 65.6 77.1 Moderately Poor 88.9 89.4 89.8 87.0 Poor 75.1 76.0 76.1 83.9 Non-poor 93.8 93.7 94.3 95.8

Urban areas 94.3 94.2 93.9 95.8 Extreme Poor 81.6 81.2 74.6 85.6 Moderately Poor 95.2 96.1 94.8 92.9 Poor 89.8 88.2 88.0 91.5 Non-poor 96.3 97.9 96.0 97.6

Rural areas 70.5 75.4 75.7 83.0 Extreme Poor 60.4 63.9 63.3 74.5 Moderately Poor 75.7 81.7 80.2 82.4 Poor 65.4 72.5 69.1 79.3 Non-poor 84.5 86.2 88.9 90.6

Socio-economic Quintiles Poorest quintile 66.2 68.0 67.2 78.3 Q2 77.1 79.5 81.5 87.3 Q3 88.3 90.5 88.0 92.1 Q4 93.0 90.4 94.7 94.9 Richest quintile 96.8 97.1 98.1 99.0

Source: World Bank using 1993, 1998, 2001 and 2005 LSMS data

In extremely poor households in Nicaragua, one out of four young people between the ages of 15 and 24 years is illiterate. It is evident that literacy rates among young people have raised an average of 8 points during the period between 1993 and 2005, particularly among extremely poor sectors.

180

Nevertheless, it is alarming that while 99% of young people from the richest quintile can read and write, only 78% from the poorest quintile can do so. Duryea and Pagés (2002) argue that while providing adult education has been a low priority in most countries, research suggests that bringing adults back to school can be an effective policy for increasing productivity.

Enrollment

Children from richer households are much more likely to be enrolled in preschool, secondary, and tertiary education as compared to poorer ones. Figure 6 displays differences in enrollment rates by age between rich and poor children and between boys and girls. The Figure provides several messages: i) differences in enrollment rates between children in the poorest and richest quintiles are more pronounced at the tails (preschool and upper secondary), ii) among the poor enrollment rates for girls are higher than those for boys at almost every age group, iii) in the poorest quintile, enrollment rates are highest between ages 8 and 9 while in the richest quintile enrollment rates peaks since age 6 (this suggest some signs of late enrollment among the poor), iv) enrollment rates drop rapidly after age 12 (which is the age at which children should complete primary school), especially among the poor, and v) differences in enrollment rates after age 17 between rich and poor boys and twice as large as between rich and poor girls. 90

Figure 6.6. At age 17 enrollment rates for males in richest quintile are 2.6 times larger than for males in the poorest quintiles91.

0

10

20

30

40

50

60

70

80

90

100

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Per

cent E

nro

lled

Girls, Poorest Quintile

Boys, Poorest Quintile

Girls, Richest Quintile

Boys, Richest Quintile

Preschool Primary Secondary Tertiary

Source: World Bank using the 2005 Nicaragua LSMS. The curves have been smoothed by eliminating some unexplained enrollment spikes in secondary and tertiary education.

Children living in the Caribbean Atlantic region display much lower enrollment rates than in other regions up to age 14. Figure 6.7 displays differences in enrollment rates by age between children across

90 By the age of 13, some 44% of boys in the poorest quintile are working and by age 17 the equivalent proportion reaches 85%. 91 When calculating the share of boys and girls who are enrolled in the class they should be at according to their age, one can observe than only half of all children of age 7 are in first grade (this share is even lower in urban areas due to high rates of early enrollment). By secondary education, only 1 out of every 5 children is enrolled in the class they should be enrolled in according to their age (and the situation is even worse in rural areas where only 7 out of every 100 children finish secondary education at the expected age, which is 17 years old). As it will be mentioned bellow, this phenomenon happens due to late enrollment in rural areas and among the poor and to early enrollment in urban areas and among the rich.

181

regions in Nicaragua. Not surprisingly, enrollment rates are higher at almost every age group in Managua and in the Pacific Region. At the primary level (ages 7 to 12) enrollment rates by age in the Caribbean Atlantic Region are approximately 10 percentage points bellow those in other regions (averaging between 70 and 80 percent). After age 12 (age at which children are suppose to start secondary education) enrollment rates in Managua are much higher than in other regions. After age 14, enrollment rates by age in the Atlantic region catch up with those in the Pacific region and even surpass those in the Central Region. Enrollment gaps across regions narrow significantly after age 17. Between ages 17 and 18, ages at which children usually finish secondary education, enrollment rates drop dramatically in Managua; such a drop is less pronounced in all other regions.

Figure 6.7. Enrollment rates by region differ significantly up to age 17 but catch up after that.

0

10

20

30

40

50

60

70

80

90

100

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Per

cen

t E

nro

lled

ManaguaPacíficoCentralAtlántico

Preschool Primary Secondary Tertiary

Source: World Bank using the 2005 Nicaragua LSMS. The curves have been smoothed by eliminating some unexplained enrollment spikes in secondary and tertiary education.

Poor children, especially indigenous and those living in households engaged in agriculture display much lower preschool and secondary enrollment rates than average. As illustrated by Figure 6.8, differences in preschool and secondary enrollment rates – contrary to what happens in primary – display great variation by socio-economic group, to the detriment of the poorest quintile. School coverage rates in the highest quintile are double and triple the coverage rates for preschool and secondary school in the lowest quintile, respectively. Analysis of preschool coverage of children between age 4 and 6 shows that half of children in the highest quintile go to preschool, whereas only 1 out of every 4 children from the 20% of poor households attend that same education level. As for high school, gaps become even more critical. While 7 out of every 10 youths between age 13 and 17 in the highest quintile attend high school, only 2 out of 10 youths from the poorest 20% of households are in the school system.

Figure 6.8 also illustrates that preschool and secondary net enrollment rates are significantly lower for indigenous children (vs. non indigenous)92, for children living in a household engaged in agricultural

92 This can partly be explained by the fact that only 13% and 14% of schools providing preschool and high school are located in the Atlantic region of the country.

182

production (vs. non-agriculture producer households), and for boys as compared to girls (above all in secondary school).

Figure 6.8. While differences in net enrollment rates by socio-economic group are mild for primary education, they are quite substantial for pres-school and secondary education.

64.2

27.5

52.2

64.3

45.7

57.250.4

39.935.6 28.530.6

17.4

74.2

0

20

40

60

80

100

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-In

dige

nous

Indi

geno

us

Non

Ag.

hous

ehol

d

Ag.

hous

ehol

d

Fem

ale

Mal

e

Per

cent

Secondary Net Enrrollment RatesPrimary Net Enrrollment RatesPreschool Net Enrrollment Rates


Children in rural areas and those living in the Central and Atlantic regions display lower than average enrollment rates in preschool and secondary education. There are large differences in enrollment across region in Nicaragua, especially for secondary education. As presented in Table 6.4, net secondary enrollment in rural areas is half of that in urban areas (28.1 vs. 61.1. percent)93. The difference in net enrollment rates between Managua and the Atlantic region is striking (27 vs. 66.2 percent for primary and 28.7 vs. 48.1 percent for secondary). On the contrary, primary net enrollment rates are rather flat across strata and across regions94.

Table 6.4. Enrollment rates in the Atlantic Region fall behind nationally, especially for preschool and secondary school.

% Children 4 to 6 enrolled in CICO/CDI

Preschool Net Enrollment Rates in %

Primary Net Enrollment Rates in %

Secondary Net Enrollment Rates in %

By Strata Rural 3.6 32.9 84.0 28.1 Urban 2.5 42.7 84.3 61.1

By Region Managua 3.4 48.1 82.9 66.2 Pacific 0.6 40.2 86.1 51.0 Central 4.6 34.7 84.9 35.9 Atlantic 3.2 28.7 80.8 27.0


Socio-economic conditions, characteristics of the parents, gender, geographical location, and employment opportunities have a significant influence on the probability of children being enrolled

93 According to MECD (2006a), 40% of secondary schools are located in rural areas. 94 Gutiérrez and Laguna (2006) indicate that there are significant differences in primary school coverage at the departmental level: while the RAAN region Net Enrollment Rate (NER) is at 81%, in the department of Granada the NER is over 100%.

183

in secondary or post-secondary education. The following results summarize regression results for the determinants of secondary and post-secondary enrollment in Nicaragua (see Table A1 in the annex) for children between 12 and 23 years or age95:

Individual characteristics: After the age of 12, the probability that a student stays in school falls by 24 percent per year nationally (19 percent in urban areas and 29 percent in rural areas). Individuals between 12 and 23 years who have a job are 20 to 22 percent less likely to attend secondary or post-secondary education. Being male is associated with a 3 percent lower probability of enrollment after age 12 in urban areas and with an 8 percent higher probability of being enrolled in rural areas. Controlling for other factors, indigenous children display a higher probability of being enrolled after age 12 (5 percent higher nationally and 11 percent higher in urban areas compared to non-indigenous individuals).

Education of the household head and spouse: The level of education of the household head is a strong determinant of school enrollment after age 12. As displayed in Figure 9, children living in a household having a head who attained post secondary are 20 to 40 percent more likely to be enrolled as compared to children living in households having a head with incomplete primary or no education96. Similarly, this phenomenon holds for children living in a household with a spouse who attained post-secondary education.

Socio-economic conditions and geographical location: Controlling for other factors, socio economic condition (proxied by the household consumption quintile) constitutes an important determinant affecting children’s probability of being at school after age 12. In particular, results indicate that children living in households belonging to the richest quintile are 24 to 40 percent more likely to be enrolled in secondary school as compared to households in the bottom quintile. The age of the household head also influences the probability of children’s enrollment. In particular, results indicate that for every additional year of age of the household head (for those heads above age 50) children are 1.5 percent more likely to be at school after age 12. Results suggest that urban children in the Central and Atlantic regions, controlling for other characteristics, display respectively a 5 to 13 percent higher probability of being enrolled after age 12 as compared to children with similar characteristics residing in Managua. The contrary occurs in rural areas where children in the Central, Pacific, and Atlantic regions display respectively a 29 to 35 percent lower probability of being enrolled after age 12 as compared to children with similar characteristics residing in Managua.97

95 See Kruger (2001) for additional information on child labor and education outcomes for children between ages 6 and 14. 96 In this respect, findings of the Impact Assessment Program on Basic Education for Youths and Adults in Nicaragua, carried out by Handa et al. (2006), reveal that the training received through this program has helped participants to have a more effective participation in their children’s education, acquiring greater awareness and interest about their children’s access to school. 97 Regression analysis controls for the probability of a given outcome conditional on a vector of characteristic X. That the expected “conditional” probability that an urban child in the Atlantic region stays in school after age 12 exceeds that of a child with similar characteristics in Managua does not mean the equivalent unconditional probability (i.e. the share of children 13 to 24 who is at school) is higher in the Atlantic region as compared to Managua. Indeed, results indicate (see figures 4 and 11) that post-secondary enrollment rates in Managua are higher than in any other region.

184

Figure 6.9. The education of the household head has a strong influence in the probability of children being enrolled in secondary school.

8.9%

18.0%19.7%

33.7%

15.7%

20.0%

27.7%

10.0%

27.3%

20.8%

5.9%

17.0%

0%

5%

10%

15%

20%

25%

30%

35%

40%

Head received degreeprimary or adult educ.

Head received degreesecondary

Head received degreetechnical

Head received degree,higher education

% In

crea

se in

the

prob

abilit

y of

bei

ng in

se

cond

ary

scho

ol [c

hild

ren

12 to

23]


Source: World Bank using the 2005 Nicaragua LSMS. [Reference group: children living in households with a head with incomplete primary or no education]

Socio-economic conditions and geographical location: Controlling for other factors, socio economic condition (proxied by the household consumption quintile) constitutes an important determinant affecting children’s probability of being at school after age 12. In particular, results indicate that children living in households belonging to the richest quintile are 24 to 40 percent more likely to be enrolled in secondary school as compared to households in the bottom quintile. The age of the household head also influences the probability of children’s enrollment. In particular, results indicate that for every additional year of age of the household head (for those heads above age 50) children are 1.5 percent more likely to be at school after age 12. Results suggest that urban children in the Central and Atlantic regions, controlling for other characteristics, display respectively a 5 to 13 percent higher probability of being enrolled after age 12 as compared to children with similar characteristics residing in Managua. The contrary occurs in rural areas where children in the Central, Pacific, and Atlantic regions display respectively a 29 to 35 percent lower probability of being enrolled after age 12 as compared to children with similar characteristics residing in Managua.98

Gross primary (secondary) enrollment rates in Nicaragua are low (normal) for Latin American standards given its level of development. Figure 6.10 plot respectively the average 1995-2004 gross enrollment rates for primary and secondary education in a pool of LAC countries against the natural logarithm of GDP per-capita (average 1995-2004). The Figures illustrate that given Nicaragua’s level of economic development, gross enrollment in primary school is low for regional standards while gross enrollment in secondary school is aligned with regional standards.

98 Regression analysis controls for the probability of a given outcome conditional on a vector of characteristic X. That the expected “conditional” probability that an urban child in the Atlantic region stays in school after age 12 exceeds that of a child with similar characteristics in Managua does not mean the equivalent unconditional probability (i.e. the share of children 13 to 24 who is at school) is higher in the Atlantic region as compared to Managua. Indeed, results indicate (see figures 4 and 11) that post-secondary enrollment rates in Managua are higher than in any other region.

185

Figure 6.10. Gross enrollment rates in Primary [Nicaragua vs. LAC, period 1995-2004]

ARG

BOL

BRA

CHL

CRIDOMECU

SLV

GTM

MEX

PRY

PER

URY

COL

NIC

PANVEN

LAC

25

40

55

70

85

100

2,000 4,000 6,000 8,000 10,000 12,000


Gro

ss E

nro

llmen

t R

ates

in S

eco

nd

ary

ARG

BRA

CHL

COL

GTM

HNDMEX

NIC

PER

LACBOL

CRI DOM

ECU

SLV PAN

PRY

URY

90

100

110

120

130

140

2,000 4,000 6,000 8,000 10,000 12,000


Gro

ss E

nro

llmen

t R

ates

in P

rim

ary


Permanence

To achieve the goal of universal primary education completion, children need not only to enroll in the educational system but also to remain on it (UNESCO, 2005). As such, it becomes important to analyze the behavior of other “permanence” indicators such as first-grade enrollment rates as well as dropout and repetition rates.

Late enrollment (also known as “over-age”) in first grade is common among children in the poorest quintiles, and especially among boys and in rural areas. Children in Nicaragua are supposed to enter the first year of primary education at age 7. Figure 6.11 indicates that first-grade enrollment is Nicaragua (i.e. the share of all children who are 7 years old and enrolled in first grade) is rather low at 20 to 30 percent (the goal is 100 percent). Low first-grade enrollment rates can be explained mainly because poor children in rural areas (and especially those living in the Central and Atlantic regions) generally enroll late (after age 7) while richer children (generally those from wealthier families and in urban areas) generally enroll early (at age 6) – see Table A3 in the annex –. Furthermore, using administrative enrollment records, Laguna and Gutierrez (2006) find that the departments of Jinotega, RAAN, and RAAS display nationally the highest rates of late enrollment in first grade.

First-grade net enrollment rates in Nicaragua are low for international standards, mainly because children enter the education system one year in advance. Compared to other countries in the LAC region, Nicaragua has a low rate of first grade net-enrollment given its level of development. Nevertheless this rate would be higher if the official age for first grade enrollment were at 6 years. Figure 6.12 indicate that the official age for first-grade enrollment (7 years of age) does not reflect the actual age at which children are entering primary education. Indeed, about 50 percent of all children enrolled in first-grade in Nicaragua are 6 years old. In this respect, and in order not to convey a misleading message, the official age for primary enrollment (as presented in Ministry Agreement No. 094-2005) should be lowered from 7 to 6 years.

186

Figure 6.11. The rate of children who are “over-age” for their grade from the poorest quintile is seven times greater than in the richest quintile

46.7

7.8

52.8

19.7

26.5

36.131.3

56.3

24.3

60.3

24.9

46.8

37.329.8

50.056.1

33.5

27.4

0

10

20

30

40

50

60

70

Poo

r

Non

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Fem

ale

Mal

e

Rur

al

Urb

an

Man

agua

Pac

ific

- U

rban

Pac

ific

- R

ural

Cen

tral

- U

rban

Cen

tral

- R

ural

Atla

ntic

- U

rban

Atla

ntic

- R

ural

Net Intake Rate Primary Over-Age


Figure 6.12. Nicaragua has the lowest net rate of first grade enrollment

ARG

BOL

DOM

ECU MEX

NIC

PAN

PER

VENLAC

COLSLVGTM

20.0

40.0

60.0

80.0

100.0

2,000 4,000 6,000 8,000 10,000 12,000GDP per capita (average 99-04 in 2000 US$ constant PPP)

Net

Inta

ke R

atio

Pri

mar

y

ARG

BOLCOL

MEX

NIC

LAC

DOM

ECU SLV

PERVEN

0

10

20

30

40

50

2,000 4,000 6,000 8,000 10,000 12,000


Net

Inta

ke R

atio

(-1

) P

rim

ary


Although drop-out rates in primary are much higher among the poor than the non-poor, drop-out rates in secondary are equally high among poor and non-poor individuals. Figures 6.13 and 6.14 illustrate differences in drop-out rates between poor and non-poor, between boys and girls, and between regions for individuals in the age groups 16 to 29.99 Figure 13 indicates that boys drop-out from school faster than girls, especially in secondary education and among the poor.

99 We select this age group because at age 16 the majority of individuals have finished their education but are still young enough to enter post-secondary education. At age 29 most individuals who pursue post-secondary education have finished their education cycle.

187

Figure 6.13. Only 1 out of every 100 boys in the poorest quintile attains 11 years of education

0.00

20.00

40.00

60.00

80.00

100.00

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Years Attained

Per

cen

t

Girls - Richest Quintile

Boys - Richest Quintile

Girls - Poorest Quintile

Boys -Poorest Quintile

PRIMARY SECONDARY TERTIARY

Source: World Bank using the 2005 Nicaragua LSMS. [Sample: individuals between 16 and 29 years old]

While 2 (3) of every 10 boys (girls) in the poorest quintile attain 6 years of education (i.e. complete primary) roughly 9 out of every 10 boys/girls in the highest quintile do so. As expected, among boys and girls in the richest quintiles differences in drop-out rates are mild below 6 years of education attained (primary). After 6 years of education attained, drop-out rates accelerate among boys and girls in the richest quintiles (averaging 10 percent per year) which are similar to those experimented by individuals in the bottom quintile. The Figure illustrates that while roughly only 1 out of every 100 boys (or girls) attain 11 years of education in the bottom quintile, 30 (37) out of every 100 boys (girls) do so in the highest quintile (see Figure 6.13 above). Finally, Figure 6.14 illustrates that every education level, attainment is much lower in the Atlantic and Central regions than in Managua and in the Pacific regions.

Children from the poorest households in rural areas, and especially those living in household engaged in agriculture, display higher than average repetition rates in primary school. Figure 6.15 indicates that poverty, beyond influencing access to education, has an important impact on other indicators on educational achievement. In particular, results indicate that repetition rates for primary education are at 14 percent among poor children vs. 9 percent among non-poor children. Repetition rates in primary are lower for indigenous children v.s. non indigenous ones (9 vs. 12 percent) and higher for children living in households engaged in agriculture vs. those living in households not engaged in agriculture (13 vs. 11 percent). Repetition rates in secondary school are about half of those in primary school (approx. 12 percent in primary vs. 6 percent in secondary nationally). Furthermore, contrary to what occurs in primary, repetition rates in secondary school are rather homogenous across different socio-economic groups, strata, and regions.

188

Figure 6.14. At every education level, attainment is much lower in the Atlantic and Central regions than in Managua and the Pacific regions

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Years Attained

Per

cen

t

Managua

Pacifico

Central

Atlantico

PRIMARY SECONDARY TERTIARY

Source: World Bank using the 2005 Nicaragua LSMS. [Sample: individuals between 16 and 29 years old]

Figure 6.15. Primary school repetition percentages in the poorest quintile are twice the values observed in the highest quintile

9.3

13.6 14.1

12.711.6

9.6

7.7

11.9

9.1

10.9

12.7 12.9

10.4

5.6

6.16.7

5.85.2

6.45.96.66.76.06.0

4.76.6

0

3

6

9

12

15

18

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Agr

icul

tura

l pr

oduc

er h

ouse

hold

Agr

icul

tura

l pr

oduc

er h

ouse

hold

Rur

al

Urb

an

Per

cent

Repetition in Primary (%) Repetition in Secondary (%)


Private primary schools without subsidies and secondary non-autonomous school have lower repetition rates. Figure 6.16 indicates that private schools display lower average repetition rates than public schools (3 vs. 13 percent). As expected, children living in households in the upper quintiles and those residing in Managua are more likely to have access to private primary education.

189

Figure 6.16. Primary Repetition Rates by Type of school

14.813.0

6.9

11.9

4.6

12.7

5.2

3.0

6.8

0

4

8

12

16

Primaria Secundaria

Per

cent

Centro educativo comunitario Escuela / centro autónomoPúblico no autónomo Privado subvencionadoPrivado no subvencionado


According to MECD official statistics, repetition rates almost doubled between years 2000 and 2005 (from 5.1 percent in 2000 to 9.9 percent in 2005). There is a high intrinsic cost of having high repetition rates if one considers that resources invested on students who repeat certain grade are a waste. In 2005 the MECD estimated that the annual cost of repetition for primary and secondary education was at approximately US$12.0 and US$1.2 million per year respectively (Laguna y Gutiérrez, 2005). Although some policymakers have pointed out that the rise in grade repetition was due to the elimination of automatic promotion, Castro (2005) indicates that this policy had no effect because of the poor communication strategy used initially, which was meant to empower key actors for the implementation of such policy (the teachers). In the same way, Castro (2005) insists that the phenomenon of grade repetition is a complex problem influenced by several factors, such as the quality of teachers, the capacity of school principals to monitor, supervise and advise teachers, the decisions made by households to withdraw their children from school, and the MECD’s own support and monitoring structures, among others. Therefore, solving this problem requires several strategies and actions that consider all of these aspects.

Repetition rates for primary (secondary) education in Nicaragua are normal (higher) for Latin American standards given its level of development. Figure 17 illustrate that given Nicaragua’s level of economic development, repetition rates in primary school are aligned with regional standards while repetition rates in secondary school are higher in Nicaragua respect regional standards. Also, when comparing these results to other countries in the region with equivalent income levels, such as like Ecuador, Honduras, and Bolivia, repetition rates in Nicaragua are in the high-side.

Figure 17: Repetition rates in Primary [Nicaragua vs. LAC, period 1995-2004]

ARG

BRA

CHL

COL

ECU

SLV

GTM

MEX

NICPRYPER

URYVEN

BOL

CRI

DOMPANLACHND

0

5

10

15

20

25

2,000 4,000 6,000 8,000 10,000 12,000


Rep

etiti

on R

ates

Prim

ary

ARG

BOLCHL

COL

CRI

DOM

ECU

SLVGTM

MEX

NICPAN

PRY

PER

URY

VEN

LAC

0.0

3.0

6.0

9.0

12.0

2,000 4,000 6,000 8,000 10,000 12,000


Rep

etiti

on R

ates

Sec

onda

ry


190

C. SCHOOL ATTAINMENT

Young individuals who are poor and especially those living in households engaged in agriculture attain less than 5 years of education on average. As illustrated in Figure 6.18, poor individuals between 23 and 29 years old (age at which most should have finished studies) have attained less than 5 years of education on average. This suggests that these individuals are likely to be primary-school dropouts. Indigenous individuals attain on average 7 years of education, which is at pace with non-indigenous individuals and much above attainment among individuals living in agriculture producing households (who on average attain only 4.7. years of education). Non-poor individuals in this age group, as well as those in the upper quintiles, attain on average 9 to 11 years of education, which is lower than the necessary to complete secondary school.

Figure 6.18. On average, young individuals between 23 and 29 years old in Nicaragua have attained only primary school

9.2

3.6

6.8

8.4

11.3

7.4 7.4

8.7

4.64.7

5.3

1

2

3

4

5

6

7

8

9

10

11

12

13

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-In

dige

nous

Indi

geno

us

Non

Agr

ic.

hous

ehol

d

Agr

ic.

hous

ehol

d

Yea

rs o

f com

plet

ed e

duca

tion

[coh

ort 2

3 to

29

year

s ol

d]

Primary

Secondary

Source: World Bank using the 2005 Nicaragua LSMS

Late enrollment, high dropouts, and high repetition rates all together are preventing poor children to complete primary education. Figure 6.19 shows the evolution of primary completion rates by age group. Results suggest that in the last two decades the proportion of children from the poorest quintile who have finished primary school has doubled; even though the primary completion gap between the richest quintile and the poorest is still large (92 vs. 42 percent). In addition, Figures 6.20 show that 70 percent of children between 15 and 19 years old have managed to finish primary school, whereas only 32 percent of young people between 20 and 24 years old have completed secondary education. If this trend were to continue, we could estimate that almost 81 percent of the population currently between 0 and 4 years of age is expected to finish primary education and only 55 percent secondary education.

191

Figure 6.19: The gap in completion rates between poor and non-poor individuals is closing for the newer generations, mainly due to significant progress among the poor

Primary Completion Rates - Poorest and Richest Quintiles

0

20

40

60

80

100

65+

60 -

64

55 -

59

50 -

54

45 -

49

40 -

44

35 -

39

30 -

34

25 -

29

20 -

24

15 -

19

Age Group

%

Poorest Richest


Figure 6.20: Primary and secondary completion rates by five-year age groups.100

Primary

80.8

70.0

0

10

20

30

40

50

60

70

80

90

100

65 a

nd +

60-6

4

55-5

9

50-5

4

45-4

9

40-4

4

35-3

9

30-3

4

25-2

9

20-2

4

15-1

9

10-1

4

5-9

0-4

Age group

Per

cen

t

Secondary

55.3

31.7

0

10

20

30

40

50

6065

and

+

60-6

4

55-5

9

50-5

4

45-4

9

40-4

4

35-3

9

30-3

4

25-2

9

20-2

4

15-1

9

10-1

4

5-9

0-4

Age groups

Per

cen

t


At a regional level, Nicaragua displays the lowest primary school completion rates and the lowest average school attainment (with the exception of Guatemala and Haiti). Figure 6.21 indicates that other countries with rather similar income levels, such as Bolivia, Honduras, and Ecuador, have nearly universal completion for primary school. Nicaragua, however, is still below the regional average only surpassing Guatemala and Haiti.

100 In issues related to education, grouping by age helps to observe positive changes over time.

192

Figure 6.21. Primary Completion Rates [Nicaragua vs. LAC, period 1996-2004]

ARG

BRA

COL CRIDOM

ECU

GTM

NIC

PANPER

VENHND

BOL CHL

SLV

MEX

PRYURY

LAC

50

60

70

80

90

100

110

120

2,000 4,000 6,000 8,000 10,000 12,000


Co

mp

leti

on

Pri

mar

y R

ates

(%

)


D. CONSTRAINTS TO SCHOOLING

Lack of access and affordability are the main reasons why children ages 7 to 12 are not enrolled in primary school. Figure 6.22 illustrates the share of children ages 7 to 12 that do not attend primary school due to lack of money and due to lack of facilities/personnel. Lack of affordability constitutes the main reason why children are not enrolled in primary school. The share of children not attending school due to lack of money is higher among children who are poor than among children who are non-poor (43 vs. 35 percent). Indeed, lack of affordability seems not to be an exclusive problem of the poor, but one affecting children in the upper quintiles. This can be explained by the fact that richer households prefer to send their children to private schools, which are often associated with more expensive tuitions. As illustrated in Figure 6.23, while less than 1 percent of the overall costs paid by households in the bottom quintile for their children who attend primary school are related to tuitions, the same proportion is at 60 percent for households in the highest quintile. Lack of education facilities and personnel, on the contrary, constitute a more important reason keeping poor children away from primary school, and especially among children living in households engaged in agriculture (some of which live in isolated rural areas with little public infrastructure).

Figure 6.22. Affordability constraints are more common among middle-class children who are not enrolled in primary school than among poor ones

10.4

18.4 17.119.7

14.39.2

0.0

16.8

22.7

7.0

24.232.7

46.5

37.2

23.7

43.2

11.8

41.7

49.8

42.333.9

40.5

0

10

20

30

40

50

60

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r ho

useh

old

Ag.

pro

duce

r ho

useh

old

Per

cent

% Not enrolled in Primary because there is no place, no class, or school is too far

% Not enrolled in Primary due to lack of money


193

Figure 6.23: Tuitions constitute a more important share of the cost of primary education for the non-poor than for the poor.

0

10

20

30

40

50

60

70

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r ho

useh

old

Ag.

pro

duce

r ho

useh

old

Fem

ale

Mal

e

Per

cent

% Would come back to school if supplies w ere free% Would come back to school if there w ere scholarship progs.

% Would not come back to school


Lack of access to facilities/personnel is an important reason for not attending primary school among children living in the Central and Atlantic regions. Table 6.5 indicates that there is variation in the reasons why children are not enrolled in school across regions in Nicaragua. In particular, lack of access to school and long distances to the nearest school (i.e. school are too far) constitute an important reason explaining why children are not enrolled in primary school in rural areas (and especially in the Atlantic and Central regions101) while other reasons such as lack of interest and family problems are more recurrent in urban areas (and especially in the Pacific region) and in Managua.

Table 6.5: Reasons why children ages 7 to 12 are not enrolled in Primary school by region and strata.

RURAL URBAN MANAGUA PACIFIC CENTRAL ATLANTIC % No interest 8.4 15.3 5.6 17.9 11.3 7.1 % Had to work 4.7 0.8 0.0 1.9 4.5 4.4 % No place/no class/no teacher

6.8 2.2 0.0 2.2 6.9 7.3

% school is too far 15.8 0.0 0.0 0.0 9.6 23.6 % Family problems 6.0 16.3 31.7 8.0 6.5 4.6 % Lack of money 38.9 47.0 40.2 52.6 39.0 38.2 % Other 19.4 18.4 22.5 17.6 22.1 14.8


Comparing with 2001 data, results in Table 6.6 indicate that while the weight of economic reasons and access (i.e. proxied by distance) have decreased as factor for non-enrollment, lack interest and family problems have become more important reasons explaining why children 7 to 12 stay out of school. Note that the number or urban non-poor children claiming not to be in school due to lack of interest (family problems) increased from 14 (5) percent in 2001 to 22.3 (24) percent in 2005. Lack of interest and family problems also increased as reasons for non-enrollment among the urban poor, but significantly less than

101 Data from the 2005 LSMS indicate that 17 percent of all individuals in the Atlantic region benefited from investments on school infrastructure (either construction or improvement) between years 2001 and 2005. This proportion was larger in the Central and Pacific regions (29 y 35 percent respectively).

194

among the non-poor. Table 6.5 also indicates that the share of rural students claiming to be away from school due to extensive distance to school decreased from 23 to 7 percent among the non-poor. Note that the share of children between 7 and 12 years who claim to not attend school due to family problems displays a significant increase in Managua (from 8 to 32 percent). In this regard, a study conducted by MECD and Universidad de Córdoba (2004a) finds that about one third of the student in primary display low levels of self-esteem; about 1 out of every 4 children has been abused by their parents or have seen their mother being beaten by their fathers; 15 percent if all children claim not to have any friends; 7 percent claim to live in a households that is more violent that those in the neighborhood; and percent claim to have used drugs; among others.

Table 6.6. Lack of interest and family problems have risen in importance as a factor explaining school non-attendance among boys between ages 7 and 12.

% FAMILY PROBLEMS

% NOT INTERESTED

% DISTANCE 102 % MONETARY PROBLEMS

2001 2005 2001 2005 2001 2005 2001 2005 Managua 8.1 31.7 0.0 5.6 0.0 0.0 73.2 40.2 Rest Urban 3.6 6.2 14.6 20.0 1.5 0.0 58.4 54.5 Poor 4.8 5.1 14.2 14.2 9.4 10.5 43.4 40.9 Urban 1.8 2.1 13.0 18.6 0.0 0.0 73.6 50.3 Rural 5.8 6.2 14.6 12.7 12.7 14.2 32.9 37.5 Non-Poor 2.5 15.8 10.1 14.3 12.5 3.0 35.7 36.0 Urban 4.8 24.0 14.0 22.3 2.9 0.0 49.5 37.2 Rural 0.0 4.7 6.0 3.5 22.5 7.1 21.4 34.4 Source: World Bank using the 2005 Nicaragua LSMS.

About 20 percent of all children who are not enrolled in primary claim they would not come back to school. As illustrated in Figure 6.24, about half of all children who are not enrolled in primary school would resume their education if there were scholarship programs and if school supplies were given for free. This is especially true among children from vulnerable groups, such as indigenous and children living in households engaged in agriculture.

102 Gasparini et. al. (2007) claim that in Nicaragua, a significant proportion of individuals from the intermediate consumption quintiles (or even the most affluent quintiles) lack some kind of basic infrastructure, such as safe drinking water, sanitation, or access to a paved road. For this reason, investments in infrastructure often do not have a pro-poor character.

195

Figure 6.24: Roughly 50 percent of all children 7 to 12 who are not enrolled in primary would come back to school if there were scholarship programs and if supplies were free

0

10

20

30

40

50

60

70

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r

hous

ehol

d

Ag.

pro

duce

r

hous

ehol

d

Fem

ale

Mal

e

Per

cent

% Would come back to school if supplies w ere free% Would come back to school if there w ere scholarship progs.

% Would not come back to school


Work, lack of money and lack of interest constitute the main reasons why individuals are not enrolled in secondary or post-secondary school. Table 6.7 presents the main reasons why students age 13 to 23 are not enrolled in school. About 30 percent of all individuals (poor and non-poor) claim to be out of school because they need to work; about 25 to 30 percent (at all socio-economic levels) claim that lack of money is the main reason keeping them away from school; and about 16 to 20 percent (a considerable share) claim that they are not interested to be at school. Other reasons, such as family problems, pregnancy, and child care add up to about 12 percent

Table 6.7: Reasons why individuals ages 13 to 23 are not enrolled in secondary/post secondary school by socio-economic condition

Non Poor Poor Poorest Quintile

Q2 Q3 Q4 Richest Quintile

% No interest 18.19 20.92 20.83 21.68 18.91 17.85 16.61 % Finished studies/other 9.05 3.38 3.84 2.32 6.77 7.68 15.48 % Domestic work 4.07 5.42 5.82 4.87 5.02 4.33 2.33 % Had to work 28.29 28.63 32.88 25.04 24.59 29.71 31.09 % No place/school too far 2.33 3.23 3.19 3.47 2.84 2.45 0.73 % child care or pregnancy 9.48 8.45 7.27 9.74 8.07 10.96 9.94 % Family problems 3.22 2.21 1.89 2.19 3.55 2.61 4.11 % Lack of money 25.37 27.76 24.28 30.68 30.24 24.41 19.71 Source: World Bank using the 2005 Nicaragua LSMS.

Family problems, child care and pregnancy constitute important reasons why females are not enrolled in secondary or post-secondary school103. Figure 6.25 displays the reasons why individuals 13 to 23 years of age are not at school by gender. Results indicate that family related constraints –

103 According to MECD (2004), about 20 percent of all students in secondary education claim to have received insults from their parents and/or teachers; 70 percent have witnessed violent fights; 13 percent have witnessed a violent death; 5 percent to be a gang member; and 25 percent to have had sexual relationships (67 percent of which did not use any contraceptive).

196

pregnancy, child care and domestic work – constitute the main reason why about 34 of every 100 young ladies are not in school.

Figure 6.25: Factors that keep individuals away from secondary and post-secondary school differ significantly by gender

19.0

10.5

3.5 3.76.3

28.0

0.1 0.4

42.3

16.312.8

25.6

5.32.11.5

22.9

0

10

20

30

40

50

% c

hild

care

/pre

gnan

cy

% D

omes

ticw

ork

% F

amily

prob

lem

s

% N

o pl

ace/

no

clas

s/sc

hool

too

far

% F

inis

hed

stud

ies/

othe

r

% N

o in

tere

st

% L

ack

ofm

oney

% H

ad to

wor

k

Per

cent

Females Males


For young men, on the contrary, lack of interest and work-related constraints constitute the main two issues keeping them away from school. In fact, while 42 out of every 100 young men claim not to be at school because of work, only 13 out of 100 young ladies claim so. About 23 percent of all young men who are not at school claim that it is due to lack of interest. The same rate is at 16 percent for young ladies.

E. AFFORDABILITY

Relative to their income level, education is more expensive for the non-poor. Table 6.8 presents statistics on average expenditures on education among education users (households with a member using the education system). Results indicate that relative to their income, expenditures on education are higher for the non-poor than for the poor: they account for about 8 percent of the income of households in the richest quintile vs. 3 percent in the poorest quintile. However, relative to their non-food consumption (since poorer households spend relatively more on food) the share is similar for poor and non-poor households (averaging 8 to 10 percent). Tuitions and transport/allowances represent a higher economic burden for the non-poor, which is not surprising given that non-poor households are more likely to use more-expensive private education services and pay for transportation (generally in private-school buses). While expenditures on tuitions eat up about 0.47 (0.17) percent of all non-food expenditures (income) among households in the poorest quintile, the same proportion is at 4.6 (3.30) percent for households in the richest quintiles.

Transportation constitutes the main expenditure item in Nicaragua. Figure 6.26 illustrates differences in the composition of education expenditures by quintile. Expenditures in transportation and allowances account for 37 to 47 percent of all school expenses at all socio-economic groups. As expected, school fees represent a much higher share of overall expenditures on education for households in the upper quintiles (since poor kids usually attend tuition-free public schools). On the contrary, expenditures on non-tuition items are larger among the poor. In particular, while expenditures on books, supplies, and uniforms account for about 40 to 55 of all expenditures education for households in the bottom two

197

quintiles, they account for about 13 to 20 percent for households in the upper two quintiles. This is not surprising as poor households usually have their children enrolled in public education (generally at the primary level) where tuitions are heavily subsidized but non-tuitions items are generally paid out-of-pocket. On the contrary, richer households are likely to invest more on private school tuitions especially for post-secondary education and preschool. Indeed, tuition-related expenditures represent about 50 percent of the overall expenditures on education for households in the highest quintile (39 percent school fees, 6 percent preschool, and 5 percent pre-registration fees).

Table 6.8: Expenditures in education relative to income and non-food consumption [education users only: households with children in the system]

Poorest Quintile

Q2 Q3 Q4 Richest Quintile

Total Education Expenditures As % of total Income 2.69% 4.08% 5.49% 6.68% 7.53% As % of total non-food consumption 7.22% 8.17% 8.88% 9.13% 10.49% Tuitions As % of total Income 0.17% 0.36% 0.91% 1.71% 3.30% As % of total non-food consumption 0.47% 0.72% 1.47% 2.33% 4.60% Transport As % of total Income 1.00% 1.58% 2.47% 3.14% 2.74% As % of total non-food consumption 2.68% 3.17% 4.00% 4.29% 3.82%


Figure 6.26: The burden of non-tuition items in overall expenditures on education, such as books and school supplies, is higher for the poor

2.6% 2.7% 4.1% 6.1%7.0%

13.7%

21.8%

38.5%

2.8%

3.7%

5.4%

32.4%21.1%

17.6%

12.6%

5.6%20.3%

12.3%10.4%

7.5%

4.4%

1.5%

16.4%7.7% 3.3%

3.8%

37.1% 38.7%45.0% 47.0%

36.4%

2.2%

4.8%

1.9%1.6%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 5

Books

supplies

Uniforms

Pre-Registration

School Fees

Preschool

Transport,allowances,and other

Source: World Bank using the 2005 Nicaragua LSMS. Sample: households having one or more users in the education system.

198

Figure 6.27: For poor households with potential users, paying for tuition for tertiary education would require investments equivalent to about 50 to 70 percent of their total per-capita income

46.4%

30.2%24.3% 23.6%

18.5%

124.7%

60.4%

39.3%

32.2%

25.8%

0%

20%

40%

60%

80%

100%

120%

140%

Poorest Quintile Q2 Q3 Q4 Richest Quintile

Tuition per student as % of total per-capita income

Tuition per student as % of total per-capita non-food expenditure

Source: World Bank using the 2005 Nicaragua LSMS [all units in per-capita per year]

Tuitions for tertiary education are not affordable by the poor and are hardly affordable by households in the middle class. Figure 6.27 calculates the average cost of the yearly tuition among students currently enrolled in tertiary education and divides it by the average per-capita non-food consumption (and income) by quintile (the sample includes households having at least one member between 17 and 23 years of age). For every student enrolled in tertiary education in the bottom three quintiles, a household needs to invest from 24 (40) to 50 (124) percent of their per-capita income (non-food consumption). Therefore, tuitions for tertiary education are not affordable by the poor and are hardly affordable by the middle class. For households in the upper quintiles, tuitions for tertiary education become somewhat more affordable but still constitute a significant investment (equivalent to 19 to 24 percent of their overall per-capita income). As suggested by Figure 6.28, tuitions for tertiary education are the main expenditure item for students enrolled in tertiary education at all socio-economic levels. This explains the low participation of poor families in the Nicaraguan university education system, in contrast to the significant economic resources allocated to public university education (C$1.051 billion cordobas in 2005)104.

104 For more details on tertiary education financing in Nicaragua see Porta (2004) and World Bank (2003). For analysis on incidence of education expenditure in Nicaragua see Gasparini et. al. (2007).

199

Figure 6.28: Tuitions are equivalent to 74 to 90 percent of all expenditures in education for students enrolled in tertiary education at all socio-economic levels

84.99 87.6981.33 81.13

89.20

15.01 12.3118.67 18.87

10.80

0%

20%

40%

60%

80%

100%

Poorest Quintile Q2 Q3 Q4 Richest Quintile

Tuition

Non-Tuition


INEQUITIES IN QUALITY

The majority of children in Nicaragua are enrolled in public schools (72 percent in public not-autonomous schools and 20 percent in public autonomous ones, see Box 6.2 bellow). Only a very small fraction of all children is enrolled in private schools. In Managua, however, the relative importance of autonomous schools and private schools is much larger, reaching 40 and 14 percent respectively. There is also a much larger fraction of children in private schools among the richest quintiles, reaching 30 percent. As opposed to what happens in other school systems in Latin America, public schools do serve children from the richest quintiles. Figure 6.29 shows that the share private enrollment in primary and secondary education in Nicaragua is aligned with regional standards.

