The Demographic Transitionmedia.virbcdn.com/files/dd/a3b6b1e276a56f69-DTBeamer-Dec...The Demographic...

EvidenceTheories

The Demographic Transition

Oded Galor

Department of EconomicsBrown University

November 29, 2015

Oded Galor The Demographic Transition

EvidenceTheories

Patterns Across the GlobeCorrelations with Economic Outcomes

The Demographic Transition

A rapid decline in fertility, mortality and population growth thatmark the transition to modern growth:

The positive relationship between income per capita and pop-ulation growth is reversed

Gains in total output are not counterbalanced by populationgrowth


EvidenceTheories


The Demographic Transition in Western Europe: Crude Birth Rates


EvidenceTheories


The Demographic Transition in Western Europe: Total Fertility Rates


EvidenceTheories


Mortality Decline Western Europe: 1730-1920


EvidenceTheories


The Evolution of Total Fertility Rate across Regions, 1960-1999


EvidenceTheories


Decline in infant mortality rates across regions, 1960-1999


EvidenceTheories


Years Elapsed since the Demographic Transition


EvidenceTheories



NPL

BTNLAO

MLIPRY

AGO

GINNGA

TGOLBR

BDI

RWA

HTIALB

GTM

HND

SYR

NICKHM

IRN

NAM

MDGBOL

MNG

ERI

COM

FIN

LSO

CHL

ETH

CAF

ROM

SAU

BGD

MEX

NOR

JOR

DNK

KWT

BHS

GBR

VNM

BGR

NLDITA

ECU

BRA

PER

NER

CHN

AUT

GEO

VEN

UZB

KGZ

PRT

DOMGMB

AZETKMARM

BEL

BLZ

CHE

PANDJI

GUY

DEU

COLSUR

LBY

BENSLV

CAN

GHA

CIV

BHR

LBN

CMRTTO

SWE

FRA

MRT

ATG

ESPCRI

KOR

ZAF

HUN

DZA

ZMBSDN

SWZ

SEN

MWI

BWAPRIUSA

ISR

ARG

JPN

TUNTHA

MARKENIDN

IND

ZWE

EGY

URY

TZA

MYS

JAM

LKAPHL

MUS

SGP

SYC

20

24

Log

GD

P Pe

r Cap

ita in

the

year

200

0

1.5 1 .5 0 .5 1

Log Years Elapsed since the Demographic Transition

controlling for the absolute latitude and continental fixed effectsOded Galor The Demographic Transition

EvidenceTheories



NPL

MLI

TGO

PRY

LBRRWA

SYR

HTI

IRNGTM

HND

NIC

BOL

BGD

CAFLSOFIN

JORKWTMEXCHL

NOR

CHN

DNK

NER

BGR

GBR

ITA

NLD

GMB

AUT

PRT

ECU

BHR

PER

BRABENBEL

DOM

CHE

PAN

GHA

VEN

DEUCMRGUY

KOR

SLV

COL

ESP

FRA

TTO

CRI

SWE

ZAF

SDN

ZMB

DZA

SEN

HUN

CAN

MWI

SWZ

ISR

BWA

THA

IDN

JPN

BRB

IND

KEN

TUN

USA

ZWE

MYS

ARGEGY

LKA

PHL

URY

JAM

SGP

CUB

MUS

64

20

24

Aver

age

Year

s of

Edu

catio

n in

200

0

2 1 0 1

Log Years Elapsed since the Demographic Transition

controlling for the absolute latitude and continental fixed effectsOded Galor The Demographic Transition

EvidenceTheories

The Rise in IncomeThe Decline in Child MortalityThe Old-Age Security HypothesisThe Decline in the Gender Wage GapThe Rise in Human Capital Formation

Theories of the Demographic Transition

The Rise in Income

The rise in income increased in the opportunity cost of raisingchildren ⇒ reduction in fertility (Becker, 1960)

The income elasticity of child quality is larger than that ofchild quantity ⇒ substitution of child quality for quantity⇒ reduction in fertility (Becker and Lewis, JPE 1973)

The Decline in Child Mortality

Mortality decline enabled families to attain the same numberof children with lower fertility rates


EvidenceTheories



The Old-Age Security Hypothesis (Caldwell, 1976)

Development of financial markets reduced the demand forchildren as an investment good

⇒ reduction in fertility

The Decline in the Gender Wage Gap (Galor-Weil, AER 1996))

The rise in the relative wages of women increases theopportunity cost of raising children more than family income

