n ational t ransfer a ccounts 1 empirical models david canning harvard university

NNational ational TTransfer ransfer AAccountsccounts11

Empirical ModelsEmpirical Models

David CanningHarvard University


I. Effects of Fertility I. Effects of Fertility RedcutionRedcution

National Transfer Accounts3

Economic Stories for Fertility Economic Stories for Fertility ReductionReduction► Technological change has increased the Technological change has increased the

returns to education/ lowered cost of returns to education/ lowered cost of education. education.

► Falling child mortality has reduced the Falling child mortality has reduced the “insurance” demand for children and reduced “insurance” demand for children and reduced wasted educational costs.wasted educational costs.

► Falling contribution of children to household Falling contribution of children to household finances with urbanization.finances with urbanization.

► Rising value of time (higher wages and Rising value of time (higher wages and human capital)– children are time expensive.human capital)– children are time expensive.


Desired Fertility and Actual Desired Fertility and Actual FertilityFertility►Economic theory concentrates on Economic theory concentrates on

changes in desired fertility.changes in desired fertility.►Desired fertility may vary between Desired fertility may vary between

male and female partners. Household male and female partners. Household bargaining model versus unitary bargaining model versus unitary household.household.

►Actual Fertility may differ from desired Actual Fertility may differ from desired fertility – role of contraceptionfertility – role of contraception


Empirical Effects of Low FertilityEmpirical Effects of Low Fertility

►Fertility is chosen. May be a Fertility is chosen. May be a consequence of female work decisions consequence of female work decisions and desired education of childrenand desired education of children

►We want to find structural effect of We want to find structural effect of lower fertility. Evidence for this lower fertility. Evidence for this requires “exogenous” change in requires “exogenous” change in fertility.fertility.


Instruments for Fertility:Instruments for Fertility:Sex RatioSex Ratio►Sex ratio of previous births is random Sex ratio of previous births is random

and affects future fertility.and affects future fertility. Son preferenceSon preference Mixed sex preferenceMixed sex preference

►Problem – sex ratio may affect Problem – sex ratio may affect household income (dowries) and household income (dowries) and desired investment in kids giving direct desired investment in kids giving direct effect of child health and education effect of child health and education investments. investments.

►Randomness?Randomness?


Instruments for Fertility:Instruments for Fertility:TwinsTwins►Twins are random. Increase family Twins are random. Increase family

size more than planned.size more than planned.►Problem – twins are less healthy than Problem – twins are less healthy than

average due to sharing of mother’s average due to sharing of mother’s resources during fetal development. resources during fetal development. Timing/spacing of births may have Timing/spacing of births may have effects on resource availability.effects on resource availability.

►Randomness? Randomness?


Instruments for Fertility:Instruments for Fertility:Abortion LawsAbortion Laws►Abortion laws affect fertilityAbortion laws affect fertility

About 26% of pregnancies end in abortionAbout 26% of pregnancies end in abortion Large fertility effect – e.g. US state lawsLarge fertility effect – e.g. US state laws

►Abortion laws may be endogenousAbortion laws may be endogenous Control for country fixed effects, time Control for country fixed effects, time

trend and country x time trend effects.trend and country x time trend effects. Precise timing may be exogenous Precise timing may be exogenous Some reversals of trend to more liberal Some reversals of trend to more liberal

laws.laws.


DataData

Female Labor Market Participation (ILO 2007)Female Labor Market Participation (ILO 2007) By cohort (15-19, 20-25,…,60-64)By cohort (15-19, 20-25,…,60-64) 1950-20001950-2000 97 countries97 countries

Fertility (WDI 2006)Fertility (WDI 2006) Total fertility rateTotal fertility rate

Abortion Index (United Nations Population Division Abortion Index (United Nations Population Division 2002)2002) Abortion Health Index: physical and mental health of the Abortion Health Index: physical and mental health of the

mother, rape, fetal impairmentmother, rape, fetal impairment Abortion Availability Index: economic hardship, on requestAbortion Availability Index: economic hardship, on request


Empirical SpecificationEmpirical Specification

ijt jt jt jt f jt m jt j t ijtP TFR k urban fschool mschool

Estimated Equation

Pijt female labor force participation of age group i, country j, year t

kjt capital stock per working age person

urbanjt population living in urban area (%)

fschooljt average years of schooling for females >15 years of age

mschooljt average years of schooling for males >15 years of age


Female Labor Force Participation Female Labor Force Participation (25-29) (25-29)

OLS, FE 2SLS, FE 2SLS, FE, LP 2SLS, FE, LP

Fertility (TFR) -3.095 -16.961 -8.280 -5.502(0.643)*** (4.771)*** (3.549)** (2.663)**

Capital per working age person 0.453 0.730 0.422 0.328(0.041)*** (0.104)*** (0.090)*** (0.092)***

Urban population (% of total) -0.353 -0.613 -0.360 -0.316(0.085)*** (0.142)*** (0.100)*** (0.104)***

Average years of schooling, male -1.780 1.294 1.415 1.340(0.701)** (1.567) (0.864) (0.809)*

Average years of schooling, female 3.540 -4.458 -3.140 -2.515(0.906)*** (2.955) (1.852)* (1.441)*