Table 6.9: Type of primary school by region, strata and socio-economic group

% Public, not

autonomous

% Autonomous school/center

% Community education

facility

% Private with

voucher

% Private without voucher

% Multigrad

o

Region Rural 84.02 12.42 2.65 0.60 0.31 60.81 Urban 57.00 25.53 0.44 4.67 12.36 5.08 Managua 44.12 35.72 0.31 4.29 15.56 9.04 Pacific 70.65 19.97 0.30 2.53 6.55 21.94 Central 82.11 12.25 1.86 1.73 2.05 51.84 Atlantic 79.00 11.16 4.50 2.28 3.07 46.23

Socio-economic Condition Non Poor 57.04 23.81 1.01 5.15 12.99 20.25 Poor 82.45 14.54 2.07 0.44 0.50 45.56 Poorest Quintile 83.84 13.45 2.53 0.00 0.18 55.12 Q2 84.79 12.73 1.71 0.67 0.10 39.90 Q3 69.40 25.15 1.32 1.44 2.70 30.78 Q4 61.10 23.22 1.34 4.41 9.93 16.64 Richest Quintile 33.87 22.74 0.29 11.19 31.91 12.36

200

Vulnerable group Indigenous 80.24 9.74 2.71 3.51 3.79 28.25 Ag. Prod.* 83.07 13.24 2.87 0.23 0.59 60.51

Total 71.16 18.66 1.60 2.54 6.05 34.34 Source: World Bank using the 2005 Nicaragua LSMS.

Figure 6.29: Private Enrollment Share, Primary and Secondary Rates [Nicaragua vs. LAC, 1999-2004]

ARG

BRA

CHL

COLDOM

ECU

URYBOL

CRISLV

GTM MEX

NIC

PAN

PRY

PERVEN

LAC

0

10

20

30

40

50


Priv

ate

Enr

ollm

ent S

hare

(%

), P

rimar

yR

ates

ARG

CHL

CRI

GTM

PANBOL

BRA

COL

DOMECUSLV MEX

NIC PRY

PERURY

VEN

LAC

0

20

40

60

80


Priv

ate

Enr

ollm

ent S

hare

(%

),S

econ

dary

Rat

es


Children living in poorer households are more likely to be enrolled in multigrado schools105. The ministry of education in Nicaragua defines multigrado schools as those having fewer teachers and classrooms than the number of grades offered. Generally, multigrado schools have high student-teacher ratios, which obligate teachers to spend less time teaching and interacting with students. As a consequence, multigrado schools are likely to offer a somewhat lower quality of learning than normal schools do (as lower student/teacher ratios are often associated with better quitlity).106 Results in Figure 6.30 indicate that about 6 of every 10 children in rural areas are enrolled in a multigrado schools and that there is a much higher concentration of this type of schools in the Atlantic and Central regions as compared to Managua and the Pacific. Also, the figure illustrates that poor children and especially those living in a household engaged in agriculture, are more likely to be taking classes in multigrado schools. On the contrary, children living in richer households are more likely to be enrolled in private primary schools, often associated with better quality of education.

There is not so much variation in the type of secondary school children attend across socio-economic groups. Table 6.10 indicates that most children enrolled in secondary take classes in public autonomous and non-autonomous school. As expected, children from richer households are more likely to be enrolled in private secondary schools but differences in private school enrollment rates between poor and non-poor children are not as large as in primary and preschool education. By region, results indicate that the share of children attending secondary autonomous and private schools is larger in Managua as compared to other regions, while in the Atlantic and central regions (and especially in rural areas) there is a larger share of students attending public non autonomous schools.107

105 According to the MECD about 40 percent of all multigrado schools do not offer courses beyond 4th grade; which leaves some of their students out of the education system afterwards. 106 Results on test-scores from the 2002 education quality survey conducted by the Ministry of Education do not indicate that Multigrado schools display lower scores than non-multigrado schools. However, results in test scores in Nicaragua did not display much variation as a whole (more on this on section 2.B bellow). 107 The Atlantic and Central regions have a lower share of autonomous schools due to the pre-defined criteria to consider schools as autonomous (usually large and financially sustainable schools). Furthermore, the regional

201

Figure 6.30: Six out of every 10 children living in households engaged in agriculture attend a “multigrado” primary school

21.2

44.8

54.6

37.7

29.8

17.712.6

36.4

28.9

14.2

60.5

0

10

20

30

40

50

60

70N

on P

oor

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r ho

useh

old

Ag.

pro

duce

r ho

useh

old

Per

cen

t in

Mu

ltig

rad

o -

Pri

mar

y

0

10

20

30

40

50

Per

cen

t in

Pri

vate

sch

oo

l - P

rim

ary

% Multigrado % Private


Table 6.10: Type of Secondary school by socio-economic group % Public, not

autonomous % Autonomous

school/center % Private, with

voucher % Private, no

voucher Socio Economic Condition

Non Poor 24.57 45.51 10.31 19.61 Poor 43.51 47.47 5.87 3.16 Poorest Quintile 43.88 49.24 5.58 1.29 Q2 42.58 47.74 6.52 3.16 Q3 37.00 49.06 6.39 7.55 Q4 26.25 48.65 8.76 16.33 Richest Quintile 17.28 38.91 13.89 29.93

Vulnerable groups Non-Indigenous 29.74 47.15 8.76 14.35 Indigenous 45.41 24.42 12.55 17.62 Non Agric. producer household 26.75 45.54 9.80 17.91 Agric. producer household 40.78 47.73 6.50 4.99

Total 29.74 47.15 8.76 14.35 Source: World Bank using the 2005 Nicaragua LSMS.

Private preschools (generally associated with better learning indicators than public ones) are less accessible to the poor. As Figure 6.31 illustrates, the majority of all children enrolled in preschool attend public facilities run by the Ministry of Education. Private enrollment at preschool is at 26 to 49 percent for households in the richest quintile while it reaches only 2 to 9 percent among children living in more vulnerable households, such as those in the poorest quintiles, indigenous, and those living in households

government in the Atlantic region has opposed that schools become autonomous. 2002 data on test scores does not find significant differences in quality between autonomous vs. Non-autonomous schools

202

engaged in agriculture. Table 11 indicates that private preschools are associated with slightly better learning indicators: the time children spend at private preschools is more than the time they spend at public schools (roughly half an hour more per day), 96 percent of the children in private preschool are taught vs. 90 percent in public ones and 40 percent of all children in private schools claim to receive care vs. 32 percent in public ones. Nevertheless, private preschools do not provide other non-learning services (such as healthcare) as often as Communal and other public preschools do. Access to preschool education is generally associated with better outcomes during primary and post-primary school (such as lower repetition rates and higher permanence). In Nicaragua 4 out every 10 children have the opportunity to attain preschool education before entering first grade. As illustrated in Figure 6.29, this share is bellow Latin American given Nicaragua’s level of development, and lower than that in other countries with similar levels of income such as Ecuador and Bolivia.

Figure 6.31: The majority of the children enrolled in preschool at all socio-economic levels attend public facilities

3.8 2.26.4 6.4

25.7

46.5

19.0

4.6

25.3

3.1

28.6

0

10

20

30

40

50

60

70

80

90

100

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-In

dige

nous

Indi

geno

us

Non

Agr

ic.

hous

ehol

d

Agr

ic.

hous

ehol

d

Per

cent

chi

ldre

n (4

to 6

) en

rolle

d in

pr

ivat

e pr

esch

ool

0

10

20

30

40

50

60

70

80

90

Per

cent

chi

ldre

n (4

to 6

) en

rolle

d in

a

ME

CD

pre

scho

ol

MECD

Source: World Bank using the 2005 Nicaragua LSMS. [MECD schools are public schools run by the ministry of education]

Table 6.11: Children enrolled in private preschools generally spend more time being taught % children who get

healthcare % children who get

taught % children who get

care Hours at preschool

MIFAMILIA 8.24 80.82 38.15 3.38 MECD 2.61 94.06 31.35 3.43 Private 3.93 94.65 39.75 3.88 Communal 18.08 91.84 31.08 3.72 Source: World Bank using the 2005 Nicaragua LSMS.

203

Box 6.2. Autonomous schools in Nicaragua

As part of the process of reform to improve efficiency and effectiveness of service delivery in Nicaragua, the so called autonomous schools were introduced in the in the education sector in 1993. Greater participation and decision-making among parents and teachers was regarded as central to this end. The main difference between autonomous and non-autonomous schools is that the former sets a participative management structure whereby parents, teachers, students (only at the secondary level), and school directors participate in decision making on general management and budget allocation. Autonomous schools divide responsibilities among different actors, mainly the MECD, the municipal delegate of the MECD, and School councils (consejos directivos).

It is important to note that the data on autonomous schools obtained from household surveys, differ form official data. According to MECD, Autonomous schools in Nicaragua show an important increase in enrollment, the number of public autonomous schools has doubled during the last 6 years, together with enrollments. By 2006, 70 percent of all public schools were autonomous, covering 83 percent of the students in the pubic system. This discrepancy might be related to the fact that while parents know whether or not their children attend a private versus public school, within the latter, in many cases they are not aware of the status of the school. Much better statistics will be needed to have a clearer picture of the type of student these schools are catering.

Source: Nicaragua Poverty Assessment; World Bank (2007)

Basic Statistics Autonomous schools 2001 2002 2003 2004 2005 2006 Number of public “autonomous” schools

2,952 2,978 3,033 4,064 4,108 5,211

Number of students registered in autonomous schools

697,297 748,293 755,425 903,739 926,876 1,012,663

% of public “autonomous” schools 50 49 47 62 61 69 % of students registered in autonomous schools

68 69 68 79 79 83

Source: MINED 2006

To date, there have been 4 evaluations of the impact of autonomous schools (see Arcia, Porta, and Laguna, 2004 for the most recent one). Results indicate that a) autonomous schools have a slight but significant impact of student performance (as proxied by test-scores), but such an impact is highly dependent on the quality of the teachers as well; b) there is linear relationship between performance in Spanish and the years that schools have been autonomous both in primary and secondary; c) autonomous schools displayed lower dropout and repetition rates up to year 2004. Furthermore, autonomous schools have more decision power in regards to investments on infrastructure, administration, and evaluation/supervision of teachers. Directors, teaches, and members of the school council perceive the autonomous regime as positive and beneficial for the education sector.

204

Figure 6.32: Only 4 out of every 10 children in Nicaragua have attained pre-school education before entering to first grade [Nicaragua vs. LAC, period 1999-2004]

ARG

BOL

CRI

ECU

GTM

NIC

PANPRY

URY

LAC

20

40

60

80

100

2,000 4,000 6,000 8,000 10,000 12,000


New

en

tran

ts t

o p

rim

ary

edu

cati

on

w

ith

EC

CE

exp

erie

nce


A. QUALITY AND SERVICES

Public Primary schools are important sponsors of nutrition programs, especially in poorer regions. Figure 6.32 illustrates that about 81 percent of all children enrolled in primary school in rural areas and about 56 percent in urban areas receive food at Primary School. This share is higher in the pacific, central, and Atlantic regions (averaging 70 to 80 percent) than in Managua (47 percent). Poor children benefit more from food provision at school. Figure 6.33 indicates that 77 of every 100 children enrolled in primary in the bottom quintile get food at school while only 39 out of every 100 in the highest quintile do so. As expected food transfers are more relevant to the poor. Poor families with children enrolled in primary schools claim that food transfers per student per month are equivalent to up to 40 percent of their total per-capita income per month. As expected, food transfers, albeit significant, represent a lower share of total income per-capita for the non-poor (about 15 percent)

Figure 6.32: About 81 percent of all students in rural areas receive food at primary school

82.3

54.747.4

79.3

71.1 72.5

0

20

40

60

80

100

Rural Urban Managua Pacif ico Central Atlantico

Per

cent

% Children getting Food at Primary School Cost of food given every month as % of total monthly income per capita


205

Figure 6.33: Food transfers per student per month given to primary school children in the bottom quintiles

are equivalent to 40 to 50 percent of their total per-capita household income

56.7%

76.8%69.6%

56.7%

38.6%

69.9% 68.8%

58.9%

77.9% 79.8% 82.2%

0%

15%

30%

45%

60%

75%

90%

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r ho

useh

old

Ag.

pro

duce

r ho

useh

old

Per

cent

0

20

40

60

80

100

Per

cent

% Children getting Food at Primary School Cost of food given every month as % of total monthly income per capita


The student-teacher ratio is lower in primary and secondary education for private schools, particularly among those that do not receive voucher. Urban/rural, nor regional differences are marked. Contrary to what could be expected, the ratios are higher for autonomous schools than for non autonomous ones. The overall student-teacher ratio has been recently falling at private institutions, reaching 24.7 and 23.1 in primary and secondary. Among public institutions, on the country, the student-teacher has been increasing, reaching 36.3 in primary and 39.7 in secondary.

Table 6.12: Ratio Pupil/Teacher by Region and Area

PRIMARY SECONDARY Auto-

nomous No

Auto-nomos

Private with

subsidy

Private with-out subsidy

Total Auto- nomous

No Auto-nomos

Private with

subsidy

Private with-out subsidy

Total

National 36.9 34.5 30.2 19.2 33.6 43.3 30.8 27.5 21.4 33.8 Rural 35.8 34.4 33.3 20.7 34.8 37.1 27.1 24.9 22.1 32.1 Urban 38.7 34.7 28.5 19.0 32.0 45.9 33.9 28.1 21.2 34.4

Zone Managua 42.9 36.5 29.6 19.1 32.2 47.8 30.4 27.7 19.4 32.9 Pacific 35.3 37.3 24.5 16.5 29.6 37.6 27.6 30.3 21.6 32.4 Central 35.5 33.6 28.5 19.5 33.9 40.7 33.1 24.4 22.8 34.4 Atlantic 37.6 35.6 33.1 22.9 35.5 31.7 29.0 22.4 22.3 26.8

Source: Author's own elaboration based on the 2005 Teacher's Labor Force Census and Initial Enrollment statistics for year 2005

Nicaragua is the Latin American country with the highest pupil-teacher ratio in the region, both in primary and secondary schools. Compared to regional standards, a Nicaraguan primary (secondary) teacher attends, on average 35 (32) students. This is higher than the corresponding regional average of 27 (21) students per teacher. While lower student teacher ratios are generally associated with better quality of education; Hanushek (1995) finds that the impact of this variable in quality might be low under some circumstances. Given Nicaragua’s scarce resources, its current student-teacher ratio seems to be a

206

reasonable one. As it will be discussed below, at this point in time, it is preferable for Nicaragua to invest on increasing the quality of the existing stock of teachers than on expanding the teaching force.

Figure 6.34: Pupil/Teacher Ratio Primary and Secondary [Nicaragua vs. LAC, period 1999-2004]

ARG

BOL BRA

CHL

COLCRI

DOM

ECU

GTM

MEX

NIC

PANPRY

PER

URY

LAC

10

20

30

40


Pu

pil/

Tea

cher

Rat

io P

rim

ary

BOLBRA

CHL

COLCRI

DOM

ECUGTMMEX

NIC

PANPRY

PERURY ARG

LAC

0

10

20

30

40


Pu

pil/

Tea

cher

Rat

io S

eco

nd

ary


Nicaragua’s teacher work force is less qualified than expected given the country’s level of development. As illustrated in Figures 6.35, Nicaragua has the lowest share of trained teachers in the Latin American region, especially in secondary education. Data suggest that 25 of every 100 of teachers in primary are not properly trained to teach, whereas the same proportion reaches more than 50 percent in secondary. Teacher training is likely to influence student’s learning, especially in secondary education.

Figure 6.35: Percentage of Primary and Secondary Teachers Trained [Nicaragua vs. LAC, 1999-2004]

BOL CRI

DOM

GTM

NIC

LACECU

PAN

60

70

80

90

100

2,000 4,000 6,000 8,000 10,000 12,000


Pri

mar

y E

du

cati

on

, Tea

cher

s (%

T

rain

ed)

BOL

ECU

GTM

NIC

CRIDOMPAN

LAC

40

60

80

100

2,000 4,000 6,000 8,000 10,000 12,000


Sec

on

dar

y E

du

cati

on

, Tea

cher

s (%

Tra

ined

)

Source: International Education Statistics (2007). Data on teachers trained displays the 1999-2004 average.

About one fifth of all children enrolled in secondary education do not have books. As illustrated in Figure 6.36, about 23 percent of all children enrolled in secondary school and about 4 percent of all children enrolled in primary school do not have books. The share of students without books in primary school is slightly higher in rural areas than in urban areas. The Central region displays a lower than average share of children without books in secondary education, while the opposite occurs in the Atlantic region. MECD (2006b) argues that these positive outcomes in terms of access to books are explained by the fact that between 2002 and 2006 a total of 7.1 million text books and notebooks were distributed to children between 1st and 6th grades (in regular and multigrado schools).

207

Figure 6.36: Managua and the Atlantic region display the highest share of students with no access to books

22.2 23.1

28.1

23.3

19.1

28.9

0

5

10

15

20

25

30

35


% enrolled in Primary w ith no books % enrolled in Secondary w ith no books


Indicators for primary display higher variation by socio-economic quintile than in secondary. Table 6.13 indicates that the percentage of students without books is much higher in secondary than in primary (5 vs. 23 percent). Furthermore, students from poor households have slightly higher school absence rates (for primary and secondary) than students from households in the higher quintiles.

Table 6.13: Indigenous children are less likely to “not” have books for Primary Education

Primary Secondary % enrolled with

no books Days absent % enrolled with

no books Days absent

Socio-economic group Non Poor 4.5 4.6 25.6 3.5 Poor 5.0 4.0 21.0 3.0 Poorest Quintile 6.0 4.1 23.2 3.9 Q3 4.0 4.6 23.6 2.7 Richest Quintile 4.8 4.0 25.2 3.3

Vulnerable group Indigenous 11.5 4.1 14.8 3.6 Agric. producer household 5.7 4.3 25.5 2.3


About 20 to 25 percent of all parents with children in primary school consider that their education is either regular or bad. Figure 6.37 and Table 6.14 display perceptions on quality of education gathered from households with children enrolled in primary education. Households in the poorest quintiles are less likely to rate their children’s education as excellent as compared to households in the highest quintiles. Indigenous households and those engaged in agriculture are less likely to rate their children’s education as excellent and more likely to rate it as regular or bad. Finally, households in Managua and in urban areas are more like to consider that their children’s education is excellent as compared to those in other regions and in rural areas.

208

Figure 6.37: Non-poor households are more likely to rate their children’s education as excellent.

13.0

6.95.1

8.5 7.7

12.59.8

6.5 7.1

20.8

11.512.7

19.8

26.124.2

20.219.7

22.1

18.8

20.920.020.5

0

5

10

15

20

25

30

Non

Poo

r

Poo

r

Poo

rest

Qui

ntile Q2

Q3

Q4

Ric

hest

Qui

ntile

Non

-Ind

igen

ous

Indi

geno

us

Non

Ag.

pro

duce

r ho

useh

old

Ag.

pro

duce

r ho

useh

old

Per

cent

% Excelent % Regular or bad


Table 6.14: Indicators on household subjective perceptions about the quality of education of their children [Primary school]

% Excellent % Good % Regular % Bad By strata

Rural 6.82 71.03 21.3 0.84 Urban 12.63 69.25 17.59 0.52

By region Managua 13.26 62.53 24.05 0.16 Pacific 8.14 69.82 20.68 1.36 Central 10.27 73.87 15.6 0.26 Atlantic 6.67 71.56 20.74 1.03


Households in the central region display better perceptions about the quality of their children’s secondary education than in other regions. Figure 6.38 and Table 6.15 present differences in household perceptions in regards to the quality of education of their children who are enrolled in secondary. Households in the Atlantic region and in rural areas less likely to rate their children’s education as excellent as compared to households in other regions and in urban areas. Contrary to what occurs in the case of primary education, households in urban areas and in Managua are more likely to consider that the quality of their children’s secondary education is regular or bad as compared to households in other regions and in rural areas. Poor households and households living in households engaged in agriculture are associated with less favorable perceptions about the quality of their children’s secondary education (see Table 6.14).

209

Figure 6.38: One of every 4 children enrolled in secondary education in Managua think that their education is either regular or bad

15.9

18.7

16.114.9

25.6

12.5

11.29.2

16.0

16.7

17.315.8

0

5

10

15

20

25

30


Per

cent

% Regular or bad % Excelent


Table 6.15: Indicators on household subjective perceptions about the quality of education of their children [Primary school]

% Excellent % Good % Regular % Bad By socio economic group

Non Poor 12.98 66.52 19.69 0.82 Poor 6.88 73.11 19.42 0.59 Poorest Quintile 5.09 76.11 18.26 0.54 Q2 8.45 71.4 19.43 0.72 Q3 7.73 68.05 23.72 0.5 Q4 12.52 65.35 21.05 1.09 Richest Quintile 20.75 66.59 11.98 0.68

By vulnerable group Non-Indigenous 9.81 70.38 19.1 0.71 Indigenous 6.46 67.48 25.64 0.41 Non Agric. producer household 11.5 68.81 18.96 0.73 Agricultural producer household 7.07 71.99 20.3 0.64


B. QUALITY BASED ON TEST SCORES108

Less than 14 percent of all students in 3rd and 6th grade are found to be proficient in their curriculum. This sub-section is based on results from the education proficiency quality survey (i.e. academic tests) conducted by the Ministry of Education of Nicaragua in year 2002. Test results indicate that between 60 and 90 percent of all students in 3rd and 6th grade have a basic (of bellow than expected) knowledge about their curriculum (mathematics and Spanish). Only a minority (10 to 25 percent) of the student population was found to have normal or proficient knowledge on their curriculums. Test scores

108 We are especially grateful to the MECD’s Division on the Evaluation of Policies, Programs and Projects for their valuable collaboration in supplying the data necessary to produce this section.

210

indicate that proficiency rates were generally higher in Managua and in the Central region with the exception of mathematics among 6th graders (which was roughly similar across regions).

Figure 6.39. Less than 6 (1) of every 100 student in 6th grade is found to be proficient in Spanish (mathematics)

Resultados por Nivel de Rendimiento Académico - Pruebas 2002

71.261.7

69.7

88.1

21.1

24.4

25.2

10.87.713.9

1.15.1

0%

20%

40%

60%

80%

100%

Español Matemáticas Español Matemáticas

3er grado 6to grado

Básico Intermedio Proficiente

Source: Tests of Academic Performance (MECD 2002)

The lowest levels of knowledge acquisition are found in rural areas, in multigrado schools, among girls, among grade repeaters, and among those students who speak a language other than Spanish. Table 6.15 indicates the learning gaps demonstrated in academic performance tests; it should be noted that the lowest percentages of students with levels of knowledge below the minimum level set by the MECD are found among students attending private subsidized schools, and also attending schools in urban areas of the Managua and Central regions, and among students whose parents have attended university or graduate school. In this respect, Box 6.3 presents the results obtained from an analysis of factors associated with academic performance.

Box 6.3: Factors associated with academic performance in Nicaragua

Results of standardized tests in Spanish and applied mathematics during 2002 were used to analyze internal and external factors associated with academic performance of students in the 3rd and 6th grades in Nicaragua. The main findings show that the most important factors associated with improvements in academic performance are: the principal’s pedagogic leadership, teacher motivation, high education levels among teachers, safe school facilities, as well as student and family motivation. In contrast, aspects such as grade repetition, child labor, school absenteeism and speaking a language other than Spanish have a negative impact on the academic performance of Nicaraguan children.

With respect to school administration, it appears that private subsidized schools have shown the best results, regardless of the good student effect (self selection), and the student’s socio-economic background.

With regard to the implementation of Computer Technologies, it appears that increasing access to computers and Internet connection systems has mixed results. While the results in the 3rd grade are highly positive, there was virtually no result whatsoever at the 6th grade level.

Source: Arcia, Porta and Laguna (2004)

211

Table 6.15: Percentage of students with lower knowledge levels than the established Minimum Level

Spanish Mathematics Spanish MathematicsNational 71.2 61.7 69.7 88.1 Geographic area

Urban 66.4 62.2 64.3 86.4Rural 75.5 61.2 77.4 90.6

RegionManagua 66.9 63.6 63.8 87.1Managua Urban 62.7 59.0 60.5 85.4Managua Rural 79.5 77.4 74.8 93.1Pacific 72.3 69.2 71.5 89.6Pacific Urban 70.9 67.6 68.4 88.8Pacific Rural 73.5 70.5 75.2 90.5Central 71.1 53.3 70.7 86.4Central Urban 63.8 58.6 61.8 83.4Central Rural 75.4 50.2 79.3 89.4Atlantic 77.9 61.6 81.1 91.8Atlantic Urban 77.7 69.7 75.7 92.7Atlantic Rural 77.8 58.7 86.1 91.3

Type of schoolPublic non-autonomous 77.6 64.6 74.0 89.7Private with subsidy 57.7 55.3 53.3 75.5Private without subsidy 59.6 55.1 50.4 84.5Public autonomous 71.7 62.0 74.0 90.1

Mode (Modalidad)Regular 69.6 63.7 67.0 87.7Multigrado 76.1 55.5 84.3 90.5

Shift (Turno)Morning 71.2 60.5 74.6 87.7Afternoon 71.4 65.0 65.3 88.6

Student's gender (sexo)Male 73.4 60.3 68.8 87.0Female 69.0 63.1 71.3 90.2

Another language 79.0 67.2 74.8 88.3Spanish 70.7 61.4 69.4 88.1

Repeater (Repitente)Repeater 78.5 70.0 76.6 91.3Non Repeater 69.7 60.0 69.3 88.0

Over-ageOver-age 73.4 59.8 79.0 91.9Normal 69.1 63.6 62.0 85.1

Parent's education levelNo studies (sin estudios) 71.6 60.3 68.9 88.5Adults education 74.7 61.0 77.9 89.6Primary 74.0 62.7 74.2 89.7Secondary 68.1 63.8 66.5 87.5University 51.1 49.7 53.2 80.9Postgraduate 50.0 37.5 54.8 70.0

Speaks another language than Spanish

3rd level (grado) 6th level

Source: MECD 2002 Academic Performance Tests.

212

CONCLUSIONS AND KEY MESSAGES

As a result of the overall analysis presented in this chapter, it is worth noting that the education system in Nicaragua faces several challenges that must be faced in a systemic way and require important investments in key sub-sectors, such as preschool education (as an important mechanism for human development) and secondary school (as an important mechanism to prepare students for joining the labor market). While most sector investments and external aid has been focused on improving and enhancing primary education in recent years, a more comprehensive approach is needed to build a sustainable education system in the medium and long run. In this respect, it becomes necessary to revise and adapt the curriculum in order to provide young Nicaraguans with the skills necessary to ensure that are able to perform and get involved successfully in the labor market. Next, we present the main conclusions and recommendations of this report:

- Education outcomes in Nicaragua have significant links with poverty. Investing in education is very profitable for individuals. Indeed, estimates indicate that a Nicaraguan is expected to earn 10 percent higher wages for each additional year of schooling attained. However, despite all the advantages that education has to offer, simple projections show that probably 20 percent of all children today will not finish primary and 45 percent will not finish secondary school. Many of the children currently enrolled in the system, as well as the current cohorts of youth that have already passed the years of primary and secondary schooling and are joining the labor force, have accumulated very little human capital. Despite larger social returns to investments in basic education, adult education and technical training will continue to be a challenge and a need for Nicaragua. Without investments focused on improving the quality of education and on increasing access to schools, 50 percent of the population is expected to remain in poverty during the next decade. The low education levels attained over the years by the Nicaraguan population explains, in large part, the labor force’s low levels of productivity and the poor quality of work.

- Education programs need to be oriented towards increasing school access and attainment as well as towards reducing drop-out rates, especially among the most disadvantaged sectors of the population and in rural areas. Even though there has been progress in increasing the percentage of Nicaraguan children enrolled in school, there are still significant enrollment gaps between socio-economic groups and regions that negatively impact the poorest population. Only 2 of every 10 children in Nicaragua’s from the poorest segments of the population attain complete primary school. While most children enter the system, a significant share drops out early from the system due to access and affordability constraints. In this sense, it becomes necessary to implement and develop programs to stimulate enrollment and permanence, where students are accompanied and monitored during their education life-cycle.

- Both supply-side limitations as well as affordability constraints hamper access to school. While lack of access to facilities and financial constraints constitute important reasons why poor children do no attend primary school (especially in the Central and Atlantic regions), lack of interest and family problems have risen in importance as factors explaining school non-attendance among urban children. Work, lack of money, and lack interest are the main reasons for boys not to be enrolled in secondary/post-secondary school; family problems, child care, and pregnancy are the main reasons for girls not to be enrolled. For children in the bottom two lowest quintiles, tuition costs are not a constraint precluding attendance since school fees are generally low or free. Nevertheless, out of pocket expenses related to sending children to school such as transportation, uniforms, and supplies are not always affordable to the poor. On the supply side, decisions on investments on infrastructure should be considered with care in order to maximize the used of some school infrastructure that remains underused (many schools, for instance, are able to open evening education programs). Investment should prioritize full access to primary schools nationally (Elvir, et.al., 2006). These investments should not be devoted only to increases access to facilities, but also to textbooks,

213

supplies, and qualified teachers; especially among the poorest segment of the population. The education sector may consider investing in programs to lower transportation costs for poor students, such as subsidized public bus fares in some urban centers. Also, scholarships targeted to the poorest segment of the population could be created to offset some non-tuition expenses (such as CCTs). Primary school fees, levies, and contributions should be minimized and preferably fully abolished.

- Improving the quality of teaching is a priority for generating better results in education. Laguna (2005) finds that the public sector can potentially eradicate the current share of untrained teachers by 69 in the short and medium run by instituting the “right” training programs. Training programs should include up-to-date pedagogical education methods and techniques linked to the curriculum and adapted to needs of the new world and which promote the use of technology resources for teaching. Untrained teachers are on average younger than certified teachers; this provides an opportunity to develop policies to target these teachers, which are likely to remain in the system for a longer period of time. As such, a professional training strategy needs to be defined, planned, and implemented; while also promoting economic incentive programs into practice aimed at keeping the best teachers within the educational system109.

- There are substantial inequities in access and quality of preschool, secondary and post secondary education between richer and poorer households, urban and rural areas, and regions. Inequities are smaller for primary education. Late enrollment, high dropouts, and high repetition rates altogether are preventing children, and especially those from poor families, to completing primary and secondary education. Young individuals who are poor, indigenous, and who live in households engaged in agriculture attain less than 5 years of education on average during their education life-cycle. As such, it becomes necessary to design and implement policies that are geographically focused, such as allocating and (or) attracting more and better teachers (or providing training to the existing ones) in rural areas and in poorer regions.

- Improving access and quality or early education programs should be a key priority to lower primary repetition rates and to improve education quality. There is abundant literature arguing the importance of early education in children’s intellectual and emotional development. Preschool education can make a difference throughout children’s life (Reimers, 1992; Young, 1996; Carneiro and Heckman, 2003; Schweinhart, 2004; among others), particularly in countries with high levels of social and economic inequality, where it could offset some of the unfavorable pre-determined conditions at birth faced by the poor (MECD, 2005). As such, the government should implement a set of policies related to early childhood to improve and expand existing infrastructure and to generate incentives to enhance quality and preparation of teachers at this education level. Doing so will probably contribute to decrease drop-out and repetition rates in primary and secondary education in the medium and long run. Furthermore, high quality preschool programs can benefit children by improving their nutrition, by helping them develop basic cognitive skills, and by inculcating in them at an early age intellectual curiosity and the desire to learn (Reimers, 1992). Furthremore, the government should strengthen the first three grades of the primary education cycle by (i) harmonizing the curriculum in these three grades to emphasize reading, writing and mathematics logic and comprehension, (ii) assigning the most experienced teachers to the first three grades, and (iv) ensuring an adequate supply of classroom and learning materials.

- Efforts should be targeted towards eliminating analphabetism among the youth. Indeed, estimates indicate that 1 out of every 4 individuals between 15 and 24 years of age (a period during which many

109 Elvir et.al. (2006) highlight the neeed to reform the existing salary structure for teachers in Basic and middle education so that wages satisfy at least a minimum básquet, and then defining incentives according to education, training, experience, performance, specialization, and geographic location.

214

enter the labor force) is illiterate. These efforts should be linked with primary completion programs as well as programs that improve the student’s abilities according to the needs of the labor market.

- It is a growing concern that young people between the ages of 12 and 23 years who live in the urban areas of Managua are less likely to be enrolled in the educational system than their counterparts in other regions. This highlights the fact that education investments in the capital city continue to be needed, mainly among the most vulnerable sectors and in marginalized neighborhoods. Furthermore, greater efforts should be placed on investments to ensure youth have opportunities for further schooling (secondary) and on skill-development programs to help them join the labor market effectively.

- Given that crossing the poverty line requires successfully concluding at least a secondary education, greater efforts should be focused on investing in intermediate level education. As such, it is necessary to prioritize secondary education programs, such as finalizing the process of curriculum reform at this level and promoting new investments with involvement of the private sector. Important challenges exist providing greater access to secondary education while improving the quality of the teaching and learning processes. New programs on technical vocational education and training should be demand driven (according to the needs ot the labor market) to ensure positive results and higher earnings among their participants. Furthermore, there is a need to reduce the cost of vocational and intermediate programs so that accessible to the poor (a good example is given by distance education programs, such as Telesecundaria). Nicaragua needs to invest seriously in developing the education level of its youth so that they can be adapted to the demands of the labor market and to technological progress. This becomes necessary to achieve greater levels of productivity in the context of globalizing region (De Ferranti et.al., 2003).

- The poorest sectors of society cannot afford to pay the high costs of tertiary education; as such, they do not have access to this education level, which is the one that brings the highest returns to households. Therefore, it is recommended that those valuable resources allocated to public universities be focuses on the provision of scholarships to young people from extremely poor households.

- At the central level, the government should i) strengthen the institutional capacity of the Ministry of Education by introducing a monitoring and evaluation system that provides timely and accurate information on education indicators, outcomes and impacts, as well as on the implementation of the Common Work Program, which at this moment is too fragmented; and ii) disseminate key education indicators on school and student performance to promote social accountability and generate support to improve service delivery. A first step could include an analysis of the 2006 national student assessment, which could be useful to identify schools that require additional assistance to reach a desirable standard of quality.

- It is important for the public education system to maintain the core principles of the decentralization program based on the experience acquired during the past 15 years. The model of decentralization should have a series of characteristics that allows it to function more harmoniously than it has in the past. It is important to promote more communication between the center, departments, and municipalities. The previous experience did not include harmonic connections among those instances or levels, which left some decentralized schools without monitoring and/or accountability mechanisms. The model should emphasize more in the quality of the service. There is a need to support initiatives linking resource allocation to specific school quality improvements projects/activities, while strengthening local monitoring and implementation capacity in schools and municipalities.

215

REFERENCES

Arcia, Gustavo; Emilio Porta y José R. Laguna (2004a). Análisis de los Factores Asociados con el Rendimiento Académico en 3º y 6º Grados de Primaria. PREAL, UNICEF, CARE y MECD. Managua, Nicaragua.

Arcia, Gustavo; Emilio Porta y José R. Laguna (2004b). Otro vistazo a la autonomía escolar de Nicaragua: aceptación y percepción. MECD. Managua, Nicaragua.

Banco Mundial (2003). Nicaragua Poverty Assessment: Raising Welfare and Reducing Vulnerability. Report No. 26128-NI. Central America Department. Managua, Nicaragua.

Callejas, Patricia, et. al. (1998). “Autonomía escolar en Nicaragua: Segundo vistazo a la reforma 1995-1997” Ministerio de Educación Cultura y Deportes, Managua.

Castro, Vanesa (2005). Informe Valorativo de la Política de Promoción Automática. MECD. Managua, Nicaragua.

Carneiro, P. and J. Heckman (2003). "Human Capital Policy". in J. Heckman and A. Krueger, eds., Inequality in America: What Role for Human Capital Policy?, (MIT Press).

De Ferranti, David, Guillermo E. Perry, Indermit Gill, J. Luis Guasch, William E Maloney, Carolina Sánchez - Páramo y Norbert Schady (2003). Closing the Gap in Education and Technology. World Bank Latin American and Caribbean Studies. Washington, D.C.

Di Gropello, Enmanuela (2006). Meeting the Challenges of Secondary Education in Latin America and East Asia: Improving Efficiency and Resource Mobilization. Human Development, World Bank. Washington, D.C.

Duryea and Pagés (2002). Human Capital Policies: What they Can and Cannot do for Productivity and Poverty Reduction in Latin American. Inter American Development Bank, Working Paper No. 468. Washington D.C., United States.

Gutiérrez, Roberto y Laguna, José Ramón (2006). Cobertura y Eficiencia Escolar a Nivel Municipal. MECD. Managua, Nicaragua.

Elvir, Ana Patricia; Josefina Vijil, Melba Castillo y Vanesa Castro (2006). Siete Prioridades de la Educación Nicaragüense para el período 2007-2012. CIASES. Managua, Nicaragua.

Handa et al. (2006). Non-formal basic education as a development priority: Evidence from Nicaragua. Mimeo. Managua, Nicaragua.

Hanushek, E. (1995), “Interpreting Recent Research on Schooling in Developing Countries”, World Bank Research Observer, 10(2), Agosto. Washington D.C.

Kruger, Diana (2001). Child Labor as an input in household production. Preliminary Draft. Universty of Maryland, Washington D.C, United States.

Laguna, José Ramón y Gutiérrez, Roberto (2006). Estado de la Educación Básica y Media 2005. MECD. Managua, Nicaragua.

216

Laguna, José Ramón (2005). Radiografía del Empirismo Docente: un análisis de la situación de los docentes nicaragüenses. MECD. Managua, Nicaragua.

Laguna, José Ramón (2003). Años de Estudio y superación de la pobreza en Nicaragua. El caso de la Red de Protección Social. Estudio de caso para optar al título de Magíster en Gestión y Políticas Públicas, Universidad de Chile. Santiago, Chile.

Laguna, José Ramón y Porta, Emilio (2004). Análisis de la Rentabilidad de la Educación en Nicaragua. MECD. Managua, Nicaragua.

MECD (2006b). Informe del MECD: Logros 2002-2006. Managua, Nicaragua.

MECD (2006a). Indicadores de la Educación Básica y Media en Nicaragua. Borrador. Managua, Nicaragua.

MECD (2004a). Investigación sobre Niñez y Juventud en Riesgo: Victimización y Violencia. Grupo de Investigaciones Psicopedagógicas (HUM- 298). Universidad de Córdoba, España. Managua, Nicaragua.

MECD (2004b). Evaluación del Rendimiento Académico de los Estudiantes de 3° y 6° Grado de Primaria: Informe de Resultados 2002. MECD. Managua, Nicaragua.

MECD (2004b). Plan Común de Trabajo del MECD: Prioridades Estratégicas para el período 2005-2008. Managua, Nicaragua.

MINEDUC (2005). Educación Preescolar Estrategia Bicentenario. Santiago, Chile.

Porta, Emilio (2004). Financiamiento de las Instituciones de Educación Superior En Nicaragua. IESALC, UNESCO. Venezuela

Porta, Emilio y Laguna, José Ramón (2007). Educational Equity in Central America: A pending issue for the public agenda. Education Quality Improvement Program 2 (EQUIP2), USAID, AED. Washington D.C., United States.

Reimers, Fernando (1992).Towards a Policy for Early Childhood Education in Latin America and the Caribbean. Working Papers. No. 1. Education and Human Resources Division, Bureau for Latina America and the Caribbean, U.S. Agency for International Development. Washignton, D.C.: Acadamey for Educational Development.

Schweinhart, L. (2004). “How the High/Scope Perry preschool study has influenced public policy” Third International, Inter-disciplinary Evidence-Based Policies and Indicator Systems Conference, July 2004.

UNESCO (2005). Panorama Educativo 2005: progresando hacia las metas. Proyecto Regional de Indicadores Educativos, PRIE. Santiago, Chile.

World Bank (2004). Nicaragua Development Policy Review: Sustaining Broad – Based Growth. Report No. 29115-NI. Washington, United States.

Young, Mary Eming (1996). Early Child Development: Investing in the Future. Human Development Department, World Bank. Washington, United States.

217

ANNEX

Table A1 (annex): Determinants of secondary and post-secondary enrollment in Nicaragua [sample: children between 12 and 23 years of age].

National Urban Rural Enrolled this year in formal education dF/dx dF/dx dF/dx Characteristics of the individual

Male dummy 0.024 N.S. 0.082 Age in years -0.225 -0.180 -0.292 Square of age 0.004 0.003 0.006 Employed dummy -0.205 -0.227 -0.200 Indigenous dummy 0.052 -0.106 0.116

Characteristics of the household Household works in agriculture, 12 months 0.037 N.S. N.S. Natural log of household size 0.169 N.S. 0.438 Square of the natural log of household size -0.052 N.S. -0.118 Female-headed household N.S. N.S. N.S. Head of household is single or divorced -0.039 -0.066 N.S. Age of head 0.015 0.016 0.011 Square of age of household Head 0.000 0.000 0.000

Characteristics of the household head Head has a job N.S. N.S. N.S. Head of household works in blue collar N.S. -0.129 N.S. Head of household works as entrepreneur or chief N.S. -0.156 N.S. Head of household works as self-employment N.S. -0.141 0.068 Head received degree primary or adult educ. 0.091 0.105 0.061 Head received degree secondary 0.160 0.176 0.174 Head received degree technical 0.209 0.186 0.395 Head received degree, higher education 0.277 0.267 0.229

Characteristics of the spouse Spouse has a job 0.082 0.146 N.S. Spouse works in blue collar N.S. -0.114 N.S. Spouse works as entrepreneur or chief N.S. N.S. N.S. Spouse works as self-employment -0.051 N.S. -0.060 Spouse received degree primary or adult educ. N.S. -0.045 0.032 Spouse received degree secondary N.S. -0.064 N.S. Spouse received degree technical 0.213 0.168 0.330 Spouse received degree, higher education N.S. N.S. N.S.

Location and socio-economic level Pacific region N.S. N.S. -0.269 Central region 0.054 0.133 -0.290 Atlantic region N.S. 0.127 -0.352 Quintile = 2 0.075 0.114 0.064 Quintile = 3 0.178 0.215 0.163 Quintile = 4 0.256 0.298 0.220 Quintile = 5 0.350 0.400 0.237 Urban dummy 0.065

Observations 10451 4944 5507 Source: World Bank using the 2005 Nicaragua LSMS. Reference categories: employment of the head/spouse: unemployed or inactive, working as unpaid-agricultural worker; education of the head/spouse: no education; department: Managua. NS: not significant. Underlined coefficients are significant at al 10 % confidence level. All other coefficients are significant at a 5 % confidence level.

218

A2. Male participation in labor or study, by age, at the extremes of income distribution.

0.0

20.0

40.0

60.0

80.0

100.0

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Q1 Trabaja Q5 Estudia Q5 Trabaja Q1 Estudia

A3. First grade enrollment distribution by poverty classification.

0

10

20

30

40

0

10

20

30

40

5 7 9 11 13 15 17 5 7 9 11 13 15 17

Source: 2005 EMNV. Source: 2005 EMNV.

Source: 2005 EMNV. Source: 2005 EMNV.