⇒ reduction in fertility


EvidenceTheories



The Rise Human Capital Formation:

Increased industrial demand for human capital induced humancapital formation ⇒ substitution of child quality for qunatity⇒ reduction in fertility⇒ (Galor-Weil, AER 1999, 2000)

Reinforced by: increased life expectancyReinforced by: decline in the profitability of child labor(Hazan-Berdugo, EJ 2002, Doepke-Zilibotti, AER 2005)

Evolution of preferences for child quality (Galor-Moav QJE,2002)

Natural selection favors the quality strategy

Religious indoctrination toward literacy (Judaism;Protestantism)


EvidenceTheories


The Rise in Income - Main Hypothesis

The rise in income increases in the opportunity cost of raisingchildren

⇒ fertility declines (Becker, 1960)

The income elasticity of child quality is larger than that of childquantity

⇒ substitution of quality for quantity as income increases

⇒ fertility declines (Becker and Lewis, JPE 1973)


EvidenceTheories


The Rise in Income: Mechanism

Child rearing is time-intensive

Budget constraint

yτn + c ≤ y

y ≡ household’s incomec ≡ household’s consumptionn ≡ household’s (surviving) childrenτ ≡ time cost per childyτ ≡ opportunity cost of raising a child


EvidenceTheories



The rise in income generates two conflicting effects:

An income effect:yτn + c ≤ [y ] ⇑

More income can be devoted to raising childrenoperates towards n ⇑

A substitution effect:

⇑ [yτ ]n + c ≤ y

The opportunity cost of raising children increasesoperates towards n ⇓


EvidenceTheories



The substituting effect dominates at a higher level of income

As income increases fertility declines

Fertility declines in the process of development


EvidenceTheories


The Rise in Income - Theoretical Evaluation

Preference-based theory (unattractive)

Innate bias against child quantity beyond a certain level ofincome - non-refutable

Non-robust (e.g., the class of homothetic preferences will nottrigger a fertility decline)


EvidenceTheories


The Rise in Income - Homothetic Preferences

Preferences:u = γ ln n + (1− γ) ln c

Budget constraintyτn + c ≤ y

Optimization: (fraction γ of income is spent on children and(1− γ) on consumption)

yτn = γyc = (1− γ)y


EvidenceTheories



Optimal number of children

n = γ/τ

The rise in income has no effect on fertility, i.e.,

|Income effect| = |Substitution effect|

Fertility is unaffected by the process of development


EvidenceTheories



Time Devoted toRaising Children

γ Income ExpansionPath

Consumption

1


EvidenceTheories


The Rise in Income: Testable predictions

Across countries that are similar in sociocultural characteristics(and thus in noneconomic factors that may affect fertility deci-sions), the timing of the fertility decline is inversely related tothe level of income per capita.

Within an economy, the number of (surviving) children acrosshouseholds is inversely related to their levels of income.


EvidenceTheories


The Rise in Income: Refuting Cross Country Evidence

Cross Section of Countries (1870-2000) - Income per worker waspositively associated with fertility rates, accounting for mortalityrates and education (Murtin (2009).

Western Europe (1870s) The DT occurred among countries thatdiffered significantly in their income per capita.


EvidenceTheories


Simultaneous DT across European Countries that Differ in Income per Capita


EvidenceTheories


The Rise in Income: Refuting Evidence from Individual Countries

France (1876—96) Income per capita had a positive effect onfertility rates during France’s demographic transition, accounting foreducation, the gender literacy gap, and mortality rates (Murphy 2009)

England (During the DT):The force associated with the rise inincome would have led to an increase in fertility rates (Fernandez Villaverde(2001)

England (1630s) Reproductive success increases with income (Clark

(JEH 2006)


EvidenceTheories


The Decline in Child Mortality - Main Hypothesis

Parents generates utility from the number of surviving children

A decline in child mortality permit parents to reach a givenlevel of surviving children with lower fertility

The decline in mortality triggered the subsequent decline infertility


EvidenceTheories


The Decline in Mortality —Mechanism

Preferences:u = γ ln n + (1− γ) ln c

c ≡ household’s consumptionn ≡ household’s surviving children

Survival childrenn = θnb

θ ≡ probability of a child to survive infancynb ≡ household’s children born


EvidenceTheories



Budget constraintyτn + c ≤ y

y ≡ household’s incomec ≡ household’s consumptionτ ≡ time cost of raising a surviving child0 ≡ time cost of raising a surviving childyτ ≡ opportunity cost of raising a surviving child