Lagged Labor force participation 0.768 1.045(0.072)*** (0.255)***

Observations 771 771 771 695R-squared 0.63 0.26 0.74 0.76


First Stage RegressionsFirst Stage RegressionsOLS, FE 2SLS, FE 2SLS, FE, LP 2SLS, FE, LP

Abortion Health Index -0.091 -0.072 -0.008(0.027)*** (0.028)*** (0.031)

Abortion Availability Index -0.029 -0.039 -0.128(0.055) (0.057) (0.065)**

Capital per working age person 0.022 0.024 0.024(0.002)*** (0.002)*** (0.002)***

Urban population (% of total) -0.019 -0.021 -0.021(0.006)*** (0.006)*** (0.006)***

Average years of schooling, male 0.235 0.195 0.189(0.058)*** (0.057)*** (0.061)***

Average years of schooling, female -0.569 -0.504 -0.520(0.059)*** (0.058)*** (0.061)***

Lagged Labor force participation -0.014(0.004)***

Lagged Abortion Health Index -0.113(0.032)***

Lagged Abortion Availability Index 0.204(0.071)***

Hansen OID test p-value 0.65 0.91 0.86Cragg-Donald F-stat 9.52 6.81 2.31Observations 771 771 695R-squared 0.80 0.80 0.80


Age Group Specific Age Group Specific Fertility EffectsFertility Effects

-35.0

-30.0

-25.0

-20.0

-15.0

-10.0

-5.0

0.0

5.0

10.0

15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64

Female Cohort

Par

tici

pat

ion

Eff

ect


Total Dynamic Effect of Total Dynamic Effect of Fertility DeclineFertility Decline

-14.0

-12.0

-10.0

-8.0

-6.0

-4.0

-2.0

0.0

2.0

4.0

15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Age Group

Par

tici

pat

ion

Eff

ect

Direct Effect Total effect


SIMULATIONSIMULATION

►Economy with Economy with no technological progress no technological progress constant survival scheduleconstant survival schedule Constant education ratesConstant education rates

► Investigate the effect of fertility Investigate the effect of fertility reductionreduction

►Calibrate to South Korea Calibrate to South Korea education/survival/fertilityeducation/survival/fertility


Simulation Framework ISimulation Framework I

► Production FunctionProduction Function

► Physical Capital Stock Physical Capital Stock

► ParameterizationParameterization

1t t t tY AH K

1 1(1 )t t tK sY K

2, 0.24, 0.08

3

0

s

A

t


Simulation Framework IISimulation Framework II

► Demographic Structure Demographic Structure

► Human Capital Human Capital

0t t i i

i

Pop b

0

f ft t i i it

i

b b fert

0 ,

j j j jt t i i i i

i j m f

H b p e


Specific Example: South KoreaSpecific Example: South Korea

19601960 20002000

Fertility (TFR)Fertility (TFR) 5.65.6 1.21.2

Life Expectancy Life Expectancy at birthat birth

55.255.2 76.876.8

Female Labor Female Labor Market Market

Participation Participation (25-29)(25-29)

26.326.3 55.755.7


Fertility and Female Labor Market Fertility and Female Labor Market Participation - KoreaParticipation - Korea

0

10

20

30

40

50

60

70

1960 1965 1970 1975 1980 1985 1990 1995 2000

Par

ticip

atio

n R

ate

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Tot

al F

ertil

ity R

ate

Female Labor Market Participation Age Group 24-29Female Labor Market Participation Age Group 30-34Female Labor Market Participation Age Group 34-39Total Fertility Rate


Simulation ScenariosSimulation Scenarios

Baseline: Initial steady state 1960. Baseline: Initial steady state 1960.

Add actual fertility reductionsAdd actual fertility reductions

1.1. Solow model: population effect on capital Solow model: population effect on capital labor ratiolabor ratio

2.2. Age Structure: Demographic Change with Age Structure: Demographic Change with fixed cohort specific participation ratesfixed cohort specific participation rates

3.3. Age Structure plus Female Labor SupplyAge Structure plus Female Labor Supply


ResultsResults

0

50

100

150

200

250

300

1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

GD

P I

nd

ex

BaselineSolow Age StructureAge Structure + Female Participation


0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 2110

Age 65+ per working age

Age 0-14 per working Age

Working Age Persons per Capita

Korea: Simulated Demographic Structure


Korea: Demographics and Workers per Capita

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 2110

0

0.1

0.2

0.3

0.4

0.5

0.6

Age 65+ per working age

Age 0-14 per working Age

Working Age Persons per Capita

Workers per Capita (right hand scale)


Simulation Results Simulation Results SummarySummary

Demographic transition has important effects on long Demographic transition has important effects on long term per capita income. The magnitude of these term per capita income. The magnitude of these effects can be sizeable and depends on…effects can be sizeable and depends on…

Aging and old age labor force participationAging and old age labor force participation The magnitude of the female labor supply response The magnitude of the female labor supply response

to fertility declines.to fertility declines.