Extreme poor Non-extreme poor

Non-poor Total

Percent

AgeGraphs by poverty

219

7. ACCESS TO AND QUALITY OF HEALTH SERVICES IN NICARAGUA AFTER A DECADE OF REFORM

Diego Angel-Urdinola and Kimie Tanabe with contributions from Rafael Cortez and Ariadna Garcia-Prado*

Nicaragua has embarked on ambitious reforms that have contributed to significant progress in health sector over the past decade. Health indicators show gradual but steady improvements: access to basic services such as improved water source and sanitation facilities has improved as well a other sartorial performance indicators such as life expectancy; infant/child mortality, immunization rates, and child nutrition among others. Despite these achievements, there are still large inequities in access and quality of health heath services across socio-economic groups and regions. Poor individuals living in rural areas (especially in the Central and Atlantic regions), the indigenous population, and individuals living in households engaged in agriculture have less access to health care services and preventive care than average. Lack of risk mitigation mechanisms, such as insurance and social security, is causing health users in Nicaragua to spend out-of-pocket a significant share of the income in health, especially to buy medications and other non-consultation items such as medical tests. Large distances, lack of medicines, and high cost as well as other demand side factors (such as self-prescription) constitute the main constraints causing poor-sick individuals to seek informal care or to not seek care at all.

INTRODUCTION

Nicaragua has been able to improve its health outcomes with a relatively efficient use of resources compared to other countries in Central America. Still, the health sector in Nicaragua faces several challenges and constraints in terms of equity, effectiveness and efficiency. While public resources are mainly used to maintain a large stock of doctors, hospitals, and clinics, and to provide low-cost consultations, the cost of other non-consultation items, such as medicines and tests are essentially paid out-of-pocket. There is a deficient access to health care facilities by the rural poor, especially in the Central and Atlantic regions. Per capita allocation of resources is concentrated in richer regions tsuch as Managua and Pacific. This set of access and affordability constraints causes poor individuals to utilize health care services less than non-poor users when they are ill.

In 2004 the Ministry of Health (MOH) established a ten-year national health plan to promote decentralization of health service delivery. This plan has been supported by different donor financing modalities through under sector wide approach with a reasonable level of donor harmonization and coordination of strategic support in key interventions. The new legislation will empower local health providers with decision-making authority, especially in relation to resource management and allocation. The MOH’s fundamental 5 year plan’s goal is to improve access to health care services among the poor and more vulnerable sectors of the population, especially in the areas of maternal and child health care.

Despite important reform efforts in the health sector in Nicaragua, poor individuals living in rural areas (especially in the Central and Atlantic regions) as well as vulnerable population groups (such as indigenous and individuals living in households engaged in agriculture) and poor urban households have less access to health care services and preventive care. Large distances, lack of medicines, and high cost

* The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP) and Rafael Cortez (Task Manager Health Sector, LCSHD) for their support and guidance. The conclusions for this paper were prepared by Rafael Cortez and Ariadna Garcia Prado. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent.

220

as well as demand side factors constitute the main constraints causing poor-sick individuals to seek informal care or to not seek care at all. Due to limited access to social insurance and social security most of the private expenditures on health in Nicaragua are paid out-of-pocket. Medicines are by far the main expenditure health, especially among the poor.

This chapter analyzes constraints to utilization of health services (mainly issues in relation to access and affordability) that adversely affect utilization of services by the poor. The chapter is structures as follows: section 1 summarizes the overall health sector performance in the past decade in the Central and Latin American context and analyzes progress achieved in relation to infant mortality and maternal health; section 2 presents the morbidity profile of Nicaragua using recent data from PAHO, WHO, and the 2005 LSMS (Living Standards Measurement Survey); section 3 analyzes inequities in access, utilization, and quality of heath services across socio-economic groups and regions as well as existing affordability constraints; a brief conclusion follows.

I. Overall Sector Performance

The performance of the health sector in Nicaragua has improved steadily over the past 10 yeas in alignment with other Central American economies. Nicaragua has embarked on ambitious reforms in key social sectors in the past decade. As a result of reforms implemented in the last 10 years, health indicators show gradual but steady improvements: access to basic services such as improved water source and sanitation facilities (which is related to hygienic-related illnesses, such as diarrhea) has improved as well a other sartorial performance indicators such as life expectancy; infant/child mortality, immunization rates, and child nutrition among others. In particular, in the last three decades, life expectancy at birth (see Figure 7.1) has increased by more than 10 years in Nicaragua. Nicaragua’s improvement in life expectancy is aligned to that achieved in its neighboring countries. Under-five and maternal mortality rates have fallen steeply, as shown in Figures 7.2. 110

Figure 7.1. Progress in life expectancy in Nicaragua since the 1970s is comparable to that achieved by other Central American economies

Increases Life Expectancy in Central America

54 57 52 5368687170

0

20

40

60

80

Nicaragua El Salvador Guatemala Honduras

Life

expec

tancy

at bir

th

1970 1980 1985 1990 1995 2000 2004

Sources: Authors using WDI Central Database: Social Indicators. 2007

110 Among 4 Central American countries, the highest infant mortality rate is found in Guatemala (33 per 1,000 live births) and the lowest is El Salvador (24 per 1,000 live births). Honduras and Nicaragua both have the same record of 31 per 1,000 live births.

221

Figure 7.2. Under-five infant mortality in Nicaragua has fallen from approximately 160 in the 1970s to 31 children per 1,000 births in 2004

0

50

100

150

200

1970 1975 1980 1985 1990 1995 2000 2004

Per

1,0

00

Nicaragua El Salvador Guatemala Honduras

Source: Authors using WDI Central Database: Social Indicators. 2007

Despite significant progress in the last decades, 31 out of every 1000 children born alive die each year in Nicaragua, half of which die within the first 28 days of life. In 2004, infant mortality in Nicaragua was at 31 per 1,000 live births. The vast majority of the cases could have been easily prevented by a combination of good care, nutrition, and medical treatment. Figure 7.3 displays average infant mortality rates in Nicaragua vs. other Central American economies disaggregated by neonatal (the probability of dying within the first 28 days of life), post-neonatal (the probability of dying between the 28th day of life and the first birthday), and under five (the probability of dying in the first 5 years of life). Estimates indicate that Infant mortality rates have declined in Nicaragua mainly due to important progress in post-neonatal mortality rates since the 1980s while neonatal mortality rate have declined only slightly in the same period. While post-neonatal mortality rates were higher than neonatal mortality rates in the late 1980s, the opposite occurs in the early 2000s. This same phenomenon holds true in other Central American economies such as El Salvador, Guatemala, and Honduras.

Figure 7.3. Trends in Under-5 Mortality

2 3 1 7 1 32 6 2 3 2 2 1 9 1 9 1 9 2 0 1 7 1 6

1 81 8

1 2

2 52 2

1 6 2 0 1 7 1 5

3 8

2 21 5

1 2

8

6

1 8

1 41 5 1 6

1 31 1

1 5

1 1

9

0

1 0

2 0

3 0

4 0

5 0

6 0

7 0

8 0

ES 198 8-1

993

ES 199 3-1

998

ES 199 7-2

002

GT 1

9 90 -199 5

GT 1

9 93 -199 8

GT 1

9 97 -200 2

HN

1986 -1

9 90

HN

1991 -1

9 95

HN

1996 -2

0 00

NI 1

98 7-1992

NI 1

99 3-1998

NI 1

99 6-2001

Dea

ths

per

1,0

00

N e o n a t a l P o s t n e o n a t a l C h i l d

S o u r c e : S t u p p , M o n t e i t h & M c C r a c k e n , 2 0 0 5

Source: Stupp et al. (2005)

Although immunization coverage remains at generally high levels (close to 90 percent coverage), it has dropped since 2004. Childhood immunization offsets the detrimental effects of poverty and low educational attainment of children. Hence, to promote immunization coverage as a strategic component of poverty-reduction program is indispensable for Nicaragua. Immunization rates of polio, measles,

222

diphtheria, pertussis and tetanus, and tuberculosis dropped from 90 percent in the late 1990s to 80, 84, 79 and 88 percent in 2004 respectively. Incomplete vaccinations can reflect flawed service delivery and inefficiency of logistics systems, as well as lack of health services and budget allocation especially in remote areas.

Fertility rates in Nicaragua declined by roughly 100 percent in the past two decades; yet there is significant regional variation. As illustrated by Figure 4, fertility rates (which proxy de average number of children women have during their life-cycle) dropped dramatically from 6 in 1985 to 3.1 in 2004 (see Figure 7.4). Since monetary welfare is often measured by per-capita income/consumption, a decrease in fertility rates (which usually pushes down the average number of members in the household) is beneficial for poverty reduction. Also, declining fertility rates may be a proxi for successful family planning and contraception campaigns.111 Compared to its neighboring countries, fertility rates in Nicaragua and El Salvador have declined faster than in Guatemala. Among the 4 Central American countries included in this analysis, El Salvador has the lowest fertility rate of (2.8 children per woman), followed closely by Nicaragua (3.1), Honduras (3.6), and Guatemala (4.3).

Figure 7.4: Fertility Rates in Nicaragua have declined in pace with those in other Central American economies

0

1

2

3

4

5

6

7

8

1960

1967

1972

1980

1985

1990

1995

2000

2003

Tot

al b

irth

s pe

r w

oman

El Salvador Guatemala Honduras Nicaragua

Source: Authors using WDI Central Database: Social Indicators. 2007

Survey estimates indicate large differences in fertility rates by region: poorer regions display higher fertility rates. Estimates from the 2005 LSMS suggest that in 2005 total fertility rates nationally were at 2.2 (1.8 in urban areas vs. 2.7 ion rural areas). As shown in Figure 7.5 below, rural fertility rates vary significantly by region. In particular, rural fertility rates vary from 1.8 in Managua to 3.4 in the Atlantic region.

111 Fertility rates are related with mother’s educational level. As expected, the total fertility rate of women with less formal education exceeds that of more educated women, with a difference of more than four children between the lowest and highest levels of education in Nicaragua and Guatemala. In all these countries, the desired total fertility rate is lower than the observed fertility rate, which clearly indicates that women are not meeting their reproductive needs.

223

Figure 7.5: Fertility Rates are significantly higher in rural areas

1.8

2.7

1.8

2.42.8

3.4

00.5

11.5

22.5

33.5

4

Urban

Rural

Man

agua

-rura

l

Pacif

ic_rur

al

Centra

l-rur

al

Atlant

ic-ru

ralBir

ths

per

wom

an (a

ge 1

5-49

)

Source: Authors using the 2005 Nicaragua LSMS.

Roughly 1 out of every 3 children born in Nicaragua is unwanted/not planned: this rate is high for Latin American Standards. The desired fertility rate reflects the number of children women want, while the observed total fertility rate (TFR) is the number of children they actually have. Figure 6 displays observed and desired total fertility rates for years 2001 to 2003 in Nicaragua, Honduras, Guatemala and El Salvador. Fertility rates of unwanted or mistimed births in Nicaragua is a little more than one third, suggesting that only 2 out of every 3 children born are wanted by their mothers while the remaining is unwanted or not planned.

Figure 7.6: El Salvador and Nicaragua display high undesired fertility rates

2 . 23 . 7 3 .1 2 .3

0 . 8

0 . 7 1 .30 .9

0123456

E l S a l v a d o r2 0 0 3

G u a te m a l a2 0 0 2

H o n d u r a s2 0 0 1

N ic a r a g u a2 0 0 1

TF

R

W a n te d / M is t im e d U n w a n te d

Source: Stupp, Monteith and Mc Craken, 2005

224

Figure 7.7: About 117 of very 1,000 young women between 15 and 19 years old in Nicaragua has at least 1 child: this indicator is the highest in all LAC

ARG

BOL

BRA

CHL

COL CRI

DOM

ECU SLV

GTM

HND

MEX

NIC

PAN

PRY

PER

URY

VEN

40

50

60

70

80

90

100

110

120

7.50 8.00 8.50 9.00 9.50

Log GDP per capita

Fer

tilit

y Rat

e Age

betwee

n 15

-19 (1

,000

pop

)

Source: Authors using WHO, PAHO Core Health Data System 2007

Although Nicaragua has the highest share of young women (15 to 19 years old) with kids in Latin America, the share of young mothers has been decreasing steadily in the past decade. Teen pregnancy is closely linked to a number of other critical issues, including overall child poverty and family well-being. Needless to say teen mothers are less likely to complete primary or secondary education that is necessary to qualify for a better employment opportunity. Continuing to reduce teen age pregnancy will help sustain the recent decreases in poverty, especially persistent child poverty in Nicaragua.

Figure 7.8: The share of young mothers (between 15 and 19 Years) in Nicaragua has declined from 142 young mothers per 1000 population in 1997 to 117 in 2004

0

20

40

60

80

100

120

140

160

1997 1998 1999 2000 2001 2002 2003 2004 2005

Per

1,0

00 p

opul

atio

n

Source: Authors using WHO, PAHO Core Health Data System 2007

Although overall health outcomes and resources indicators in Nicaragua are aligned with those of other Central American economies, Nicaragua spends relatively more or its public resources on health. Public health expenditure in Nicaragua as in percentage in GDP is highest among neighboring countries in Central America. In the year 2000, it is recorded to be 6.8 percent of GDP. Similar to other three Central American countries, between 2000 and 2001 the percentage share of public health expenditure in GDP declined to 3.8 percent. Despite this significant drop, Nicaragua still spends on health significantly more that its neighbor countries as well as in the regional context given its level of development (see Figures 7.10A and 7.10B). While Nicaragua spends more of its public budget on health, basic health outcomes (as described above) and sector resources do not surpass those of its neighboring

225

countries. Nicaragua’s health system displays physical and human resources that are according to its level of development in the Latin America context (but achieved at higher levels of expenditure). While the number of physicians per every 10,000 habitants is high given Nicaragua’s level of development, the number of nurses per 10,000 habitants is low. The number of hospital beds and hospital discharges per 1,000 habitants is aligned to its level of development. This suggests that the sector may not be spending resources as efficiently as other countries are (i.e. Nicaragua is achieving similar outcomes than some of its neighbors who are paying half as much).

Figure 7.9: Human and Physical Resources in Nicaragua vs. Latin America Physicians Nurses

ARG2004

BOL2001

BRA2001

COL2003CRI2000

ECU2001

SLV2002

GTM2003

NIC2003

PAN2003

PRY2004

PER2000

URY2003

VEN2001

0

5

10

15

20

25

30

35

40

7.50 8.00 8.50 9.00 9.50

Log GDP per capita 2004

Phy

sici

ans

rati

o (1

0,00

0 ha

b.)

ARG2004BOL2001

BRA2001

CHL2003COL2003

CRI2000

ECU2001

SLV2002

GTM2003

NIC2003

PRY2004

PER2000

URY2003

PAN2003

0

2

4

6

8

10

12

7.50 8.00 8.50 9.00 9.50


Pro

fess

iona

l nur

ses

rati

o (1

0,00

0 ha

b.)

Hospital Beds Hospital Discharges

ARG2000

BOL2004

BRA2002 CHL2003

COL2004CRI2003

DOM2004

ECU2003

SLV2004GTM2003

MEX2003

PAN2004

PRY2002PER2004

URY2003

VEN2003HND2002NIC2004

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

7.50 8.00 8.50 9.00 9.50


Hos

pita

l bed

s ra

tio

(per

1,0

00 p

op)

ARG2003BOL2004

BRA2002

CHL2003

COL2000

CRI2003

DOM2004ECU2003

SLV2004

GTM2003

HND2003

MEX2003

NIC2004

PAN2004

PRY2003

PER2004URY2003

VEN2003

0

50

100

150

200

250

7.50 8.00 8.50 9.00 9.50


Hos

pita

l dis

char

ges

rati

o (p

er 1

,000

pop

)

Source: Authors using WHO, PAHO Basic Health Indicator Database, and the World Bank, WDI Central

226

Figure 7.10A: Although investments on health have decreased since year 2000, Nicaragua still spends roughly twice as much as its neighbors

Public Expenditure in GDP

1.8

3.1 3.53.8

1.42 2

6.8

01

23

45

67

8

Nicaragua 2000 /2001

Guatemala2001/2003

Honduras1998/2001

El Salvador2001/2003

% in

GD

P

Source: Authors using WHO, PAHO Basic Health Indicator Database

Figure 7.10B: Public and Private Expenditure on health as % of GDP

ARGBRA

CHL

COLCRI

DOM

ECU

SLV

GTM MEX

NIC

PAN

PRY

PER

URY

VEN

0

1

2

3

4

5

6

7

8

8.00 8.20 8.40 8.60 8.80 9.00 9.20 9.40 9.60

Log (GDP per capita 2004)

Pub

lic E

xpen

ditu

re o

n H

ealt

h as

% o

f GD

P

ARG

BOL

BRA

CHL

CRI

DOMSLV

GTMHNDMEX

NIC

PAN

PRY

PERVEN

COL

ECU

URY

0.5

1.5

2.5

3.5

4.5

5.5

6.5

7.5

7.50 8.00 8.50 9.00 9.50 10.00


Pri

vate

Hea

lth

Exp

endi

ture

in G

DP


MATERIAL CARE

In the last ten years the government of Nicaragua has made great advances in maternal health policy. In 2004 Nicaragua just updated its National Health Policy (from 2004 to 2015), paying especial attention to issues related to maternal health. The promotion of healthy mothers and children needs effective (cross-sartorial) interventions and various policies such the improvement women’s and children’s social and economic status, social programs with families and communities, educations for all, universal access to basic health care, access to family planning services, ensuring skilled attendance during childbirth, adequate neonatal and child health care, and domestic violence free environment, among others. As such, the country has developed laws in relation to childhood and adolescence health promotion, breast-feeding, delivery and pre-post natal care, as well as other regulations preventing nontraditional professionals that participate in delivery care and protecting relations between parents and Children.

In the LAC context, maternal mortality rates in Nicaragua are aligned to the country’s level of development. The maternal mortality ratio (per 100,000 live births) in Nicaragua was at 230 in year 2000. As indicated in the Figure 7.11, Nicaragua is just at Latin America average given its level of

227

economic development. According to the Ministry of Health, of total maternal deaths in 2004, 11 percent were due to homicide (69 percent of which were linked to the mother’s suicide, mainly among adolescents). Both physical and sexual violence against young mothers is having serious consequences for the physical and mental health of both mothers and children in Nicaragua and thereby has become major public health issue in Nicaragua. A recent study in the city of Leon in Nicaragua indicates that physical and sexual aggression against mothers (either before or during pregnancy) increases substantially the risk mortality of their children all the way to age 5 (Asling-Monemi, Pena, Ellsberg&Persson, 2003).

Figure 7.11: Maternal Mortality Ratio in the year 2000 (per 100,000 live births)

ARG

BOL

BRA

CHL

COLDOM

ECU

GTM

HND

MEX

NIC

PANPRY

PER

URY

VEN

0

50

100

150

200

250

300

350

400

450

7.50 8.00 8.50 9.00 9.50

Log GDP per capita

Mater

nal M

ortalit

y 10

0,00

0 pe

r liv

e birth

Source: Authors using WHO Core Health Indicators 2007

While the share of deliveries attended by train personnel in Nicaragua is slightly above Latin American standards, there are large disparities across socio-economic groups and regions. While 95 percent of all deliveries in the richest quintile are attended by a trained doctor, the equivalent rate is at only 56 percent among poor women (the remaining deliveries are attended by midwifes). While the proportion of births attended by a doctor in Managua is as high as 97 percent, the equivalent rate is at 87, 72, and 46 percent in the Pacific, Central, and Atlantic regions respectively. While in rural Managua the proportion of deliveries attended by doctors exceeds 90 percent, the equivalent proportion is at 33 percent in the rural areas within the Atlantic region. As expected, the opposite occurs for the proportion of deliveries attended by midwives: while less than 10% of mothers living in rural Managua are attended by midwives, more than half of all mothers in rural Atlantic are attended by midwives.

Figure 7.12: Proportion of Deliveries Attended by Trained Personnel in LAC

ARG2003

BOL2003

BRA2000

CHL2003

COL2002CRI2002

DOM2002

ECU1999

SLV2004GTM2002

MEX2003

NIC2003

PAN2003

PRY2003

PER2004

URY2002VEN2002

HND2001

50

60

70

80

90

100

110

7.60 7.80 8.00 8.20 8.40 8.60 8.80 9.00 9.20 9.40 9.60

Log GDP per capita

Pro

portion of

deliver

ies at

tend

ed by traine

d

pers

onne

l (%

)

Source: Authors using WHO, PAHO Core Health Data 2007

228

Figure 7.13: Births Attended by Trained Personnel by Quintile and Region

Birth Attended by Trained Personnel by Quintile

55.674.3 82.8 90.5 95.1

35.218.3 12.3 6.1

0

20

40

60

80

100

120

Poorest Q2 Q3 Q4 Richest

perc

ent

Doctors Nurses Midwife

Birth Attended by Trained Personnel by Geographic Location

96.9 8771.9

45.7

93.6

59

0.93.1

4.2

5.1

2.4

4.2

2.2 9.320.4

41.3

3.8

31.4

0

20

40

60

80

100

120

Managua Pacific Central Atlantic Urban Rural

perc

ent

Doctors Nurses Midwife

0

20

40

60

80

100

Doctors 90.5 76.1 60.3 33

Midwife 9.5 17.8 30 52.1

Rural Managua

Rural Pacific

Rural Central

Rural Atlantic


MORBIDITY

Overall, respiratory illnesses are the most popular disease in Nicaragua followed by chronic illnesses and diarrhea (see Figure 7.11 below). About 59 percent of all individuals claiming to have been sick in the 4 weeks previous to the 2005 LSMS survey claim to have suffered from respiratory illnesses. Chronic ant other multiple deceases account for about 32 percent of all illnesses reported by those individuals claiming to have been sick. While respiratory diseases are relatively more common among individuals in the poorest quintiles, chronic illnesses are more common among individuals in richest quintile. Not surprisingly, Indigenous people and individuals living in agricultural households are more vulnerable to suffer from diarrhea and other multiple type of illness.

229

Figure 7.11: Morbidity in Nicaragua 2005

59%

6%

13%

19%3%

respiratory diarrheachronic illness other/multipleskin/accident/violence


Respiratory infections and diarrhea are the primary causes of infant and child morbidity and mortality are in Nicaragua. Prevalence of diarrhea and acute respiratory infections (ARI) are high among children, especially in rural areas. Prevalence rates highest among children aged 12 to 23 months, reflecting in part the protective effect of breastfeeding among babies bellow one year old. As older babies begin to be exposed to other foods, the prevalence of diarrhea begins to increase among babies between 6 and 11 months old. Between 41 and 50 percent of all mothers with a child who suffered a recent episode of diarrhea consulted someone about the disease. Not surprisingly, urban mothers were more likely to seek consultation than rural mothers. As shown in Figure 7.12 below, in Nicaragua and in El Salvador the highest percentage of children with diarrhea who were treated with oral dehydration salts (ORS) reside in urban areas: 54 vs. 50 percent in Nicaragua and 55 vs. 51 percent in El Salvador. In Guatemala and Honduras, the opposite occurs.

Figure 7.12: About half of all children under 5 years old in Nicaragua are treated with Oral Dehydration Salts (Children Less than 5 Years Old)

Error! Objects cannot be created from editing field codes. Source: Stupp, Monteith and McCraken, 2005

Managua and the Pacific region as well as urban areas display a larger share of their population suffering from chronic illnesses than average. The Central and Atlantic regions display a larger share of their population suffering from multiple illnesses. Diarrhea and respiratory diseases are more popular among male population while chronic and multiple illnesses are more common to female population.

Figure 7.13: Indigenous people and individuals living in agricultural households are more vulnerable to suffer from diarrhea

230

0%

5%

10%

15%

20%

25%

30%

Poor

est Q

uinti

leQ2 Q3 Q4

Riches

t Quin

tile

Non-In

digen

ous

Indige

nous

Non A

gricult

ural

Agricu

ltural

chronicillness diarrhea

7%6% 5%

8%

14%16%

11%8%

17%15%

22%24%

0%

5%

10%

15%

20%

25%

30%

Managua Pacifico Central Atlantico

diarrhea chronic illness other/multiple


About 1 out of very 3 children between 12 and 59 months old in Nicaragua is expected to suffer from Anemia. Iron-deficiency anemia in children is associated with impaired cognitive performance, motor development, coordination, language development and scholastic achievement. Anemia increases morbidity from infectious diseases because it adversely affects several immune mechanisms. Among other feasible factors that cause anemia, nutritional deficiency due to lack of dietary iron, is a major cause of anemia in Central America. If anemia remains undiagnosed, it can lead to infertility in women of childbearing age and premature delivery among pregnant women. The percentage of children aged 12-59 months with anemia in Nicaragua is at 28.4 percent. The prevalence of anemia is higher in rural areas than in urban areas. Anemia prevalence also varies somewhat among income quintiles. Both age of children and mother’s education level seems to be negatively correlated with the percentage of children who have anemia. As the age of the children increases, the percentage of children who have anemia decreases. Similarly, the higher the mother’s education level, the less likely their kids to have anemia112.

Nicaragua has the lowest rate of HIV prevalence in Central America. It is estimated that 191,000 people are living with HIV/AIDS in Central America. The majority of HIV/AID positive population lives in Honduras and Guatemala (see Figure 7.14). In an effort to prioritize promotion and prevention of the disease, Nicaragua has implemented campaigns to promote the use of condoms, to avoid early sexual relations, and to identify the symptoms of AIDS in both urban and rural areas. Similar to what occurs in the Caribbean and South America, the HIV epidemic is mostly concentrated in the urban/commercial areas and transmission is primarily due to heterosexual contact. The illness is more common among men than among women. However, the gender gap of HIV/AIDS population is shrinking in Central America, including Nicaragua. In spite of the efforts made by the government in response to the epidemic, there are

112 The World Bank: Key Issues in Central America Health Reforms: Diagnosis and Strategic Implication, p33.

231

still important challenges to face. There is a need, for instance, to promote dissemination of information and knowledge about this deadly infectious disease, especially among women, and within the rural and indigenous population.

Figure 7.14: HIV Prevalence in Adults 15 to 49 years Old in Nicaragua was at 0.2 percent in year 2001 HIV Prevalence Among Adults 15-49 Years Old, 2001

0.20%

0.60% 0.60%

1.00%

1.50%

1.60%

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

1.60%

1.80%

Nicaragua El Salvador Costa Rica Guatemala Panama Honduras

Source: National HIV Programs in Guatemala, Honduras, El Salvador, and Nicaragua. 2001

ACCESS TO HEALTHCARE

Utilization

The Nicaraguan Ministry of Heath (MOH) provides health care through its network of more than 1000 facilities, including 33 hospitals, 177 health centers, and 872 health posts. The MOH administers the system through 18 departmental offices (SILAIS). The Nicaraguan Social Security Institute (INSS) is the second most important health care provider: about 19 percent of all people 20 and 39 years old went to the INSS for consultation. The INSS purchases a defined package of services from 48 health provider organizations called Empresas Medicas Previsionales (EMPS) (see PAHO, 2002). Being used by 43 percent of population, health centers are the most common type of medical facility used in Nicaragua followed by private clinics (16 percent), public and private hospitals (13 percent), the INSS (11 percent), health posts (9 percent), and other facilities (8 percent). These results were obtained using estimates from the 2005 LSMS. In 2005, about half of the poor population received consultation in health center, 13 percent in health posts and 12 percent in public hospital. Non poor individuals (generally those users in 4th and 5th quintile and those living in urban Managua and/or the Pacific region) are more likely to use private clinics and the INSS (besides health centers) for consultation.

About half of all individuals who get sick receive/seek medical treatment. In year 2005, according to the LSMS, approximately half of all individuals (50 percent of non-poor and 40 of poor) who claim to have been sick 4 weeks prior to the survey received some type of consultations or medical treatments. At the national level, health care utilization as measured by outpatient health care visits ratio in per 1,000 populations had increased from 1,674 to 2,153 between years 2001 and 2004. Despite this large increase in utilization, outpatient care facilities have decreased slightly from 1,129 in 2002 to 1,122 in 2004 (see Figure 7.15). By 2004 (see Figure 7.16) outpatient visits per 1,000 population in Nicaragua were slightly above Latin American Standards given the countries level of development.

232

Figure 7.15: Despite increasing utilization since 2001, the number of health facilities has remained unchanged.

Utilization and Provision of Health Service 2001-2004

0

500

1,000

1,500

2,000

2,500

Number of outpatient care facilities Outpatient health care visits ratio[per 1,000 pop.]

2001 2002 2003 2004

Source: Authors using WHO, PAHO Basic Health Indicator Database

Figure 7.16: Outpatient Care Visits in Nicaragua are slightly above Latin American Standards given its level of development

ARG2003BOL2004

CHL2003

CRI2003

DOM2004ECU2003SLV2004GTM2003HND2003

MEX2003

NIC2004

PRY2003

URY2003VEN2003

BRA2002

PAN2002PER2004

0.00

2,000.00

4,000.00

6,000.00

8,000.00

10,000.00

12,000.00

14,000.00

16,000.00

18,000.00

20,000.00

7.50 8.00 8.50 9.00 9.50


Out

patien

t he

alth

car

e visits rat

io (pe

r 1,00

0

pop)


Utilization rates among the non-poor are about 13.2 percent higher than among the poor. Table 7.1 presents health care utilization rates among individuals who claim to have been sick 4 weeks prior to the survey. Results indicate that utilization rates vary significantly across socio-economic groups, region, and strata. In particular, while utilization among non-poor urban users is 55.8 percent, utilization among poorer and more vulnerable users (such as those living in the rural areas, in the poorer regions, and from a household engaged in agriculture) are at 42.6 percent. Utilization rates are higher among women than among men (52.3 vs. 47.1 percent). Not surprisingly utilization rates are higher among elders and the infants below 1 year of age (at 54.33 and 81.02 percent respectively) and lower among the youth (at 32.7 to 41.7 percent).

233

Table 7.1: Utilization of Health Service in Nicaragua Utilization of Health Service

Nicaragua 2005 among those reported to have been sick (past 4 weeks)

Ordinary consultation

%

Emergency Consultation

%

No consultation %

Socio economic group Non Poor 50.59 5.18 44.23 Poor 40.41 2.23 57.36 Poorest Quintile 36.48 2.37 61.15 Q3 46.87 2.75 50.37 Richest Quintile 53.85 5.88 40.27

Vulnerable group Indigenous 48.78 3.45 47.77 Agricultural producer 40.31 2.53 57.16

Strata Rural 41.90 2.55 55.55 Urban 49.53 4.97 45.50 Managua 48.18 6.86 44.95 Pacific 48.88 3.19 47.93 Central 44.29 3.04 52.67 Atlantic 41.03 2.09 56.88

Gender Female 48.44 3.81 47.75 Male 43.15 3.90 52.96

Age group age 0-1 74.47 6.55 18.97 age 2-12 52.26 3.62 44.13 age 13-19 30.01 2.63 67.36 age 20-29 38.05 3.60 58.35 age 30-39 39.93 3.16 56.91 age 40-49 47.27 4.16 48.57 age 50-59 45.00 4.40 50.60 age 60-69 46.98 3.95 49.08 age 70-79 49.91 4.42 45.67 age 80-97 48.68 7.12 44.20


Level of education, access to medical insurance, socio-economic group, distance to health facility, and region are important determinants of health care utilization in Nicaragua. Regression analysis (using a probit model) is useful to quantify the main determinants of healthcare utilization conditional on a set of individual and households characteristics (for full regression results see Table A1 in the annex). Estimates from the 2005 LSMS suggest that that education level of household head and spouse, size of household, region, the coverage of insurance, and at most distance to health care facilities are significant factors for utilization of health care service, and preventive medical treatment in Nicaragua:

Type of illness and age group: While Patients in the age group 20-39 are 15 percent less likely to seek medical consultation with respect to those younger than 20 years old. Patients in the age group 50-59, 60-69, and older than 70 years are 17, 31, and 56 percent more likely to receive treatment when sick as compared to patients younger than 20 years old. Patients with respiratory diseases (diarrhea) are 55 (100) percent less (more) likely to receive a medical treatment when ill as compared to individuals ill from accidents or skin problems (the omitted category).

Socio-economic group and education level: Individuals living in a household with a head/spouse who completed primary and secondary education are 6 and 9 percent more likely to receive medical treatment when sick as compared to individuals in household with a head/spouse with no education.

234

Higher the income quintiles, greater the chance to seek medical consultations: individuals whose incomes are in Q2, Q3, Q4 and Q5 are 11, 20, 23 and 32 percent higher probability to receive treatment with respect to individuals in poorest quintile.

Insurance and distance to health facility: Individuals with health insurance are 56 percent more likely to get medical treatment when sick as compared to individuals with no insurance coverage. The longer the distance to the nearest health facility, the less likely users are to seek medical consultations. For every one additional kilometer away from a consultation facility offers individuals the probability seeking consultations decreases by 0.2 percent.

Strata and Region: Individuals living in Pacific and Central regions (and in urban areas) display a 3 to 5 percent higher probability to receive medical consultations with respect to individuals living in Atlantic region.

Only 3.7 percent of all individuals seek preventive health care in Nicaragua (see Figure 7.17). As expected, urban households as well as those in the highest consumption quintiles utilize more preventive care services. As life expectancy in Nicaragua has been increasing year by year, the population gets to live longer and in some cases with one or more chronic conditions. This trend places new, long-term demands on health care systems. Not only are chronic conditions projected to be the leading cause of disability (if not successfully prevented and managed) but also they may become the most expensive challenge faced by Nicaragua’s health systems in the near future. Many diseases can be prevented, yet current health care systems in Nicaragua do not make the best use of their available resources to support this process.113

Figure 7.17: Preventive care utilization among individuals in the highest quintile is up to 300 percent higher than among individuals in the bottom quintile

Utilization of Preventive Care

4.88

2.331.88

2.59

3.43

4.263.73

2.61

4.24

2.62

3.7

6.3

01234567

Non P

oor

Poor

Poore

st Q

uint

ileQ2

Q3

Q4

Riche

st Q

uintil

e

Non-Ind

igen

ous

Indig

enou

s

Non A

gricu

ltura

l

Agricu

ltura

lAll

Per

cent


113 According to WHO, most current health care systems in general, are based on responding to acute problems, urgent needs of patients, and pressing concerns. Testing, diagnosing, relieving symptoms, and expecting a cure are hallmarks of contemporary health care. While these functions are appropriate for acute and episodic health problems, a notable disparity occurs when applying this model of care to the prevention and management of chronic conditions since preventive health care is inherently different from health care for acute problems, and in this regard, current health care systems fall remarkably short

235

Given that many conditions are preventable, every health care interaction should include prevention support. When patients are systematically provided with information and skills to reduce health risks, they are more likely to reduce substance use, to stop using tobacco products, to eat healthy and balanced diet, to maintain mental and spiritual well-being, and to engage in physical activity. These risk reducing behaviors can dramatically reduce the long-term burden and health care demands of chronic conditions. To promote prevention in health care, investments on awareness-rising are key to promote a change in thinking and to stimulate commitment and actions among patients and families, health care teams, communities, and policy-makers in Nicaragua.

Level of education of the household head and spouse, gender, household size, region, and socio-economic condition are characteristics that influence the probability that individuals seek preventive care. Findings of a probit regression (for full regression results see Table A2 in the annex) identify that education level of household head and spouse, size of the household, region, level of income, and gender are significant factors of preventive care. Individuals living in a household with a head who completed primary and secondary education are 8 and 2 percent more likely attain preventive care as compared to individuals living in a household with a head with no education. Moreover, individuals living in a household with a spouse who completed primary, secondary and tertiary education have a 12, 16 and 23 percent higher probability to seek preventive care as compared to individuals having a spouse with no education. Individuals living in Managua, Pacific and Central region are 30, 36 and 12 more likely to receive preventive medical care with as compared individuals living in the Atlantic region. Controlling for other factors, consumption levels do appear to be a significant determinant of preventive healthcare utilization. With respect to individuals in poorest quintile, individuals in the third, fourth, and fifth quintile are 17, 16 and 19 percent more likely to utilize preventive medical treatment. Finally, estimates indicate that with respect to female population, male individuals are 6.2%less likely to seek preventive medical care.

PROVIDERS

Health centers are the most popular type of facility in Nicaragua (see Table 7.2). Health posts, health centers, public and private hospitals, the Social Security Institute (INSS), and private clinics are the main providers of health services in Nicaragua. Health posts are more commonly used in rural areas, and in the Central and Atlantic regions. Private clinics, usually associated with better quality of service delivery, and INSS are more used by patients in urban areas and in the Managua and Pacific regions (Table 7.2). Health centers and health post are the main health providers among the poor (about 71 percent of all poor patients rely on these facilities when sick). About 16 percent of all poor patients use health posts when sick; these are no well equipped facilities that provide very basic services (generally in rural areas and small towns). Only about 10 percent of the poor use INSS or private clinics to get service. On the contrary, about 4 of every 10 non-poor patients use INSS or private clinics when sick: the richer the patient the more like to use private clinics and INSS facilities and the less likely to use health posts.

The majority of the population receives medical treatment from trained health professionals when sick. Yet, between 17 and 25 (6 and 10) percent of all individuals in poorest quintile as well as those who belong to vulnerable sectors of the population (such as indigenous and those living in agricultural producer households) receive medical treatment from nurses rather than doctors for ordinary (emergency) consultations. While receiving care from nurses is not necessarily a signal of lower quality for some ordinary consultations (when symptoms/pathologies are mild), in cases of life-threatening emergencies (or for some more complicated pathologies that require prescription) receiving medical care from nurses is less than desirable. As expected, the share of consultations attended my nurses is higher in the poorest regions and in rural areas.

236

Table 7.2: Poor households generally use health posts and health centers when sick Sample: those reported to have received consultations (past 4 weeks)

Health post

Health center

Public or private hospital

INSS Private clinic

Other

Socioeconomic group Non Poor 4.33 33.76 15.68 14.67 23.9 7.66 Poor 15.72 55.51 10.16 4.01 5.54 9.06 Poorest Quintile 19.78 56.24 8.31 0.71 3.5 11.46 Q2 14.04 54.95 10.89 6.6 6.33 7.2 Q3 7.5 50.49 14.6 7.8 12.19 7.42 Q4 5.39 34.08 16.06 16.49 20.3 7.67 Richest Quintile 1.88 23.97 15.57 16.52 34 8.05

Vulnerable group Indigenous 27.61 26.8 18.89 6.01 11.63 9.06 Agric. household 16.86 47.14 10.51 1.84 13.71 9.94

Strata Rural 16.79 47.9 10.28 3.03 12.32 9.68 Urban 3.06 38.1 15.85 15.85 19.99 7.16 Managua 3.63 29.79 14.83 23.81 20.29 7.65 Pacific 2.93 47.91 13.72 9.84 17.43 8.17 Central 10.44 51.32 12.24 3.43 15.55 7.03 Atlantic 27.81 31.88 13.8 2.72 11.55 12.25


Figure 7.18: About 27 percent of all indigenous patients receive medical care from nurses when sick

0%

20%

40%

60%

80%

100%

Poo

r

Non

Poo

r

Poo

rest

Quin

tile Q2

Q3

Q4

Ric

hes

t

Quin

tile

Indig

enou

s

Non

-

Indig

enou

s

Agr

icultura

l

pro

duce

r

Non

Agr

icultura

l

Doctors Nurses


237

Table 7.3: Poor individuals residing in rural areas are less likely to receive care from doctors, even in emergency cases

Sample: those reported to have received consultations (past 4 weeks)

Ordinary consultation by

doctors

Ordinary consultation by

nurses

Emergency consultation by

doctors

Emergency consultation by

nurses Socioeconomic group

Poor 77.67 14.92 89.97 7.45 Non Poor 91.73 3.82 97.93 1.67 Poorest Quintile 67.90 22.11 88.53 5.86 Q3 82.86 11.61 88.63 11.37 Richest Quintile 87.71 6.76 92.52 7.18

Vulnerable group Indigenous 67.13 24.49 82.42 9.96 Agric. household 74.91 16.79 89.72 6.92

Strata Urban 93.40 2.74 97.86 1.97 Rural 76.22 15.72 91.31 5.90 Managua 95.14 1.02 97.62 2.38 Pacific 91.97 3.42 97.13 1.81 Central 81.01 13.61 95.11 3.70 Atlantic 69.00 19.21 85.47 9.46


SOCIAL SECURITY

Access to health insurance is low in Nicaragua for international standards, especially in rural areas. Being covered by health insurance reduces the probability that individuals spend more that what they can afford when facing health shocks and enhances higher service utilization. The social security system in Nicaragua is the primary source of health insurance in the country. As illustrated in Figure 7.19 health insurance coverage in Nicaragua is low as compared to that in El Salvador and Guatemala, especially in rural areas.

Figure 7.19: Only 9 out every 100 individuals in Nicaragua are covered by some type of health insurance [Estimates include all types of insurance, including Social Security]

27%

34%

20%19%

12%

19%

8%

10%

6%7%

3%

1%

11%

19%

9%

6%

0%

5%

10%

15%

20%

25%

30%

35%

40%

Guatemala 2000 El Salvador 2000 Nicaragua 2001 Honduras 1999

Capital Other Urban Rural NationwideSource: LSMS Surveys Source: World Bank 2006a

238

The majority of individuals with access to health insurance live in non-poor urban households. Access to insurance is highly concentrated in urban areas and in wealthier regions. While 13 and 24 percent of the overall population living in the Pacific and Managua regions respectively have access to health insurance, the same proportion is only at 4 percent in the Central and Atlantic regions (see Figure 7.20). The current condition of low coverage illustrates that the social security system provides inadequate financial protection, and it also creates inequitably distribution of health service delivery in Nicaragua. The major factors that prevent the broadening of social security coverage in Nicaragua are: (i) lack of institutional presence in rural areas where vulnerable populations such as seasonal workers or agricultural producers and indigenous people inhabit; (ii) lack of knowledge of social security benefits among low income workers, as well as a negative image of the social security that associates with high costs but fewer benefits than they expect; and finally (iii) lack of political will to improve the system and expand the coverage in particular, dealing with the informal labor force. Indeed, only 5.22 percent of the poor population (2.5 percent in poorest quintile and 3.4 percent in rural areas) has access to insurance.

