pro-poor growth. cross-country evidence: high correlation b/w growth & poverty dollar &...
TRANSCRIPT
Pro-Poor Growth
•Cross-country evidence: high correlation b/w growth & poverty
•Dollar & Kraay (2000): - Growth is good for the poor irrespective of the nature of growth.
•More microeconomic approach is required rather than cross-country regression method.
Three Scenarios
1: Trickle-down Growth
- Economic growth reduces poverty but increases inequality.
2: Pro-poor Growth
- Economic growth reduces poverty but the poor receive proportionally greater benefits. Growth reduces
poverty but inequality also reduces.
3: Immiserizing Growth (Bhagwati)
- Economic growth increases poverty.
-What is Pro-Poor growth ?
-How can we measure its degree ?
-Poverty reduction depends on:
(i) growth rate(ii) how the benefits of growth are distributed to the poor
-Pro-poor growth index takes into account the two.
Poverty Growth Curve
z
dxxfxzP0
)(),(
z
xzxzP ),( xzxzP loglog),(
L(p) : % share of income (exp) of the bottom p % of pop.
)( pL
: mean income of the society
: generalized Lorenz curve
Atkinson’s Theorem
0)( pL (for all p)
0(for all poverty line and entire class of poverty)
P -Given and
p
pL p )(
-Taking log and first difference in both sides:
)(loglog pLp
If 0)( pL
0)(log pL
0log p
0log
)(loglog pLp
ppg log)(
Lemma 1: If 0)( pg for all p, then poverty reduces
unambiguously.
)(log)(
)(loglog)(
pLpg
pLpg
Lemma 2:
)(0 pg
0 & g(p) < 0
Pro-poor
Trickle-down
Immiserizing
)( pg
Table 1: Poverty growth curve: Thailand
p 88-90 90-92 92-94 94-96 96-98 98-00 88-00
10 6.31 2.51 8.89 7.27 -2.55 -4.39 3.01
20 6.10 3.21 8.72 7.30 -2.46 -3.11 3.29
30 5.84 3.61 9.16 7.14 -2.20 -2.67 3.48
40 5.85 4.05 9.41 6.99 -2.14 -2.34 3.64
50 5.89 4.48 9.58 6.81 -2.13 -2.10 3.75
60 5.95 4.80 9.68 6.72 -2.14 -1.85 3.86
70 6.00 5.19 9.69 6.59 -2.07 -1.55 3.97
80 6.05 5.76 9.36 6.54 -1.96 -1.21 4.09
90 6.29 6.52 8.49 6.48 -1.63 -0.77 4.23
100 9.06 7.49 7.65 5.75 -1.00 -0.85 4.68
10 20 30 40 50 60 70 80 90 1005
6
7
8
9
10
Fig1:Poverty growth curve: 1988-1990
10 20 30 40 50 60 70 80 90 1002
3
4
5
6
7
8
Fig2:Poverty growth curve:1990-1992
10 20 30 40 50 60 70 80 90 1007.5
8
8.5
9
9.5
10
Fig3:Poverty growth curve:1992-1994
10 20 30 40 50 60 70 80 90 1005.5
6
6.5
7
7.5
Fig4:Poverty growth curve:1994-1996
10 20 30 40 50 60 70 80 90 100-3
-2.5
-2
-1.5
-1
-0.5
Fig5: Poverty growth curve:1996-1998
10 20 30 40 50 60 70 80 90 100-5
-4
-3
-2
-1
0
Fig6:Poverty growth curve:1998-2000
Table 2: Summary of international poverty growth curves
Positive growth Negative growth Total Pro-poor 84 11 95Not pro-poor 71 23 94Immiserizing 9 0 9Inconclusive 35 8 43Total 199 42 241
Ravallion’s growth incidence curve:How is it different from our poverty growth curve?
x(p) : per capita income (exp) at the pth percentile
)('log)(
)('loglog)(
pLpr
pLpr
))(ln()ln())(ln(
)()('
'
pLpx
pLpx
r(p): Growth rate of the per capita income at the pth percentile
r(p) > , then growth is pro-poor.
-Ravallion’s GIC is based on the first order dominance, whereas our PGC is based on the second order dominance.
-Since r(p) is the growth rate of income at the pth percentile, its estimation from the unit record data will be subject to more errors because the data are discrete.
-Our dominance requirement is based on the estimation of the growth rates of the mean income up to the pth percentile. Thus it is subject to less errors. Our approach only requires the decile shares and the mean income.