ConclusionsConclusions

Empirical results suggest that the decline in fertility Empirical results suggest that the decline in fertility leads to a significant increase in female labor force leads to a significant increase in female labor force participationparticipation

This increase in female labor force participation This increase in female labor force participation compounds the positive long term growth effects compounds the positive long term growth effects induced by the demographic transitioninduced by the demographic transition

The magnitude of this effect depends on the The magnitude of this effect depends on the participation behavior of the 65+ age group – the participation behavior of the 65+ age group – the focus of our complementary studyfocus of our complementary study


II. Heath and Life Span Effects


Health and Life Span EffectsHealth and Life Span Effects

►Value of Health/Lifespan Value of Health/Lifespan ImprovementsImprovements

►Health and Worker ProductivityHealth and Worker Productivity►Life spans and life cycle behaviorLife spans and life cycle behavior

retirementretirement Consumption/savingsConsumption/savings InstitutionsInstitutions

►Health lifespan and educationHealth lifespan and education


Value of Health Value of Health ImprovementsImprovements►Welfare Gain from Lifespan Welfare Gain from Lifespan

ImprovementImprovement►Value of life span gain in money units.Value of life span gain in money units.►Vale of a statistical lifeVale of a statistical life►What money gain would give the same What money gain would give the same

welfare benefit as the gain in life welfare benefit as the gain in life expectancy?expectancy?


Individual UtilityIndividual Utility

Life time welfareLife time welfare

Budget constraintBudget constraint

0

U (C, S)= max exp(- t) S(t) u(c(t)) dt

0 0

exp(-rt) S(t) c(t) dt= exp(-rt) S(t) y dt


Indirect UtilityIndirect Utility

►AssumingAssuming

► Annuity of $1 for life has valueAnnuity of $1 for life has value

0

V (Y, S)= u(y) exp(-rt) S(t) dt ( ) ( )u y A S

r

0

( ) exp(-rt) S(t) dtA S


Equivalent VariationEquivalent Variation

►Survival rates rise from SSurvival rates rise from S00 to S to S1 1 while while income rises from yincome rises from y00 to y to y1 1

►The equivalent variation e (rise in The equivalent variation e (rise in annual income) of the health annual income) of the health improvement solvesimprovement solves

►OrOr

1 0 1 1( , ) ( , )V y e S V y S

1 0 1 1( ) ( ) ( ) ( )u y e A S u y A S


Approximation of EVApproximation of EV

►Using a Taylor series expansionUsing a Taylor series expansion

►The equivalent variation depends on The equivalent variation depends on the discounted growth in survival and the discounted growth in survival and the level of income the level of income

1 01

1 0

( ) ( )( )

( ) ( )

A S A Su ye

u y A S


Value of Life Span IncreasesValue of Life Span Increases

► We can estimate the equivalent variation if we We can estimate the equivalent variation if we know the age specific survival function before know the age specific survival function before and after, the level of income and the shape and after, the level of income and the shape of the utility function.of the utility function.

► The utility function needs to be determined The utility function needs to be determined both in terms of its slope ( higher order terms both in terms of its slope ( higher order terms may be important as well) and its intercept – may be important as well) and its intercept – notice we implicitly take the utility of being notice we implicitly take the utility of being dead to be zero. dead to be zero.

► U(c) is the utility of being alive and having U(c) is the utility of being alive and having consumption c. We can find the intercept from consumption c. We can find the intercept from value of life studies. value of life studies.


Health and Full IncomeHealth and Full Income

►Health adds directly to welfare as well Health adds directly to welfare as well as acting as an input into production.as acting as an input into production.

►Value of Life studies put a high Value of Life studies put a high monetary valuation on small risks of monetary valuation on small risks of death.death.

►Over 50% of welfare gain in US from Over 50% of welfare gain in US from 1900 has been lifespan (Nordhaus).1900 has been lifespan (Nordhaus).

►Calibration of utility function as in Calibration of utility function as in Becker, Philipson and Soares (2005). Becker, Philipson and Soares (2005).


Health and Worker Health and Worker ProductivityProductivity

IssuesIssues►No consensus on how to define healthNo consensus on how to define health►Health status indicators have large Health status indicators have large

measurement errors.measurement errors.►Effect is bi-directional – we cannot Effect is bi-directional – we cannot

infer causality from correlation. infer causality from correlation.


Measuring Health CapitalMeasuring Health CapitalMultiple Indicators of HealthMultiple Indicators of Health

►Self assessed health statusSelf assessed health status►Morbidity RatesMorbidity Rates►Physical function limitationsPhysical function limitations►Physical growth outcomesPhysical growth outcomes


Health Human CapitalHealth Human Capital

► We are interested in health that comes as a We are interested in health that comes as a result of health and other investments – result of health and other investments – controlled health.controlled health.

► Uncontrolled health , e.g. due to genetic Uncontrolled health , e.g. due to genetic differences will affect productivity but is not differences will affect productivity but is not health capital.health capital. Compare with IQ and schooling as human capital.Compare with IQ and schooling as human capital.

► Ideally we would measure the effect of a Ideally we would measure the effect of a health input on health status and then trace health input on health status and then trace out the effect of the improved health status out the effect of the improved health status on productivity –but this is rare.on productivity –but this is rare.