Figure 7.20: Access to health Insurance is concentrated among the urban non-poor living in the Managua and Pacific regions

Coverage of Health Insurance by Socio-economic Status

18.17

5.002.29

6.4410.11

16.68

12.30

8.01

17.11

3.39

25.05

0

5

10

15

20

25

30

Non P

oor

Poor

Poore

st Quin

tile Q2 Q3 Q4

Riches

t Quin

tile

Non-In

dige

nous

Indige

nous

Non A

gricu

ltura

l

Agricu

ltura

l % c

over

ed b

y in

sura

nce

Coverage of Health Insurance by Geographic Location

4.09

18.41

12.94

4.88 4.42

24.92

0

5

10

15

20

25

30

Rural Urban Managua Pacifico Central Atlantico

% c

over

ed b

y in

sura

nce


239

CONSTRAINTS

Large distances, lack of medicines, and high cost constitute the main reasons why poor individuals not seek medical care when sick. Figure 7.21 and Table 7.4 present information why individuals did not seek care when sick by consumption quintile. Results indicate that that approximately 51 (80) percent of all patients who did not seek care when needed in the poorest (richest) quintile did so because their case was mild of because opted to self-medicate. Interestingly, the share of self-medication is highest among intervals in the richest quintiles (perhaps because they have more access to medicines, which are generally expensive for the poor). About 16 percent of all users in the first quintile did not seek medical care when needed because the nearest health facility was too far (these individuals probably resided in rural areas and in the Atlantic or central regions). The equivalent share was at 1.0 percent among users in the richest quintile. Affordability constraints and lack of medicines, as expected, are a more important factor for non-utilization among poor users than among non-poor users. In particular, while 10.2 (13.26) percent of all users who did not seek care in the bottom quintile did so because services are too expensive (there were no medicines available), the same share was at only 4.5 (3.4) percent for users in the highest quintile.

Figure 7.21: About 16 of every 100 patients in the poorest quintile do not seek health care when sick because the nearest health facility is too far

16.2510.9

5.942.22 1.06

7.52

7.89

8.91

7.205.93

13.26

10

8.49

4.723.43

10.16

6.46

8.55

6.56

4.51

05

101520253035404550

PoorestQuintile

Q2 Q3 Q4 RichestQuintile

%

distance poor quality no medicine expensive


Table 7.4: 6 to 8 percent do not seek health care when because of poor quality of health services provided Reasons Not to Seek care when sick

Mild case %

Distance %

Poor quality

%

No medicine

%

Too Expensive

%

Self-medicated

%

Other %

Poorest Quintile 21.62 16.25 7.52 13.26 10.16 29.17 2.02 Q2 26.86 10.9 7.89 10 6.46 35.46 2.43 Q3 28.91 5.94 8.91 8.49 8.55 36.29 2.91 Q4 35.77 2.22 7.20 4.72 6.56 41.38 2.15 Richest Quintile 33.45 1.06 5.93 3.43 4.51 46.89 4.73 Source: Authors using the 2005 Nicaragua LSMS.

240

The average distance (time) to get to the nearest health care facility in the Atlantic region is 17.4 km (1.2 hours). Poor and vulnerable individuals and especially those living in rural areas and in the poorest regions have less access to healthcare facilities. Poor users, especially indigenous one and those living in households engaged in agriculture, need to travel on average 13 to 14 Km to the nearest health facility (vs. less than 6 km among the urban non-poor). The time users spend to reach the nearest health facility follows the same trend: poorer and vulnerable individuals spend up to 1.3 hours to get to the nearest health care facility vs. less than half and hour among the urban non-poor.

Figure 7.22: Users in rural areas spend 3 times more time to get to the nearest health facility than users in urban areas

02468

101214161820

Urban Rural Managua Pacifico Central Atlantico

kilo

met

ers

0

0.2

0.4

0.6

0.8

1

1.2

1.4

hour

s

Distance (kms)to get the place of consultationTime to reach (hrs) the place


About 20 percent of all individuals in Nicaragua spend more than 10 percent of their income on health care. As illustrated in Figure 7.23, Nicaragua displays the highest share of individuals spending more than 10 percent of their income in health-care (at 18 percent). This share is significantly higher than that in Honduras and El Salvador (6 and 4 percent respectively). According to PAHO 2004 health indicators, about 95 percent of all private expenditures on health in Nicaragua are paid out of pocket vs. 80 to 85 percent in countries with similar income levels such as Bolivia, Ecuador, Honduras, Guatemala and El Salvador. In countries with more developed insurance and social security systems such as Colombia, Chile, and Uruguay, the same indicator is below 50 percent. All this suggests that out0of-pocket expenditures in health in Nicaragua are likely to constitute a heavier burden for households as compares to other countries in the region.

Figure 7.23: The share of households spending more than 10 percent of their income on health care in Nicaragua is three times higher than that in Honduras or El Salvador

18%

7%

3%

15%

8%

5%5%

1%1%

4%

1%1%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

>10% >20% >30%

Health Care Expenditures as a Percent of Income

Pro

port

ion o

f Popula

tion S

pen

din

g the

Iden

tified

Pro

port

ion o

f In

com

e on H

ealth C

are

Nicaragua 2001 Guatemala 2000 Honduras 1999 El Salvador 2000

Source: World Bank 2006a

241

Medicines constitute the main out-of-pocket expenditure on health, especially among the poor. Decompositions of average monthly per capita expenses for health in Nicaragua indicate the following distribution of expenses nationally: expenditures on consultations, 7 percent of total monthly expenses on health; medications, 55 percent (the highest among all other expenses); medical tests (including x-rays and other diagnostic tests), 11 percent; hospitalization, 8 percent; medical insurance, 8 percent; and other health care related expenses, 11 percent. Although medicines are the main expenditure on health at all quintiles, poor households spend relatively more on them (80 percent in the bottom quintiles vs. 41 percent in the highest quintile). Consultations account for a small share of overall expenditures on health at all income quintiles (between 3 and 7 percent); this result is not surprising as medical consultations are heavily subsidized in Nicaragua. Not surprisingly, users from non poor households spend more on items related to better quality services, such as insurance, tests, and hospitalization. Expenditures on insurance are only a significant fraction of overall expenditures on health (about 17 percent) among households in the highest quintile.

Figure 7.24: Expenditures on insurance and Hospitalization are a significant fraction of overall expenditures on health among households in the upper quintiles

3.37 5.13 6.50 9.04 6.79

81.28 79.4771.48

59.29

41.36

5.03 6.4011.92

12.67

11.29

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Poorest quintile Q2 Q3 Q4 Richest quintile

Consultation

Medication

Other

Insurance

Hospitals

Test

Source: Authors using the 2005 Nicaragua LSMS. Sample: Households with positive expenditures on health.

Expenditures on health account for 16 to 19 percent of overall non-food expenditures. Table 7.5 presents statistics on expenditures on health as a share of income and food consumption (in per capita per month) among households with positive expenditures on health. Results suggest that while richer households spend more of their income on health than poorer ones (10 percent in the highest quintile vs. 4 percent in the lowest quintile), the share of non-food consumption allocated on health is rather similar (19 percent in the highest quintile vs. 16 percent in the lowest quintile). Medicines are more expensive for the poor relative to their non-food consumption. Results in Table 5 suggest that while 8 percent of overall non-food consumption among households in the richest quintile is used to pay for medicines, the same share is at 14 percent for households in the poorest quintile. Finally, Figure 7.25 indicates that Expenditures on health are relatively more expensive in the Atlantic and Central Regions.

242

Table 7.5: Medicines are more expensive for the poor relative to their non-food consumption Poorest

Quintile Q2 Q3 Q4 Richest

Quintile Total health Expenditures As % of total Income 3.8% 4.6% 6.7% 8.2% 10.1% As % of total non-food consumption 16.06% 15.16% 17.37% 17.84% 19.33% Medicines As % of total Income 3.22% 3.76% 4.97% 5.06% 4.40% As % of total non-food consumption 13.65% 12.41% 12.81% 10.98% 8.39% Source: Authors using the 2005 Nicaragua LSMS. Sample: Households with positive expenditures on health.

Figure 7.25: Expenditures on medicine is relatively more expensive in the Atlantic and Central Regions

7.3%

10.0%

12.1%

14.1%

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

Managua Pacifico Central Atlantico

As of % of Non-foodConsumption

As of % of Income

Source: Authors using the 2005 Nicaragua LSMS. Sample: Households with positive expenditures on health.

243

CONCLUSIONS

The Nicaraguan health care system faces several challenges that need to be addressed in order to improve the health status of its population: (i) inefficiencies in the allocation and utilization of resources, (ii) low level of financial protection (iii) high out-of-pocket expenses made by the poor, (iv) difficulties in access and poor utilization of health care services, and (v) an unregulated private sector and limited capacity of MINSA to perform its stewardship role to ensure pro-poor strategies and an efficient health system. Efforts to face these problems should be made within an equitable framework, since the poor and indigenous population has not been benefited as much the population on average from general improvements in health outcomes. Even for immunization and reproductive health services, which are free of charge, there are differences in utilization between better-off and poor households. Current health disparities in Nicaragua will grow wider unless action is taken to address the needs of the most disadvantaged and vulnerable sectors of the population. Access, utilization, and financing essential Health Services should be expressed explicitly as a policy objective of the national Poverty Alleviation Strategy.

Nicaragua is not likely to achieve its country-specific health targets under the Millennium Development Goals, specifically in maternal and child mortality and in child malnutrition. The trend of progress in health indicators is slowing as the stage of “easy” gains have been overcome and achieved. Reaching child and infant mortality goals will require more attention and resources to reduce neonatal mortality and to the delivery of an Integrated Health Care Model (Modelo de Atencion de Salud, MAIS). Since the most effective measures for reducing neonatal mortality are also effective in reducing maternal mortality, the delivery of integrated health care service package with a multi-sectoral approach will be doubly beneficial. Specifically, the strategies for maternal and child care should include the following:

• Promote child, and maternal health care preventive services, with focus in earlier (first trimester) and more frequent prenatal visits (at least five), as well as broader coverage of postpartum care for women.

• Expand access to institutional births, as the share of women delivering in health care centers is still low for poor and rural women (being especially low in the Atlantic region). This will require steps on both the demand side and the supply side. Nicaragua’s strategy of establishing Casas Maternas have shown promising results and seem to be a good and expanding intervention.

• Prevent discontinuities in immunization coverage in CA4, particularly in the last doses of DPT and the measles vaccine. Further research will be important to identify the factors explaining the discontinuity of immunization.

• Integrate key interventions into basic packages that are managed and financed by Ministry of Health (MOH). Up to date, most of those key health interventions have been partially supported by donor financing, e.g. family planning services. It is essential that Nicaragua integrates these interventions within their MOH budgets to ensure their sustainability.

Furthermore, Nicaragua could improve health outcomes of the poor by addressing the marked inefficiencies in current health spending:

• In allocating fiscal and other resources Nicaragua should move away from historical budgeting to a system based on health needs, especially of vulnerable populations, and avoid concentrating resources in Managua and wealthier regions. More resources need to be targeted towards prioritizing primary care, prevention, and health promotion interventions. In Nicaragua, locally-driven results-based budgeting should be strengthened to reverse the allocation process which favored metropolitan areas and hospital care.

• The country should also move away from basing deployment of human resources on historical patterns. Doing so has contributed to having few health workers in poor rural and urban areas. While Nicaragua deploys social workers based on an assessment of each region’s health risks and needs, it

244

has not yet extended this practice to health workers. Similar to other countries if the region, redistribution of health personnel through a centralized agency using health needs criteria, has produced some promising results.

• Over-reliance on physicians is another important source of inefficiency and Human resources imbalances. The existing scarcity of nurses needs to be addressed by hiring more nurses and auxiliary personnel, especially for primary health care, revising the current medical education system, which emphasizes physician training and places less value on nursing, and creating greater incentives to enter the nursing profession by improving their salaries, which are often less than half of physician salaries.

Since a high proportion of the population in Nicaragua remains completely uninsured, poor families are vulnerable to external health shocks that take them into poverty. Even the poor, who are typically seen as the target of government financed actions, often opt to pay a substantial proportion of total Health consultations, diagnostic services, and medicines out of pocket. Out of pocket (OOP) expenditures represent a high proportion of the poor’s income. Reducing OOP among the poor is a priority task and requires more than a supply side intervention. The pro-poor strategy should increase access to health care services, especially in rural and remote poor areas with low utilization of public health care facilities. High cost of transportation and lack of medicines are important factors explaining why poor individuals do not seek medical care when sick. In this context, INSS have to play a key role to improve health equity. Currently, government subsidies to social insurance are shown to benefit the richest groups.

Addressing all these issues will request coordinated actions on both the demand and supply side. Measures are required to increase supply of health care services, especially in poor and underserved rural areas. Alternative models of services delivery to improve access among for the most vulnerable segments of the population could be financed and regulated by the Government. These modalities can include different options with different comparative advantages such as: subcontracting services to non-governmental organizations, strengthening MOH health care centers, including flexible human and other resource management, deploying MOH mobile teams, and implementing decentralized community management models. Nicaragua will have to design their own strategy to reinforce supply and improve access to health and nutrition services in the poorest and remote areas based on successful local experiences. Experiences of PROCOSAM, Casas Maternas, and NGO’s contracting in family planning and reproductive health services are valuable. Demand side strategies could be also implemented to help achieving the objectives of the National Health Plan. For instance, conditional cash transfers (CCTs) could be implemented to overcome some financial and cultural barriers that prevent full access to services as part of the extension service coverage strategy. Existing CCTs could be used as a complementary tool to target public subsidies towards the most vulnerable sectors of the population and as an opportunity to improve simultaneously access to nutrition and primary health services. This intervention also requires an appropriate exit strategy and effective health and nutrition counseling components, which promotes long-term healthy behaviors.

Since the concepts of universal coverage, free services, and priority health care interventions are not necessarily matching to the current public resource allocation and health outcomes, introducing health equity dimensions into the health system performance is essential. Nicaragua has the political and organizational challenge of improving the effectiveness of the Anti-Poverty strategy that covers poor’s needs and of increasing their health outcomes. This requires selective and effective coverage to the poor, including: the promotion of social accountability schemes, technical quality, organizational improvements; availability of trained health workers and medicines; multi-sector coordination to improve health and nutrition outcomes; targeting investments to less served areas; increasing outreach of core interventions; strengthening private and public partnerships to reach remote areas; improving drug management; focusing government health expenses on core services; and encouraging the private sector to invest and complement in other services.

245

REFERENCES

Asling-Monemi, Kajsa, Pena, Rodolfo, Ellsberg, Carroll Mary, Persson, Ake Lars. Violence Against Women Increases the Risk of Infant and Child Mortality: a Case-Referent Study in Nicaragua, 2003, Bulletin of World Health Organization.

FESAL (Encuesta Nacional De Salud Familiar), El Salvador’s National Family Health Survey, 98/99 and 2002/3. Www.Fesal.Org.Sv

PAHO, 2002a. Estudio Delphi: problemas presentes y futuros de los recursos humanos en salud. Programa de Desarrollo de Recursos Humanos. División de Desarrollo de Sistemas y Servicios de Salud.

PAHO, 2002b. La salud en las Américas. Publicación científica y técnica No. 587. Volumen II. Washington DC.

World Bank, 2005a. Guatemala Public Expenditure Review. Latin America and the Caribbean Region. Washington DC.

World Bank, 2005b. Nicaragua Health services Extension and Modernization (2nd APL).

World Bank, 2005b. Project Appraisal document to the Republic of Honduras for a Nutrition and Social Protection Project. Project Appraisal Document, report number 31740-NI, Washington DC.March.

World Bank, 2005c. El Salvador, Poverty Assessment. .Latin America and the Caribbean Region. Washington DC.

World Bank, 2005d. El Salvador Social Protection Project, Project Appraisal Document. Report No.32648-SV, September 26.

World Bank, 2006. Project Appraisal Document on Guatemala.

World Bank, 2006a. Key Issues in Central America Health Reforms: Diagnosis and Strategic Implications, Volume II. Washington DC.

World Development Indicators, 2006, World Bank, Washington DC.

World Health Organization (WHO), 2005.Core Indicators. Geneva. http://www.who.int/whr/2005/whr2005_en.pdf

World Health Organization, 2004. The World Medicines Situation. Geneva, Switzerland. http://w3.whosea.org/LinkFiles/Reports_World_Medicines_Situation.pdf

World Health Organization, 2006. Geneva

246

8. ACCESS TO SAFE DRINKING WATER AND BASIC SANITATION IN NICARAGUA

By Simon Zbinden and Diego Angel-Urdinola*

Access to safe drinking water and basic sanitation is a key basic service that has a direct and significant impact on human development. As with other basic services like education and healthcare, the lack of access to basic water and sanitation services is both a cause and a symptom of poverty114. Because causality runs both ways, inequality in access to water and sanitation is a good indicator to describe and reflect poverty levels among different population groups. Inequality of access to safe drinking water mirrors to a large extent the existing inequalities of opportunities for human development in general. In Nicaragua, like in other parts of the world too, the impact of deficient water and sanitation services falls primarily on the poor and extremely poor in rural and peri-urban areas. Excluded from the basic public services, these population groups make their own inadequate arrangement or pay excessively high prices to water vendors for meager water supplies, as it has increasingly occurred in recent years in peri urban areas in Managua. By not having access their poverty is further aggravated and their productivity impaired. Yet besides the overwhelming human development arguments, there are also powerful economic and environmental reasons. The World Health Organization estimates the long term return on investments in water and sanitation to be at between $5 and $28 per dollar invested115. In Nicaragua, important productive sectors, such as tourism and agriculture, not only depend heavily on water and sanitation services but also on a healthy environment. Nowadays however, poor sanitation infrastructure and the lack of waste water treatment is a serious threat to Nicaragua’s sensitive ecosystem. Not addressing the problem appropriately may lead to further irreversible damage at the detriment of future generations.

Progress in access to basic water services at the national level has been stagnant over the last decade. In terms of access to safe drinking water116 among Nicaraguans, the data from the three last

* The authors are with the World Bank and the Water and Sanitation Program (WSP) Nicaragua. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP), Nelson Antonio Medina Rocha (Consultant, ETWAN) and Ulrich Schopmeyer (KfW) for their valuable comments and suggestions. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent. 114 Jarman, J., 1997. “Water supply and sanitation” in Beall, J., (ed.), A City for All: Valuing the Difference and Working with Diversity, Zed Books, London. 115 WHO, 2004. Evaluation of the costs and benefits of water and sanitation improvements at the global level. Report prepared by Guy Hutton and Laurence Haller. World Health Organization, Geneva, Switzerland. 116 Neither census data nor LSMS data allows determining the true figures of access to safe drinking water based on the criteria set by the Joint Monitoring Program of the United Nations (www.wssinfo.org). Access to water in the present analysis refers to the following water sources: households with i) pipes inside the house, ii) pipes outside the house (within the yard), iii) public standpipe, iv) private well, and iv) public well. However, the shown figures may actually overestimate the true access figures. The category “private well” includes both protected/improved but also unprotected self-dug family wells. The latter do generally not provide water quality sufficiently safe for human use. According to the definition of the Joint Monitoring Program unprotected wells are not considered a safe drinking water source and hence do not count towards the Millennium Development Goals. Since unprotected “private wells” are common in rural Nicaragua and count for between 16% and 19% of all indicated water sources - in rural areas even up to 40% - the presented access figures may not be directly equated with real coverage of safe drinking water. On the other hand, the present analysis does not count the category “from other house/neighbor/company” as access. Yet, this category may in fact include a certain percentage of households, who do indeed enjoy access to safe drinking water. For these reasons, the shown figures are of indicative nature only. The national water authority, CONAPAS, publishes official coverage figures. It estimates access to water - based on the rough assumption that

247

censuses shows that significant progress has been made between the seventies and nineties, yet progress appears to have come to a halt over the last decade (Figure 8.16). According to 2005 figures, water coverage (= access117) reached 93.1 percent in urban areas118, 63.4 percent in rural areas, and on average 80.3 percent. A major increase had been achieved in rural areas between the seventies and nineties. The stagnation (or even slight decrease) during the last decade does not necessarily mean that no investments were made in new water service infrastructure to expand access. It rather means that investments were merely keeping up with population growth; insufficient to expand coverage. On the other hand, investments tended to be allocated to improve water service levels (Figure 8.17). Whereas the overall water coverage remained stagnant, access through piped systems (pipes inside or outside the house, but within the yard) has seen an expansion in both rural and urban areas over the last thirty years. Although no data is available on investments in rehabilitation of failed water systems, it is probable that a significant amount of sector resources has been allocated to rehabilitate existing systems; most of them water systems that prematurely failed due to inappropriate management practices. Based on these findings, it seems to be unlikely that Nicaragua will reach the Millennium Development Goals in access to safe drinking water set for the year 2015, unless existing investment levels and patterns are altered and management practices improved.

Figure 8.16: Access to water in Nicaragua119. Figure 8.17: Access to water in Nicaragua (through piped system)120.

69.9%

49.2%

91.3%

81.4%

64.9%

94.3%

80.3%

63.4%

93.1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Urban Rural Total

1971

1995

2005

72.7%

5.9%

38.7%

83.8%

18.4%

55.5%

86.2%

26.9%

60.7%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Urban Rural Total

1971

1995

2005

Substantial disparities in water coverage persist between poor and non-poor population groups. Water coverage decreased slightly among poor and extremely poor in recent years. The lack of access to an improved water source and income poverty are highly correlated in Nicaragua. Among the extremely poor, only 62 percent have access to an improved water source, whereas 92 percent of the non-

only half of the category “private wells” may be counted as a safe water source - to be at 76.7% for the national level, 95.5% for the urban, and 52.8% for rural area (www.conapas.com.ni). 117 This chapter uses equally “access” as “coverage” to indicate the share of the population served with water and/or sanitation services. 118 Definition of urban areas: Localities with 1,000 or more inhabitants and certain characteristics such as routes, electricity services, commercial settlements, etc.. Source: National Statistics and Census Institute (Instituto Nacional de Estadística y Censos, INEC). 119 Access to drinking water refers to households with: i) pipes inside the house, ii) pipes outside the house (within the yard), iii) public standpipe, and iv), public or private well. Data sources: Nicaragua National Censuses 1971; 1995; 2005. 120 Access to drinking water through piped system refers to households with: i) pipes inside the house, ii) pipes outside the house (within the yard). Data sources: Nicaragua National Censuses 1971; 1995; 2005.

248

poor did so (Figure 8.18). According to the last three LSMS121, water coverage even slightly decreased among the poor and extremely poor during the last eight years. The decrease may be attributed to insufficient capital investments in new water systems in order to keep up with the relatively high population growth among these groups, let alone any increase in coverage. Indirectly, it may be also be the result of sector resource allocation policies and modalities ineffective to target the poor. A recent World Bank financed study on public spending in Nicaragua backs this presumption122. The study, which presents a more detailed analysis of distributional effects of public spending, reveals that in 2005 the two richest income quintiles benefited most from sector capital investments for water systems expansion. The poverty gaps are further mirrored in the disparities in access to water through piped systems, which present a better and therefore more costly water service. In 2005, some 88 percent of the households in the richest income quintile received their water from a tap inside or outside the house, but only 28 percent of the households in the lowest income quintile (Figure 8.19). However, the data also shows that more people of the lower income quintiles nowadays enjoy water from a tap than did eight years ago. These findings underpin the results from the census data showing stagnation in the overall coverage, but increases in water coverage through piped systems.

Figure 8.18: Access to water in Nicaragua among different income groups123.

Figure 8.19: Access to water in Nicaragua among income quintiles (through piped

systems)124.

64.5%

73.7%

89.9%

63.5%

73.7%

90.2%

61.7%

70.8%

91.7%

50%

60%

70%

80%

90%

100%

Extreme Poor Poor Non-poor

1998

2001

2005

20%

30%

40%

50%

60%

70%

80%

90%

100%


2005

1998

Disparities with respect to water coverage not only prevail between poor and rich, but also exist across regions and between urban and rural zones within the regions (Table 8.33). While in Managua 95 percent of all households enjoyed access to water in 2005, some 89 percent did so in the Pacific region, 74 percent in the Central, and only a mere 56 percent of the households in the Atlantic region. Similar to the figures at the national level, rural areas are generally poorly endowed with secure water services. The lack of access to a safe water source is particularly marked in rural areas in the Central and the Atlantic regions, with only 61 and 42 percent coverage, respectively. These figures, together with the poor overall progress made in the last decade, point to the problem of marginality of areas with low water coverage. Whereas more accessible and more developed areas such as the Pacific region also tend to have higher water coverage, poorer and marginal areas such as the Atlantic region are much more likely to be

121 Due to the survey methods used for the LSMS, coverage figures may slightly differ from those stemming from censuses. 122 Gasparini, L., et al., 2007. La Distribución del Gasto Social en Nicaragua. Centro de Estudios Distributivos, Labores y Sociales (CEDLAS). Universidad Nacional de la Plata, La Plata, Argentina. 123 Data sources: LSMS 1998; 2001; 2005. 124 Data sources: LSMS 1998; 2005.

249

disregarded and excluded. Whether implicitly or explicitly, water authorities, donors, and NGOs alike seem to have preferred better and more accessible locations to build or finance water systems. Given the clear geographical characteristic of the actual coverage, it is likely that marginal costs will increase for further water coverage expansion in future.

Table 8.33: Access to water across regions in Nicaragua in 2005 (in % of total households)

Managua Pacifici) Central and Northii) Atlanticiii) Main water source: Total Urban Rural Total Urban Rural Total Urban Rural Total 1. Pipes inside the house 65.2 65.8 16.5 45.7 59.9 8.2 29.5 30.0 5.9 14.2 2. Pipes outside the house 25.7 23.9 21.5 22.9 25.6 15.7 19.8 10.8 5.7 7.5 3. Public source 0.8 0.6 3.5 1.8 3.1 8.7 6.4 5.1 1.9 3.0 4. Public or private well 3.5 4.6 37.8 18.1 3.4 28.5 18.1 37.9 28.4 31.6 5. Spring 0.3 0.1 2.8 1.2 0.5 24.2 14.4 1.4 35.3 23.7 6. River/stream/lake 0.4 0.0 2.5 1.0 0.5 8.7 5.3 0.9 19.0 12.8 7. Truck/oxcart 0.7 0.2 1.8 0.8 0.7 0.2 0.4 0.3 0.0 0.1 8. From another house 2.8 4.3 12.5 7.7 5.7 5.1 5.4 12.7 3.3 6.5 9. Other 0.5 0.4 1.2 0.7 0.7 0.7 0.7 0.8 0.5 0.6 Access to water (Σ1;2;3;4) 95.2 94.9 79.3 88.5 91.9 61.1 73.8 83.8 41.9 56.2 Source: LSMS 2005. i) Includes the departments of Chinandega, León, Managua, Masaya, Granada, Carazo, Rivas ii) Includes the departments of Nueva Segovia, Jinotega, Madriz, Estelí, Matagalpa, Boaco, Chontales iii) Includes the departments of Río San Juan, R.A.A.S, R.A.A.N.

The majority of people without access to safe drinking water lives in rural areas, in particular in the Central and Atlantic region (Table 8. 34). With respect to absolute numbers, the vast majority of people without access to water lives in rural areas. Of an approximated total of 1,089,000 Nicaraguans who lack access to water, 77 percent (839,000 persons) live in rural areas. Of these, more than 80 percent live in the Central and Atlantic regions, (397,000 and 286,000, respectively). Urban dwellers without access live mainly in the Pacific region (103,000), with the majority in the city area of Managua (60,000). In the urban context, the lack of water coverage principally prevails in poor urban neighborhoods and settlements125. The figures indicate that not only water coverage in percentage is low in rural areas; they also show that - in absolute numbers - the majority of people without water access lives in rural areas. The findings call for a more determinate strategy with respect to water coverage in rural areas, if the government is to achieve the MDG. They call for investment policies and sector resource allocation decisions that take much more into account the considerable backlog in rural marginal areas.

Table 8. 34: Number of people without access to water across regions in Nicaragua.

Managua Pacific Central Atlantic Without access to water Total Urban Rural Total Urban Rural Total Urban Rural Total

% 4.8 5.1 20.7 11.5 8.1 38.9 26.2 16.2 58.1 43.8 Households

No. 11,572 19,833 31,210 62,615 10,224 72,612 82,836 6,870 48,011 54,881

People No. 60,074 102,689 156,688 319,451 50,558 397,107 447,665 36,461 285,631 322,092

Sources: LSMS 2005, National Census 2005, author’s estimations.

125 Whereas in most big cities and capitals around the world poor urban neighborhoods are mainly peri urban areas scattered at a peripheral belt around the city, Managua is different. In Nicaragua’s capital the poor neighborhoods and settlements are scattered throughout the city area, whereas the richer, newly build urbanizations tend to be located at the periphery of the city.

250

Some progress has been made in access to basic sanitation infrastructure over the last decade, but only little in terms of new connections to a public sewage system. Most small towns remain without sewage systems and waste water treatment. In 2005, about 85 percent of all households had access to basic sanitation infrastructure (Figure 8.20). In rural areas, where the most widespread form of sanitation infrastructure is the latrine, the coverage has been steadily increasing from 56 percent in 1995 to almost 70 percent in 2005. However, findings from the 2001 LSMS suggest that a considerable part of this increase must have taken place before 2001126 Already in 2001, the coverage was at about 85%, which implies that little progress has been achieved between 2001 and 2005. Besides the many latrine projects implemented in the past decade by cooperation agencies, NGOs and municipalities alike, a considerable part of the boost before 2001 may be attributed to the vast Hurricane Mitch relief efforts in 1998 and 1999, during which myriads of relief and cooperation agencies temporarily increased their sector spending in an attempt to reinstall basic water and sanitation infrastructure. Little progress has been made in terms of connections to a public sewage system (Figure 8.21). The coverage only increased by about two percent over the last decade and reached 19 percent in 2005. Sewage systems almost exclusively exist in urban areas. Both capital investments and operation spending in sanitation tend to be pro-rich, because richer households are more likely to have a connection to a sewage system. The cited analysis of public spending in Nicaragua estimates that 80 and more percent of the benefits generated through public spending in sanitation is reaped by the two richest income quintiles. According to the Nicaraguan Water and Sanitation Sector Analysis127, only 29 urban locations including Managua count on a public sewage systems, and in only 20 cases the waste water gets some form of treatment. Although there is no reliable data available, it is likely that the quality of the effluents is low. Whereas in Managua, a waste water treatment plant is currently built with funds from the German KfW, the vast majority of small towns remains without any form of sewage system, let alone any form of waste water treatment. In an estimated 160 localities with between 2000 and 50’000 inhabitants, the waste water from households and local industry often enters rivers and lakes untreated or infiltrates directly into the soil, with far-reaching consequences for Nicaragua’s sensitive ecosystem and groundwater, respectively.

Figure 8.20: Access to basic sanitation infrastructure in Nicaragua128.

Figure 8.21: Access to sanitation infrastructure connected to a public sewage

system

90.7%

17.6%

53.5%

94.3%

55.6%

77.6%

95.7%

69.5%

84.8%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Urban Rural Total

1971

1995

2005

15.3%

0.2%

31.0%

17.2%

0.6%

29.9%

19.4%

0.0%

33.1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Urban Rural Total

1971

1995

2005

126 World Bank, 2003. Nicaragua Poverty Assessment: Raising Welfare and Reducing Vulnerability. World Bank Report No. 26128-NI. (Page 12). 127 “Análisis Sectorial de Agua Potable y Saneamiento de Nicaragua”, published in 2004 by the Nicaraguan government with the support of the World Health Organization. 128 Access to basic sanitation refers to households with: i) toilet connected to a sewage system, ii) toilet connected to a septic tank, iii) toilet with discharge into a river, iv) latrine. Sources: Nicaragua National Censuses 1971; 1995; 2005. Published by INEC.

251

Water service quality (continuity of water supply) has been deteriorating appreciably in recent years. Mirrored in an increased frequency of media covered water related incidents such as protests, assaults and water piracy, the continuity in the water supply has worsened appreciably in recent years. In the 2005 LSMS, only about two thirds of Managua’s population indicated to have permanent water supply, slightly more in the Pacific and the Central regions, and only a mere 50 percent did so in the Atlantic region (Table 8.9). Water supply did not seem to be significantly better among richer income groups. Besides, it is likely that the true situation may actually be worse than the LSMS figures imply. The LSMS interviews were conducted between August and October 2005, during the period of highest rainfall. Yet water availability and hence water supply varies considerably over a year. The annually recurrent water stress during the dry season between January and May affects water supply negatively. In Managua, which accounts for a quarter of Nicaragua’s population, but also in smaller cities such as Juigalpa or Jinotepe, extended water shortages are now common, especially but not only during the dry season. Although richer neighborhoods do not seem to be spared by water supply cuts, the poor are over proportionally more affected. While the rich have water tanks to bridge over water cuts, the poor are forced to make their own inadequate arrangements. Recently, cases have been reported where private well owners have taken advantage of the desperate situation and started to sell water the poor in barrels at prices that surpass the official tariff by up to several hundred percent129. The precarious water situation in the urban areas is to a large extent the result of the deteriorating state of the water and sanitation utility, ENACAL.

Table 8. 9: Continuity of Water Supply in 2005 Regions Income Quintiles Water Supply

Managua Pacific Central Atlantic I II III IV V

Permanent (% of househ.) 67.1 77.5 74.1 49.4 68.2 74.9 69.3 70.0 72.2

Partial (%) 33.0 22.5 25.9 50.6 31.8 25.1 30.7 30.0 27.8

Days / week 6.0 5.2 4.3 5.0 4.9 4.9 5.4 5.4 5.5

Hours / day 9.2 11.0 10.6 8.6 10.7 10.2 9.9 9.6 9.8

Source: LSMS 2005.

Reflected in decreasing service quality, the state of Nicaragua’s urban water and sanitation utility, ENACAL, has been deteriorating steadily over recent years. The reasons for the deterioration of the water service quality have to be found in a mix of mismanagement and structural factors such as low labor productivity, low micro metering rates, leakages and hence high percentage of unaccounted-for water losses, as well as water tariffs far below production cost level due to a politically motivated tariff freeze since 2001130. In several larger cities including Managua however, the water supply cuts are also to be blamed to rapid population growth. New urbanizations and settlements were connected to the existing system without the necessary increases in water production, due to a lack of resources. In Managua, the hydraulic balance of the water system seems to be seriously hampered, leaving several neighborhoods without water for months131. On the income side, the unaccounted-for water losses surpassed the 50

129 An estimation based on data from the 2005 LSMS show that cubic meter prices for water delivered in barrels can be up to five times as high as the official water tariff. The average “barrel” price indicated in the 2005 LSMS was 11.2 Cordobas (= $0.66), whereas the official water tariff oscillated between 2.1 and 6.5 Corbobas per cubic meter depending on the residence area and the total consumption. 130 Water tariffs have been freezed in nominal Cordoba terms in 2003 by the national water regulator (INAA). The tariffs are estimated now to be between 30% and 72% below production cost levels, due to the steady decline of the tariff’s purchasing power in real terms and also due to a dramatic increase in the company’s energy bill. 131 ENACAL, 2007. Statement and figures made and presented by Ruth Herrera Selma at the “National Water Day”, in Las Piedresitas, Managua, March 22, 2007.

252

percent threshold in 2006131. Low labor productivity and generous collective labor agreements continue to be a heavy financial burden to the company. ENACAL’s thirty unions managed to prevent virtual all recent reform attempts to increase labor productivity. As a result, ENACAL has been a deficit and loss making public enterprise in recent years, finding itself today at the verge of illiquidity131. The company’s net income through collection stagnated in real terms in recent years, while operational costs have increased substantially, largely due to dramatic increases in the energy bill (Table 8.10). Preliminary estimations reveal that the deficit for 2006 amounts to about 382.6 million Cordobas (equivalent to US$21 million)131. Maintenance costs were reduced by skipping ordinary maintenance tasks, with corresponding detrimental effects on the infrastructure’s lifespan. The tariff freeze was initially meant not to hurt the poorest, yet today the poor and extremely poor suffer most from the company’s incapacity to provide a minimum service level. The increasing unrests and the poverty dimension of the problem have pushed the problem up in the political agenda. Although the Ortega administration has called water its top priority after energy, it hasn’t taken visible action yet. In the long run, ENACAL will not get around profound structural reforms and a tariff increase and unless the government is willing to adopt a more explicit policy to subsidize water services for the poor. At all events, a significant cash injection will be required for visible improvements in service quality in the very short run in order to restore clients’ confidence.

Table 8.10: Operational cash flow of ENACAL (in thousand of US$)

2002 2003 2004 2005 Operational cost 16,729 22,663 24,888 24,714 Maintenance cost 12,196 6,814 2,863 3,019 Registration and collection cost 5,798 5,858 6,455 6,759 Administration cost 7,325 10,987 6,605 5,701 Financial cost 6,828 9,597 8,046 4,637 Total cost 48,876 55,919 48,857 44,830 Collection 41,660 38,832 38,801 38,577 Deficit -7,216 -17,087 -10,056 -6,253 Cost coverage ratio 0,85 0,69 0,79 0,86 Source: ENACAL, 2006 (elaborated by Carlos Diaz, Consultant BID).

Water service quality remains also critical in other urban zones across the country, in particular in small towns. Whereas Managua seems to be the most pronounced case of poor water service delivery, the issue is equally critical in many small towns across Nicaragua. The widespread poor water service quality offsets to a considerable extent the potential benefits that the relatively high water coverage rates in urban areas could generate. As in Managua, the burden of poor service quality in these small towns falls primarily on the poor and extremely poor. Many of these small towns struggle with high population growth, including a strong inflow from surrounding rural areas132. Nicaragua counts 73 water systems in small towns between 5,000 and 50,000 inhabitants, from which 60 are run by ENACAL, seven are run by the service provider of Matagalpa and Jinotega (AMAT and EMAJIN) and six are run by municipal-owned service providers133. In addition, 105 systems exist in localities between a 1,000 and 5,000 inhabitants. 77 of these systems are run by ENACAL, eight by AMAT and EMAJIN, and 20 by municipal service providers. The underlying problems and structural shortcomings encountered among many of these systems are essentially similar to those of ENACAL in Managua. Although these service providers (other than ENACAL) do enjoy some degree of autonomy, their entrepreneurial scope is limited due to a legal framework that does not promote a higher degree of private sector participation. Tariffs, for example, cannot be freely determined in accordance to the specific marginal production cost in each

132 In 14 out of 17 departments (including the autonomous regions RAAS and RAAN) the percentage of people living in urban areas has increased during the last decade (INEC, 2006. Resumen Censal). 133 Source: ENACAL, 2004. Management Planning Unit (Gerencia de planificación).

253

town, but are administered centrally by INAA. Major changes in the legal framework and better management capacities would be required to increase the long term financial and technical sustainability of such systems in small towns.

In rural areas, water systems often lack long term sustainability. Insufficient water quality is a widespread problem. In addition, a lot of systems are vulnerable to natural disasters. Under the former government, Nicaragua’s social investment fund (FISE) was assigned the responsibility of the promotion of water and sanitation in rural areas. In the framework of the decentralization process of the recent past, FISE has collaborated with municipalities for its social investments in water and sanitation infrastructure in rural areas. However, although the municipal law - in contradiction to the sector laws - does assign some competences to the municipalities, their involvement has been limited so far. In its activities, FISE has focused mainly on infrastructure investments rather than on sustainability of the systems. The operation and maintenance is delegated to communal water committees (CAP)134, after they receive some ad hoc training in operating the system during the construction period. After that, the CAPs do not receive support on a more continuous basis. Yet, as the high rate of premature failures among rural water system implies, the CAPs often lack the capacity to sustainably operate and maintain the water system over their entire technical lifespan without external support. With a few expectations, FISE so far did not succeed in establishing Municipal Water and Sanitation Units in a large number of municipalities, as initially planed. These technical units were meant to be a support to the CAP in its operation and maintenance activities, help to monitor the water quality on a continuous basis, and supply chlorine for permanent water disinfection. Widespread problems with water quality due to bacterial contamination and an unacceptably high rate of water system failures are largely the result of the lack of technical assistance and monitoring in rural areas135. The institutional incapacity to deliver a minimum level of attendance in rural areas, which includes FISE and the municipalities, is particularly poverty relevant as the problems are aggravated the more marginal and hence the poorer the communities become. The rural infrastructure, especially in water and sanitation, remains highly vulnerable to natural disasters such as floods, hurricanes and earthquakes. Risk mitigation measures are seldom taken into account when constructing new water systems.

Figure 8.22: Infant Mortality in Latin America136

CHL

MEX

CRI

URY

BRA

DOM

COLPAN

VEN

PERELSPRY

GTM

ECU

NIC

HON

BOL

0

10

20

30

40

50

60

70

80

7.50 8.00 8.50 9.00 9.50 10.00

Natural Logarithm of GDP percapita in 2004

Infa

nt M

orta

lity

Rat

e pe

r 1,

000

live

birt

hs in

200

4

134 Comité de Agua Potable (CAP) 135 “Análisis Sectorial de Agua Potable y Saneamiento de Nicaragua” (p. 222ff) 136 WDI dataset.

254

Figure 8.23: Under-five mortality in Latin America137.

BOL

HON

NIC

ECU

GTM

PRYELS PER

VENPAN COL

DOM

BRA

URY

CRI

MEX

CHL

0

20

40

60

80

100

120

140

7.50 8.00 8.50 9.00 9.50 10.00


Und

er-fiv

e M

orta

lity

Rat

e pe

r 1,

000

live

births

in 2

004

Poor hygiene practices remain a serious problem mainly among the poor in peri urban and rural areas. Together with the poor water quality, inadequate practices contribute to the poor sector related health outcomes. Water-borne infectious diseases are a main determinant to MDG health indicators such as infant mortality, under-five mortality and malnutrition. Diarrhoeal diseases are amongst the top three killers of children in the world today138. As Figure 8.22 and Figure 8.23 illustrate, both infant mortality and under-five mortality, respectively, are relatively high in Nicaragua (NIC) compared to other countries in the region.

Figure 8. 24: Access to water and prevalence of Acute Diarrhea Diseases in 15 departments of Nicaragua139.

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800 1000 1200

Acute Diarrhea Diseases per 10'000 habitants

wat

er c

ove

rag

e in

%

137 WDI dataset. 138 WHO, 2000. World Health Report. Geneva, World Health Organization, 2000: 164. 139 Acute Diarrhea Disease figures from 2001, 2003, 2005 in 15 departments regressed with census water coverage figures. Sources: National Epidemiological Surveillance System (Sistema Nicaragüense de Vigilancia Epidemiológica Nacional SISNIVEN), National Census 2005.

255

Acute diarrhea diseases (ADD) is largely caused by the consumption of unsafe drinking water and poor hygiene practices, such as not washing the hands with soap before eating and after going to the bathroom140. Although the prevalence of ADD does not exclusively depend on access to safe drinking water, the relationship is evident. Health data from the years 2001, 2003 and 2005 in fifteen departments show a clear negative correlation between water coverage and the prevalence of ADD, meaning that the lower the water coverage the higher the prevalence (Figure 8. 24). The departments with the highest prevalence of ADD are also the ones with the highest share of poor and extremely poor. The data further provides an indication that adequate water handling and effective treatment of in-house stored water for human use, such as filtering, solar or chlorine disinfection, is little widespread. Adequate water disinfection methods could offset to a large extent the lack of access to a safe drinking water source. Although no ample data is available on hygiene practices, a series of surveys presented in grey literature shows that hygiene practices and habits generally cause anxiety, especially among the poor and extremely poor141. It is likely that improvements in hygiene practices would deploy a significant and positive impact on health outcomes. Hygiene practices have apparently received little attention in the past and hence relatively little sector funds were provided for promotion of better practices. A more integrated approach to water, sanitation and hygiene behavior that puts considerable more weight to the latter will be required in future.