Poverty equivalent growth rate (PEGR)
z
dxxfxzP0
)(),(
H
dppxx
P
0
)(1
Growth elasticity of poverty (Kakwani, 1993)
=
H
dpprpxx
PdLn
0
)()(1
)(
where r(p) = dln(x(p)), the growth rate of income at pth %
Using )()( ' pLpx
,
/)(dLn = total poverty elasticity
= growth elasticity of poverty when inequality does not change
H
dppLdLnpxx
P
0
))('()(1
= inequality effect of
poverty reduction
Pro-Poor Index: Kakwani & Pernia (2000)
/
> 1: Pro-poor
0 < < 1: Trickle-down
< 1: Immiserizing
Poverty Equivalent Growth Rate is the growth rate that would result in the same level of poverty reduction as the actual growth rate if the growth process was not accompanied by any change in inequality.
*
* )1,(),( ffd
is the poverty equivalent growth rate (PEGR)
0).(
f
0),(
fand,
0
0 *
*
and 0*
: Pro-poor growth
: Trickle-down
: Immiserizing
* = dppx
x
P
dpprpxx
P
H
H
)(
)()(
0
0
- Foster-Greer-Thorbecke
* =
dppxz
pxz
dpprpxz
pxz
H
H
)())(
(
)()())(
(
0
1
0
1
for 1
- Poverty Gap Ratio
*1 =
dppx
dpprpx
H
H
)(
)()(
0
0
- Watts Measure (Ravallion)
*W =
H
dppgH0)(
1
Estimation of PEGR
(z,L(p))
= (Ln [ (z, 2 , L2(p)] – Ln[ (z, 1 , L1(p)])/ where is given by = Ln (2) – Ln (1) *=()/
ˆˆˆ
where is growth elasticity of poverty and should satisfy
ˆ = ˆ/)(,,(ln)(,,(ln)(,,(ln)(,,(ln2
121221112 pLzpLzpLzpLz
a n d
ˆ = ˆ/)(,,(ln)(,,(ln)(,,(ln)(,,(ln2
112221121 pLzpLzpLzpLz
-Based on Kakwani’s (2000) poverty decomposition:
will be always negative.
< 0: growth is pro-poor
> 0 & < 0: growth is trickle down > 0 & > 0: growth is immiserizing
Poverty Equivalent growth rate
Year Percentage Poverty Severity of
Actual Growth
Rate Of poor gap ratio poverty ratio
1990-91 9.6 10.7 10.4 10.0
1991-92 4.0 4.1 3.7 3.6
1992-93 4.8 5.8 6.6 6.8
1993-94 7.3 7.2 7.3 7.5
1994-95 8.2 9.7 9.5 8.9
1995-96 5.8 5.1 5.0 4.6
1996-97 1.8 9.0 8.3 9.6
1997-98 -7.6 -9.0 -10.0 -10.9
1998-99 9.8 9.6 10.5 11.5
-10
-5
0
5
10
15
1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99
Growth Rate
PEGR (Headcount)
Year Actual Poverty Equivalent Growth Rate
Growth rate
Headcount ratio
Poverty gap ratio
Severity of poverty ratio
1988-90 9.06 5.5 5.9 6.1
1990-92 7.49 4.3 3.4 3.0
1992-94 7.65 8.8 8.7 8.8
1994-96 5.75 7.4 7.2 7.2
1996-98 -1.00 -2.7 -2.5 -2.5
1998-00 -0.85 -2.3 -3.8 -4.4
1988-2000 4.68 3.6 3.3 3.1
-4
-2
0
2
4
6
8
10
1988-1990
1990-1992
1992-1994
1994-1996
1996-1998
1998-2000
1988-2000
Actual growth rate
PEGR (Headcount)
Where Do We Go From Here ?
-The analysis presented here is ex-post attempting to answerquestions such as:
-Is economic growth pro-poor? And if so, what is its degree?
-Other important questions are: -Why is growth pro-poor in one period and not in another period? -What are the factors that affect pro-poorness of growth? -Is migration from rural to urban areas pro-poor? -Has the pro-poorness of growth increased or decreased because of differential growth rates in rural and urban areas? -Does the improvement in the educational levels lead to more or less pro-poor growth? -Has the increase or decrease in inequality within urban and rural areas contributed to pro-poorness of growth? -Answer to these and many related questions can give us an insight on policies to achieve pro-poor growth .
-We can answer these and related questions by utilizing the poverty decomposition proposed in the paper.
-Son (2003) “A New Decomposition” Journal of Economic Inequality, Vol.1, pp 1-7
-What are the policies that make growth pro-poor?
-How can we measure whether or not a given government policy is pro-poor? How can we improve these policies so that we achieve a maximum reduction in poverty with fixed resources?
-Which of the government services pro-poor and which are not?
-Our future research will focus on these issues.