Feedbacks from Income to Feedbacks from Income to HealthHealth►Model has three functions which occur Model has three functions which occur

simultaneously. simultaneously. ►Health is a function of health and other Health is a function of health and other

inputs (including shocks).inputs (including shocks).►People decide, based on their income, People decide, based on their income,

on inputs and activities that affect on inputs and activities that affect health. health.

►Health affects productivity and income Health affects productivity and income


Analytical FrameworkAnalytical Framework

►Health production functionHealth production function Health HHealth H Health inputs lHealth inputs l Exogenous health factors (genetic etc.) g Exogenous health factors (genetic etc.) g

usually unobservedusually unobserved Random error e1Random error e1

( , , 1)H h l g e


Input determinationInput determination

►The level of inputs depends on The level of inputs depends on household characteristics, such as household characteristics, such as wage earnings W, and the availability wage earnings W, and the availability of inputs Xof inputs X

► We “solve out” for the effect of We “solve out” for the effect of current health h on input demands.current health h on input demands.

( , , , 2)l d X W g e


ProductivityProductivity

►Wages W depend on health H, Wages W depend on health H, education E, other factors Z and an education E, other factors Z and an error term error term

( , , , 3)W w H E Z e


Estimation ProblemsEstimation Problems

►We have measurement error in health We have measurement error in health –biases results downwards.–biases results downwards.

►Health affect wages but wages also Health affect wages but wages also affect health via their effect on health affect health via their effect on health inputs – we have reverse causality.inputs – we have reverse causality.

►We only want the human capital We only want the human capital element of health’s contribution to element of health’s contribution to wages, not the genetic component. wages, not the genetic component.


Problems can be overcome Problems can be overcome using an instrumental using an instrumental variablevariable► Suppose instead of health we use predicted Suppose instead of health we use predicted

health based on the local availability of health based on the local availability of health services and factors that can used as health services and factors that can used as policies to affect health.policies to affect health.

► This removes measurement errorThis removes measurement error► This removes the reverse causality since the This removes the reverse causality since the

predicted health is independent of an predicted health is independent of an individual’s wage.individual’s wage.

► The predicted health measure is pure The predicted health measure is pure “controlled health” and does not include any “controlled health” and does not include any individual specific uncontrolled health. individual specific uncontrolled health.


Empirical Results on Wages Empirical Results on Wages DeterminantsDeterminantsAll these variables are instrumented- for All these variables are instrumented- for

example by local food prices or distance to example by local food prices or distance to a health facility when younga health facility when young

► Calories important (below 2000 kcal).Calories important (below 2000 kcal).► Proteins importantProteins important► BMI importantBMI important► Height importantHeight important► Days ill/working days lost importantDays ill/working days lost important


Lifespan, Retirement, and Lifespan, Retirement, and SavingSaving►Mis-match between time path of labor Mis-match between time path of labor

income and consumption.income and consumption.

Cash and in kind transfers within the Cash and in kind transfers within the household and between generations household and between generations through bequests.through bequests.

Transfers through the social security Transfers through the social security system.system.

Private Saving/borrowing.Private Saving/borrowing.


Savings RatesSavings Rates

0

10

20

30

40

50

60

70

1950 1960 1970 1980 1990 2000Year

per

cent Japan

Saingapore

Hong Kong

USA

Source:PWT6.1


Micro to Macro Micro to Macro ► Macro focus on age structure effects.Macro focus on age structure effects.► In micro data savings rates vary by age In micro data savings rates vary by age

with a peak at around 55 but these age with a peak at around 55 but these age effects on household savings are modest.effects on household savings are modest.

► Most of the savings boom in East Asia was Most of the savings boom in East Asia was due to higher savings at every age with due to higher savings at every age with only a modest contribution from age only a modest contribution from age structure effects.structure effects.

► Accounting effects of demographic Accounting effects of demographic change can only explain a small fraction change can only explain a small fraction of variation in savings.of variation in savings.

► We need to explain changes in saving We need to explain changes in saving behavior at each age.behavior at each age.


Savings BoomsSavings Booms

► Increase in individual savings due to Increase in individual savings due to improvements in health and longevity?improvements in health and longevity?

► Major alternative theory is habit formation Major alternative theory is habit formation in consumption.in consumption.

► Effect of new financial institutions is also Effect of new financial institutions is also possiblepossible


Why Longevity Could Raise Why Longevity Could Raise Savings RatesSavings Rates

► Possible Arguments Possible Arguments Unhealthy life span increase.Unhealthy life span increase. Effect of longer lifespan on compounding when Effect of longer lifespan on compounding when

interest rates and income growth are positive.interest rates and income growth are positive. Effect on returns to saving. Without annuities, Effect on returns to saving. Without annuities,

a high mortality rate reduces effective returns.a high mortality rate reduces effective returns.► Our ArgumentOur Argument

Social security system incentives restrict labor Social security system incentives restrict labor supply of the elderly and effectively limit the supply of the elderly and effectively limit the retirement age.retirement age.


Critique Critique Compression of Morbidity Compression of Morbidity► If longer life spans are associated with If longer life spans are associated with

healthy aging (compression of morbidity), healthy aging (compression of morbidity), optimal response is to extend the working optimal response is to extend the working life with little impact on savings rates.life with little impact on savings rates.