The relatively poor overall performance of the sector in the past decade is to be predominantly related to a) the political and institutional shortcomings in the sector, and b) to insignificant resources from the state budget allocated to the sector142. The sector’s relative inefficiency and inefficacy is largely to blame for the lack of sustainability and the slow coverage increase, especially among the poor peri urban and rural areas. A serious commitment at the political level is inevitable if the sector performance is to be improved significantly. The water and sanitation sector comprises of tree major state institutions: i) the national urban water utility (ENACAL), which operates water and sanitation systems in urban areas143; ii) the water regulator (INAA), which is responsible for the regulation, including surveillance of service quality and tariff adjustments; and iii) the Social Investment Fund (FISE), which is in charge of the promotion of water and sanitation systems in rural areas. NGOs play an important role mainly in rural areas144. The National Water and Sanitation Commission (CONAPAS) is the sector’s policy body. Its board comprises members from seven state institutions145. However, this sector architecture, in particular the status of CONAPAS and FISE, remains on unstable grounds as it based on presidential decrees rather than on a solid legal basis. Since decrees may be easily overturned by the next government, this lack of a solid legal framework hampers the long term sustainability of the entire sector. Apart from this, the sector’s effectiveness and efficiency also suffers from insufficient institutional capacities and continuous discretionary political interference in regulatory and operational matters. In particular INAA, the regulatory authority responsible for tariff setting, is often the object of political interference. CONAPAS has initiated a sector information system that gathers information from the existing information systems of some of its member institutions. Despite this recent

140 Curtis V, and S. Cairncross, 2003. Effect of washing hands with soap on diarrhoea risk in the community: a systematic review. Lancet Infect Dis 2003; 3: 275-81. 141 Example: Base line survey for the blue star campaign (ECEAZ-2000): Results Report, by the Center for Communication Programs of the John Hopkins University (2001). Not published. 142 Further details will be elaborated on and presented in the 2006 Public Expenditure Review Nicaragua. 143 In Matagalpa and Jinotega water and sanitation services are provided by AMAT and EMAJIN, two local state-run service providers. 144 A WSP study to be published in early 2007 examines the role and importance of NGOs in the water and sanitation sector (“Contribución y capacidades de las ONGs y otros actores de la sociedad civil, en el sector agua y saneamiento de Nicaragua”) 145 Including the Presidential Secretariat (SETEC), the Health Ministry (MINSA), the Environment Ministry (MARENA), the Institute for Terrestrial Studies (INETER), as well as members of ENACAL, INAA, and FISE

256

initiative, the quality (and quantity) of the sector information remains modest and is a constraint to both monitor sector outcomes and also to the process of policy decision making. RASNIC146 is the sector’s knowledge network. It is joined by state institutions, cooperation agencies and NGOs alike. Notwithstanding its weak legal legitimacy, CONAPAS has recently taken a strong leadership in sector policy setting and sector governance. In October 2005, it elaborated and approved a coherent sector strategy in line with the National Development Plan. The sector strategy also gave rise to the sector round table as a coordination forum between government and the donor community147. In October 2006, the government and the donors agreed on a roadmap to complete a Sector Wide Approach (SWAP). In addition, a Code of Conduct on alignment and harmonization was signed. The general roadmap of the SWAP outlines several work lines that intend to address the mayor legal, institutional, and coordination challenges the sector is confronted with. The principal objective of the SWAP is to make the sector more coherent, effective and efficient. So far the Ortega government hasn’t decided neither whether it will continue with this strategy nor about the future sector architecture.

Key Policy Recommendations

• Rural areas: Achieving the MDGs in water and sanitation is the sector’s top priority. It is essentially a challenge in rural areas. Taking into account increasing marginal costs, substantial social infrastructure investments will be required in rural areas, where the vast majority of the poor lives without access to water and sanitation, especially in the Atlantic and Central/Northern region. Poor and extremely poor population groups benefit most from such investments. Appropriate co-financing and local participation policies will be necessary to ensure adequate technology and service levels that can be managed and afforded by the community in the long run. Infrastructure investments need to be accompanied with effective decentralization and capacity building strategies to strengthen local capacities, in particular at the municipal level, in order to provide technical and organizational assistance to CAPs and communities. The municipalities should be given a more determinate role in water and sanitation service provision.

• Urban areas (Managua and larger cities): One of the most urgent tasks in urban areas is to secure water provision and restore clients’ confidence. A cash injection for service quality improvements in the short run will likely be inevitable to prevent a virtual collapse of the water provision in some areas. However, a profound structural reform of the urban service provider ENACAL needs to be initiated in parallel to prevent that investments turn into de facto consumption subsidies. Once visible service improvements have achieved, a plan for a gradual adjustment of tariffs has to be elaborated, unless the government is willing to subsidize water services to poor on the basis of a more explicit subsidy policy. Any tariff adjustment needs to maintain a pro-poor orientation for poor urban dwellers. Eventual loans and grants to ENACAL should be linked to measurable outcomes in service improvements, key management and technical efficiency figures. An output based modality for grants and subsidies delivery is imperative. Additional funds will be required for expanding the urban sewage system in particular in peri urban areas and waste water treatment infrastructure in larger cities.

• Urban areas (small towns): Small towns need special attention because of their specific context and problem setting. The regulative and normative framework should be adjusted to allow for more autonomy and local public and private participation among water and sanitation service providers in small towns, including public-private-partnerships and micro enterprises. They also need specific

146 Water and Sanitation Network of Nicaragua (Red de Agua y Saneamiento de Nicaragua RASNIC). 147 Including IDB, WSP-World Bank, SDC, UNICEF, PHO, CIDA, EU, JICA, Netherlands/SNV, and the German Cooperation KFW which currently heads the forum.

257

investment plans appropriate to their size. The unsolved waste water pollution problem in these localities demands for laying more emphasis on public sewage systems and waste water treatment solutions in these localities.

• Sanitation and hygiene: In order to impact more on health related MDGs, sanitation and hygiene promotion deserves considerably more attention than it received in the past. A more integrated approach is imperative. Sector resources should not only be allocated to sanitation infrastructure (hardware) but also to the promotion of better hygiene behavior (software) as hygiene practices may be as determinant to health outcomes as access to water and sanitation infrastructure. Hygiene habits may be improved through a set of different approaches such as health education in schools, media campaigns, house to house visits, etc.

• Sector sustainability: Sustainability remains a key challenge and is highly relevant to poverty in the long run. Both the sector as a whole and the water and sanitation infrastructure in rural and urban areas still widely lack the desirable sustainability in aspects related to governance, management, organization (including participation), long term financial stability, environment, technology, and risk prevention. Although sustainability is the outcome of many different factors, the following must be named as the most urgent and most important ones: Major adjustments in the legal framework will be required not only for the sector institutions to become more efficient and effective but also to lay a more robust basis for the current sector architecture, in particular to strengthen the role of the sector’s governing body (CONAPAS or other). The sector as a whole depends on a more sustainable financial basis, alimented with sufficient fiscal resources in order to become more independent from fluctuating donor funds, especially for investments in rural areas. Apart from previously mentioned profound structural reforms, the three major sector institutions, ENACAL, INAA, and FISE, need significant capacity building in several of the aspects mentioned. In rural areas, FISE (or another institution) requires to be strengthened in its activities to promote sustainability and reduce the vulnerability of water and sanitation systems. Water quality must be guaranteed permanently and countrywide through a mechanism that not only monitors the quality but provides the means and methods.

• Sector information: The sector information system needs to be strengthened. Adequate policy making, sector management and monitoring of sector advances are seriously hampered by the current lack of reliably and actualized information. The sector information system may need a special legal and institutional basis in order to ensure its financial and operational sustainability.

• Sector coordination: An effective coordination mechanism between the donor community and the government will be inevitable. The process should be leaded by the sector’s policy body (CONAPAS or other). A certain consensus on sector policies and outcomes will have to be established to facilitate sector coordination. Although the previous roadmap may be changed under the new government, a SWAP approach should be aspired in order to increase the efficiency and efficacy of the sector resources. The roadmap needs further elaboration on the strategies how to achieve the MDGs.

258

9. PREVALENCE TRENDS AND PREDICTORS OF NUTRITIONAL STATUS AMONG NICARAGUAN CHILDREN 0-59 MONTHS, 1998, 2001 AND 2005

By Janet Irene Picado, Rafael Flores and Jose Ramon Laguna*

The purpose of this report was to describe trends in nutritional status among children 0-59 months in Nicaragua from 1998 to 2005, to identify predictors of nutritional status and to describe current and potential public sector interventions in nutrition in order to inform policy decisions regarding actions and programs to significantly reduce malnutrition.

The analysis of nutritional status of children were based on data from the Living Standards Measurement Surveys (LSMS) carried out in 1998, 2001 and 2005 and the analysis of the predictors of malnutrition was for the 2005 survey only.

Approximately 1 in every five children under five years suffers from growth retardation (21.5%) Stunting decreased 5.9 percentage points from 27.4% in 1998 to 21.5% in 2005 (most of this reduction was from 1998-01), a modest 0.84 percentage points per year. If this rate of decline were to continue, it will take Nicaragua 22 years to eliminate stunting as a public nutrition problem.

An important finding is that among the poor, stunting declined steadily and significantly during the period under study. The poorest and second quintiles had the largest reductions (11 and 10.6 percentage points respectively), when compared to the three upper consumption quintiles. What is remarkable is that stunting increased among children from families in the fourth and fifth consumption quintiles between 2001 and 2005. This would suggest that among these groups, stunting is being caused by other factors not directly related to consumption or income, such as feeding practices or care. These results need to be interpreted with caution, and further research is needed to better understand the phenomenon.

To identify the predictors of stunting, multiple logistic regression was used to model the relationship of stunting (the outcome) to several explanatory variables related to care, health and sanitation and household food security. Only children 0-35 months were included in the determinants analysis, because this is the vulnerable period when growth falters and interventions are needed.

From this cross sectional data we can infer that the main increases in stunting occur between 0-5 mo. and 6-11 mo. (6.5 percentage points) and between 6-11 mo. and 12-23 mo. (8.3 percentage points). After about 24 months children start to grow normally and mean z-scores become more stable. Children 0-11 mo. are much less likely (60%) to be stunted when compared to children 12-23 mo; however, children 24-35 mo. are only slightly more likely (29%) to be stunted when compared to the same group because stunting is already established by 23 months and does not increase significantly in the older groups. This confirms that children in Nicaragua are at the highest risk for stunting from birth to 23 mo. and prevention for stunting should start during gestation and continue through 23 mo. because after that stunting is not reversible.

Stunting was slightly higher in male children than in females and gender is an independent determinant of stunting, although not a strong one. Males are 29% more likely to be stunted than females. In 2005 the

* The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP) and Aline Coudouel (Senior Economist, LCSHS) for their valuable comments and suggestions. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent.

259

Central region had the highest levels of stunting overall. Among the extreme poor, stunting levels in the Central region were above 45%. In terms of total stunting the Atlantic region followed closely. It appears that differences in stunting across regions are due to differences in poverty and education. Consumption quintiles were used as a proxy variable for socio-economic status and this variable was found to be a strong and independent predictor of stunting; for example children from the poorest quintile were 2.5 times more likely to be stunted when compared to the richest quintile; for the second, third and fourth quintiles the same pattern was evident. Therefore, even though malnutrition increased among the fourth and fifth quintiles from 2001-2005, poverty is still a strong predictor of malnutrition.

Ethnicity (indigenous or not) shows a slight association with stunting, even after taking into account other variables; children of indigenous families about 20% more likely to be stunted, relative to non-indigenous children. Maternal education is another important determinant of stunting. Children 0-36 mo. of mothers with no education at all are 1.5 times more likely to be stunted when compared to children whose mothers had primary or adult education. Children of mothers with secondary or higher education are protected from stunting, when compared to the latter group.

Per capita food expenditure was also a powerful and independent predictor of stunting. Stunting was 2 times more likely among children from families with low per capita food expenditure when compared to families with medium level spending; children from families with high per capita food expenditure were protected from stunting when compared to the latter group.

Two other variables emerged as independent predictors of stunting, related to disease and sanitation. The presence of a toilet or latrine in the household was a protective factor for stunting (children were about 22% less likely to be stunted in relation to children from household without stunting). Children from households without a safe water supply were more likely to be stunted (29% more) than children from households with safe water. Birth order (first born, second, third or higher) and mother/child ratio (one, two, three or higher), are variables related to care and are independent determinants of stunting. First born children are protected from stunting relative to second born children. Third born or higher children are slightly more likely to be stunted relative to the latter group. A mother/child ratio of three or more means a higher chance of being stunted, when compared to a mother/child ratio of two.

Because many of the causes of malnutrition are directly related to poverty, poverty reduction is one of the long-term strategies or interventions to reduce malnutrition. However, direct nutritional interventions/programs are needed to reduce levels of malnutrition, and should remain permanent unless there are major improvements in economic development that alleviate poverty significantly.

Nicaragua has highly effective programs and programs that are potentially highly effective. To date, there are two nutrition programs with documented success in reducing malnutrition. One is the Ministry of Health’s National Micronutrient Program with two components: supplementation (vitamin A and iron) and fortification (salt with iodine; sugar with vitamin A and flour with iron and folic acid). The second program is the Red de Protección Social (RPS), a conditional cash transfer program implemented from 2000-06. The RPS was targeted to some of the poorest municipalities of the country. The impact evaluation of the first two years of the program (2000-02) showed a decline in stunting among children 0-59 mo. of 5 percentage points.

The Programa Comunitario de Salud y Nutrición (PROCOSAN) is MINSA’s community-based growth promotion program (CBGP). PROCOSAN is a preventive health and nutrition program that actively engages families of children under two and their communities in maintaining the adequate growth of young children. For sick children under five years old, the program extends its treatment and referral services. PROCOSAN is a program with a high potential to be effective.

260

If Nicaragua has highly effective and potentially effective nutrition programs that are well targeted to vulnerable populations, why is malnutrition declining at such a slow rate?

The simple answer is that although targeting of programs has been adequate, investment in highly effective or potentially effective nutrition programs is insufficient to produce a greater reduction of stunting. There is a gap in the coverage of these programs of approximately 30% based on information from 2005-2006. Another issue is that in practice it does not appear that prevention of malnutrition is high on the list of priorities. During 2005 the total budget allocated to highly effective or potentially effective programs (MINSA and RPS) was US$7,137, 611.17. The total budget for programs that are not highly effective to reduce malnutrition was US$26,359,954.78 (school feeding and other programs).

Based on 2007 population estimates, it would cost Nicaragua approximately US$44,166,310.00 per year to provide complete coverage of the vulnerable population with PROCOSAN, and other MINSA interventions as well as the RPS, to significantly reduce malnutrition.

In summary, Nicaragua has the knowledge and experience to reduce malnutrition in the short term. But first, the government must treat the reduction of malnutrition as a true priority, in practice. What is needed is: 1) an adequate combination of highly effective programs (such as PROCOSAN, RPS and other MINSA interventions) and 2) institutional strengthening to increase capacity required to bring programs to scale. In all prioritized municipalities with a high level of poverty, 100% coverage must be achieved and maintained among the vulnerable population.

INTRODUCTION

The purpose of this paper is to describe trends in nutritional status among children 0-59 months in Nicaragua from 1998 to 2005, to identify predictors of nutritional status and to describe current and potential public sector interventions in nutrition in order to inform policy decisions regarding actions and programs to significantly reduce malnutrition.

This effort is part of the Poverty Assessment 2006-07 that has been prepared by the Nicaraguan government, through the Technical Secretariat to the Presidency (SETEC) and the World Bank.

The analysis of nutritional status of children are based on data from the Living Standards Measurement Surveys (LSMS) carried out in 1998, 2001 and 2005 and the analysis of the predictors of malnutrition was for the 2005 survey only. All anthropometry analysis was based on the World Health Organization International Growth Standards 2006 (de Onis et al. 2006).

WHAT ARE THE CAUSES OF MALNUTRITION?

Malnutrition is a complex and widespread problem with major consequences for society. About half of the mortality in children under five due to infectious diseases is caused by mild and moderate malnutrition. Children who survive malnutrition have lower learning capacity and are less productive as adults, reducing the quality of life and affecting the economies of families, communities and nations. (Sangvhi 1999; Pelletier et al. 1993; World Bank 2005).

The causes of malnutrition are multi-sectoral. At the child level, disease and inadequate dietary intake are the immediate causes. These two factors work synergistically. For example, a child who does not eat well is more susceptible to illness. Disease increases nutrient loss and suppresses appetite therefore sick children do not eat well so the cycle continues and accounts for most of the child morbidity and mortality in the world. (Unicef 1998; Sangvhi 1999)

261

The underlying causes of malnutrition occur at household/family and community level and include insufficient access to food, inadequate maternal and child caring practices (actions or behaviors that translate available food into good child growth and development) and poor water/sanitation and inadequate health services. These underlying causes are related to inadequate or inappropriate knowledge/education as well as discriminatory attitudes that limit household access to actual resources (Unicef 1998).

The cycle of malnutrition and illness continues when small adult women who were malnourished as children give birth to small babies that are at higher risk of mortality and malnutrition. (Sanghvi 1999)

Most programs that address malnutrition in the short term work at the child/family/community level to change immediate and underlying causes of malnutrition in the short-term. The analysis and discussion in this paper pertains to those levels.

WHAT ARE PRIORITY NUTRITION INTERVENTIONS?

One of the myths regarding nutrition programs or interventions is that most malnutrition is caused by inadequate access to food, and although quality of food is important, most malnutrition in small children is caused by inadequate care and feeding practices, bad sanitation and disease . Program experience has shown that improving food supplies or income among the poor without changing the way young children are cared for is not effective in improving malnutrition. (World Bank 2005)

Contrary to common belief, there are many cost-effective interventions/programs to combat malnutrition in women and children. There are both short-term and long-term approaches to improve nutrition, and it is important for countries to have a balance between the two. Examples of interventions that take the long route are: primary health services that include family planning and infectious disease control; policies on marketing of breast milk substitutes; employment creation; improving incomes of the poor; increasing women’s educations; marketing regulation of unhealthy food; food and agricultural policies to increase supply of safe and healthy food; safe water and sanitation, among others. (World Bank 2005)

Short-term or direct/specific nutrition interventions include: (Sanghvi 1999; World Bank 2005)

• Promotion of exclusive breastfeeding for the first six months of life (includes counseling, comprehensive communications campaigns, legislation, promotion and support for The Baby Friendly Hospital Initiative (BFHI), community-support groups etc.)

• Integration of nutrition interventions in maternal health services (prenatal iron/folate supplements; prenatal counseling to promote adequate diets and reduced workloads during pregnancy and after delivery; monitor weight gain during last two months of pregnancy; distribution of vitamin A during postpartum period; treatment for intestinal parasites and malaria prophylaxis; screening for severe anemia and treatment and referral for severe anemia)

• Integration of nutrition components in basic health services for children < 2 years of age (counseling to support appropriate complementary feeding and continued breastfeeding for two years; growth monitoring and promotion; micronutrient supplements) adequate nutritional care during illness and severe malnutrition)

• Micronutrient supplements; micronutrient fortification

• Community based-growth promotion and health care (Integrated Management of Childhood Illnesses-IMCI); interventions described above can be delivered to the community through these programs

262

• Conditional cash transfers

HOW HAS MALNUTRITION EVOLVED IN NICARAGUA FROM 1998 TO 2005?

Anthropometric indicators of nutritional status in children

Stunting is the most common form of protein-energy malnutrition in children under five in Nicaragua, similar to other countries of the region. Stunting refers to gaining insufficient length/height relative to age and implies long-term malnutrition and poor health. Stunted children are small in size and have diminished learning capacity. Length/height for age is the nutritional index used to measure stunting. Z-scores for length/height for age (HAZ)148 were calculated based on WHO Child Growth Standards 2006 for 1998, 2001 and 2005, and children with HAZ < -2 standard deviations were classified as stunted (de Onis et al. 2006)WHO 1995).

Children’s nutritional status was also classified based on other indicators and these results are in the Appendix.

Other nutritional problems among small children

Anemia is the most frequent nutritional problem in Nicaragua. In countries with a high prevalence of anemia the main cause is iron deficiency and children 6-24 mo. and women 15-49 years are most vulnerable. The National Micronutrient Survey 2000 (Encuesta Nacional de Micronutrientes 2000) found that anemia (hemoglobin <11 g/DL) affected 61.8% of children 6-11 mo. and 54.2% of those aged 12-23 mo (Ministerio de Salud 2002). Anemia in children under two years of age can be prevented by exclusive breastfeeding for six months followed by iron supplementation through 23 months.

Low-birth weight (<2500 grams) affects 8.9% of births; in developing countries one of the major causes is intrauterine-growth retardation, which is associated with maternal nutrition and infections during pregnancy. Children born with low-birth weight are at higher risk of mortality and chronic diseases (Ministerio de Salud 2007).

Zinc nutritional status has never been measured among Nicaraguan children. However, stunting and iron deficiency anemia are both associated with zinc deficiency. Given such high levels of stunting among small children, there is probably a high risk for zinc deficiency. Zinc supplementation improves growth in areas with high prevalence of growth retardation, and also improves nutrition by reducing the frequency and severity of diarrhea (Brown et al. 1998).

As discussed above, care practices are important determinants of health and nutrition in young children, and how mothers feed their children in a given context is related to care. Formative research conducted by the Nicaraguan Ministry of Health with support from the World Bank in the year 2001 identified feeding practices among families of children less than two years of age. Following analysis of these feeding practices, the researchers carried out “trials of improved practices” to assess families’ capacity to improve these practices. For example, mothers of children < 6 months were willing to improve breastfeeding practices so that breastfeeding was exclusive. Among children between 6-23 mo., most mothers were able to: increase the amount of food their children consumed; increase the frequency of meals; make foods thicker, as well as other practices. An important finding of this research was that “many rural families can improve their children’s nutrition significantly, with the resources they already have” (Picado et al. 2001).

148 Z-scores are the deviation of the value for an individual from the median value of the standard population, divided by the standard deviation of the reference population for a specific index.

263

Breastfeeding is one of the most important of all “care” practices. Exclusive breastfeeding for six months and continued breastfeeding for 2 years saves infant lives and prevents malnutrition, among many other benefits. (Saadeh et al. 1993) Most mothers in Nicaragua breastfeed their infants although mean duration of exclusive breastfeeding and continued breastfeeding (2.5 mo. and 17.6 mo. respectively) are less than optimal (INEC 2002).

Overweight among children under five is still not a major problem in the country, with a national prevalence of 5.4%, which is higher among the non-poor and urban residents (Appendix 8). However, it is above the expected level for a healthy population, and given that 48% of women 15-49 are overweight (INEC 2002) in Nicaragua, it is likely that the proportion of overweight children will rise in the future.

Nicaragua has two great success stories in nutrition. Iodine deficiency disorders are no longer a public health problem in the country, due to fortification of salt. Vitamin A deficiency has also been eradicated through two programs: distribution of massive doses of vitamin A for children via national immunization days, and fortification of sugar with vitamin A, since the late 1990's.

MINSA’s National Surveillance System for Nutritional Interventions (SIVIN for its acronym in Spanish) monitors several nutrition indicators including anemia and iron deficiency. Recently SIVIN has reported important gains in the reduction of anemia and iron deficiency. However, these results must be interpreted with caution, since other programs have not shown the same success recently (Ministerio de Salud 2007 b; Maluccio and Flores 2004)

Prevalence and determinants of stunting 2005

To identify the determinants of stunting, multiple logistic regression was used to model the relationship of stunting (the outcome) to several explanatory variables. Variables were limited to those available in the survey, therefore the results are not meant to be comprehensive; for example, the survey does not provide any information on feeding behaviors, one of the major determinants of nutritional status. Variable selection for the analysis was guided by Unicef’s conceptual framework on the causes of malnutrition (Unicef 1998), and a review of similar studies (Chawla 2000; Martorell et al. 2002) (Appendices 11 and 12). Only children 0-35 months were included in the determinants analysis, because this is the vulnerable period when growth falters and interventions are needed. The regression coefficients in multiple logistic regression are odds ratios, and they provide a measure of the strength of the association of the outcome to each explanatory variable. An odds ratio of 1 signifies there is no association.

The following variables were independently associated with stunting among children 0-35 mo. (Table 9.1) from LSMS 2005 and were kept in the final model:

-Demographic variables: age; gender; ethnicity; area of residence (urban/rural); region of residence.

-Variables related to care, food and health: consumption quintiles; maternal education; per capita food expenditure.

-Variables related to care: birth order; mother/child ratio.

-Variables related to disease/sanitation: quality of household water supply; presence of toilet or latrine in home.

These results are consistent with the literature on the subject (Martorell et al. 2002) and reinforce the idea of the multi-causality of malnutrition.

264

Table 9.1. Odds ratios (95% confidence intervals) from logistic regression models predicting stunting in children 0-35 mo.

1491 National, LSMS 2005

Odds Ratio Std. Err. 95% Confidence Interval Age (mo.) - 0-11 0.39 0.01 0.38 0.40 -12-23 1.00 - - - -24-35 1.29 0.02 1.26 1.32 Gender -Female 1.00 - - - -Male 1.27 0.01 1.24 1.307 Ethnicity -Not -indigenous 1.00 - - - - Indigenous 1.20 0.03 1.15 1.26 Geographic area -Urban 1.00 - - - -Rural 0.89 0.01 0.87 0.92 Maternal education -None 1.51 0.02 1.47 1.55 -Primary/adult education 1.00 - - - -Secondary or higher 0.63 0.01 0.61 0.6 Consumption -Quintile 1 2.52 0.09 2.34 2.72 -Quintile 2 2.02 0.07 1.88 2.16 -Quintile 3 2.99 0.01 2.79 3.19 -Quintile 4 3.45 0.11 3.25 3.66 -Quintile 5 1.00 - - - Per capita food expenditure -Low 2.09 0.04 2.03 2.16 - Medium 1.00 - -High 0.67 0.01 0.64 0.70 Presence of toilet or latrine -None 1.00 - - - -Yes 0.78 0.01 0.76 0.81 Household water supply -Safe 1.00 - - - -Unsafe 1.29 0.02 1.25 1.34 Birth order -First born 0.72 0.02 0.69 0.75 -Second born 1.00 - - - -Third born or higher 1.16 0.03 1.11 1.21 Mother/child ratio -One 1.07 0.02 1.03 1.11 -Two 1.00 - - - -Three or more 1.54 0.03 1.48 1.59 Region -Managua 1.00 - - - -Pacific 0.95 0.02 0.92 0.99 -Central 1.16 0.02 1.12 1.20 -Atlantic 0.62 0.01 0.59 0.65

Source: Nicaragua LSMS 2005

149 Reference group for variables with two dummies is middle group; for consumption quintiles reference is the fifth quintile and Managua is reference for region. An odds ratio greater than 1 is indicative of a higher chance/likelihood of being stunted, while an odds ratio less than one means protection from stunting

265

Stunting and age

Prevalence of stunting among Nicaraguan children 0-59 mo. was 21.5% in 2005, when the expected is 3% in a healthy population. Although stunting decreased 5.9 percentage points from 27.4% in 1998 to 21.5% in 2005, stunting continues to be the main form of malnutrition in Nicaragua with approximately 1 in every five children under five years suffering from growth retardation (Table 9.2). Stunting levels are high in relation to Latin America and the Caribbean; (De Onis et al. 2000). If this rate of decline (0.84 percentage points per year) were to continue, it will take Nicaragua 22 years to eliminate stunting.

Table 9. 2 Stunting national and by age groups (mo.), gender and geographic area; 1998, 2001 and 2005.

1998 2001 Change 2005 Change Change

% % 98-01 % 01-05 98-05 National 27.4 22.5 -4.9 21.5 -1.0 -5.9 Age groups (mo.) 0 - 5 8.3 14.4 6.1 6.8 -7.6 -1.5

6 - 11 14.4 20.8 6.4 13.3 -7.5 -1.1

12 - 23 27.4 22.0 -5.4 21.6 -0.4 -5.8

24 - 35 30.0 21.0 -9.0 23.2 2.2 -6.8

36 - 47 32.0 25.3 -6.7 27.1 1.8 -4.9

48 - 59 32.8 25.7 -7.1 24.5 -1.2 -8.3

Gender

Male 28.7 24.1 -4.6 22.0 -2.1 -6.7

Female 26.0 20.8 -5.2 20.9 0.1 -5.1

Geographic area

Urban 22.8 16.6 -6.2 16.5 -0.1 -6.3

Rural 31.8 28.9 -2.9 27.0 -1.9 -4.8 Source: Nicaragua LSMS 1998; 2001 and 2005

About 7% of infants between 0-5 mo. were already stunted in 2005 (Table 9.2). Since it is estimated that 8.9% (Ministerio de Salud 2007a) of infants are born with low birth weight in Nicaragua, this may reflect in part intra-uterine growth retardation as well as post-natal causes. Intra-uterine growth retardation is related to maternal nutrition and infections during pregnancy. Adequate prenatal care with integrated nutrition services is needed for prevention. Since infants up to 6 months need only breast milk for optimum nutrition and growth, inadequate breastfeeding practices may contribute to the development of stunting in this age group.

From this cross sectional data we can infer that the main increases in stunting occur between 0-5 mo. and 6-11 mo. (6.5 percentage points) and between 6-11 mo. and 12-23 mo. (8.3 percentage points) (Table 9.2). After about 24 months children start to grow normally and mean z-scores become more stable (Figure 9.1).

266

Source: Nicaragua LSMS 1998; 2001 and 2005

Analyses to assess the determinants of stunting are consistent with these results. Children 0-11 mo. are much less likely (60%) to be stunted when compared to children 12-23 mo; however, children 24-35 mo. are only slightly more likely (29%) to be stunted when compared to the same group (Table 9.1), because stunting is already established by 23 months and does not increase significantly in the older groups. This confirms that children in Nicaragua are at the highest risk for stunting from birth to 23 mo. and prevention for stunting should start during gestation and continue through 23 mo. Because after that stunting is not reversible.

Stunting and gender

Stunting was slightly higher in male children than in females and gender is an independent determinant of stunting, although not a strong one. Males are 29% more likely to be stunted than females (Tables 9.1 and 9.2). A recent study among indigenous communities in Guatemala found that mothers’ perceptions of the nutritional needs of male and female children were different and that son preference by mothers has an effect on growth for both genders. Further research is required to determine if male children are treated differently than females in ways that affect growth, as this has implications for the counseling component of maternal-child nutrition programs. (Tumilowicz et al. 2006)

Stunting and place of residence

As expected there was a large difference in stunting (10 percentage points) between rural and urban areas (Table9 .2). Further analysis to understand the determinants of stunting shows that much of this variation is explained by differences in education and consumption quintiles (proxy for socio-economic status) (Table 9.1). Persistence of such large inequities between urban and rural territories underscores the need for more adequate targeting to populations most in need.

267

In 2005 the Central region150 had the highest levels of stunting, for urban (19.1%), rural (32.2%) and total (27.6) (Table 9.2). In terms of total stunting the Atlantic region followed closely (total 24.5%). Managua and the Pacific regions had similar levels of stunting, and Managua rural had less stunting than Managua urban and all other rural regions. Managua urban had an unexpectedly high level of stunting (17.5%) (Table 9.3).

Table 9.3. Stunting in children 0-59 mo. 1998, 2001 and 2005; national and by regions, urban/rural

1998 2001 Change 98-01

2005 Change 01-05

Change 98-05

% % % National 27.4 22.5 -4.9 21.5 -1.0 -5.9 Managua 14.9 9.7 -5.2 16.9 7.2 2.0 Urban 16.4 10.2 -6.2 17.5 7.3 1.1 Rural 6.7 4.8 -1.9 9.5 4.7 2.9 Pacific 27.7 17.7 -10 16.5 -1.2 -11.2 Urban 25.3 16.6 -8.7 15.1 -1.5 -10.2 Rural 30.1 18.8 -11.3 18.3 -0.5 -11.8 Central 35.1 33.3 -1.8 27.6 -5.7 -7.4 Urban 29.7 26.0 -3.7 19.1 -6.9 -10.6 Rural 37.2 37.5 0.3 32.2 -5.3 -4.8 Atlantic 28.5 25.1 -3.4 24.5 -0.6 -4.0 Urban 25.6 18.1 -7.5 10.2 -7.9 -15.4 Rural 31.0 29.9 -1.1 29.3 -0.6 -1.7 Source: Nicaragua LSMS 1998; 2001 and 2005

As above, it appears that differences in stunting across regions are due to differences in poverty and education. Region of residence was an independent but not strong predictor of stunting; a child from the Central region was a bit more likely (17%) to be stunted when compared to a Managua child; and being from the Atlantic region was slightly protective after adjusting for all other variables, but this is due to unknown factors, and would require further investigation (Table 9.1).

Stunting and poverty

Stunting is strongly associated with both poverty group and consumption quintiles. Stunting is more than double among children from families in the extreme poverty group when compared to children from non-poor families (Table 9.4). Among moderately poor families, the prevalence of stunting is reduced by 14.8 percentage points when compared to families in extreme poverty (37.2% versus 22.4%). Prevalence of stunting is double among poor families when compared to non-poor (Table 9.4). The same pattern holds when stunting is stratified by consumption quintiles. Stunting decreases with each increasing consumption quintile but is still relatively high even in the upper quintiles, which supports the idea that malnutrition is caused by many different factors, not necessarily inadequate access to food.

150 Includes the departments of Madriz, Estelí, Nueva Segovia, Matagalpa, Jinotega, Boaco and Chontales.

268

Table 9.4. Stunting in children 0-59 mo 1998-2001 and 2005; all children and by poverty groups, income quintiles

1998 2001 Change 8-01

2005 Change 01-05

Change 98-05

% % % All children 27.4 22.5 -4.9 21.5 -1.0 -5.9

Poverty group Extreme Poor 46.4 43.8 -2.6 37.2 -6.6 -9.2

Moderately Poor 30.0 24.9 -5.1 22.4 -2.5 -7.6 Poor 36.2 31.5 -4.7 27.4 -4.1 -8.8

Non-poor 14.8 10.7 -4.1 14.6 3.9 -0.2 Quintile Poorest 46.4 40.7 -5.7 35.4 -5.3 -11.0

II 31.4 23.7 -7.7 20.8 -2.9 -10.6 III 21.1 18.3 -2.8 17.0 -1.3 -4.1 IV 15.2 9.0 -6.2 16.6 7.6 1.4

Richest 6.0 6.4 0.4 11.7 5.3 5.7 Source: Nicaragua LSMS 1998; 2001 and 2005

Among children of families living in extreme poverty, levels of stunting varied widely across regions. Among the extreme poor, stunting levels were above 45% in the Central region, both urban and rural. These are the highest levels of stunting in the country, overall. Out of a list of 80 municipalities prioritized by the Ministry of Health (MOH) in 2006, 47 are located in the Central region.151 The Atlantic rural follows, with 36.9% stunting in the rural region among the extreme poor.152 Prevalence of stunting is much lower among extremely poor families living in Managua, both urban and rural followed by Pacific, both urban and rural (Table 9.5). In general poverty is lower in the Managua and Pacific regions, and families have more access to services than those living in the Central and Atlantic regions which may in part moderate the effects of extreme poverty. Major efforts to prevent stunting must be targeted to extremely poor families in the Central and Atlantic regions.

Table 9.5. Stunting among children 0-59 mo. in 2005; by region and poverty level

Poverty level Stunting Region Extreme poverty Moderate poverty Poor Not- poor Managua 13.0 21.5 20.6 15.5 Urban 14.6 23.3 22.3 15.8 Rural 0.0 10.2 9.3 9.8 Pacific 22.9 19.3 20.1 12.5 Urban 20.2 21.4 21.3 9.7 Rural 24.3 16.7 18.9 17.4 Central 45.3 25.1 34.0 15.2 Urban 46.2 21.1 27.7 11.8 Rural 45.2 26.9 36.0 19.6 Atlantic 35.6 23.7 29.0 15.4 Urban 14.5 14.3 14.3 7.5 Rural 36.9 26.3 31.5 22.1 Source: Nicaragua LSMS 2005

151 Within the Central region, the departments with the highest levels of stunting are Jinotega, Matagalpa and Madriz (Endesa 2001) 152 Within the Atlantic region, the Región Autónoma del Atlántico Norte (RAAN) has the highest levels of stunting and one of the poorest municipalities in the country: Waspam, on the Wangki river, where the main population is Miskito (Endesa 2001).

269

Socio-economic status is a powerful predictor of stunting and is related to a family’s ability to obtain food, care and health (Martorell et al. 2002). In the determinants analysis, consumption quintiles were used as a proxy variable for socio-economic status. This variable was found to be a strong and independent predictor of stunting; for example children from the poorest quintile were 2.5 times more likely to be stunted when compared to the richest quintile; for the second, third and fourth quintiles the same pattern was evident (Table 9.1).

Therefore, even though stunting increased among children from families in the fourth and fifth quintiles from 2001-2005, poverty is still a strong predictor of malnutrition.

Other predictors of stunting

Ethnicity (indigenous or not) shows a slight associated with stunting, even after taking into account other variables; children of indigenous families are about 20% more likely to be stunted, relative to non-indigenous children (Table 1). However, this is probably due to other factors related to care or health that are not accounted for in the regression model, and does not mean that indigenous children are genetically predisposed to be smaller (Martorell et al. 2002).

Maternal education is another important predictor of stunting. This variable is related to care and socio-economic status. (Martorell et al. 2002) Children 0-36 mo. of mothers with no education at all are 1.5 times more likely to be stunted when compared to children whose mothers had primary or adult education. Children of mothers with secondary or higher education are protected from stunting, when compared to the latter group (Table 9.1).

Per capita food expenditure was also a powerful and independent predictor of stunting. Stunting was 2 times more likely among children from families with low per capita food expenditure when compared to families with medium level spending; children from families with high per capita food expenditure were protected from stunting when compared to the latter group (Table 9.1). One of the limiting factors in the diets of small children from poor families is variety, that is, diets are lacking in animal foods and fruits and vegetables. Families who are able to spend more money on food probably buy more of these items and children’s diets are more varied. The addition of animal foods in children’s diets promotes growth (Brown et al. 1998).

Two other variables emerged as independent determinants of stunting, related to disease and sanitation. The presence of a toilet or latrine in the household was a protective factor for stunting (children were about 22% less likely to be stunted in relation to children from household without stunting). Children from households without a safe water supply were more likely to be stunted (29% more) than children from households with safe water (Table 9.1). This underlines the importance of water and sanitation in the prevention of disease and malnutrition of young children, which remain strong predictors of stunting even when variables such as consumption and education are taken into account.

Birth order (first born, second, third or higher) and mother/child ratio (one, two, three or higher), are variables related to care and are independent determinants of stunting (Table 9.1).

First born children are protected from stunting relative to second born children. Third born or higher children are slightly more likely to be stunted relative to the latter group (Table 9.1).

A mother/child ratio of three or more means a higher chance of being stunted, when compared to a mother/child ratio of two (Table 1).

270

As more children are born into a family, mothers have less time and resources to provide adequate care and attention to their children, so with each subsequent birth, the new child suffers. Both of these variables point to the need of family planning services, so mothers can space pregnancies more adequately.

Trends in stunting 1998-2005

Between 1998 and 2005, stunting among children 0-59 mo. was reduced by 5.9 percentage points from 27.4% to 21.5% at the national level, a modest 0.84 percentage points per year. Most of this reduction took place between 1998 and 2001 (27.4% to 22.5%-1.63 percentage points per year) and between 2001 and 2005 stunting was reduced by 1 percentage point only (22.5% to 21.5%), an average of .25 percentage points per year (Table 9.6). Overall, the net reduction in the total number of children stunted over the period was 62, 741 children however this was also due in part to a reduction in the population growth during the same period (Table 9.6).

Table 9.6. Change in numbers of children 0-59 mo. with stunting (1998-2005)

Year Estimated population 0-

59 mo. Prevalence of stunting (%)

Total number of stunted children

1998 802,507 27.4 219,887

2001 801,056 22.5 180,238

2005 730,912 21.5 157,146

Change 62,741

Source: Ministry of Health and Nicaragua LSMS 1998-2001 and 2005

In both urban and rural areas, stunting declined from 1998 to 2001, although the decline was stronger in urban areas (Table 9.6), indicating that people in urban areas had more resources at their disposal to improve/protect the health and nutrition of their families. This pattern did not continue, as stunting did not improve in urban areas between 2001 and 2005, and improved slightly in rural areas, indicating there may have been better targeting to the rural areas.

During the same period, across regions, trends in stunting varied widely. Stunting in the Pacific region153 declined overall by a huge 10 percentage points (3.33 percentage points per year); and declined 11.3 percentage points in the rural area, the highest in the country. This was followed by Managua (-5.2 percentage points) and modest reductions in the poorest regions, Atlantic (-3.4 percentage points) and Central (-1.8 percentage points-all in the urban area) (Table 9.3).There was no change in malnutrition among children of the Central rural region, the region with the highest levels of stunting/poverty in Nicaragua. The region also suffered the devastation caused by Hurricane Mitch, and endured the coffee crisis. It appears social programs in the Central region prevented a deterioration of the malnutrition rate, but social investment was not enough to improve malnutrition. Therefore, the reduction in the prevalence of stunting from 1998 to 2001 is mainly explained by the improvements in the Pacific and Managua regions.

However, from 2001 to 2005 there was a different pattern of variation within and across regions. Stunting increased in Managua, declined slightly in the Pacific region; in the Central region, stunting decreased in both urban and rural areas, overall 1.4 percentage points per year, better than the national average. This is a very positive change and may indicate that targeting of programs improved during this

153 Masaya, Carazo, Granada, Rivas, León and Chinandega.

271

period in that part of the country. In the Atlantic region, stunting decreased in urban areas only, but did not change in the rural areas, where some of the poorest and most remote communities in the country are located. Because the population is so disperse and difficult to reach, special efforts need to be made so that health and nutrition services reach these communities in both quantity and quality. (Table 9.3)

The fact that stunting increased so much in Managua must be interpreted with caution and requires further investigation. Levels of stunting appear high even among the non-poor, suggesting causes other than lack of income.