►We can regard a longer life as We can regard a longer life as “stretching” time, which stretches the “stretching” time, which stretches the retirement age but does not affect retirement age but does not affect savings rates.savings rates.

►The empirical effect of longevity on The empirical effect of longevity on savings lacks a theoretical foundation.savings lacks a theoretical foundation.


National Life Expectancy and National Life Expectancy and Health, 2000Health, 2000

20

30

40

50

60

70

80

30 40 50 60 70 80

life expectancy

hea

lth

y lif

e ex

pec

tan

cy H = Healthy life expectancyL=Life expectancy

H = -7.062 + 0.979 L (0.615) (0.010)

R2 = 0.982

Data from World Health Report 2001


Life Expectancy and Healthy Lifetimes

0.7

0.8

0.9

1

30 40 50 60 70 80 90

Life Expectancy

Lif

e H

eal

thy

Rat

io

Data for 2000, from World Health Report 2001


HypothesisHypothesis

► Under complete markets the effect of longer Under complete markets the effect of longer life spans on savings rates is zero or even life spans on savings rates is zero or even negativenegative

► A positive effect of longevity on savings rates A positive effect of longevity on savings rates depends on the presence of institutions that depends on the presence of institutions that prevent or discourage longer working lives.prevent or discourage longer working lives.


Life Cycle TheoryLife Cycle Theory

►Maximize lifetime utility with a budget Maximize lifetime utility with a budget constraintconstraint

0

( ) ( , )tt tU e u c v z t dt

t t t t

dWw r W c

dt


AssumptionsAssumptions

►Full insurance – annuitiesFull insurance – annuities►Exogenous health and mortalityExogenous health and mortality►Constant death rateConstant death rate►Disutility of work rises with age but Disutility of work rises with age but

depends on life expectancy – depends on life expectancy – compression of morbiditycompression of morbidity

/( , ) t z tv z t de de


Optimal Consumption and Optimal Consumption and Retirement Retirement ►Two conditions – optimal consumption Two conditions – optimal consumption

over time.over time.►Optimal retirement – wage times Optimal retirement – wage times

marginal utility of consumption equals marginal utility of consumption equals the disutility of working.the disutility of working.

►General solution defines retirement General solution defines retirement and consumption by an implicit and consumption by an implicit function. function.


Figure 1Retirement and Consumption


Assume Log UtilityAssume Log Utility► Use implicit function theorem to find optimal Use implicit function theorem to find optimal

retirement and consumptionretirement and consumption

2 2

1 1(1 ) log( ) 1 1 log( )1

log(1 ) (1 )

d ddd d dR z r z z

d d d

02 2

0

11 log( )1 1

( )1 (1 ) (1 )

ddc dr z z

w d d d

( )c

rc


Wage Level EffectsWage Level Effects

► In a model with log utility the wage In a model with log utility the wage level does not affect the retirement level does not affect the retirement decision – income and substitution decision – income and substitution effects balance.effects balance.

►With a general utility function (CRRA > With a general utility function (CRRA > 1), rising wages promote earlier 1), rising wages promote earlier retirement and a lower retirement and a lower consumption/wage ratio, i.e., a higher consumption/wage ratio, i.e., a higher savings rate. savings rate.


Preliminary Empirical Preliminary Empirical Results MicroResults Micro► Use HRS survey. Gives good measures Use HRS survey. Gives good measures

of household wealth and subjective of household wealth and subjective survival probabilities for individuals.survival probabilities for individuals.

► Question: do people who expect to live Question: do people who expect to live longer save more and so hold more longer save more and so hold more wealth?wealth?

► Problem: all current variables depend Problem: all current variables depend on wealth. Subjective survival on wealth. Subjective survival probabilities depend on wealth and probabilities depend on wealth and have large measurement error (lots of 0 have large measurement error (lots of 0 or 1 in replies).or 1 in replies).


EstimationEstimation

Model current Wealth as Model current Wealth as depending ondepending on

► Inheritances.Inheritances.► Planned accumulation to date as proxied Planned accumulation to date as proxied

by a function of age, schooling, and by a function of age, schooling, and height.height.

► Probability of survival to 75, instrumented Probability of survival to 75, instrumented with parents’ current age, or age at death.with parents’ current age, or age at death.

► Unplanned accumulation to date, the error Unplanned accumulation to date, the error term. term.


ResultsResults► Wealth increases with inheritances.Wealth increases with inheritances.► Wealth increases with height and Wealth increases with height and

education, probably reflecting higher education, probably reflecting higher income.income.

► Wealth increases with age in the HRS Wealth increases with age in the HRS sample (primarily between 40 and 60).sample (primarily between 40 and 60).

► Instrumented subjective survival Instrumented subjective survival probabilities have a positive and probabilities have a positive and significant effect on wealth holdings. significant effect on wealth holdings. (Passes instrument validity test).(Passes instrument validity test).


Macro SavingMacro Saving

►Aggregation over cohorts is not stright Aggregation over cohorts is not stright forward – depends on the distribution of forward – depends on the distribution of income.income.

►Age specific savings rates may rise Age specific savings rates may rise while average savings rates fall when while average savings rates fall when life span increases.life span increases.

►Zero savings over lifespan means zero Zero savings over lifespan means zero saving on average in steady state – saving on average in steady state – saving is a disequilibrium phenomenon.saving is a disequilibrium phenomenon.


Increasing Longevity and Increasing Longevity and SavingSaving

0

T T’

SavingRate

Age


Aggregate Savings RatesAggregate Savings Rates► In a stable population, with a fixed life In a stable population, with a fixed life

expectancy, net life cycle savings are expectancy, net life cycle savings are zero.zero.

► A A riserise in life expectancy with a in life expectancy with a fixedfixed age structure increases aggregate age structure increases aggregate savings; the saving of the young and savings; the saving of the young and middle aged for retirement is larger than middle aged for retirement is larger than the dis-saving of the older generation.the dis-saving of the older generation.

► This saving boom is temporary; it This saving boom is temporary; it disappears when the age structure disappears when the age structure adjusts to a stable structure given the adjusts to a stable structure given the higher lifetimes.higher lifetimes.


Table 5Effects on Steady-State Saving

Rate

Effect on Steady-State Effect on Steady-State Saving RateSaving Rate

(percentage points)(percentage points)

Old/ Working Age Ratio rises by 0.01Old/ Working Age Ratio rises by 0.01 -1.336-1.336(4.07)(4.07)

Life expectancy rises by 1 year with universal Life expectancy rises by 1 year with universal coverage, mandatory retirement, and a fully coverage, mandatory retirement, and a fully

funded systemfunded system

0.4240.424(2.27)(2.27)

Life expectancy rises by 1 year with universal Life expectancy rises by 1 year with universal coverage, mandatory retirement and a pay-as-you-coverage, mandatory retirement and a pay-as-you-

go system with replacement rate of 0.5go system with replacement rate of 0.5

0.0030.003(0.02)(0.02)

Life expectancy rises by 1 year with universal Life expectancy rises by 1 year with universal coverage, mandatory retirement and a pay-as-you-coverage, mandatory retirement and a pay-as-you-

go system with replacement rate of 1.0go system with replacement rate of 1.0

-0.418-0.418(1.90)(1.90)

Effect of introducing a retirement incentive with Effect of introducing a retirement incentive with life expectancy at 66 years.life expectancy at 66 years.

2.4892.489(2.23)(2.23)

Effect of introducing a retirement incentive with Effect of introducing a retirement incentive with life expectancy at 81 years.life expectancy at 81 years.

3.0553.055(3.48)(3.48)

Effect of moving from a pay-as-you-go system Effect of moving from a pay-as-you-go system (replacement rate 1.0) to a fully funded system (replacement rate 1.0) to a fully funded system

with life expectancy 66 years.with life expectancy 66 years.

0.0050.005(0.16)(0.16)

Effect of moving from a pay-as-you-go system Effect of moving from a pay-as-you-go system (replacement rate 1.0) to a fully funded system (replacement rate 1.0) to a fully funded system

with life expectancy 81 years.with life expectancy 81 years.

13.14813.148(2.93)(2.93)


Health and EducationHealth and Education

►Health and cognitive developmentHealth and cognitive development► Incentive effects from longer working Incentive effects from longer working

lifelife►Lower depreciation of human capitalLower depreciation of human capital►Heath and education Heath and education

complementaritiescomplementarities Fewer working days lostFewer working days lost


III. The Demographic Dividend


Crude Birth Rate vs. Crude Death Rate, 1800-2000

0

10

20

30

40

50

0 10 20 30 40 50

Crude Death Rate (per 1,000)

Cru

de B

irth

Rat

e (p

er 1

,000

)

Sweden

Japan

India

2000

2000

1950

1800

1900

Population grow th rate = 0%/yr



1900


Environment and PopulationEnvironment and Population

► CO2 Emissions rise with population but are CO2 Emissions rise with population but are more elastic with respect to incomemore elastic with respect to income

► Log CO2=1.12 Log Pop + 1.48 Log YLog CO2=1.12 Log Pop + 1.48 Log Y(0.01)(0.01) (0.03)(0.03)

R2= 0.897R2= 0.897Panel data (5 year)1960-2000 time dummies includedPanel data (5 year)1960-2000 time dummies included

► Major threat is rising incomes in India and Major threat is rising incomes in India and ChinaChina

► Acid rain and ozone deletion have had an Acid rain and ozone deletion have had an effective global institutional response. effective global institutional response.

► Global institutional response to global Global institutional response to global warming has been weak.warming has been weak.


Capital and Land EffectsCapital and Land Effects

►Free movement of capital prevents Free movement of capital prevents declinei n capital/labor ratios.declinei n capital/labor ratios.

►Effects may be large if capital markets Effects may be large if capital markets are closed/imperfect.are closed/imperfect.

►Land effects may be large in Land effects may be large in agricultural societies. agricultural societies.

►But land shortage may spur But land shortage may spur industrialization. industrialization.


Components of Population Components of Population GrowthGrowth► Kelly and Schmidt 1995Kelly and Schmidt 1995► While growth of population numbers does While growth of population numbers does

not matter, components do matter.not matter, components do matter.► Population growth = birth rate- death rate. Population growth = birth rate- death rate.