Figure 9.2. Trends Stunting by Poverty 1998-2005

46.4

30.036.2

14.8

43.8

24.9

31.5

10.7

37.2

22.427.4

14.6

0

10

20

30

40

50

Extreme Poor Moderately Poor Poor Non-poor

1998 2001 2005

Source: Nicaragua LSMS 1998; 2001 and 2005

Among the poor in general, stunting declined steadily and significantly during the period under study (Table 9.4 and Figure 9.2). The poorest and second quintiles had the largest reductions (11 and 10.6 percentage points respectively), when compared to the three upper consumption quintiles. Among extremely poor families, stunting decreased 9.2 percentage points and slightly less (7.6 percentage points) among the moderately poor. This is a very positive finding that is consistent with results that show that consumption increased among the poor (see Nicaragua Poverty Assessment) even though poverty was not reduced overall. It may be in part a result of the increased social investment after Hurricane Mitch that occurred in regions other than the Central rural region (where there was no improvement from 1998 to 2001)) and increased targeting of nutrition programs to the Central region from 2001-2005, when stunting was reduced. What is remarkable is that stunting increased among children from families in the fourth and fifth consumption quintiles between 2001 and 2005, probably due to Managua where malnutrition increased. This would suggest that among these groups, stunting is being caused by other factors not directly related to consumption or income, such as feeding practices or care, but these results need to be interpreted with caution. Regardless, poverty is still a strong and independent predictor of stunting.

ARE CURRENT NUTRITION PROGRAMS IN NICARAGUA POSITIONED TO BE EFFECTIVE IN REDUCING MALNUTRITION?

Because many of the causes of malnutrition are directly related to poverty, poverty reduction is one of the long-term strategies or interventions to reduce malnutrition. But reducing poverty is a slow and difficult process. In the meantime, programs with direct nutrition interventions are needed to reduce levels of malnutrition, and should remain permanent unless there are major improvements in economic development that alleviate poverty significantly. This discussion will focus on the adequacy of current

272

programs with regards to targeting, and effectiveness or potential for effectiveness in reducing malnutrition. Coverage of the vulnerable population with current programs will also be addressed.

A recent review of nutrition programs in Central America proposed criteria for defining a program’s potential for effectiveness in reducing malnutrition (Neufeld et al. 2006); the following criteria have been adapted for Nicaragua:

a. Program benefits are targeted to pregnant/lactating women and children <2 years of age, in geographic areas with the highest poverty.

b. The program actively promotes exclusive breastfeeding for six months and continued breastfeeding up to two years as well as optimal feeding practices for children 6-23 mo.

c. The program distributes a fortified food supplement containing at least one animal product, or micronutrient supplements or provides micronutrients for home fortification

d. The program has a health care component that detects illness in a timely fashion and provides appropriate treatment for infectious diseases.

Nicaragua has the know how, the experience and the success to significantly reduce malnutrition. Current programs with a specific nutrition component154 are carried out by three different government institutions: MINSA, Ministry of the Family (MIFAMILIA), Ministry of Agriculture (MAGFOR).

To date, there are two nutrition programs with well documented success to reduce stunting or micronutrient malnutrition. One is Minsa’s National Micronutrient Program with two components: supplementation (vitamin A and iron) and fortification155 (salt with iodine; sugar with vitamin A and flour with iron and folic acid). Supplements are targeted to pregnant and lactating women and children 6-23 mo.; fortification targets the general population156. Vitamin A and Iodine deficiencies have been controlled and there have been reports in reduction of anemia, but these results must be further confirmed (Ministerio de Salud 2007b).

The second program is the Red de Protección Social (RPS), a conditional cash transfer program carried-out by MiFamilia from 2000-06. This program worked by supplementing the household income of beneficiaries for up to three years to: - increase family spending on food; -increase primary enrollment during the first four years; and to -improve the health and nutrition of children under five. Beneficiaries received a transfer for the purchase of more and better quality food; those with eligible school age children received additional transfers for school attendance, supplies and for the teacher. To receive these cash transfers, beneficiaries had to: attend health education workshops every two months; bring children up to five to preventive health care appointments and ensure enrollment and 85% school attendance record for children between 7 and 13 who had not yet completed fourth grade; and deliver the transfer to the school. The RPS was targeted to poor families living in municipalities of Madriz, Matagalpa and Jinotega (one municipality only) and 30,000 families were benefited during the 6 years. The impact evaluation of the first two years of the program (2000-02) showed a dramatic decline in stunting among children 0-59 mo. of 5 percentage points -from 42% to 37 %157 through improved diets and increased preventive health care for beneficiary children. This is considered a remarkable reduction in stunting for

154 The Programa Integral de Nutrición Escolar is the Ministry of Education’s school feeding program targeted to school age children; it was not included above as a nutrition program, because its objective is school retention-however it benefits 897,000 students and the annual budget for 2005 was US$9,773,954.78-although this type of program does not contribute to reducing malnutrition (Ruiz and Reyes 2007) 155 Fortification is the addition of micronutrients to processed foods. 156 Small children are not beneficiaries because they do not consume enough food to perceive benefits. 157 Prevalence of stunting based on the NCHS 1977 reference population.

273

such a short period (Maluccio and Flores 2004; Maluccio et al. 2005). RPS was not successful in reducing anemia, despite having distributed iron supplements, possibly because mothers did not give the supplements as advised. The total cost of the program was approximately US$30,000,000.00 or $1000 per family/3 years158. Despite this success, the current government has so far expressed their decision to discontinue the RPS (personal communication with Orlando NúΖez January 2007), although it appears this decision is not final yet.

The Programa Comunitario de Salud y Nutrición (PROCOSAN) is MINSA’s community-based growth promotion program (CBGP). This program was designed in 2000-01 with the community and World Bank technical assistance. Lo bueno es que puede comer lo que nosotros tenemos (Picado et al. 2001) the formative research project carried out by MINSA in 2000 was an important input for the program design.

PROCOSAN is a preventive health and nutrition program that actively engages families of children under two and their communities in maintaining the adequate growth of young children. For sick children under five years old, the program extends its treatment and referral services.

There are certain key concepts that make CBGP like PROCOSAN different from other growth monitoring programs.

• The communities or potential program users participate from the start in building the program vision along with program personnel from participating institutions.

• Program activities are all conducted within communities, with community volunteers.

• The program focuses on children 0-2 y, the group most likely to benefit from nutrition interventions.

• Children are classified monthly based on the adequacy of their growth or weight gain.

• The priority is to detect and address initial growth faltering to prevent malnutrition.

• Data collection is used for decision making at every level, beginning with the family.

• When there are problems, the search for solutions begins in the family, but also in the community.

• At the family level, the principal solution is improvements in behavior to address inadequate feeding practices.

• At the community level, activities are programmed that aim to make it easier for families to maintain the growth of their children, for example, by addressing problems of food shortages, poor water condition, or collective child-care needs that go beyond a single household.

• Education is participatory, based upon individual counseling and negotiation. Suggestions for improvements are negotiated with the mothers.

PROCOSAN delivers most services during monthly growth promotion sessions. Children are weighed monthly in the community for the first two years of life to determine the adequacy of weight gain between visits. At each session trained community volunteers talk with mothers and listen to determine the causes of problems or the reasons for successes over the past month

Mothers are counseled by community health workers based on their individual situations; that is, counseling is tailored. Mothers are also counseled regarding early childhood stimulation. Children are referred to the health post or health center and are followed-up with a home visit if necessary. The home treatment and referral protocols from community Integrated Management of Childhood Illnesses (IMCI) have been integrated into the program as well as distribution of iron tablets. 158 $333 per family/year (5.2 persons on average and 1 child <24 mo.)

274

PROCOSAN had 82,000 beneficiaries as of 2005-06, from 2057 communities in 64 municipalities with a budget of US$208,765.00 for 2005159 (BID 2006). Of these municipalities, 47 are located in the Central region, and 5 are in the Atlantic. The pilot program of PROCOSAN was not evaluated, as the authorities at the time (2003) viewed that expenditure as a non-essential expenditure rather than an essential investment in the effective use of future nutrition program resources. This failure to evaluate the pilot has deprived subsequent decision makers of needed information to help them guide resource allocation to improve nutrition results. A process evaluation was conducted in 2006, and results are expected shortly. According to the criteria presented above, PROCOSAN is a program with a high potential to be effective but all programs need to be evaluated; precisely so we can learn if they are working as designed, and if not what needs to be improved to be effective.

Communication materials from PROCOSAN have been integrated into the health component of RPS and are also used in PAININ (see below). It is highly likely that the reduction in stunting observed in the Central rural region from 2001-05 was a result of the benefits from the RPS and PROCOSAN.

The Programa de Atención Integral a la NiΖez (PAININ) is a psychosocial, cognitive and physical development program run by Mi FAMILIA that benefits children under 6 years (with emphasis on children <3 years) and pregnant and lactating women. The services provided include preschool education, maternal counseling on early childhood stimulation, growth monitoring and promotion and supplementary feeding at program sites within communities. Counseling is also provided to pregnant and lactating mothers on health and nutrition. PAININ has not been considered a nutrition program and previous evaluations have not measured nutrition outcomes (Rafael Flores, personal communication, 2007). In 2005 the program had a coverage of 87,000 children <6, 13,926 pregnant women in 1545 communities located in 64 poor municipalities with a total budget for that year of US$15,486,000.00160 (BID 2006). A new phase of the program is scheduled to initiate this year, and a micronutrient component has been added, that consists of home fortification with micronutrients. The potential effectiveness of this program to reduce malnutrition is not clear, except for the micronutrient component, because by design it is not a nutrition program. The program needs to be integrated with health care interventions in this area to enhance its nutrition impact.

Programa de Atención a Grupos Vulnerables is a vulnerable group feeding program that is carried out by MAGFOR with technical assistance and funding from the World Food Program (WFP) and participation of MINSA at the local level. The objective of the program is to contribute to the dietary needs of pregnant and lactating women and children (through distribution of food rations) and to promote positive practices related to health, sanitation and nutrition. The program is implemented through the basic mother/child programs (prenatal care, growth monitoring etc.) at the health center/post level and the food distribution is conditional to a beneficiary’s participation in the MINSA programs. In 2005 the program had 17, 875 pregnant and lactating women and 20,661 children 7-24 mo. and a budget of US$1,100,000.00. It operated in 36 highly vulnerable municipalities. Documents reviewed did not provide information on a program evaluation, therefore any nutritional impact is unknown (BID 2006; Ruiz and Reyes 2007). Based on the criteria established above, this program is not considered of high potential impact on the nutritional status of children. Although this program has the potential to increase family food availability, small children’s dietary intake depends more on feeding practices, intra-family food distribution and other socio-environmental factors (Neufeld 2006). That is why group feeding programs do not result in an

159 This is US$2.6 per child/year; however these are operating costs and do not include costs for incremental investments that are needed along the life of the program. Ruiz and Reyes (2007) have estimated that the cost to take the program to scale with 100% coverage in the poorest municipalities is approximately US$20 per beneficiary per year. 160 The cost per beneficiary is US$153/year.

275

improved dietary intake for children 0-23 mo. and instead, it is recommended to use foods that are specifically targeted to small children, such as fortified weaning food.

From this review of current government nutrition programs (or programs with a nutrition component) it is very clear that interventions overlap and some programs duplicate functions of other ministries. For example, the RPS provided health services via private providers instead of working in coordination with MINSA. According to Leslie Castro, previously the program manager for the RPS (personal communication 2007), the program costs would be greatly reduced if MINSA provided the health services, since this is the most expensive component. PAININ duplicates MINSA functions with several components: micronutrient distribution; growth promotion activities and counseling activities. Clear policies and clear roles for involved institution and better coordination are needed to avoid this duplication

If Nicaragua has highly effective and potentially effective nutrition programs that are well targeted to vulnerable populations, why is malnutrition declining at such a slow rate?

The simple answer is that investment in highly effective or potentially effective nutrition programs is insufficient to produce a greater reduction of stunting.

Table 9.7. Vulnerable population in municipalities with high and severe poverty Nicaragua 2006

Vulnerable population National level Prioritized population % Pregnant women 174,098 77,121 44 Lactating women 156,845 69,478 44 Children < 2 years 297,332 130,327161 44 Source: Ruiz and Reyes 2007

In 2006 MINSA had classified about 80 priority municipalities,162 based on the poverty map and other criteria, and these geographic areas represented about 44% of the total population. As a result, the vulnerable population was estimated at 130,327 children under 2, and 146,599 pregnant and lactating women163 (Table 9.7) (Ruiz and Reyes 2007). This means that to prevent malnutrition, this vulnerable population must be enrolled in effective programs on a permanent basis (this population will vary yearly due to population growth). Program coverage has never been that high. After start-up in 2001, PROCOSAN was expanded slowly and reached coverage of approximately 82,000 children < 2 by the year 2005-06. This represents 63% of the target population, which leaves a gap of 37% (although the gap was much higher in previous years). Additional coverage provided by the RPS did not increase these figures much. Over a period of 6 years the RPS covered 30,000 families (3 years/family), including an estimate of 30,000 children under two years (about 10,000/year). If these children are included, the gap for 2005-06 becomes 34%, still very high. This situation is aggravated by the fact that MINSA coverage with the basic health package tends to be lower among this same population as well.

As discussed above, targeting within programs has been adequate, and this probably contributed to the reduction in stunting observed in the Central region from 2001-05. However, in terms of government policies, in practice it does not appear that prevention of malnutrition is high on the list of priorities. During 2005 the total budget allocated to highly effective or potentially effective programs (MINSA 164 and RPS) was US$7,137, 611.17. The total budget for programs that are not considered

161 This figure will probably be lower in 2007 due to adjustments of population estimates based on 2005 population census. 162 These figures may have changed with the new government. 163 Although this paper is focused on child malnutrition, maternal nutrition should have the same priority. 164 Includes PROCOSAN, micronutrients, breastfeeding promotion, extension of coverage of basic health package.

276

highly effective or potentially effective to reduce malnutrition based on the established criteria (above) was US$26,359,954.78165 (school feeding, PAININ and vulnerable groups feeding program) (BID 2006; Ruiz and Reyes 2007).

How much does Nicaragua need to invest each year to reduce malnutrition significantly?

For 2007, the vulnerable population in the prioritized municipalities is estimated as 118,348 pregnant and lactating women, and 107,000 children under two years (MINSA statistics office). To provide complete coverage of this population with PROCOSAN, and other MINSA interventions as well as the RPS,166 would cost approximately US$44,166,310.00 per year

In summary, Nicaragua has the knowledge and experience to reduce malnutrition in the short term. But first, the government must treat the reduction of malnutrition as a true priority, in practice. What is needed is: 1) an adequate combination of effective programs (such as PROCOSAN, RPS and other MINSA interventions) and 2) institutional strengthening to increase capacity required to bring programs to scale. In all prioritized municipalities with a high level of poverty, 100% coverage must be achieved and maintained among the vulnerable population. Stunting can be reduced through actions. Nicaragua will never develop to its full potential with such a large stunted population; therefore, stunting must be reduced in the short-term.

CONCLUSIONS

- Stunting is the main nutritional problem due to protein-calorie malnutrition in Nicaragua. Wasting is not a public health problem, and children’s weights are normal, suggesting that energy is not a limiting factor in children’s diets. Other problems are low-birth weight, anemia and possibly zinc deficiency. Inadequate feeding practices contribute to malnutrition among children 0-23 mo.

- Stunting begins early in life, probably due to intrauterine growth retardation, and develops by the age of 24 mo. After this, children grow normally. Programs targeted toward older children or school age children will not decrease stunting.

- Nicaragua has high levels of stunting relative to Latin America as a whole whole (de Onis et al. 2002). Although comparisons are difficult because available information from studies in other countries are from other time periods, age groups etc. However, within Central America, Costa Rica has the least amount of stunting. Guatemala and Honduras have the highest levels. Nicaragua, El Salvador, Panama and Belice have similar levels (WHO global database 2007)..

- The highest levels of stunting are among children in extreme poor households from the Central rural region followed by the Atlantic rural regions.

- Stunting declined 5.9 percentage points from 1998-2005 (most of this from 98-01); the total number of children stunted declined by 62,741 during the period.

- Among the poor, stunting declined steadily and significantly from 1998-2005, at a rate higher than the national average, which suggests that targeting has improved.

165 Includes US$9,773,954.78 for school feeding; the new Sandinista government has plans to increase the school feeding budget to US$17,173,702.80 per year (Hambre Cero, 2007). 166 The RPS is a comprehensive social protection program that includes other short and long term benefits, for example in education; the cost of US$333.33 per family includes all program components since it is not possible to separate out nutrition. Overall costs would be reduced if the health component was assumed by MINSA.

277

- Stunting in Nicaragua is caused by different factors related to care, household food security and access to health services and sanitation. Poverty reduction is a long-term strategy to reduce malnutrition, and meanwhile, shorter, integrated approaches are needed that address the underlying causes of malnutrition.

- Nicaragua has programs that are highly effective or potentially effective in reducing stunting and micronutrient malnutrition. However, coverage among vulnerable populations is still low. Also, nutrition programs have not been the priority, since spending is much higher on other programs that are not effective in reducing stunting. With a true commitment to reduce malnutrition that practice translates into increased spending on well targeted effective programs as well as institutional strengthening, Nicaragua could significantly reduce stunting in a short-time.

Based on 2007 population estimates, it would cost Nicaragua approximately US$44,166,310.00 per year to provide complete coverage of the vulnerable population with PROCOSAN, and other MINSA interventions as well as the RPS, to significantly reduce malnutrition

RECOMMENDATIONS

- The best approach to reduce stunting is to prevent stunting from occurring. Programs should focus on mothers and children under two to prevent stunting. This is the so-called “window of opportunity” to intervene. Interventions targeted to children in other age groups, may have other benefits, but they will not reduce stunting.

- Targeting to the poor is important, and Central rural and Atlantic rural regions should continue to be prioritized for prevention of malnutrition and the goal should be to reach full coverage of vulnerable populations. In other regions of the country, the municipalities with the highest levels of severe poverty should be prioritized. Interventions need to be sustained among poor populations as long as overall conditions are not improved (i.e. poverty is reduced).

- Long term solutions to malnutrition include the improvement of maternal education, improving incomes among the poor, improved water and sanitation. These interventions are necessary and complementary and should not be sacrificed to short term approaches. Both are necessary and there should be a balance. Interventions that increase school enrollment and retention of young girls is a long term step in helping to reduce malnutrition.

- The basic cornerstone of the country’s short-term efforts to reduce malnutrition should be PROCOSAN and the RPS (in extreme poverty municipalities only) as well as MINSA’s micronutrient and breastfeeding promotion programs. These programs have all the elements that make programs effective.

- Increased coverage of nutrition programs should go hand in hand with increased coverage of the basic integrated health care package that should include family planning services.

- Breastfeeding is one of the most important interventions to prevent stunting during the first two years. More investment is needed to promote exclusive breastfeeding for six months and continued breastfeeding up to two years as an effective way to reduce stunting.

- Current programs need to be reviewed in light of their effectiveness to reduce malnutrition; budget allocations should be based on effectiveness criteria and goals for the reduction of malnutrition. Goals should be set to increase coverage of programs to gradually close the gap among the vulnerable population and those levels of coverage should be sustained.

278

- Current supplementation programs for children under two should be enhanced by the addition of zinc (for improving growth and infections) and vitamin C (to improve iron absorption). The best way to deliver these micronutrients along with iron, folic acid and vitamin A is via Sprinkles167 that could easily be distributed through PROCOSAN.

- The maternal basic health care package should be revised and updated especially with regards to nutrition and other services needed to prevent intra-uterine growth retardation.

- Further research is needed to understand why male children are more likely to be stunted than females.

- Because of its importance and relevance, PROCOSAN should undergo an impact evaluation.

167 Sprinkles are sachets containing blends of micronutrients in powder form that is easily added to foods prepared in the home.

279

REFERENCES

Brown KH, Peerson JM and Allen LH. Effect of zinc supplementation on children’s growth: a meta

analysis of supplementation trials. Bibliotheca Nutritio et Dieta 54:76-83, 1998.

Chawla, M. Malnutrition among preschool children in Nicaragua in 1998: prevalence, determinants and

policy implications. November 1999.

de Onis M,. Frongillo EA and Blössner M. Is malnutrition declining? An analysis of changes in levels of

child malnutrition since 1980. Bulletin of the World Health Organization, 78: 1222-1233, 2000.

de Onis M, Garza, C, Onyango, AW, and Martorell, R. (editors). WHO Child growth standards. Acta

Paediatrica International Journal of Pediatrics. Volume 95 April 2006 Supplement 450.

Instituto Nicaragüense de Estadísticas y Censos (INEC), Ministerio de Salud. Nicaragua: Encuesta

nicaragüense de demografía y salud 2001 (ENDESA 2001). Octubre 2002.

Maluccio, J.A. and Flores R. Impact evaluation of a conditional cash trasfer program: The Nicaraguan

Red de Protección Social, Food Consumption and Nutrition Division Discussion Paper No. 184.

International Food Policy Research Institute, Washington, D.C.; 2004.

Maluccio, J.A.; Adato, M.;Flores, R. and Roopnaire, T. Nicaragua Red de Protección Social-Mi Familia:

breaking the circle of poverty. International Food Policy Research Institute, Washington, D.C.;

2005.

Martorell, R.; Flores, R. and Hickey, M. Stunting in Guatemala: Analyses of Change Over 15 Years.

Agricultural Policy Development Project Research Report No. 8. Abt Associates Inc., Bethesda,

Maryland. August 2002.

Ministerio de Salud Nicaragua. Datos preliminares-indicadores básicos de salud 2005. World Wide Web:

http://www.minsa.gob.ni/estadisticas/ind2005/ind2005.html. Consulted May 2007a.

Ministerio de Salud Nicaragua. Sistema integrado de vigilancia de intervenciones nutricionales. Informe

de progreso 2003-2005. Managua, 2007 b.

Ministerio de Salud, República de Nicaragua. Encuesta nacional de micronutrientes (EMN 2000).

Informe completo. Mayo 2002.

Neufeld LM, Hernández Cordero S and Fernández Gaxiola AC. Desnutrición crónica en Centroamérica e

inventario de intervenciones Nutricionales. Dirección de Epidemiología de la Nutrición. Instituto

Nacional de Salud Pública, Cuernavaca, Morelos. Septiembre 2006

Núñez O. Programa para combatir la pobreza y alcanzar la seguridad y la soberanía alimentaria. Consejo

de seguridad y soberanía alimentaria. Managua, enero 2007.

Banco Interamericano de Desarrollo (BID), Gesaworld sociedad anónima. Oportunidades de inversión en

nutrición preventiva para Guatemala, Honduras, Nicaragua y El Salvador. Borrador de plan de

inversiones en nutrición preventiva para Nicaragua, 2006.

280

Pelletier, D.L.; Frongillo, E.A. and Habicht, J.P. Epidemiological evidence of a potentiating effect of

malnutrition on child mortality. American Journal of Public Health, 1993.

Picado, JI; Reyes, CM; Carrillo, P; Yllescas, LT y de Grijalva, Y. “…Lo bueno es que puede comer lo

que nosotros tenemos”: improvement of feeding practices in children under two years of age in

the municipalities of Madriz, Chontales and Masaya. Ministry of Health Nicaragua. March 2001.

Piwoz, E. Undernutrition in Nicaraguan Pre-school Aged Children. Prevalence and Determinants.

Prepared for the World Bank. February 1994.

Ruiz ME, Reyes CM. Programa nacional hacia la erradicación de la desnutrición crónica infantil en

Nicaragua 2007-2015. Nicaragua, March 2007.

Saadeh R, Labbok M, Cooney K, Koziz-Booher P., eds. Breast-feeding. The technical basis and

recommendations for action. Geneva: World Health Organization, 1993.

Sanghvi T. Nutrition essentials. A guide for health managers. BASICS, WHO, UNICEF 1999.

SPSS Inc. SPSS Base 15.0 for Windows User’s Guide. SPSS Inc. Chicago, Illinois, 2007.

StataCorp. Stata Statistical Software: Release 9.0. College Station, TX: StataCorp, 2005.

The World Bank. Repositioning Nutrition as Central to Development. Washington D.C. 2005.

Tumilowicz Torres, AT; Pelletir, D; Habicht, JP; Pelto, G. Timing of growth stunting differs between

indigenous Guatemalan boys and girls and maternal education attenuates stunting in girls.

FASED Journal, 20 (5, Part 2): A 1052, 2006.

Unicef (1998) The State of the World’s Children 1998. Oxford: Oxford University Press.

WHO Expert Committee on Physical Status: the Use and Interpretation of Anthropometry Physical

Status: the use and interpretation of anthropometry: report of a WHO expert committee. (WHO

technical series; 854). Geneva: World Health Organization 1995.

WHO Anthro 2005, Beta version Feb. 17th, 2006: Software for assessing growth and development of the

world’s children. Geneva: WHO, 2006 (http://www.who.int/childgrowth/software/en/).

281

10. OPPORTUNITIES FOR INCOME GENERATION IN NICARAGUA: ACCESS TO

INFRASTRUCTURE, INVESTMENT INPUTS, AND RURAL PRODUCTIVITY

Diego Angel-Urdinola, Ezequiel Molina and Maria Victoria Fazio*

This paper analyzes inequality of opportunities for income generation, mainly looking at roads, electricity, telecommunications, credit, titling, access to networks, and rural productivity. Thus, this section will specifically address the following sets of questions: (i) what are the existing inequities in access to productive services and infrastructure (such as roads, electricity, telecommunications, credit), nationally, by income groups, across rural and urban areas, and across geographical regions? (ii) what are the existing inequities in intangible assets (mainly access to networks and titling) nationally, by income groups, across rural and urban areas, and across geographical regions? (iii) how are income generation outcomes, measured by productivity, related to inequality of opportunities in Nicaragua? (i.e. inequalities in access to productive services, capital, labor, and intangible assets). The paper puts especial emphasis in inequities and determinants of agricultural production. The rational behind this choice is that agricultural production is a direct function of access and quality of productive assets (or so called factor of production); namely capital, labor, and land. Indeed, as will be explained in detail bellow, access to factors of production in Nicaragua (especially to capital and land) displays large inequities across regions and large concentration among the urban non-poor. Furthermore, agricultural productivity is key engine of growth in the agriculture sector. Previous literature (World Bank, 2002; Krueger, 2000; Nadim, 2002) indicates that poverty reduction in Nicaragua is highly responsive to growth in agriculture. This occurs because the agricultural sector (see Gutierrez and Ranzanni, 2007) represents about one fifth of Nicaragua’ total output and one third of Nicaragua’s total employment. There are several channels through which agricultural productivity (and growth in the agricultural sector) can affect poverty: higher agricultural productivity can translate into higher income for producers, more employment, production of cheaper food, and higher tax revenue from agricultural activity, among others.168

INEQUITIES IN ACCESS TO PRODUCTIVE SERVICES AND INFRASTRUCTURE

This section analyses differences in access and quality of income-generating infrastructure services by socio-economic group in Nicaragua using the 2005 Living Standards Measurement Survey (LSMS) for roads, electricity, telecommunications, and credit. The section also provides comparisons of infrastructure-related indicators in Nicaragua vs. other countries in Latin America. The analysis highlights differences within regions and strata as well as disadvantages in access and quality of infrastructure among venerable groups such as poor, indigenous, and agricultural-producer households.

Better Infrastructure is favorable for economic growth. It has long been argued that the development of infrastructure (roads, tunnels, bridges, railways, airports, harbors, telephone exchanges and networks, electricity, access to credit, etc) is associated to favorable possibilities for growth and poverty reduction in a country. Economies with better and broader access to roads, electricity, transportation, credit, and telecommunications area associated with higher growth rates and lower income inequality and poverty

* The authors are with the World Bank. This work was prepared as Background Paper to the Nicaragua Poverty Assessment Report No. - 39736 - NI. We thank Florencia Castro-Leal (Task Team Leader Poverty Assessment, LCSPP) and Jaime Saavedra (Sector Manager, LCSPP) for their support and guidance. The views expressed here are those of the authors and need not reflect those of the World Bank, its Executive Directors, or the countries they represent. 168 Haggblade et al (1989), Haggblade et al (1991), Hazell and Ramasany (1991) and Delgado et al (1994) among others.

282

(de la Fuente and Estache, 2004). 169 The impacts of infrastructure on “development” could be understood as a force that works via both households and enterprises (see Prud’homme, 2004). For households, infrastructure-related services improve welfare by improving their quality of life. A significant share of poor in developing countries, and especially in rural areas, lack of good and reliable infrastructure services. As a consequence of low supply, they generally pay high prices for low-quality services. Access to infrastructure-related services makes it possible to provide inputs used by enterprises at a lower cost (this phenomenon acts just like technological progress, since allows firms to produce at lower prices and increases their incentives to competition). Table 10.1 illustrates a selective review of the literature that addresses the links between access to infrastructure, growth, inequality and poverty. As this is a selective review one disclaimer should be made about the impossibility of generalization of these set of results.

Table 10.1. Links between Access to Infrastructure, Growth, Inequality and Poverty

Source: Author’s compilation

169 Fuente and Estache (2004) find that 53 percent of all studies in the sample support a positive impact of infrastructure investment on productivity or growth. The authors also find a positive impact of investments in infrastructure on growth. The elasticities estimated for Latin America region in the 1990s suggest that a 10 percent increase in infrastructure stocks increase output (GNP) by 1.4 to 1.6 percent (for every percentage point increase in per capita income the authors find that the share of people living in poverty declines by 0.5 of a percentage point).

Infrastructure- related service Author Evidence

Levine, Loayza, and Beck (2000)

Using traditional cross-section, instrumental variable procedures and state-of-the-art econometrics (GMM techniques), the paper finds that the exogenous components of financial intermediary development – measured by (i) liquid liabilities, (ii) assets of deposit money banks divided by assets of deposit money banks plus central bank assets and (iii) credit by deposit money banks and other financial institutions to the private sector divided by GDP – is positively associated with economic growth.

Loayza, Fajnzylber, and Calderón (2005)

Using GMM techniques, the paper finds that financial depth (the ratio to GDP of the stock of claims on the private sector) and public infrastructure (as measured by telephone lines connecting a customer's equipment to the public switched telephone network per 1,000 people) contributes positively to economic growth.

Barham, Carter and Useche (2004)

This paper deals with land market reforms and programs to facilitate access to land among poor households in Guatemala, Honduras, Mexico, and Nicaragua. The authors find that access to capital is crucial for poor households to improve the usage of their relatively abundant labor and scarce assets. Through increasing their efficiency of inputs utilization, households improve their welfare and escape poverty.

Beck, Demirguc-Kunt, and Levine (2004)

Using a broad cross-country sample, the authors find that better financial development (as measured by the value of credit accessible by the private sector divided by GDP) reduces income inequality and poverty.

Financial Depth

Li, Squire, and Zou (1998)

Using a panel data set, the authors conduct empirical analysis that shows that financial depth is an important determinant of inequality.

283

Table 10.1 Links between Access to Infrastructure, Growth, Inequality and Poverty (cont.)

Source: Author’s compilation

Infrastructure- related service Author Evidence

Transportation Gannon and Liu (1997)

The paper discusses the theoretical and empirical links between transportation access, growth and poverty reduction. The major findings are: (i) better transportation boosts economic growth and enhances poverty reduction, (ii) there is a vast range of mechanisms by which transportation affect growth and poverty, so general recipes may not work and country-based policies should be pursed.

Roads Yepes (2004) Using data from a rural panel of households in El Salvador the author estimates two simple measures: the average distance from households to paved roads and to the closest market place in rural areas. These two variables improved in some municipalities more than in others between 1999 and 2001. Results suggest that extreme poverty fell by 8.8 percent in municipalities with little improvements in the two indicators, while it fell by 13.9 percent in those where improvements were significant.

Telecommunications, Electricity and Transportation.

Calderón and Servén (2004)

Using a large panel data set of 121 countries, spanning the years 1960 to 2000, the paper focuses on the links between infrastructure, growth and income inequality. Infrastructure stock is measured as an aggregate index using data from the telecommunications sector (number of main telephone lines per 1,000 workers), the power sector (the electricity generating capacity of the economy —in MW per 1,000 workers), and the transportation sector (the length of the road network —in km. per sq. km. of land area). Infrastructure quality is measured by indicators available in the telecommunications sector (waiting time for telephone main lines —in years), in the power sector (the percentage of transmission and distribution losses in the production of electricity), and in the transport sector (the share of paved roads in total roads). In order to account for the endogeneity issues GMM estimators are used. The authors find that growth is positively affected by the stock of infrastructure assets, and that income inequality declines with higher infrastructure quantity and quality.

Roads, Railways, Telecommunications and Energy

Chong and Calderón (2004)

Applying cross country and GMM dynamic panel methods the paper studies the links between infrastructure development and the distribution of income for the period 1960-1995 in Latin America. In order to asses these relationship quantity and quality proxies are used: For telecommunications, (quantity) the number of telephone main lines connected to local exchanges and (quality) the percentage of unsuccessful local calls (cross-section) together with the waiting list for telephones (panel) are used. For electricity the volume indicator is the electricity generating capacity, while the quality indicator is the transmission and distribution losses of electricity as a percentage of total output. Findings indicate that infrastructure development is negatively linked with income inequality.

284

Infrastructure-based Welfare

This section quantifies household welfare based on an index that captures household access to basic infrastructure and housing. Many of the basic results on poverty and poverty dynamics can be obtained through traditional analyses of consumption can also be obtained using data on household access to assets, housing, and basic infrastructure (Sahn and Stifel, 2000 and 2003; Filmer and Pritchett, 1998; Hammer, 1998). Access to household infrastructure gathers the necessary properties for proper welfare analysis: it retains transparency in construction and rank individuals credibly in terms of welfare (we can safely assume that individuals with more access to infrastructure are better off than individuals with less access to it). Filmer and Pritchett (1998) argue that an index that serves as a proxi for access to household infrastructure may be a better proxi for long-run household wealth than per-capita consumption. This is because changes in per-capita consumption, especially among the poor, are usually reflected on changes in food as opposed to infrastructure consumption. In other words, we can not strictly assume that what happens in the household infrastructure dimension is a good reflection of what happens in the consumption dimension. Measures of welfare based on access to infrastructure, nevertheless, pose a limitation as they treat household access to infrastructure as giving similar utility without allowing for differences in unobserved quality. Despite these limitations, in a poor country facing positive economic growth, such as Nicaragua, analyzing dynamics in infrastructure-based welfare among the poor is relevant to proxi dynamics in long-term welfare and constitutes an important exercise. Access to household infrastructure is likely to foster more relevant multiplier effects within the economy than increases in food consumption.

Box 10.1. A Model to Estimate an Infrastructure Index

In order to estimate a measure of well-being based on assets, we rely on principal component analysis (Lawley and Maxwell, 1971). The principal component constitutes a linear index capturing the most information (variance) which is common to all the variables. Denote by Aij the observation for household i and infrastructure item j (for example, whether the household has a access to electricity or not). Principal component analysis finds a small number of n factors, denoted by the letter f, which can be used to reconstruct the original variables (in this case the original information on housing and infrastructure) as linear functions of the q factors, such that :

Aij = fi1 β1j + fi2 β2j + ... + fiq βqj + εij (1)

In (1), Aij is known since it is one of the values describing whether household i has access to infrastructure item j or not. The term fik represents the observation for household i of the value of factor k which needs to be estimated. The term βkj is the coefficient indicating the dependence of the observed asset variable j upon the factor k (a shadow price); this coefficient being also estimated. The term εij can be as statistical error. In other words, factor analysis produces an index representing (through the vector of common factors F) the data generating process underlying the actual observations Aij. This is done by finding the one dimension of the space in which the original observations are represented with the largest variance, from j = 1, ..., p to k = 1, ...,n with n<p.

Source: Angel-Urdinola and Mete (2007)

Table 10.2 provides descriptive statistics of household infrastructure items considered for this analysis and which were fully comparable in both surveys, namely: access to paved roads, access to electricity, access to fixed phone, access to cellular phone, access to piped water, and trash collection services. Typically, to estimate a better infrastructure index, we also include other dwelling infrastructure characteristics in the analysis, such as the type of floors and walls. Results in Table 10.2 indicate that between 2001 and 2005 most infrastructure and housing variables displayed improvement. For instance,

285

the share of households with access to piped water, fixed telephone, cellular telephone, and trash collection services increased by 5 percent, 40 percent, 696 percent, and 26 percent respectively. Other housing related variables related to the dwelling, such as the share of dwellings with access to toilet inside, good-quality floor, and good-quality walls grew by 29 percent, 6 percent, and 2 percent respectively.

Table 10.2. Descriptive Statistics on Household Infrastructure

Share of households having access to:

1998 %

2001 %

2005 %

% change 2001-2005

Infrastructure Variables Paved road 22.3 42.2 52.2 23.7 Electricity 68.7 72.2 73.8 2.2 Fixed telephone line 9.7 10.1 14.2 40.6 Cell phone n.a. 2.9 23.1 696.6 Piped water 60.8 61.5 64.6 5.0 Trash collection services 31.1 33.7 42.5 26.1

Housing Variables Toilet in dwelling 22.5 22.9 29.4 28.7 Good quality walls* 55.3 60.7 61.7 1.6 Good quality floors** 51.5 56.6 60.0 6.0 House with kitchen room 62.0 65.6 71.9 9.6 Own house *** 77.8 76.9 76.4 -0.7

Source: Authors using the 2005 Nicaragua LSMS. * Walls made from cement, stone, or concrete. ** Floors made of concrete, tile, stone, or brick. *** With or without a title.

Infrastructure-based poverty decreased in Nicaragua by 10.2 percent between years 2001 and 2005. We construct an infrastructure index, denoted by A, using principal component techniques as explained in Box 10.1 above. This index has a log-normal distribution similar to that of other welfare measures such as income or consumption (see Figure A1 in the annex). The infrastructure poverty headcount describes the proportion of the population for whom infrastructure-based-welfare A is less than the poverty line z. We arbitrarily set the poverty line as the median of the infrastructure index A nationally in 2001 (so that the infrastructure headcount is set at 50 percent in 2001). For estimating poverty dynamics, we keep shadow prices of infrastructure items fixed as well as the poverty line. As presented in Table 3, the headcount decreased by 10 percent nationally, by 26 percent in urban areas, and by 5 percent in rural areas. Most of the decrease in the headcount was due to growth (i.e. a positive shift in average household access to particular infrastructure items) and to a lower extent to redistribution.170

170 Changes in poverty rates can be decomposed into changes due to economic growth in the absence of changes in inequality, and changes in inequality in the absence of growth (Datt and Ravallion, 1992). Denoting by P(µt , Lt) the poverty measure corresponding to a mean level of assets in period t of µt and a Lorenz curve Lt , the decomposition

is rrrrir RLPLPLPLPP +−+−=Δ )],(),([)],(),([ 122 μμμμ The first component is the change in poverty that would have been observed if the Lorenz curve had remained unchanged. The second component is the change that would have been observed if mean income had not changed. The last component is a residual.

286

Table 10.3. Changes in infrastructure-based poverty between 2001 and 2005 [base year, 2001]

Infrastructure Index National Urban Rural Headcount 2001 50.0 30.55 76.453 Headcount 2005 44.9 22.94 72.677 Change in the headcount

In percent -10.20 -25.92 -4.94 In percentage points -5.10 -7.61 -3.78

Growth-Inequality decomposition Growth component -5.80 -6.72 -6.21 Redistribution component 0.17 -0.74 0.001 Interaction 0.86 -0.15 2.43


Roads

Gannon and Liu (1997) discuss the link between investment in transportation, growth and poverty reduction.171 The mechanism through which transportation generates higher growth and lower poverty are vast and to a certain extent rest upon the kind of transportation investment being made. Regarding the link with economic growth, the authors argue that investments in transport reduce the cost of assembling intermediate inputs for production. This in turn, reduces production costs and thereby prices. Fall in prices promotes regional and international trade, making it possible for agriculture to commercialize, for industry to specialize, and for production and employment to expand by exploiting economies of scale.

Table 10.4. In rural areas, only 26 out of every 100 households have access to a paved road.

Main way to access dwelling % Paved road

% Unpaved road

% Trocha

% River, Sea, Lake or Other

Total 52.19 31.29 12.56 3.96 Area

Urban 75.99 21.49 0.89 1.63 Rural 18.97 44.97 28.86 7.21

Region Managua 78.62 20.31 1.06 0.00 Pacific 60.42 27.40 8.96 3.23 Central 39.58 37.62 20.66 2.14 Atlantic 9.94 47.17 24.39 18.50


Access to good roads is limited in rural areas (especially in the Atlantic and Central regions) as compared to that in Managua and other urban centers. Compared to international standards (see Figure 10.1), the share of paved roads as a percentage of total roads is low in Nicaragua given its level of development. As presented in Table 10.4, about half of all households in Nicaragua have access to a paved road from their dwelling of residency. In Managua and in the Pacific region, 60 to 80 percent of all households have access to a paved road. In the Central and Atlantic regions this proportion is considerably lower at 40 and 10 percent respectively. In the Atlantic (Central) region, about 70 (60) percent of all households access their dwellings by using unpaved roads and trochas, which are usually poorly maintained informal roads difficult to access by regular vehicles). In the Atlantic region, about 18.5 percent of all households claim that the main way to access their dwelling is by river or sea. The

171 For a review on the Asia experience on the impact of investment in transport and energy on poverty see (Cook et al, 2005)

287

Central and Atlantic regions clearly fall behind in terms of connectivity through roads, which is likely to be hindering commerce, tourism, and flows of investments between them and the rest of the country.

Figure 10.1. The share of paved roads as a percentage of total roads in Nicaragua is below regional standards given its level of development [Comparison for year 2000]

ARG

BOL BRA

CHL

COL

CRI

DOM

ECUSLV

GTM

HND

MEX

NI

PAN

PRY

PER

VEN

0.0

10.0

20.0

30.0

40.0

50.0

60.0

7.50 8.00 8.50 9.00 9.50 10.00

Log (GDP 2004)

% P

aved

Roa

ds o

f Tot

al R

oads

Source: WDI dataset. Year 2000 is the last available year with a good sample of countries for this series.

Although poor households in Nicaragua still have limited access to paved roads, significant progress has been achieved in this front since 1998. As illustrated in Figure 2, household access to paved roads from their dwelling of residency has improved significantly since 1998, especially among the poor. In 1998 only about 8 of every 100 households in the bottom quintile had access to a paved road from their dwelling of residency. The same proportion reached 24 percent in 2005. Similar progress was achieved at all quintiles. As far as the process through which transport investments lead to improvements in the standard of living of the low-income households often involves many links. In addition to improving accessibility, transport investment affects employment providing income-earning opportunities for the poor. Furthermore transport plays a crucial role in promoting agricultural production and commercialization by reducing the cost of shipping agricultural products to market. This also helps expanding the area of land under cultivation and as a result the production of exports expands too.

Poor households, indigenous, and households engaged in agriculture have limited access to paved roads. As illustrated in Figure 10.3, access to paved roads from the dwelling of residency differs greatly by socio-economic condition. In particular, while 64 out every 100 non-poor households have access to a paved road from their dwelling, only 32 out of every 100 poor households do so. Access to paved roads is particularly low among households in the lowest quintiles, among indigenous households, and among households engaged in agricultural production (18 to 21 percent). However, significant investments to improve roads infrastructure are being make. In particular, data suggests that about 25 percent of all households benefited by a roads-program in the year prior to the survey. This share was 33 percent for indigenous households.