High birth and death rates both seem to High birth and death rates both seem to have a negative effect on growth.have a negative effect on growth.

► Lagged (15 year) birth rate sometimes Lagged (15 year) birth rate sometimes positivepositive

► handouthandout


Importance and ProblemsImportance and Problems

►Components of population growth may Components of population growth may matter even if pop growth is neutral matter even if pop growth is neutral on average.on average.

►Fixed effects estimators are biasedFixed effects estimators are biased►Does not distinguish behavioral effects Does not distinguish behavioral effects

from accounting effects from accounting effects


Age Structure EffectsAge Structure Effects

Y Y L WAN LWA N

log , log , log , logY Y L WA

y z p cN L WA N

y z p c

y z p c


Age Structure EffectsAge Structure Effects(continued)(continued)

0( )y X p c y p c

0( )z X p c y

0( * ), *z z z z X


Estimating EquationEstimating Equation

►Working age pop growth Working age pop growth

►Population growth Population growth

0( )y X p c y p w n

w

n


ConstraintsConstraints

► Equation is derived from an identity.Equation is derived from an identity.► Some parameters are fixed by the identitySome parameters are fixed by the identity► Coefficients on log initial participation rate p Coefficients on log initial participation rate p

and and working age share of pop c should be and and working age share of pop c should be equal size and opposite to coefficient on equal size and opposite to coefficient on initial income.initial income.

► Coefficients on participation rate growth, Coefficients on participation rate growth, work age pop growth and pop growth should work age pop growth and pop growth should be 1, 1, -1.be 1, 1, -1.

► Why estimate these? Why estimate these?


11 22 33OLSOLS 2SLS2SLS 2SLS2SLS

log working log working age over age over total poptotal pop

00.257**.257**(0.118)(0.118)

00.284**.284**(0.125)(0.125)

00.180.180(0.143)(0.143)

log log participatioparticipation raten rate

--0.186***0.186***(0.058)(0.058)

--0.166***0.166***(0.058)(0.058)

--0.164***0.164***(0.062)(0.062)

growth of growth of participatioparticipation rate n rate

-0.570*-0.570*(0.302)(0.302)

-0.280-0.280(0.527)(0.527)

-0.212-0.212(0.560)(0.560)

growth of growth of working working age over age over total poptotal pop

00.789**.789**(0.329)(0.329)

1.2221.222****(0.575)(0.575)

-0.793-0.793(0.973)(0.973)

growth of growth of working working age ratio age ratio times times opennessopenness

3.328***3.328***(1.184)(1.184)


Estimation versus identityEstimation versus identity

► Identity should be found by Identity should be found by estimator if model is correctestimator if model is correct

►Difference between estimate and Difference between estimate and identity reflects mis-specification.identity reflects mis-specification.

►Model may not adjust fully for Model may not adjust fully for changes in labor quality when changes in labor quality when labor quantity is changing rapidly.labor quantity is changing rapidly.

► Identity assumes full employment Identity assumes full employment of resources. of resources.


Demographic Dividend is not Demographic Dividend is not AutomaticAutomatic►Depends on effective policies in other Depends on effective policies in other

areas areas EducationEducation Labor marketLabor market TradeTrade GovernanceGovernance Macroeconomic management Macroeconomic management

►Latin America provides example of Latin America provides example of "unreaped dividend""unreaped dividend"


Case Study: IrelandCase Study: Ireland

►Rapid Economic Growth 1990’sRapid Economic Growth 1990’s►Celtic TigerCeltic Tiger►Approaching East Asian Growth Approaching East Asian Growth

Miracle rates.Miracle rates.


Economic Miracles and Debacles

-2%

-1%

0%

1%

2%

3%

4%

5%

6%

7%

1980s 1990s

GD

P p

er c

apit

aan

nu

al g

row

th r

ate

IrelandEast Asia & PacificSub-Saharan Africa


Sub-Saharan Africa’s Sub-Saharan Africa’s populationpopulation


East Asia's Population


Ireland's Population


Ratio, Working-Age to Non-Working-Age Population,

Ireland and Comparisons

1.00

1.25

1.50

1.75

2.00

2.25

2.50

1950

1960

1970

1980

1990

2000

2010

2020

2030

2040

2050

Year

Rat

io IrelandUnited KingdomEast Asia


Why this rapid demographic Why this rapid demographic change in Ireland?change in Ireland?►Mortality rates at European norms.Mortality rates at European norms.►Fertility rates high by European Fertility rates high by European

standards until 1979.standards until 1979.►Rapid fertility decline after 1979.Rapid fertility decline after 1979.►High levels of migration.High levels of migration.


Crude Birth and Death Rates, Ireland

0

5

10

15

20

25

1960 1965 1970 1975 1980 1985 1990 1995 2000

Year

Bir

ths

and

dea

ths

per

1,

000

peo

ple

Crude birth rate Crude death rate


Family Planning in IrelandFamily Planning in Ireland

1969: The Fertility Guidance Company Ltd. is 1969: The Fertility Guidance Company Ltd. is formed (predecessor of the IFPA) formed (predecessor of the IFPA)

1973: Irish Supreme Court legalizes the 1973: Irish Supreme Court legalizes the importation of contraceptives for personal importation of contraceptives for personal use use

1979: Health (Family Planning) Act becomes 1979: Health (Family Planning) Act becomes law: sale of contraceptives legalized for law: sale of contraceptives legalized for family planning (with prescription)family planning (with prescription)

1985: Amendment to 1979 law makes contra-1985: Amendment to 1979 law makes contra-ceptives available to those over 18 without ceptives available to those over 18 without prescription.prescription.