288

Figure 10.2. Access to paved roads has roughly doubled since 1998 for households at all socio-economic groups.

Main access to dwelling is paved road

0

10

20

30

40

50

60

70

80


% h

ouse

holds

2005 2001 1998

Source: Authors using the 1998, 2001, and 2005 Nicaragua LSMS.

Rural roads are thought to contribute to poverty reduction through access to education and health, and expansion of markets for agricultural products. Using data from improvement in rural roads in El Salvador over the 1999–2001 period Yepes (2004) found that the impact on poverty of these improvements are remarkable. In cantons where roads improved poverty fell by 5 percent more than in cantons were roads did not improve. Estache and Fay (1995) find that improved access to roads and sanitation has been a key determinant of reduction of inequality and poverty for the poorest regions both in Argentina and Brazil.

Figure 10.3. Indigenous households and households engaged in agricultural production are the ones with the lowest access rates to paved roads.

64.0

32.2

23.7

35.4

47.6

60.1

74.2

53.6

21.0

69.2

17.5

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

Non-p

oor

Poor

Pooes

t Quintile

Q2

Q3

Q4

Riches

t Quintile

Non-in

dige

nous

Indige

nous

Non agr

ic. P

rodu

cer

Agric.

Pro

duce

r

% h

ouse

hold

s

Principal Access road is paved


289

Lack of access to roads hinders access to markets and better economic opportunities, especially among households engaged in agriculture Lack of good roads makes it difficult for households engaged in agriculture to commercialize their farm production, which leaves them in the hands of intermediaries who often buy their produce at lower than market prices (due to high transportation costs). Table 10.5 displays the type of buyer of agricultural produce according to the region of residency. Note that the share of sales purchased by direct consumers by region is related to their connectivity in terms of roads. In particular, while 28 percent of all producers in Managua sell their agricultural output directly to the consumers, the same proportion is at 5 percent for farmers in the Central and Atlantic regions. Indeed, the majority of the producers in less accessible regions rely on local or outside merchants to purchase their agricultural produce. As Figure 10.4 illustrates, the average distance between farmers and the closest “commercial road” is at 0.53 Km in Managua while it reaches 18 and 136 km in the central and Atlantic regions.

Table 10.5. Farmers living in less accessible regions rely on merchants to purchase their agricultural production

Main Buyer of agricultural produce Managua Pacific Central Atlantic Total % Direct consumer 28.09 14.30 5.22 5.53 7.68 % Local merchant 28.76 27.15 31.08 40.28 32.36 % Outside merchant 43.15 56.85 61.21 53.41 58.17 % other 0.00 1.70 2.49 0.77 1.80 Total 100.0 100.0 100.0 100.0 100.00 Source: Authors using the 2005 Nicaragua LSMS. Sample: households engaged in agriculture who sale their production for profit.

Box 10.2. IDA Involvement on Roads in Nicaragua

Long-term IDA involvement has helped increase the share of the road network that is in working condition by almost 20% between 1999 and 2006 (from17% to 20% of all roads).172 IDA-financed projects restored over 3,000 km of secondary roads destroyed by Hurricane Mitch, rehabilitated the Pan American highway between Managua and San Benito – bringing back on line a key trade link with the rest of Central America – and more recently, improved 240 km of rural roads, linking the poor to markets, health centers and schools. Throughout, it has employed a contingent of local micro-enterprises in road maintenance ventures – giving very poor people stable, productive work – and built analytical, fiduciary, and maintenance capacity at ministerial and local levels. IDA’s support helped leverage additional funds for the establishment and operation of a national Road Maintenance Fund in 2005, which will help ensure long-term viability of road investments.

Source: Nicaragua IDA impact country story. World Bank (2007)

172 IDA technical assistance and funds from Danida up-graded the systems and capacity in the Ministry to monitor road quality, as such post-1998 data is much more reliable.

290

Figure 10.4. For producers in the Atlantic region, the (median) average distance between their farm and the closest “commercial” road reaches (92) 136 Km.

0.53 3.77

18.01

136.46

0

20

40

60

80

100

120

140

160

Managua Pacific Central Atlantico

Kilo

met

ers

Distance to main road used to commercialize agricultural produce -- in Kms

Source: Authors using the 2005 Nicaragua LSMS. Sample: households engaged in agriculture who sale their production for profit.

Energy

Access to electricity is restricted among the poor, and especially among indigenous household and among those engaged in agriculture. Lower access to electricity hinders welfare, especially for households working in agriculture and who deal with perishable products, such as milk and milk derivatives. Lack of access to electricity also lowers the capacity of households to run small businesses and may negatively influence their children’s education. Table 10.6 suggest that poor households, especially in rural areas and in the Atlantic region, are less likely to have access to electricity at home than non-poor and urban households. Vulnerable groups such as indigenous households and households engaged in agriculture display electricity access rates below 50 percent.

Table 10.6. Statistics on main source of lighting by socio-economic group

Main source of light at home Electricity Kerosene-Gas Candles and other Socio-economic group

Poorest quintile 33.89 48.77 17.35 Quintile 3 74.46 17.18 8.36 Richest quintile 93.91 3.79 2.30

Vulnerable group Non indigenous 74.93 18.74 6.33 Indigenous 46.53 18.74 34.72 Non-Ag. Producer 91.10 4.67 4.23 Agricultural Producer 38.13 47.47 14.40


291

Figure 10.5. There are large differences in access to electricity by strata and region.

95.5

43.3

98.6

86.8

56.8

34.1

0

20

40

60

80

100

120

Urban Rural Managua Pacific Central Atlantic

% H

ouse

hold

s

Access to Electricity at home


Differences in access to electricity are large across regions. Figure 10.5 illustrates that there are large differences in the share of households with access to electricity at home by region and strata. Data suggest that access to networks among households in urban areas (96 percent) is significantly larger than that in rural areas (about 43 percent). While in Managua almost all households have access to electricity at home, in the Central and Atlantic regions only 57 and 34 percent of all households have access to electricity respectively. As Figure 10.6 illustrates, access to electricity in urban areas surpasses regional standards given Nicaragua’s level of development. However, this is not the case in rural areas where access to electricity is below regional standards.

Figure 10.6. Access to electricity in Nicaragua in urban (rural) areas is above (bellow) regional standards given its level of development

BOL, 2004

BRA, 2004

CHL, 2003

COL, 2004ECU, 2003

SLV, 2004

GTM, 2003

HND, 2003

MEX, 2003

NI, 2005

CRI, 2004

DOM, 2004

0

20

40

60

80

100

120

7.60 7.80 8.00 8.20 8.40 8.60 8.80 9.00 9.20 9.40

Log (Per capita GDP 2004)

Acc

ess

to E

lect

ricity

in %

(R

ural

are

as)

Source: WDI dataset.

292

Figure 10.6: Access to electricity in Nicaragua in urban (rural) areas is above (bellow) regional standards given its level of development (cont)

BOL, 2004

CHL, 2003

COL, 2004

DOM, 2004

ECU, 2003

SLV, 2004

GTM, 2003HND, 2003 NI, 2005

BRA, 2004

CRI, 2004

MEX, 2003

90

92

94

96

98

100

7.60 7.80 8.00 8.20 8.40 8.60 8.80 9.00 9.20 9.40


Acc

ess

to E

lect

ricity

in %

(U

rban

are

as)


Kerosene is the main alternative source of energy for lighting in rural areas and among the poor. Kerosene constitutes, after electricity, the second main source of lighting in Nicaragua. As illustrated by Figure 10.7, between 40 and 43 percent of all households in rural areas (and mainly in the Central and Atlantic regions) use kerosene as the main source of lighting. Note that in the Pacific region and in Managua, non electricity sources of light are less used. Candles and others (such as firewood), which are very inefficient sources of light are still being used by 26 of every 100 households in the Atlantic region.

Figure 10.7. Little less than 20 percent of all households in the Atlantic region still use candles as their main source of light.

1.3

43.1

9.6

34.1

40.4

3.2

13.7

1.4

3.7

9.2

25.5

0

5

10

15

20

25

30

35

40

45

50

Urban Rural Managua Pacific Central Atlantic

% H

ouse

hold

s

Kerosene

Candes and other

Source: Authors using the 2005 Nicaragua LSMS

293

Access to electricity for households in the bottom two quintiles, albeit low, has increased by 32 and 15 percent respectively since 1998. Figure 10.8 displays household access rates to electricity at home by quintile for years 1998, 2001, and 2005. Results indicate, as mentioned before, that access rates among richer household are much higher than among poorer ones. Nevertheless, improvements in access over time have been more relevant among the poor, and especially among households in bottom quintiles. Between years 1998 and 2005 access rates increased from 26 to 34 percent among households in the first quintile and from 50 to 58 percent among households in the second quintile. Improvements among the rich, on the contrary, were rather mild, especially among households in the third and fifth quintiles. Data suggest that about 16.5 percent of all households have benefited from a “utilities” program (usually associated to the expansion networks) in the year prior to the survey. Beneficiaries were concentrated in rural areas (21 percent of all households), in the poorest quintiles (22.4 percent), and among agricultural producers (23.4 percent) and indigenous households (20 percent).

Figure 10.8. There has been significant progress in household access to electricity between years 1998 and 2005, especially among the poor.

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0


Per

cent

2005 2001 1998

Source: Authors using the 1998, 2001, and 2005 Nicaragua LSMS.

Relative to their income, electricity consumption is more expensive for households living in the Atlantic Region. Figure 10.9 plots expenditures on electricity per month as share of total household income. Results indicate that energy consumption more expensive relative to their income level in urban areas, among non-poor households, and among households living in the Atlantic and Managua regions. In particular, while energy consumption per month accounts for 2.7 percent of monthly income for households in the Central region, it accounts for 4.4 percent of monthly income for households in the Atlantic region (that is, 63 percent more; perhaps due to low economies of scale and high marginal costs of expanding networks in this region). In the LAC context, electricity consumption among households in Nicaragua is normal relative to their income. Figure 10.10 below indicates that most households in Latin America use between 3 and 5 percent of their monthly income to pay for electricity services. In Nicaragua this share is at 2.4 and 3.5 percent for rural and urban areas respectively. To control electricity prices in Nicaragua is a task difficult to achieve if international oil prices remain high given that the country is one with the highest rate of electricity production from oil sources in the region (see Figure 10.11). Keeping oil prices down in recent years has only been achieved by rationing energy supply, which in turn has deteriorated the quality of the service.

294

Figure 10.9. Relative to their income level, consumption of electricity is more expensive in rural than in urban areas

3.52

2.41 2.44

3.493.74

3.04

2.69

4.37

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Non-poor Poor Rural Urban managua pacifico central atlantico

Ele

ctric

ity a

s %

of I

ncom

e (m

onth

ly)

Source: Authors using the 2005 Nicaragua LSMS. Sample: households connected to the networks and who pay for the service.

Figure 10.10. The share of income spent by households on electricity in Nicaragua is low for international standards, especially in rural areas.

GTM, 2000

PER, 2003

BOL, urban 1999NIC, urban 2005

NIC, rural 2005

COL, urban 2003

URY, 2003

MEX, 2002ARG, 2002

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.50 8.00 8.50 9.00 9.50 10.00


Ele

ctrici

ty a

s per

cent of in

com

e (m

onth

ly)

Source: Komives et all (2006) and 2005 Nicaragua LSMS. Sample: households connected to the networks and who pay for the service.

Between 15 and 20 percent of all households with access to electricity do not pay for the service. Table 10.7 provides some statistics on payments for households with access to electricity. Electricity theft is common in Nicaragua, especially among the poor. In particular, results suggest that 20 to 25 out of every 100 users of electricity in the poorest quintile do not pay for the service (i.e. they are connected to

295

the network illegally). High rates of electricity theft constitute a burden to energy suppliers. High rates of energy theft and high oil prices force suppliers to cut the service periodically in order to save costs. Furthermore, results indicate that electricity consumption for richer households is more likely to be metered than that of poor households (83 percent in the richest quintile vs. 51 percent in the poorest quintile). Lack of metering usually forces utility companies to guess household consumption based on average patterns of the population and not based on actual consumption (which is generally associated with lower revenues).

Table 10.7: Electricity theft is common in Nicaragua, especially among the poor.

If the household has access to electricity at home % Pays, with meter % Pays, no meter % Does not pay Socio-economic group

Poorest quintile 51.70 22.64 25.67 Quintile 2 66.20 13.74 20.06 Quintile 3 66.89 14.63 18.49 Quintile 4 76.41 8.73 14.87 Richest quintile 83.51 7.76 8.72

Vulnerable group Non indigenous 74.08 10.94 14.98 Indigenous 74.27 17.64 8.09 Non-Ag. Producer 75.60 10.23 14.18 Agricultural Producer 66.68 15.52 17.79


Figure 10.11. It will hard for Nicaragua to keep prices of energy down due to the country’s high dependency on oil to produce electricity

0.25 1.09 1.31 1.762.97

9.72

19.06

29.6832.38

34.80

40.03

49.0051.66

69.53

75.18

0

10

20

30

40

50

60

70

80

COL ARG CHL CRI BRA PER BOL ECU MEX GTM SLV PAN HND DOM NIC

Ele

ctric

ity p

rodu

ctio

n from

oil so

urce

s (%

of t

otal

)


Telecommunications

There are large gaps in access to telecommunications between poor and non-poor households. Investing on telecommunications is important for economic growth. Calderón and Servén (2004) indicate that a higher stock of telecommunications (proxied by number of main telephone lines per 1,000 workers)

296

and better communications infrastructure (measured by waiting time for telephone main lines , in years) is associated with higher economic growth and lower income inequality. Figure 10.12 plots differences in household access to fixed and cellular telephones by socio-economic group. Households in the richer quintiles have access rates to fixed and cellular phone that oscillate between 10 and 30 percent while access rates among households in the bottom quintiles oscillate between 2 and 15 percent. Access to cellular telephones is larger than access to fixed telephones at all socio-economic quintiles. Indeed, access to cellular telephones is significantly larger than that to fixed telephones for groups living in more isolated areas, such as indigenous households (19 vs. 3 percent) and households engaged in agriculture (6 vs. 1 percent). Cellular phones are also more accessible to households in the poorest quintiles. While household access rates to fixed phones among households in the first and second quintiles are lower than 1 percent, access rates to cellular phones are at 1 and 6 percent respectively.

Figure 10.12: Access to telecommunications is low even among households in the richest consumption quintiles

11.2

0.70.0 0.5

2.4

7.2

18.5

7.5

2.6

10.5

0.7

23.1

4.4

1.3

5.5

13.2

22.0

27.8

16.0

18.6

21.4

5.4

0

5

10

15

20

25

30

Non-p

oor

Poor

Pooes

t Quin

tile Q2 Q3 Q4

Riches

t Quin

tile

Non-in

digen

ous

Indig

enou

s

Non a

gric.

Pro

duce

r

Agric.

Pro

duce

r

Per

cent

Hou

sheo

lds

with

acc

ess

Fixed Telephone Cellular telephone Both


Less than 1 out of every 100 households in rural areas has access to a fixed telephone at home. Results from the 2005 LSMS indicate that access rates to a fixed telephone are very low in rural areas and in the Central and Atlantic regions (averaging 2 to 4 percent). Access rates to fixed phone in urban areas and in Managua oscillate between 12 and 15 percent. Interestingly, household access to cellular phones is larger than that to fixed phones at every region and strata and especially in rural areas (6 vs. 1 percent) and in the Atlantic region (9 vs. 2 percent). This suggests that cellular technology has become the main mechanism to access telecommunications services in the more isolated regions. As Figure 10.13 illustrates, Nicaragua displays one of the lowest number of telephone mainlines per 1,000 people as compared to other countries in the region. Moreover, no much progress in expanding telephone mainlines been achieved in recent years as compared to that achieved by some of Nicaragua’s CAFTA competitors such as Guatemala, Costa Rica, and El Salvador (see Figure 10.14)

297

Table 10.8: Telecommunications through cellular technology has become the leading way to access to phone services in Nicaragua.

% with Fixed telephone % with Cell phone % with Both Urban 12.3 23.5 11.9 Rural 0.3 5.8 0.1

Managua 15.2 24.3 15.6 Pacific 5.5 19.8 5.5 Central 4.5 8.9 3.3 Atlantic 2.1 8.8 1.6


Figure 10.13: Nicaragua displays one of the lowest access rates to fixed telephone in the region.

ARG

BOL

BRA

CHLCOL

CRI

DOM

GTM

ECU

HND

MEX

NIC

PAN

PER

PRY

SLV

URY

VEN

0

50

100

150

200

250

300

350

7.50 8.00 8.50 9.00 9.50 10.00


Tel

epho

ne m

ainl

ines

per

1,0

00 p

eopl

e in

yea

r 20

04


Box 10.3. IDA Involvement on Telecommunications in Nicaragua

With IDA’s support, national coverage of fixed and mobile phones increased sevenfold from 194,000 in 1999 to more than 1.3 million in 2005, and mobile coverage in provincial capitals increased from 50 percent in 1999 to 100 percent in 2003. Prior to 2000, insufficient competition left Nicaragua with one of the lowest telephone connection rates in Latin America. IDA investment lending helped privatize the state monopoly, and created a regulatory agency. IDA’s budget operations supported passage of competition laws needed to level the playing field, and facilitated the creation of a Telecom Investment Fund to extend services in rural areas. By putting this fund to work, an IDA financed project helped half a million citizens living in 365 small towns benefit from public payphone access. Now, some of Nicaragua’s poorest can connect to doctors and relatives and obtain market data on their crops, putting them one step closer to modernity and economic opportunity.

Source: Nicaragua IDA impact country story. World Bank (2007)

298

Figure 10.14. Among all actual and potential members of the CAFTA, Nicaragua is the country with the lowest telephone mainlines per 1,000 people.

0

50

100

150

200

250

300

350

2000 2001 2002 2003 2004

Telep

hone

mainlines

(pe

r 1,

000

peop

le)

Costa Rica

Nicaragua

Honduras

Guatemala

El Salvador

Rep. Dominicana

Source: WDI dataset

CREDIT SERVICES

About 25 percent of all households claim to have received a loan in the 12 months prior to the survey. Financial intermediary development – measured by liquid liabilities and credit deposits divided by GDP – is positively associated with higher economic growth and lower inequality (Levine, Loayza, and Beck, 2000). The 2005 LSMS includes a module on credit. Unfortunately, the survey collects information on credit only if a household received a loan in the past 12 months. Therefore, the survey does not allow gathering information on households who applied for credit and did no get it, which is a better proxi for access to the service. Nevertheless, detailed data are gathered on the source of financing, amounts, interest rates, etc. As figure 16 illustrates, with the exception of indigenous households, about 25 of every 100 households in Nicaragua received a loan in the 12 months prior to the survey at all socio-economic groups. A little more than half of all loans given to households were issued by informal creditors (such family, friends, NGOs, merchants, or informal credit lines). Poor households, indigenous households, and households engaged in agriculture are likely to obtain their loan form an informal credit sources which, as we will indicate below, are associated with higher interest rates. Financial development may reduce poverty through higher growth by improving the allocation of capital. Poor households who are led to become entrepreneurs due to involuntary unemployment have no collateral, credit history, or connections. These capital market imperfections may prevent the allocation of capital to poor entrepreneurs with high-return projects. By promoting financial development and access to credit markets these projects could made possible and inequality and poverty would be reduced. 173

Loans among agricultural producer households account for up to 20 percent of their per-capita income. Table 10.9 provides some descriptive statistics on loan amounts by socio-economic group. Results indicate that loans (in per capita per year) among the poor account for a large share of their yearly per-capita income. In particular, loans are significantly large as a share of income for agricultural

173 For simplicity, in order to estimate some credit statistics (such as type of provider and use of the loan) at the household level, we restricted our sample to the largest credit received in cases where households received more than one loan. Estimates indicate that 9.7 percent of all household who had access to a credit received more than two loans.

299

producing households (roughly 20 percent). Loans to the poor (averaging 670.oo Cordobas per-capita per year) account for 11 percent of their yearly per-capita income. The same share is at 12 percent for the non-poor. Regression analysis suggests that more educated households (i.e. those having a head with tertiary education) are 11 percent more likely to have obtained a credit from a formal lender (banks, cooperatives, and micro-finance institutions) than households having a head with no education. Households in urban areas are 7 percent more likely to access formal credit while households in the Atlantic region are 25 percent more likely to borrow from informal sources (such as merchants, informal lenders, and friends). Somehow surprisingly, having a land/house property title does not influence the probability that households access formal credit (see Table A1 in the annex).

Figure 10.15. The majority of all loans received by households are issued by informal sources of credit, such as merchants or family/friends.

15 16 16 15 16 17

1215

1315 14

13

64

8 7

13

16 11

3

11

9

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

Non-poo

r

Poor

Pooe

st Q

uintile

Q2

Q3

Q4

Riches

t Quintile

Non-in

dige

nous

Indige

nous

Non agric.

Produ

cer

Agric

. Produ

cer

% h

ouse

hold

s

% receiving informal credit % receiving formal credit

Source: Authors using the 2005 Nicaragua LSMS [formal lenders: private banks, government/MAGFOR, micro-finance institutions, credit cards, cooperatives, and Non conventional banks].

Table 10.9. Descriptive statistics on loan amounts received by households

Average loan amount given to households

- Cordobas per capita per year - (1)

Average household income - Cordobas per capita per year -

(2)

Loans as % of income

(1)/(2)

Socio economic group Non poor 2,821.32 21,734.10 12.98 Poor 670.28 6,356.42 10.54 Poorest quintile 423.81 4,696.24 9.02 Quintile 3 1,187.53 9,263.26 12.82 Richest quintile 4,473.42 34,163.53 13.09 Vulnerable group Indigenous 988.16 19,083.80 5.18 Agricultural Producer 2,545.02 12,980.34 19.61


300

Loans from informal providers (often more accessible to the poor) are associated with higher interest rates than loans from formal providers. Figure 10.16 plots credit providers and the average monthly interest rate they charge on their loans. Results suggest that informal credit lines and credits from merchants are the providers of 45 out of every 100 loans in Nicaragua. These providers – generally more available to the poor (see Table 10.10) – charge very high interest rates as compared to formal credit providers such as private banks, cooperatives, and other financial institutions. Estimates indicate that while interest rates charged by informal lenders can be as high as 12 percent per month, interest rates charged by formal lenders fluctuate around 4 percent per month.

Figure 10.16. About 60 percent of all household loans are issued by financial institutions (24 percent) and informal credit lines (31 percent).

2.4

0.2

23.5

2.4

8.5

4.0 4.05.2

31.1

2.5

16.3

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

Private ba

nk

Gov

ernm

ent o

r pro

gram

MAG

FOR

Fina

nce institu

tions

Credit c

ards

Coop of sav

ings

and

cre

dit a

nd pro

ducti

Non con

ventiona

l ban

k/pr

ivate le

nder

NGO/P

ROJE

CT

Friend

s, re

lativ

es and

neigh

bors

Mer

chan

ts

Inform

al cre

dit line

Other

% lo

ans

acco

rdin

g to c

redit p

rovi

der

0

2

4

6

8

10

12

14

Ave

rage

month

ly in

tere

st


Table 10.10. Main credit providers by socio-economic group

Poorest quintile

Q 2 Q3 Q4 Richest quintile

% Private bank 0.0 0.4 0.5 0.4 6.7 % Government program 0.0 0.0 0.4 0.5 0.0 % Finance institutions 5.8 17.8 17.7 29.5 31.4 % Credit cards 0.1 1.4 1.3 1.0 5.5 % Cooperatives 10.1 11.1 7.5 8.1 7.6 % Non conventional banks 4.2 3.1 3.1 3.8 4.9 % NGO project 2.6 6.6 4.5 2.7 4.0 % Friends/relatives and neighbors 5.5 7.4 4.2 3.9 5.7 % Merchants 4.9 7.5 16.6 25.4 17.4 % Informal credit line 63.0 41.1 40.6 22.8 15.6 % Other 3.9 3.6 3.7 1.9 1.2 Source: Authors using the 2005 Nicaragua LSMS

About one third of the acquired loans are used for investment purposes. Estimates using the 2005 LSMS suggest that about 70 percent of all loans received by households are used for general household consumption (purchasing cars, houses, and other non-investment items) and 30 for business related investments. As illustrated by Figure 10.17, households engaged in agriculture are more likely to use

301

loans for investment-related purposes than the average household (40 percent vs. 30 percent). Indigenous households display the highest rate of loans used for household consumption at 76 percent. .

Figure 10.17. Poor households are more likely to use loans for investment-related purposes than non-poor households are.

31.029.0

27.0

33.0

28.0

32.0 31.0 31.0

24.0 23.0

48.0

0

10

20

30

40

50

60

Non-p

oor

Poor

Pooes

t Quintile Q2 Q3 Q4

Riches

t Quintile

Non-in

dige

nous

Indige

nous

Non agr

ic. P

rodu

cer

Agric.

Pro

duce

r

% h

ouse

hold

s us

ing

cred

it fo

r in

vest

men

t pur

pose

s

Source: Authors using the 2005 Nicaragua LSMS. Investment loans are defined as those destined to improve/expand agriculture and non-agriculture businesses.

Participation of the head in social networks, the level of education of the head, socio-economic status, and region are important determinants affecting the probability of getting a loan. Table 10.11 presents results of a probit regression that estimates marginal effects of household characteristics in the probability of getting a loan in Nicargua. The analysis was conducted for all households in the sample and for agricultural producer households only. Results for both samples convey very similar messages. Estimates indicate that having a head who participates in a social network (credit union, professional association, or local committee) increases the probability that a household receives a loan by 9 percent. Specifically, having a head who is a member of credit union (professional association) increases the probability of getting a loan by 20 to 21 (13 to 19) percent. Controlling for other observables, poorer households are 2 to 5 percent less likely to obtain a credit than non-poor households. Education is an important determinant affecting the likelihood of getting a loan: households with secondary and post-secondary education are 6 to 25 percent more likely to obtain a credit as compared to households having a head with no education.

302

Table 10.11. Marginal Effects of household characteristic in the probability of getting a loan

Whole Sample Agricultural Producers Marginal Effects (Dependent variable: 1 if head of household received a credit received a credit)

(1) (2) (1) (2)

Participation Head participates in a productive association* 0.09 0.09 Different Social Networks

Household benefits from agricultural program - 0.03 - N/S Head has participation in Credit Union - 0.21 - 0.20 Head has participation in Professional Association - 0.13 - 0.19 Head has participation in Local Committee - - - N/S

Poor Household is poor -0.02 N/S -0.04 -0.04 Household is in extreme poverty -0.05 -0.05 -0.05 -0.05 Education of the Head Head, primary 0.02 0.02 N/S N/S Head, secondary 0.06 0.06 0.07 N/S Head, technical 0.15 0.15 N/S N/S Head, tertiary 0.09 0.08 0.25 0.21 Region Managua N/S N/S N/S N/S Pacifico N/S N/S N/S N/S Central 0.07 0.07 0.06 0.06 Urban dummy N/S -0.02 N/S N/S Observations 6861 6861 2794 2794 Source: authors’ own calculations from the EHM 2005. Presented (underlined) coefficients are significant with at a 5 (10) percent confidence level. Reference categories: education of the head: no education or preschool; geographic Location: Atlantic. *credit unions, professional association, and local committees.

INEQUITIES IN INTANGIBLE ASSETS

Networks and Organizations

Access to networks and associations has become a coping mechanism for households to promote social participation, empowerment, and better to access markets and services. There is evidence (Escobar, 1995 and Scott, 1998) suggesting that service delivery and competitiveness tend to be low in countries where institutions remain weak. Participation in associations strengthens community driven development (CDD), which in turn contributes to strengthen governance, improve the targeting of social programs, enhance local capacity building, and promote the inclusion of the poor (Mansura and Rao, 2004). Community-based associations such as the Self employed Women’s Association in India, the Orangi slum association in Pakistan, and the Iringa Nutrition association in Tanzania, among others (Mansura and Rao, 2004) are examples of highly successful associations that have contributed to increase the welfare of their participants and their communities through service delivery and community programs. Since community-based development relies on the capacity of individuals and communities to self-organize and to use their social capital productively, the concept of participation in association becomes essential in the CDD process. In general, participation in association is important to access markets and inputs (e.g. producer associations); to protect individuals against other institutions (e.g. unions and consumer associations); to gain political power (community committees); and to access goods, programs, or services (e.g. religious associations and government programs).

303

Box 10.4. Association of Organic Coffee Growers of Matagalpa

Producer associations in Nicaragua are protected under the Municipality Law (Law 45) and the Social Participation Law (Law 475). The IPADE (Instituto para el Desarrollo de la Democracia) and USAID are institutions promoting development of successful programs to strengthen associations of producers in Nicaragua. One example is the “Association of Organic Coffee Growers” in the municipality of Molino Norte in the State of Matagalpa. Seeking help from USAID, the association achieved to train a group of coffee growers in how to produce and commercialize organic coffee. Nowadays organic coffee is worth 3 times as much as regular coffee in the international market and producers have been able to enjoy from higher revenues as they have been trained to access and benefit from this market. Today, about 419 hectares in the state of Matagalpa are used for the production of organic coffee. Small and medium producers have not only achieved to get a better price for their coffee production, but also to cut production costs. “Now, everybody want to produce organic coffee”, claims Mr. Reyes; the president of the Association of Organic Coffee Growers.

The organic coffee project is part of a US$157 USID broader initiative to promote sustainable development in Nicaragua. Coffee producer Arturo Jaén, owner of “La Ponderosa farm” learned to produce its own fertilizer out of chicken dung and coffee/corn residuals. Mr. Jaén spends US$ 0.53 cents to produce a “quintal” of this fertilizer as compared to US$12 he would have to spend to purchase commercial fertilizer. Besides contributing to improve productivity and the welfare of producers, the association has achieved to promote production techniques that are environmentally friendly and that have helped to reduce erosion and the exposure of workers to toxins that may be harmful to their health.

Source: USAID (http://nicaragua.usaid.gov/historia_40.html)

Participation in productive organizations increases the probability that households benefit from social programs by 15 to 16 percent. Table 10.12 estimates the impact of household participation in productive organizations (local committees, professional associations, and credit unions) on their probability to benefit from social programs (utility programs, titling programs, agriculture programs, and health and education programs). Results indicate that controlling for other socio economic characteristics (education of the head, region, socio economic group, and strata among others) households having a head who participates in a productive association are 15 to 14 percent more likely to benefit from social programs. Households having a head who attained intermediate education (secondary and technical) are 11 to 15 percent more likely to benefit from social programs as compared to households having a head with no education. Households in the Atlantic region are 15 percent less likely to benefit from social programs than households in Managua and urban households are 18 percent more likely to benefit from social programs that rural ones.

304

Table 10.12: Marginal Effect of household characteristic in the probability of benefiting from social programs [social programs: utilities, titling, agriculture, health and education]

(1) (2) (3) Dependent Variable: 1 if household benefits from programs, 0 if not

Social Programs

Social Programs

Utility Programs

Participation Dummy =1 if head participates in productive associations 0.162 0.071

Local committees and Prof. Associations only Dummy =1 if head participates in productive associations 0.154

Local committees, Prof. Associations, and Credit Union Education of the Head

Head, primary 0.073 0.073 0.018 Head, secondary 0.104 0.104 N.S. Head, technical 0.135 0.136 N.S. Head, tertiary N.S. N.S. N.S.

Indigenous head 0.080 0.081 0.083 Region

Pacific N.S. N.S. -0.077 Central N.S. N.S. N.S. Atlantic -0.153 -0.151 -0.09

Urban dummy 0.185 0.185 -0.057 Socio-economic group

Second quintile -0.033 -0.034 N.S. Third quintile N.S. N.S. N.S. Fourth quintile N.S. N.S. N.S. Fifth quintile -0.087 -0.090 -0.077

Observations 6861 6861 6861 Source: Authors using the 2005 Nicaragua LSMS. N.S.: not significant. Presented coefficients are significant with at a 5 percent confidence level. Omitted variables: households having a head with no education and households living in Managua.

Poor households and especially those engaged in agriculture are more likely to belong to local committees and professional associations. Table 10.13 presents household participation rates in Nicaragua by type of association. Results indicate that nationally about 4.4 percent of all households participate in local committees, 2 percent in professional associations, 2 percent in credit unions, 8 percent in religious associations, and about 6 percent in other type organizations (such as women organizations, clubs, etc…). Participation in productive associations (such as local committees and professional associations) is higher in rural areas and especially in the central region, among agricultural producers, and among households in the bottom 3 quintiles. Household participation in credit unions, on the contrary, is higher in Managua, in urban areas, and among households in the upper consumption quintiles. Households headed by a male head are more likely to participate in productive associations than households headed by a female head. Participation in religious associations is higher than average in Managua, among indigenous households, and among households engaged in agriculture.

More educated households as well as households having self-employed heads display a higher probability to belong to productive associations. Using regression analysis we estimate the likelihood of households to participate in productive associations conditional on their characteristics (see Table A2 in the annex). Results indicate that more educated households are more likely to participate in productive associations. In particular, households having a head who attained primary, secondary, and tertiary education are 4, 7, and 16 percent more likely to participate in a productive association than households having a head with no education. Households having a self employed head as well as those having a head working in agriculture are 2 to 4 percent more likely to participate in productive associations than households having a head working as wage earners or in non-agriculture-related activities. Households living in the central region are associated with a 4 percent higher probability of participating in productive

305

associations (vs. those residing in Managua) while urban households are 5 percent less likely to participate in productive associations as compared to rural ones.

Table 10.13: Household participation in Associations in Nicaragua

Household Participation in %

Local Committees

%

Professional Associations

%

Credit Unions

%

Religious Associations

%

Other organizations

% Total 4.47 2.30 1.76 7.57 6.35 Area

Rural 6.91 3.07 1.19 6.98 7.71 Urban 2.72 1.75 2.16 8.00 5.38

Region Managua 2.80 1.83 2.95 11.07 8.25 Pacific 4.21 1.85 1.38 4.43 4.20 Central 5.90 3.33 1.48 7.04 6.81 Atlantic 4.89 1.75 0.88 9.09 6.32

Quintile Quintile 1 4.68 2.29 0.13 5.71 4.80 Quintile 3 5.35 1.77 1.58 8.29 7.02 Quintile 5 4.37 3.16 2.40 8.60 7.06

Poverty Non poor 4.68 2.29 0.13 5.71 4.80 Poor 5.35 1.77 1.58 8.29 7.02

Employment Sector of head Non-Agricultural Producer 2.94 1.26 1.90 7.48 5.31 Agricultural Producer 7.58 4.43 1.47 7.76 8.47

Ethnicity of head Non-Indigenous 4.38 2.32 1.83 7.30 6.32 indigenous 6.42 1.92 0.22 13.66 7.04

Gender of head Male Head 4.89 2.56 1.88 7.75 6.55 Female Head 3.55 1.74 1.48 7.18 5.92


Titling

The collateral value of landholdings is generally assumed to increase with ownership rights, thereby improving credit access among landholders. The large percentage of untitled property in much of the developing world is a frequently cited contributing factor (Holden, 1997) for low access to credit. Land is considered as advantageous collateral because it cannot be removed and does not easily devalue. However, many land owner borrowers face credit barriers for lack of formally documented ownership rights (Field and Torero, 2006). Consistent with this notion, government land-titling programs are thought to be important to promote access to credit among the poor. Indeed, wide scale land-titling has become a popular policy prescription for reducing credit constraints in developing countries (Binswanger et al.,1999). Nevertheless, property titles are not necessarily sufficient to transform modest landholdings into viable collateral for commercial loans. Use of titles to securitize loans may fail in impoverished settings because transaction costs involved – such as those associated with collateral processing, foreclosure and resale – are sizable compared with the average loan sought. Such costs are even higher when political or legal factors impede repossession of property (Deninger and Freder, 1993).

306

Table 10.14: Descriptive statistics on house titling by socio-economic group

% of households who own a house

% who own without documentation

% who own with documentation

Total 76.5 34.0 66.0 Rural 77.4 44.4 55.6 Urban 75.8 26.5 73.5

Region Managua 75.4 34.0 66.0 Pacific 73.0 33.1 66.9 Central 78.3 29.8 70.2 Atlantic 82.5 46.3 53.7

Poverty Non poor 75.8 27.6 72.4 Poor 77.6 44.7 55.3

Employment Sector of head Non-Agricultural Producer 72.7 29.3 70.7 Agricultural Producer 84.3 42.4 57.6

Ethnicity of head Non indigenous 76.0 32.8 67.2 Indigenous 86.4 58.5 41.5


More than one third of all homeowners do not have a property title in Nicaragua. Titling can be important intangible asset because it allows you access to credit and thus open the doors to productive investments. Furthermore, titling can have other benefits such as allowing owners to rent out their land for a profit. The 2005 LSMS, reports that though 77 percent of all households in Nicaragua claim to own the house they live in, only 66 percent possess a property titles on their property. Absence of property titles is more common in rural areas and in the Central and Managua regions, where informal home ownership reaches 34 to 46 percent. Furthermore, even in urban areas and among households in the richest quintiles, informal house ownership is as high as 30 percent. Informal ownership is the highest among indigenous households (at 59 percent) and among households working in agriculture (about 43 percent). Despite these high rates, less than 1 percent of households overall are benefiting from titling programs in Nicaragua (LSMS 2005). Titling programs are more common in urban areas and especially Managua where about 2.5 percent of all households claim to have benefited from a titling program within a year prior to the survey.

Lack of titling is also common among agricultural producers who claim to own land. Data suggest that 21 percent of all agricultural producers do not possess a title on their land. This fact is more frequent in the Atlantic region where about 32 percent of all producers who claim to own their land do not have any official documentation. Interestingly, only 10 percent of all landowners who do not have a title on their land fear that they may have problems with their land in the future or that they may be expropriated from it, although this holds true for all socio-economic groups. There is also anecdotal evidence suggesting that households lack incentives to register their property as without a title they avoid paying property taxes and using it as credit collateral. Most landowners claim to have obtained their land by either purchasing it or by inheriting it (Table 10.15). About 11 percent of poor households in rural and 21 percent of households in urban areas claim to have obtained their land during the agricultural reform and about 6 percent of all landowners claim to live in a land they invaded. Not surprisingly, landowners who claim to live in invaded lands are generally poor ones.

307

Figure 10.18. About one third of all agricultural producers in the Atlantic region do not have a property title on their land.

21.3 21.9

16.5

20.6

17.8 17.0

32.331.0

23.622.0

9.9

6.1

0

5

10

15

20

25

30

35

Tota

l

Rural

Urban

Man

agua

Pacific

o

Centra

l

Atlant

ico

Pooes

t Quin

tile Q2 Q3 Q4

Riches

t Quin

tile% H

ouse

hold

s w

ho o

we

land w

ithout docu

men

tation


Table 10.15. About 20 percent of the urban poor claim to have obtained their land during the “reforma agraria”

Purchased Inherited Agric. Reform Invaded Given as gift Total Rural Areas

Non-poor 50.9 36.8 7.2 1.9 3.1 100.0 poor 41.1 38.5 10.8 5.9 3.8 100.0

Urban Areas Non-poor 56.7 37.0 4.6 0.0 1.7 100.0

poor 33.8 25.4 20.5 6.5 13.7 100.0 Source: Authors using the 2005 Nicaragua LSMS

Generally, having a land title is associated with better outcomes in relation to access to credit and productivity as well as with a higher probability that households rent their land for profit. In situations where land tenure insecurity is pervasive, as in Nicaragua, systematic efforts of land regularization can have positive effects on land values as well as equity. Receipt of a registered title raises land values by 30 percent and greatly increases the propensity to invest. Greater demand for regularization of land rights, especially from the poor, suggests that titling can have a positive distributional effect. Having a land title is a necessary but not sufficient condition to transform modest landholdings into viable collateral for commercial loans. Titles are as important as a well developed market for land and property in general for financial institutions to forsee associated gains above the costs involved in collateral processing, such as foreclosure and resale of land properties, and to be able to legally repossess without political impediments. In addition to efficient land markets and credit systems, titled land needs to be complemented by training, technical assistance and improved market access for increases in productivity and profits to take place.

308

Table 10.16: Share of landowners who rented their land for profit during the 12 months prior to the survey

Type of Title Landowner did not rent land during last 12

months

Landowner rented land during last 12 months

Total

Conventional title 69.9 30.1 100 Title from agriculture reform 43.6 56.4 100 Purchase letter 60.5 39.5 100 Other document 54.0 46.0 100 No document 75.8 24.2 100 Total 67.7 32.3 100 Source: Authors using the 2005 Nicaragua LSMS

AGRICULTURAL PRODUCTIVITY

Productivity gaps

This section is devoted to study equity in the distribution of production factors (capital, labor, and land), inputs, and intangible assets among agricultural producers in Nicaragua. In this section agricultural producers are defined as those households who work the land (whether owned or not) and obtained any type of yield from such a purpose (i.e. the analysis includes only those households who were able to attach a monetary value to the produce of their agricultural activities). Since not all agricultural-producer households own land, each household in the sample was attached to an UPA (Unidad Primaria Agricola) from which we obtain all necessary information about agricultural production. To facilitate the analysis, we restricted our sample to only one UPA per household (the one with the largest yields).174 Land owners who do not work the land or who rent their land out, as well as landowners engaged in “pecuary” activities (i.e. livestock, eggs production, milk production, leather production, etc.), were excluded from the analysis. According to survey estimates, about 31 percent of all households in the sample fulfill this description. As Figure 10.19 illustrates, households engaged in agricultural production are a vulnerable group of the Nicaraguan population, associated with higher poverty and lower education levels than average.

Figure 10.19: Agricultural producer households are poorer and less-educated than average.

46.1

14.8

44.2

75.4

70.6

28.0

88.9

93.4

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

% Poor % Extreme Poor % Living in ruralareas

% Heads withprimary education or

below

Per

cent

All HouseholdsAgricultural Producer household


174 Only 3 percent of all household (113 observations) declared to own more than one UPA in the year previous to the survey.

309

On average, the segment of the population dedicated to for-profit agricultural production is poorer and less educated than average. On average, the segment of the population dedicated to for-profit agricultural production is poorer and less educated than average. The majority of agricultural producers lives in rural areas and in the poorest regions. A large share of the poor population (66 and 46 percent in the first and second quintiles) is engaged in agricultural production. Moreover, income from agriculture accounts for more than two thirds of the overall income among the poor. A significant share of all agricultural income – 73 to 92 percent – comes from revenues from agricultural production (the remaining comes from wages and services). Table 10.17 also presents statistics on production size by quintile. Not surprisingly, producers in the highest quintiles are more likely to be large (in terms of the yields produced per hectare) while producers in the bottom quintiles are more likely to be small.

Table 10.17. Descriptive statistics for agricultural producers.

% households

who are agric.

Producers

% households

who are small

producers

% households

who are medium

producers

% households

who are large

producers

Yields as % of Ag.

Income

Ag. Income as % of total

income

Poorest Quintile 65.81 12.36 83.2 4.44 91.37 68.34 Q2 45.98 7.34 84.21 8.45 72.93 60.70 Q3 33.56 10.17 79.13 10.71 74.79 53.33 Q4 20.30 11.69 72.97 15.34 75.50 48.76 Richest Quintile 11.23 9.41 68.25 22.34 78.07 43.48 Source: Authors using the 2005 Nicaragua LSMS.

Figure 10.20. Share of agricultural producers by region and strata.

4.8%

24.6%

49.5%

55.8%

66.3%

6.2%

0%

10%

20%

30%

40%

50%

60%

70%

managua pacifico central atlantico rural urban

Hou

seho

lds

who

are

Agr

icul

tuta

l Pro

duce

rs


Agricultural productivity is important for economic growth. One stylized fact in the literature of development is the link between agricultural productivity, growth, and poverty. There are several channels through which agricultural productivity (and growth in the agricultural sector) can affect poverty: higher agricultural productivity can translate into higher income for producers, more employment, production of cheaper food, and higher tax revenue from agricultural activity, among others. Various studies show high multiplier effects of improvement in agricultural productivity, especially in

310

areas with good infrastructure well developed urban-rural communications and access.175 Table 10.18 presents some interesting studies in relation to this topic. Results in these papers indicate, among others, that i) growth in the agricultural sector helps poverty alleviation; ii) although growth of agricultural output help reduce poverty, this reduction may be small in absolute magnitude when compared to the effects of growth in the non-agricultural sector. However, growth in the agricultural sector is likely to help and reach the poorest and more vulnerable segments of the population.

Table 10.18. Links between Agricultural Growth and Poverty Alleviation Author Results Coxhead and Warr (1991)

Using a computable general equilibrium model, the author show that in a small open economy, technical improvements in farming are likely to benefit labor. Estimates from this model predicted almost an 8% increase in the income of landless laborers for a 10% improvement in agricultural output applicable to irrigated areas (with technology that saves land but uses labor).

Franco and Godoy (1993)

Using a computable general equilibrium model especially calibrated for Bolivia, the authors show that technical improvements to crops generate benefits in economic growth and employment for the whole economy.

Datt and Ravallion (1996)

Using a decomposition of growth by sectors in India, the authors demonstrate that agricultural productivity growth benefits the poor. The authors also show that agricultural growth is pro-poor and land yields are inversely related to different poverty measures. They estimated elasticities of poverty with respect to agricultural growth on the order of -1.2 to -1.9 nationally and -0.4 to -0.5 in urban areas.

Throrbecke and Jung (1996)

This paper finds that poverty is largely reduced by growth in agricultural sector.

Gallup, Radelet and Warner (1997)

In this paper the authors found that one-percent growth in agricultural GDP per capita leads to 1.61 percent increase in per capita income of individuals in the bottom quintile. An equivalent increased in industrial GDP increases the income of individuals the bottom quintile by only 1.16 percent.

Wodon (1999) Using simulations, the author finds that a pro-poor (agriculture-oriented) development strategy would probably bring down the poverty headcount in Bangladesh by 3 point in 2008, ceteris paribus.

Hazell and Haddad (2000)

The authors reported Rangarajan (1982) results where he estimated that a 1 percent addition to agricultural growth rate in India would generate a 0.5 percent addition to growth rate of agricultural output and a 0.7 percent addition to the growth rate of national income.

Irz et al (2001) The authors report, based on econometric evidence found in their paper, that the percentage of national population that earn less than $1 a day tends to decline with improvements in both agricultural value-added per worker and land yields.

Bravo-Ortega and Lederman (2005)

Using GMM IV system estimation to deal with the endogeneity of the sectoral GDPs, the authors find that for developing countries a 1 percent increase in a country’s agricultural output per worker is associated with an increase of 0.36 percent in the average income of the poorest households. But agricultural output per worker has a smaller effect than non-agricultural output per worker on the average income of the first quintile. Estimates imply that a 1 percent rise in this sector is associated with a 0.64 percent increase in the average household income of the poor. The Latin America and Caribbean estimates are 0.19 and 0.77 for the agricultural and non-agricultural sector respectively.

Source: Authors own elaboration.

175 Haggblade et al (1989), Haggblade et al (1991), Hazell and Ramasany (1991) and Delgado et al (1994) among others.

311

Gaps in Productivity are large, especially by producer size and region. Figure 10.21 displays differences in productivity (yields per hectare in local currency) across regions, producer size, and socio-economic condition. Large Agricultural Producers display productivity levels that are more than six times larger than among small producers. Not surprisingly, urban producers, often having better access to infrastructure, technology, and credit; are more productive than rural producers (even poor urban producers are on average more productive that non-poor rural producers). There are also large inequities in productivity by region. The Atlantic region, being the poorest region, displays the lowest levels of agricultural productivity while the Central and Pacific regions display productivity levels that are higher than the national average. Small-rural producers in the Atlantic region are likely to be one of the most vulnerable groups in Nicaragua: they display higher levels of poverty, low levels of education, low productivity, and limited access to infrastructure, equipment, and qualified labor. 176

Figure 10.21. Economies of scale are important in Nicaragua: Large agricultural producers are 6 times more productive than small ones.

0100020003000400050006000700080009000

10000

Small

Pro

duce

rs

Mediu

m P

rodu

cers

Lar

ge P

rodu

cers

Atla

ntic

Mana

gua

Pac

ific

Cen

tral

Tota

l

R

ural

Poor

Rur

al Non

-Poo

r

Urb

an P

oor

U

rban

Non

-Poo

r

Yie

lds

per

Hct

r

Source: Authors using the 2005 Nicaragua LSMS. Small Producer, yields of less 2000 CO/year; medium Producer; yields between 2000 and 30,000 CO/year; large Producer yields above 30,000 CO/year

Large producers, associated with higher average levels of productivity, are not a homogeneous group: they display very high levels of “within” inequality in productivity. Data indicate that there are high levels of inequality within groups, especially among producers of large size and within regions. In particular, as illustrated Lorenz Curves (see Figure 10.23), inequality in yields per hectare is highest among large producers, among non-poor producers, and among producers in Managua and in the Atlantic region. As such, large producers as well as non-poor producers should not be considered as homogenous groups necessarily. Interestingly, the Atlantic region is not only the less productive region on average but also the one with the highest inequality in productivity.

176 Rural productivity is usually computed product-by-product in tons (or kilograms) per hectare. Due to data limitations (small sample sizes and lack of information about land size cultivated by product) doing product-by-product analysis is not feasible using the 2005 LSMS data. Analysis in this chapter relies on information on the monetary value in local currently of the output produced from all crops by producer households (which we define as yields) divided by the land used for the production of all crops (in hectares). The main limitation of this “aggregate” analysis the impossibility to isolate differences in yields that are due to the type of crop produced (such as harvest time, water needs, seasonal demand, market prices, etc…).

312

Figure 10.22. Atlantic region is not only the less productive region on average but also the one with the highest inequality in productivity.

0

.2.4

.6.8

1Lo

renz

Cu

rve

by

Are

a

0 .2 .4 .6 .8 1Cumulative population proportion/x

Rural Urban

Equality

0.2

.4.6

.81

Lore

nz C

urv

e b

y R

egio

n


Managua Pacific

Central Atlantic

Equality

0.2

.4.6

.81

Lore

nz C

urv

e b

y P

oor


Non-Poor Poor

Equality

0.2

.4.6

.81

Lore

nz C

urv

e b

y P

rodu

cers

Siz

e


Small Producers Medium Producers

Large Producers

Inequities in Yields Per Hectare by Groups


Factors of production

Among producers, there are large inequities in access to Land and Capital. Capital, labor, and land are the main (and classical) factors of production used for agricultural production. Inequality in productivity is generally explained by inequities is quantity and quality of the aforementioned factors of production. While the 2005 LSMS provides good information on the quantity of the factors of production used by producers, information on quality is limited. Therefore, results presented here are likely to be conservative ones as inequities in factors of production are likely to be even higher once quality analysis in introduced. To proxi factors of production we used the following variables: for labor, we used the number of workers (including family members) working the land in the UPA; for capital, we used the monetary subjective value in local currency of the agricultural equipment used by the producer (including production animals, tractors, harvest and seeding equipment, water pumps and sprinklers, vehicles, fumigating equipment, and electric plants). Capital and Labor were normalized by the amount of land used for production in hectares (i.e. capital per hectare and labor per hectare). Figure 10.23 plots the Lorenz curves for all factors of production in Nicaragua (for all producers). Results indicate that capital and land are less equally distributed among producers than labor and other agricultural inputs (such as seeds, fertilizers, etc; see next section). 177

177 We construct an index for agricultural supplies using factor analysis (i.e. principal component techniques) based on whether or not producers use fertilizers, seeds, improved seeds, and plaguicidas)

313

Figure 10.23. Inequality in access to Capital and Land is much higher than inequality in access to Labor and Agricultural Supplies.

0.2

.4.6

.81

Lore

nz

Cu

rve

0 . 2 .4 .6 . 8 1

C a p ita l p er he c ta r e L a b o r p er he c ta r e

C o n su m p t io n p e r ca p i ta E qu a li ty

0.2

.4.6

.81

Lor

enz

Cu

rve

0 . 2 . 4 .6 .8 1

L an d E x p e n di t u re i n I n p u ts p e r h e cta re

C o n su m p t io n p e r ca p i t a E q u a l i t y

I n e q u itie s in C a p it a l - L a b o r - L an d - E x p e n d itu re in I n p u ts a n d Co n s u m pt io n in N ic a ra gu a


While poor and small producers use more labor and less capital and land for production, large and non-poor producers do otherwise. Table 10.19 presents some descriptive statistics on access to factors of production in Nicaragua by socio-economic group, strata, production size, and region. Results in Table 10.18 indicate that the size of the UPA is generally larger among non-poor producers, especially in urban areas. Differences in land size between poor and non-poor producers in much larger in urban areas than in rural areas: in urban (rural) areas non-poor producers have on average 7 (2) times as much land as poor producers. The average size of the UPA is also higher in the Managua and Atlantic regions (52 and 37 hectares respectively) than in the Pacific and Central regions (19 and 13 hectares respectively). Not surprisingly, large producers have land plots that are about 3 to 4 times larger than those of small and medium-size producers. Capital per hectare is significantly higher among urban non-poor producers. Note that this group has about 3 times more capital than any other group. Differences in capital by area and producer size are less notorious (which suggest a high concentration of the capital stock among urban non-poor producers). As predicted by economic theory (based on a Cobb-Douglas production function), while poor and small producers generally have less access to land and capital, they employ more labor (generally unskilled) in order to conduct activities that other producers undertake using equipment (such land irrigation, seeding, and harvesting). This occurs because labor is relatively cheaper for small and poor producers (as they rely mostly own family/unpaid labor).

314

Table 10.19. Descriptive Statistics on factors of production and output for agricultural producers in Nicaragua

Yields in Cordobas per hectare

Land Size in hectares Capital in Cordobas per hectare

Labor in Workers per hectare

Strata Urban Poor 5753 7.07 1625.56 1.80 Urban Non-Poor 6493 67.27 3461.79 1.62 Rural Poor 4584 17.09 729.05 1.44 Rural Non-Poor 4910 33.45 1167.95 1.28 Region Managua 4226 52.52 1590.40 2.28 Pacific 5484 12.91 1486.49 1.87 Central 5564 18.90 1063.20 1.55 Atlantic 2867 37.66 491.32 0.70 Producer Size Small Producers 1418 14.31 627.53 2.43 Medium Producers 4592 17.77 1074.01 1.43 Large Producers 9428 56.70 1112.04 0.64 Total 4833 23.90 1035.26 1.43 Source: Authors using the 2005 Nicaragua LSMS. Small Producer, yields of less 2000 CO/year; medium Producer; yields between 2000 and 30,000 CO/year; large Producer yields above 30,000 CO/year

Box 10.5. Relationship between fertilizers and agricultural productivity

Fertilizers are one of the most enhancing productivity agricultural inputs in the agricultural economy. Hayami et al (1970) found that fertilizers have an economically and statistically significant influence in agricultural productivity when studying a cross section of countries in the 1960s. Use of fertilizers increases productivity between 10 and 17 percent regarding the specification used or the way of measuring productivity (per firm basis, national aggregates, per capital basis, etc). Headley (1968) revises the impact of pesticides on productivity and report that the marginal value product of pesticides exceeds their marginal cost by a considerable amount. The author also finds a positive impact of fertilizers on productivity. Carrasco and Moffitt (1992) present estimates of pesticide productivity derived from alternative functional specifications than the usual Cobb-Douglas. They found that even though the magnitude of pesticide productivity could be smaller to the one find by Headley, their estimates are still indicating a positive and significant influence of pesticides and fertilizers in productivity. Antle (1983) also finds fertilizers to be an essential element to disentangle the productivity determinants when studying the links between infrastructure and agricultural productivity. Relevance of pesticides and fertilizers has become a styled fact nowadays in developing economies.

The use of agricultural inputs is important to boost agricultural productivity and to mitigate the risk of a bad harvest. Agricultural inputs enhance the quality, growth, and development of crops (mainly seeds, fertilizers, and plaguicidas). Agricultural inputs also mitigate the risk of having a bad harvest and enhance the quantity and quality of productive output. Usage of inputs in Nicaragua is generally low: only 11 percent of all producers use certified seeds, 6 percent use organic fertilizers, 37 percent use chemical fertilizers, and 67 percent use “plaguicidas”. Large and non-poor producers as well as producers in the Pacific region use more inputs that poor, small, and rural producers. Producers in the Atlantic region largely fall behind in their usage of inputs for production as compared to the average producer.

315

Table 10.20. The share of agricultural producers using certified seeds and fertilizers is generally low, especially among poor and small producers.

% producers using Non

certified Seeds

% producers using Certified

Seeds

% producers using organic

Fertilizers

% producers using chemical

Fertilizers

% producers using

Plaguicidas Area Urban Poor 25.51 8.01 2.46 45.04 63.70 Urban Non-Poor 18.95 14.05 12.23 52.75 71.66 Rural Poor 19.76 9.23 5.74 29.47 62.59 Rural Non-Poor 23.39 12.40 5.84 46.89 74.41 Region Managua 10.37 17.63 11.32 29.46 64.14 Pacific 20.02 12.53 12.77 61.42 61.95 Central 26.15 11.64 4.46 39.32 73.38 Atlantic 12.88 4.37 1.56 8.47 57.96 Producer Size Small 19.30 7.04 6.51 22.17 43.78 Medium 21.47 8.56 5.71 35.83 68.00 Large 20.20 29.03 8.02 63.04 81.79 Total 21.12 10.46 6.02 37.16 66.89


Figure 10.24. Large producers, non-poor producers, and producers in the Pacific region are more likely to use fertilizers and plaguicidas.

0

10

20

30

40

50

60

70

80

90

Urban

Poo

r

Urban

Non

-Poo

r

Rural

Poor

Rural

Non-P

oor

Manag

ua

Pacific

Centra

l

Atlanti

cSmall

Medium

Larg

e

Per

cent

% producers using chemical fertilizer% producers using organic fertilizer% users using plaguicidas


Expenditures

Expenditures on labor and agricultural inputs (mainly fertilizers) constitute the most important expenditure items among producers. Table 10.21 provides information on the expenditure composition

316

for agricultural producers by producer size and socio-economic group. Labor (mainly wages) constitutes the main expenditure item (66 to 68 percent of total expenditures) among small and medium producers and among rural non-poor producers. Large producers and urban producers spend relatively less on labor (from 32 to 45 percent of their total expenditure) but more on agricultural inputs (35 to 56 percent of their total expenditure). Fertilizers account for an important fraction of overall producer’s expenditures on agricultural inputs, especially for large and urban producers. Indeed, as suggested in Figure 26, fertilizers and certified seeds account for as much as 6 to 10 percent of overall producer’s income among large non-poor producers, and among producers in the Pacific region. Expenditures on utilities, while a relatively low expenditure components, are more important for small and poor producers (3.7 percent of their overall expenditures vs. 0.28 to 0.68 percent among large and non-poor producers). Expenditures on operations account for as much as 20 percent of overall expenditures for large and urban non-poor producers vs. 7 to 11 percent among small/medium and poor producers. The latest is not surprising because large and non-poor producers (generally more technified) spend on items such as equipment rentals, gasoline, and transpiration than small and poor producers do.

Table 10.21: Expenditure composition for agricultural producers

Producer Rural Urban Expenditures per capita

Small Medium Large Poor Non-poor Poor Non-poor

Operational

Transportation 1.34 1.49 3.31 2.59 1.78 2.04 3.46

Storage 0.19 0.00 2.38 0.01 2.21 0.00 0.57

Rent Equipment 0.08 0.79 3.55 0.89 0.57 0.59 5.66

Equipment maintenance 0.00 0.01 2.27 0.00 1.24 0.00 1.89

Rent Animals 1.20 1.05 0.15 1.54 0.46 1.91 0.13

Production 0.32 0.27 0.09 0.18 0.25 0.07 0.03

Gasoline Lubricants 1.51 0.33 4.46 0.25 2.37 0.04 4.21

Production Taxes 0.19 0.13 1.15 0.07 0.57 0.09 1.19

Land Taxes 2.19 1.64 1.01 0.98 1.15 0.71 1.57

Packaging 1.07 1.26 0.68 2.51 0.79 1.78 0.32

Others 0.32 0.32 0.10 0.34 0.25 0.59 0.01

Subtotal 8.42 7.28 19.16 9.38 11.63 7.82 19.04

Expenditures on Labor 68.27 66.17 45.31 35.15 70.85 32.40 33.72

Utilities

Water 1.17 0.24 0.09 0.20 0.11 1.67 0.22

Electricity 1.93 0.55 0.19 0.42 0.25 2.05 0.45

Subtotal 3.10 0.80 0.28 0.62 0.35 3.72 0.67

Agricultural Input

Non certified Seeds 1.94 1.07 0.21 1.68 0.42 4.18 0.21

Certified Seeds 0.79 0.55 6.54 1.62 0.70 2.22 9.86

Plants 0.77 0.07 0.01 0.03 0.07 0.49 0.00

Organic Fertilizer 1.15 0.46 0.30 0.91 0.38 0.43 0.12

Fertilizer 12.57 18.78 23.14 42.32 12.35 40.89 30.86

Plaguicidas 2.99 4.84 5.07 8.30 3.25 7.86 5.52

Subtotal 20.22 25.75 35.26 54.85 17.16 56.07 46.58

Total 100 100 100 100 100 100 100


317

Figure 10.25. Producers on average spend about 10 percent of their household income in Fertilizers and certified seeds.

0.0

2.0

4.0

6.0

8.0

10.0

12.0

Man

agua

Pacific

o

Centra

l

Atlantico

Small

Med

ium

Larg

ePoo

r

Non-p

oor

Poor

Non-p

oor

Per

cent

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Per

cent

Expensitures in Fertilizer as % if Income

Expensitures in Certified Seeds as % if Income


Determinants of productivity

This section relies on regression analysis using a specification where the natural logarithm of output (in yields per hectare) is regressed against proxies for Capital, Labor, and other characteristics of producer households that may influence their levels of productivity (such as access to infrastructure, use of fertilizers, region, level of education of the head, etc.). Our model is based on a Cobb-Douglas production function (see Box 10.6 below); log-log and log-linear regression models are used.

Box 10.6. Estimating a Cobb-Douglas production function.

The Cobb-Douglas production function is the most commonly estimated function in the economics literature. It has the following algebraic form:

χβα HKALQ =

Where Q is output, A, α , β and χ are constants, L is labor, K is capital, and H is land. The function is

said to be homogenous of degree χβα ++ , since multiplication of L, K and H by some constant will

raise output by a proportion χβα ++k . If the three exponents sum to unity, the Cobb-Douglas function is said to be homogeneous of degree one, exhibiting constant returns to scale.

It is worth to note that household survey data is less than desirable to estimate determinants of agricultural productivity (more detailed sector data with larger sample sizes, such as agricultural census, are preferable for this type of analysis) so that results presented here should be interpreted with care. Previous findings on determinants of agricultural output in Nicaragua (see Larson, 2004) conducted analysis based on household survey data from 1998. Among other findings, the author results suggest that agricultural

318

producers in Nicaragua face decreasing returns to scale, with durable capital equipment being an important part of agricultural families’ income determinants. The study finds that having access to technical assistance, product diversification, and social capital (i.e. participation in farmer’s organizations) were not statistically significant determinants of productivity. 178

Returns to labor are high, especially for poor producers. Results in Table 10.22 suggest that for every extra worker per hectare, yields per hectare increase between 50 and 70 percent. Not surprisingly, since returns to inputs are generally concave (i.e. utility of having more of the factor increases at a decreasing rate), returns top labor are higher among poor producers than among non-poor producers. Other characteristics of the household; such as the gender of the head, participating in productive organizations, socio-economic level, and having a land property title; appear not top be significant covariates affecting productivity.

Table 10.22: Determinants of Productivity in Nicaragua

Dependent variable: Logarithm of Cordobas per hectare, CO/Htr.)

OLS Q1 Q3 OLS Small and Medium

Prod.

Large Prod.

Cobb-Douglas Log-linear Factors of Production Logarithm of number of workers per hectare 0.79 0.85 0.72 - - - Number of workers per hectare - - - 0.70 0.55 0.51 Number of workers per hectare square - - - N/S N/S N/S Subjective value of land in CO/htr. N/S N/S 0.00 N/S N/S N/S Rent value of land (in CO/htr.) if rented 0.00 0.00 0.00 0.00 0.00 0.00 Dummy if land has irrigation N/S N/S N/S N/S N/S N/S Logarithm of Capital (in CO/htr.) 0.05 N/S 0.06 Capital (1000 CO/htr.) 0.10 0.07 - Capital per hectare square 0.00 - - Characteristics of the household Head of household is female N/S N/S N/S N/S N/S N/S Head participates in a Social Network N/S - - N/S N/S N/S Household is poor N/S - - N/S N/S N/S Household is extremely poor N/S - - - - - Title of the Land Household has Property Title N/S N/S N/S N/S N/S N/S Household obtained land during Ag. Reform 0.31 N/S 0.40 N/S N/S N/S Public registry Household has land signed in public registry N/S N/S N/S N/S N/S N/S Source: authors’ own calculations from the LSMS 2005. N/S: not significant. (Underlined) coefficients are significant at 10 percent (5 pecent). Reference categories: Type of title; carta de Compra Venta or No title; Education of the head; no education or preschool; Type of Land: Ondulada (Wavy); Geographic Location: Atlantic region.

Estimates suggest that for every 1000 cordobas invested in capital per hectare (about US$60 per hectare), productivity is expected to increase by 7 to 10 percent. This result does not necessarily mean that returns to capital are lower than returns to labor; it just means that investments in Capital are very low (indeed, data suggest the average capital stock among producers is about US$22 per hectare). Investing US$60 per hectare is not much; given that the average UPA size is 16 hectares, it accounts for about US$960 per UPA per year. With this magnitude of investment, producers are not able to buy the

178 Larson et al (2007) find that while among wealthier households differences in productivity cannot be accounted for by differences in livelihood strategies, this is not the case for the poor ones. In the case of the poor households, they find that non-farm activities are more profitable than self-employment, but various labor market constrains prevent them to chose freely.

319

necessary equipment (such as tractors or irrigation systems) to really boost their production. If investments were to increase ten times (averaging US$600 per hectare or US$9600 per UPA per year; which would allow producers to buy a tractor, for example) productivity would increase by almost 100 percent. 179

Use of Fertilizers is an important determinant of agricultural productivity. Estimates in Table 10.23 suggest that using fertilizers increases productivity levels by 22 to 34 percent nationally. The use of fertilizers increases productivity by more (23 to 50 percent) among small and poor producers.

Table 10.23 cont.: Determinants of Productivity in Nicaragua

(Dependent variable: Logarithm of Co/htr) OLS Q1 Q3 OLS Small and Medium

Prod.

Large Prod.

Inputs and Access to Infrastructure Producer uses fertilizer 0.22 0.23 0.13 0.34 0.50 N/S Nearest road is paved 0.17 0.22 N/S N/S N/S N/S Household has access to electricity N/S N/S N/S N/S Distance to nearest water source in Km N/S 0.16 N/S N/S N/S N/S Main Crop Produced in Land Corn N/S N/S N/S N/S N/S N/S Beans 0.48 0.57 0.47 N/S 0.63 N/S Banana N/S N/S 0.62 N/S N/S N/S Coffee - - N/S N/S 0.80 N/S Agricultural Producers Size Medium 1.92 2.15 1.61 2.17 - - Large 3.02 3.27 2.66 3.44 - - Type of Land Flat Land N/S - - 0.37 0.36 N/S Quebrada N/S - - N/S N/S N/S Pendiente o Ladera N/S - - 0.36 0.38 N/S Geographic Location Managua N/S N/S N/S N/S N/S N/S Pacific N/S N/S 0.22 N/S N/S N/S Central N/S N/S 0.24 0.45 N/S 1.07 Urban Dummy N/S N/S 0.53 N/S N/S N/S Observations 976 976 976 976 847 132 R - squared 0.76 - - 0.60 0.48 0.59 Source: authors’ own calculations from the LSMS 2005. N/S: not significant. (Underlined) coefficients are significant at 10 percent (5 percent). Reference categories: Type of title; carta de compra venta or no title; Education of the head; no education or preschool; type of land: Ondulada (Wavy); Geographic Location: Atlantic region. Small Producer, yields of less 2000 CO/year; medium Producer; yields between 2000 and 30,000 CO/year; large Producer yields above 30,000 CO/year

Having access to a paved road increases average yields per hectare by 17 to 20 percent. This may occur because, as mentioned before, producers with access to better roads generally have more access to markets and are able to sell their produce at higher prices. Medium coffee and beans producers are associated with 60 to 80 percent higher productivity than other (non-bean, non-banana) producers (see Table 10.23). Consistent with the descriptive analysis presented above, estimates indicate that medium and large producers are 2 and 3 times more productive than small producers. Richer and larger producers

179 Note also that the returns to capital presented here are likely to be underestimated because the measure of capital stock is probably undervalued by the informer reporting a subjective value of a limited stock of equipment; that is, the LSMS is not the most adequate instrument to estimate the value of the capital stock. Returns to capital found here are similar to those found in Ecuador using 2003 data (see World Bank 2004)

320

in the Pacific and Central region (the main agricultural regions in the country) are associated with 20 to 50 percent higher levels of productivity.

CONCLUSIONS

Economies with better and broader access to roads, electricity, transportation, credit, and telecommunications area associated with higher growth rates and lower income inequality and poverty. Although levels of infrastructure development in Nicaragua are still low relative to the LAC region, there has been significant progress in terms of access to basic infrastructure and productive services in the country since year 1998. The share of households with access to piped water, fixed telephone, cellular telephone, and trash collection services increased substantially between 1998 and 2005. Other housing related variables related to the household’s dwellings, such as the share of dwellings with access to toilet inside, good-quality floor, and good-quality walls also displayed a significant improvement both in rural and urban areas.

Investments in roads are likely to improve welfare among the poor: better roads are associated with higher yields per hectare among producers and better access to markets. Investments in transport reduce the cost of assembling intermediate inputs for production. This in turn, reduces production costs and thereby prices. Indigenous households and households engaged in agricultural production are the ones with the lowest access rates to paved roads. Lack of good roads makes it difficult for households engaged in agriculture to commercialize their farm production, which leaves them in the hands of intermediaries who often buy their produce at lower than market prices (due to high transportation costs).

Low access to electricity hinders welfare, especially for households working in agriculture and who deal with perishable products, such as milk and milk derivatives. Lack of access to electricity also lowers the capacity of households to run small businesses and may negatively influence their children’s education. There is an urgent need for investments on expanding electricity networks in rural areas, in the Atlantic region, and among vulnerable segments of the population. This will be difficult to achieve if prices of electricity remain stable and as long as the country continues to be highly dependent on oil as the main source of energy. Between 15 and 20 percent of all households with access to electricity do not pay for the service. High rates of energy theft and high oil prices force suppliers to cut the service periodically in order to save costs. The government (through the utility service companies) should invest in monitoring electricity theft and in diversifying its sources of power away from oil-based energy.

The government should consider strategies to continue to improve formal micro-credit programs. Results indicate that loans (in per capita per year) among the poor account for a large share of their yearly per-capita income. Among agricultural producer households, for instance, loans account for up to 20 percent of their per-capita income and are mainly used to operate and expand their business. However, the majority of all loans received by households are issued by informal sources of credit, such as merchants or family/friends. Loans from informal providers (often more accessible to the poor) are associated with much higher interest rates than loans from formal providers.

Access to networks and associations has become a coping mechanism for households to promote social participation, empowerment, and better to access markets and services. Participation in productive organizations (credit union, professional association, or local committee) increases the probability that households benefit from social programs by 15 to 16 percent. Also, estimates indicate that having a head who participates in a productive organization increases the probability that a household receives a loan by 9 percent. Investments/programs destined to strengthen (or promote) these type of associations should be considered.

321

The government needs to educate the population on the importance of having property titles and should invest more in titling programs. About one third of all homeowners in Nicaragua do not possess a property titles on their property. Absence of property titles is more common in rural areas and in the Central and Managua regions. Lack of titling is also common among agricultural producers who claim to own land. 21 percent of all agricultural producers do not possess a title on their land. Generally, having a land title is associated with better outcomes in relation to access to credit and productivity as well as with a higher probability that households rent their land for profit. In situations where land tenure insecurity is pervasive, as in Nicaragua, systematic efforts of land regularization can have positive effects on land values as well as equity. Receipt of a registered title raises land values by 30 percent and greatly increases the propensity to invest. Greater demand for regularization of land rights, especially from the poor, suggests that titling can have a positive distributional effect. Having a land title is a necessary but not sufficient condition to transform modest landholdings into viable collateral for commercial loans. Titles are as important as a well developed market for land and property in general for financial institutions to foresee associated gains above the costs involved in collateral processing, such as foreclosure and resale of land properties, and to be able to legally repossess without political impediments. In addition to efficient land markets and credit systems, titled land needs to be complemented by training, technical assistance and improved market access for increases in productivity and profits to take place

Agricultural productivity is important for economic growth. In Nicaragua agricultural production is still very rudimentary for the majority of the producers. As such, gaps in Productivity in Nicaragua are large, especially by producer size and region. Urban producers, often having better access to infrastructure, technology, and credit; are much more productive than rural producers (even poor urban producers are on average more productive that non-poor rural producers). This occurs because there are large inequities in access to land and capital, which are concentrated among a few large urban non-poor producers. Usage of inputs in Nicaragua is generally low: less than one third of all producers certified seeds, organic fertilizers, chemical fertilizers, and/or “plaguicidas”. The use of agricultural inputs is important to boost agricultural productivity and to mitigate the risk of a bad harvest. Programs should be designed to subsidize inputs or at least make them more accessible to the poor.

322

REFERENCES

Angel-Urdinola, D.; Mete, C. (2007). "Welfare Dynamics and Determinants in Tajikistan Based on Assets Analysis”. Mimeo. World Bank. Washington, D.C.

Antle, J. (1983). “Infrastructure and Aggregate Agricultural Productivity: International Evidence”, Economic Development and Cultural Change, vol. 31, No3, pp.609-619

Aschauer, D. A. (1989a). "Is Public Expenditure Productive," Journal of Monetary Economics, vol. 23: 177-200.

Aschauer, D. A. (1989b). "Does Public Capital Crowd Out Private Capital," Journal of Monetary Economics, vol. 24: 171-88.

Barham, Carter and Useche, (2004). “Towards a New Generation of Policies to Enhance Land Access for Growth and Poverty Reduction in Central America and Mexico” University of Wisconsin-Madison

Beck, T., Demirguc-Kunt, A. and Levine, R. (2004). “Finance, Inequality, and Poverty: Cross-Country Evidence.” NBER Working Paper 10979, National Bureau of Economic Research, Cambridge, MA.

Binswanger, H.P. and Deninger, K. (1999). “The Evolution of World Bank Land Policy: Principles, Experience, and Future Changes”. World Bank Research Observer, Vol. 14(2): 247-276

Bravo-Ortega, C. and Lederman, D. (2005). “Agriculture and National Welfare around the World: Causality and Heterogeneity since 1960”. Policy Research Working Paper 3499, World Bank, Washington, DC.

Calderón, C. and Chong, A. (2004). "Volume and Quality of Infrastructure and the Distribution of Income: An Empirical Investigation," Review of Income and Wealth, vol. 50(1), pages 87-106, 03.

Calderon, C. and Serven, L. (2004). “The Effects of Infrastructure Development on Growth and Income Distribution”, Washington DC., The World Bank, LACVP.

Canning, D., Fay, M. and Perotti, R. (1994). "Infrastructure and Growth," in M. Bsaldassarri, M. Paganaetto and E.S. Phelps, eds., International Differences in Growth Rates. New York: St. Martins Press.

Carrasco-Tauber, C. and Moffitt, L. (1992). “Damage Control Econometrics: Functional Specifications and Pesticide Productivity” American Journal of Agricultural Economics, vol. 74, No1, pp 156-162.

Cook, C., Duncan, T, Jitsuchon, S., Sharma, A. , Guobao, W. (2005). “Assessing the Impact of Transport and Energy Infrastructure on Poverty Reduction” Copublished by ADB, DFID, JBIC, and WB.

Coxhead, I. and Warr, P. G. (1991). “Technical Change, Land Quality and Income Distribution: A General equilibrium Analysis”, American Journal of Agricultural Economics, 73: 345-60

323

Datt, G. and Ravallion, M. (1996). “How Important to India’s Poor is the Sectoral Composition of Economic Growth?” World Bank Economic review, 10(1): 1-25

Datt, G. and Ravallion, M. (1998). “Farm Productivity and Rural Poverty in India”. Journal of Development Studies, 34(4): 62-85.

De Franco, M. and Godoy, R. (1993). “Potato-led Growth: The Macroeconomic Effects of Technological Innovations in Bolivian Agriculture”, Journal of Development Studies, vol. 29(3): 561-87.

De la Fuente, A. and Estache, A. (2004). “Infrastructure Productivity and Growth: A quick survey”, Washington DC.

Delgado, C., Hazell, P., Hopkins, J. and Kelly, V. (1994). “Promoting Intersectoral Growth Linkages in Rural Africa through Agricultural Technology and Policy Reform”. American Journal of Agricultural Economics, vol. 76:1166-71

Deninger, K. and Freder, G. (1993). “Land Policy in Developing Countries”. Rural Development Note No. 3. World Bank, Washington D.C.

Easterly, W. and Rebelo, S. (1993). "Fiscal Policy and Economic Growth: An Empirical

Escobar, A. (1995). “Encountering Development: The Making and Unmaking of the Third World”. Princeton, NY.: Princeton University Press

Estache, A., and Fay, M. (1995). “Regional Growth in Argentina and Brazil: Determinants and Policy Options.” Unpublished paper, World Bank, Washington, DC.

Field, E. and Torero, M. (2006). “Do property titles increase credit access among the urban poor? Evidence from a National Titling Program”. Mimeo, IFPRI.

Filmer, D. and Pritchett, L. (1998). Estimating Wealth Effects without Income or Expenditure Data. World Bank Policy Paper Research No. 1994. Washington, D.C.

Gallup, J., Radelet, S. and Warner, A. (1997). “Economic Growth and the Income of the Poor” CAER II Discussion Paper N36. Cambridge, MA: Harvard Institute for International Development.

Gannon, C., and Z. Liu. (1997). “Poverty and Transport.” Unpublished paper, World Bank, Washington, DC.

Greenwood, J., and Jovanovic, B. (1990). “Financial Development, Growth and the Distribution of Income.” Journal of Political Economy, vol. 98: 1076–1107.

Haggblade, S., Hammer, Jeffrey and Hazell, P. (1991). “Modeling Agricultural Growth Multipliers” American Journal of Agricultural Economics (May):361-74

Hammer, J. (1998). Health Outcomes across Wealth Groups in Brazil and India. Mimeo, DECRG, The World Bank.

Hayami, Y and Ruttan, W. (1970). “Agricultural Productivity Differences among Countries”. The American Economic Review, vol 60, No5, pp 895-911.

324

Hazell, P. and Haddad, L. (2000) “CGIAR Research and Poverty Reduction” Draft paper prepared for Technical Advisory Committee of the CGIAR. Washington, DC:IFPRI, September.

Headley, J. (1968). “Estimating the Productivity of Agricultural Pesticides” American Journal of Agricultural Economics, vol. 50(1): 13-23

Holden, P. (1997). “Collateral without Consequence: Some Causes and Effects of Financial Underdevelopment in Latin America”. The Financier, Vol. 4(1): 12-21

Irz, X., L. Lin, C. T. and Wiggins, S. (2001) “Agricultural Productivity Growth and Poverty Allevition” Development Policy review, vol. 19 (4): 449-66

Larson, D. Le’on, M. and Aguyayo, Y. (2004). “Are There Common Determinants of the Agricultural Geographies of Ecuador and Nicaragua?” Photocopy, World Bank, Washington, DC.

Larson, D. Le’on, M. and Aguyayo, Y.(2007). “On the role of self-employment and specialization in family farms and business: evidence from Nicaragua” Photocopy, World Bank, Washington, DC.

Lawley, D. and Maxwell, A. (1971). Factor Analysis as an Statistical Method. Edited by Butterworth & Co., London.

Levine, R., Loayza, N. and Beck, T. (2000). “Financial Intermediation and Growth: Causality and Causes.” Journal of Monetary Economics, vol. 46 (1): 31–77.

Li, H., Squire, L. and Zou, H. (1998). “Explaining International and Intertemporal Variations in Income Inequality.” Economic Journal, vol. 108 (446): 26–43.

Loayza, N., Fajnzylber, P. and Calderon, C. (2005). “Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations and Forecasts.” Unpublished paper, World Bank, Washington, DC.

Mansura, G. and Rao, V. (2004). “Community Based and Driven Development: A Critical Review”. World Bank Working Paper Series No. 3209. Washington D.C.

Prudhomme, R., (2004) “Infrastructure and Development”, Paper prepared for the ABCDE (Annual Bank Conference on Development Economics). Washington DC

Sahn, D. and Stifel, D (2000). Poverty Comparisons over Time and across Countries in Africa. World Development, vol. 28 (1): 2123-55

Sahn, D. and Stifel, D (2003). Exploring Alternative Measures of Wealth in the Absence of Expenditure Data. The Review of Income and Wealth, vol. 49 (4): 463-489

Sanchez-Robles, B. (1998). "Infrastructure Investment and Growth: Some Empirical Evidence,"

Scott, J. (1998). “Seeing Like a State: How Certain Schemes to Improve the Human Conditions to Have Failed”. Edited by New Heaven: Yale University Press.

Thorbecke, E. and Jung, H. S. (1996) “A Multiplier Decomposition Method to Analyze Poverty Alleviation”, Journal of Development Economics, vol. 48 (2): 279-300.

325

Wodon, Q. (1999). “Growth, Poverty and Inequality: A Regional Panel for Bangladesh”. Policy Research Working Paper 2072. World Bank, Washington, DC.

World Bank (2004). Ecuador Poverty Assessment. Report No. 27061-EC. World Bank, Washington D.C.

Yepes, T. (2004). “Infrastructure and Poverty Reduction in Rural El Salvador”. Mimeo. World Bank,

Washington, DC.

326

ANNEX I

Figure A1: Distribution of the Infrastructure Index for year 2005 [using factor analysis].

0.5

11.

52

Den

sity

0 1 2 3 4Per capita infra proxy using factor analysis

Source: Authors using the 2005 Nicaragua LSMS. Assets included in the analysis: number of rooms in dwelling, dummies for whether or not households have access to a paved road, electricity, fixed telephone, cellular phone, piped water, trash collection services, toilet and kitchen inside dwelling, good floor (concrete, tile, stone, or brick), and good walls(from cement, stone, or concrete).

327

Table A1: Determinants of having access to a formal credit All Urban Rural Dependent Variable: household received a credit from a formal creditor dF/dx dF/dx dF/dx Characteristics of the household

Natural logarithm of household size 0.105 0.117 0.093 Household head is a female -0.090 -0.097 -0.116 Household head is single N.S. N.S. N.S. Household is indigenous -0.165 -0.321 N.S. Household engaged in agriculture N.S. N.S. N.S.

Characteristics of the head Age of the head N.S. N.S. N.S. Squared of the age of the head N.S. N.S. N.S. Head is working N.S. N.S. -0.188 Head works as a blue collar worker N.S. N.S. N.S. Head is a firm owner 0.222 N.S. 0.326 Head is self employed 0.250 0.172 0.248 Head has attained primary school N.S. N.S. N.S. Head has attained secondary school N.S. N.S. N.S. Head has attained technical school N.S. N.S. N.S. Head has attained tertiary school 0.114 N.S. N.S.

Characteristics of the spouse Age of the spouse N.S. N.S. N.S. Squared of the age of the spouse N.S. N.S. N.S. Spouse is working 0.099 0.142 0.074

Collateral Household owns land and has a land title 0.112 N.S. 0.095 Household owns a house and has a title N.S. N.S. N.S.

Socio economic group Second quintile 0.135 0.431 N.S. Third quintile 0.125 0.388 N.S. Fourth quintile 0.231 0.491 0.173 Highest quintile 0.350 0.598 0.264

Region Pacific N.S. N.S. N.S. Central N.S. N.S. N.S. Atlantic -0.248 -0.224 -0.268

Urban dummy 0.067 - - Observations 1734 988 746 Source: authors’ own calculations from the LSMS 2005. N/S: not significant. (Underlined) coefficients are significant at 10 percent (5 percent). Reference categories: education of the head/spouse, no education or preschool; occupational status of the head, wage earner; socio-economic group, poorest quintile; geographic location, Managua.

328

Table A2: Determinants of being a member of a productive association [Local committees, Prof. Associations, and Credit Union]

Dependent Variable: a member of the household participates in a productive association

All Urban Rural

Socio-economic level of the household Household is indigenous N.S N.S. N.S. Second quintile N.S N.S. N.S. Third quintile 0.029 N.S. 0.049 Fourth quintile 0.038 0.046 0.032 Highest quintile 0.061 0.058 0.053

Characteristics of the head Head has attained primary school 0.04 0.04 0.044 Head has attained secondary school 0.065 0.055 0.087 Head has attained technical school 0.154 0.122 0.274 Head has attained tertiary school 0.144 0.137 N.S. Head is working N.S N.S. N.S. Head is self employed 0.025 N.S. 0.05 Head works in the agricultural sector 0.021 0.044 N.S.

Region Pacific 0.007 0.012 -0.041 Central 0.043 0.043 -0.007 Atlantic 0.01 0.035 -0.056

Urban dummy -0.054 Observations 6861 3456 3405

Source: authors’ own calculations from the LSMS 2005. N/S: not significant. (Underlined) coefficients are significant at 10 percent (5 percent). Reference categories: education of the head, no education or preschool; socio-economic group, poorest quintile; geographic location, Managua.