Migration in IrelandMigration in Ireland

►Relatively large fluctuations – Relatively large fluctuations – endogenous response to economic endogenous response to economic performance performance

►Sizable impacts on age structureSizable impacts on age structure►Out-migration mainly among 15-24 Out-migration mainly among 15-24

year oldsyear olds► In-migration mainly among 35-44 year In-migration mainly among 35-44 year

olds, plus their childrenolds, plus their children


Population Growth in Ireland

-0.01

0

0.01

0.02

0.03

1960 1965 1970 1975 1980 1985 1990 1995 2000

population growth birth rate natural population growth


Irish Female Labour Participation

0

20

40

60

80

20 25 30 35 40 45 50 55 60 65age group

rate

1950

1960

1970

1980

1990

2000


GDP per Capita Growth

-3

-2

-1

0

1

2

3

4

5

6

1950 1960 1970 1980 1990 2000

GDP per Capita Growth Fitted


Sources of Growth

-2

-1

0

1

2

3

4

1960 1970 1980 1990 2000

Age Structure Change World Growth Catch-up


Ireland ConclusionIreland Conclusion

► Demographic change has promoted Demographic change has promoted economic performance in Ireland.economic performance in Ireland.

► Fertility decline was abetted by the family Fertility decline was abetted by the family planning movement in Ireland and the planning movement in Ireland and the legalization of contraception. legalization of contraception.

► The policy environment in Ireland favored The policy environment in Ireland favored capture of the demographic dividend.capture of the demographic dividend.

► Ireland's natural demographic dividend is Ireland's natural demographic dividend is now at its peak. now at its peak.


QuestionsQuestions

►Age structure effects – accounting Age structure effects – accounting versus behavior.versus behavior.

►Conditions that interact with effects of Conditions that interact with effects of the demographic dividend.the demographic dividend.

►Link from fertility and mortality to age Link from fertility and mortality to age structure – formal demography.structure – formal demography.

► Importance of Migration. Importance of Migration.


Demographic AccountingDemographic Accounting

►Age structure effects are “accounting” Age structure effects are “accounting” to some extent.to some extent.

►Assume age specific behavior remains Assume age specific behavior remains the same and examine the effect of the same and examine the effect of age structure changesage structure changes

►But – there are also behavioral effectsBut – there are also behavioral effects


Mortality RatesMortality Rates

►Declines in age specific mortality rates Declines in age specific mortality rates change age structure.change age structure.

►Mortality decline leads to increase in Mortality decline leads to increase in longevity – may have life cycle effects.longevity – may have life cycle effects.

►Mortality decline also linked to Mortality decline also linked to morbidity decline – healthier people. morbidity decline – healthier people.


Birth RatesBirth Rates

►Changes in the birth rate affects age Changes in the birth rate affects age structurestructure

►Birth rates changes are usually linked Birth rates changes are usually linked to female labor supply.to female labor supply.

►Labor supply and fertility jointly Labor supply and fertility jointly determined (not one causes the determined (not one causes the other).other).

►Lower fertility may be linked to higher Lower fertility may be linked to higher levels of investment in human capital.levels of investment in human capital.


Policy InteractionPolicy Interaction

►What policies are most important for What policies are most important for making use of the extra labor supply making use of the extra labor supply that comes from the demographic that comes from the demographic transition?transition?

►Labor market, openness, etc.Labor market, openness, etc.► Ireland and many “good” features.Ireland and many “good” features.


Explanations for Ireland’s Explanations for Ireland’s GrowthGrowth► EU membership and subsidies EU membership and subsidies ► Increased trade and FDI, partly due to tax Increased trade and FDI, partly due to tax

incentives incentives ► Delayed convergenceDelayed convergence► Good macroeconomic management Good macroeconomic management ► Social contract between government, trade Social contract between government, trade

unions, and employers unions, and employers ► Education expansion in the 1970’s.Education expansion in the 1970’s.► (Transfer pricing)(Transfer pricing)


Age Structure and FertilityAge Structure and Fertility

►Can we rewrite of growth – age Can we rewrite of growth – age structure identity in terms of fertility structure identity in terms of fertility and mortality rates?and mortality rates?

►Clearly lagged birth and death rates Clearly lagged birth and death rates matter for current age structurematter for current age structure

►Simple relationship?Simple relationship?


MigrationMigration

►Migration is clearly endogenousMigration is clearly endogenous►Responds to domestic and Responds to domestic and

international opportunities.international opportunities.►Primarily movements of working age Primarily movements of working age

population.population.►““Loss” mitigated by remittances.Loss” mitigated by remittances.►Welfare criteria especially unclear with Welfare criteria especially unclear with

migration. migration.


The EndThe End

n ational t ransfer a ccounts 1 empirical models david canning harvard university

Documents