EvidenceTheories



Optimization: (fraction γ of income is spent on children and(1− γ) on consumption)

yτn = γyc = (1− γ)y

Optimal number of surviving children (NNR)

n = γ/τ

Optimal fertility (# of successful pregnancies - TFR)

nb = n/θ = γ/(τθ)


EvidenceTheories


The Decline in Mortality —Testable Predictions

Child mortality rate, (1− θ), has a positive effect on TFR

Child mortality rate, (1− θ), has no effect on (NetReproduction Rate) NRR


EvidenceTheories


...The Decline in Child Mortality

NRR does not decline unless:

There exists a precautionary demand for children

RA with respect to fertility > RA with respect to consumption

Replacement fertility is insignificant (empirical estimates 0.2—0.6)

Resources saved from investment in non-surviving children are notchanneled towards higher fertility


EvidenceTheories


The Decline in Mortality and Fertlity - Evidence


EvidenceTheories


The Decline in Child Mortality —Challenges to the Theory

Worldwide: NRR and TFR plummets jointly in during thedemographic transition. but the theory does not predict a decline inNRR

US, France and Some LDCs: The decline in mortality did notprecede the decline in fertility

Western Europe: No change in the patterns of mortality decline atthe time of the sharp decline in fertility

England: The decline in mortality started in England in the 1730s(140 years before the fertility decline) and was accompanied by asteady increase in fertility rates until 1800


EvidenceTheories


The Decline in Mortality: Refuting Evidence from Individual Countries

France (1876—96): Mortality rate had no effect on fertility duringFrance’s demographic transition, accounting for education, income,and the gender literacy gap. (Murphy 2009)

England (1861—1951):The force associated with the decline inchild mortality would have led to an increase in fertility rates (FernandezVillaverde (2001); Doepke (2005)


EvidenceTheories


The Old-Age Security Hypothesis

Children is a form of investment good (in the absence of capitalmarkets)

The development of financial markets reduced the demand forchildren for investment purposes and triggered a decline in fer-tility


EvidenceTheories


The Old-Age Security Hypothesis - Challenge to the Theory

The decline in the importance of old-age support is unlikely to be amajor force behind the significant reduction in fertility —at a rate of30—50% —during the demographic transition:

Rare examples in nature of offspring that support their parentsin old age

Institutions supporting individuals in their old age were formed wellbefore the demographic transition

England (16th century) Parents did not rely on support fromchildren in their old age (Pelling and Smith-1994)

Prior to the demographic transition, richer individuals whopresumably had better access to financial markets, did not havefewer surviving children


EvidenceTheories


Contributions to the field of Economic Growth

Gender and Growth

First attempt to examine the role of differences across genderin the growth process

Heterogeneity and Growth

Galor and Zeira (1993) —first attempt to explore the effect ofheterogeneity, across households, and the growth process

Galor and Weil (1996) —first attempt to explore the effect ofheterogeneity, within the households, on the growth process


EvidenceTheories


Mechanism: I. Development and Women’s Wages

Female-Biased Technical change

Mechanization and advanced technologies have complementedmental tasks more than physical tasks

Women have physiological comparative advantage in mental(rather than physical) tasks

=⇒ The process of development has (inevitably) increased theproductivity of women relative to men:

k ⇑ =⇒ (wF /wM ) ⇑

wF ≡ women′s wageswM ≡ men′s wagesk ≡ capital-labor ratio


EvidenceTheories


Evolution of the Gender Earninig Ratio - US


EvidenceTheories


Evolution of the Gender Literacy Gap - England


EvidenceTheories


Mechanism: II. Women’s Relative Wages and Fertility

Child rearing is time-intensive

Women are the prime care-takers engaged in child rearing

Budget constraint (if only women raise children)

wF τn + c ≤ wM + wF

wF + wM ≡ household’s incomec ≡ household’s consumptionn ≡ household’s (surviving) childrenτ ≡ time cost per childwF τ ≡ opportunity cost of raising a child


EvidenceTheories


Mechanism: II. Women’s Relative Wages and Fertility

The rise in women’s wages, wF , generates two conflicting effects:An income effect:

wF τn + c ≤ wM + [wF ] ⇑

More income for raising children =⇒ operates towardsn ⇑

A substitution effect:

⇑ [wF τ ]n + c ≤ wM + wF

Opportunity cost of children increases =⇒ operatestowards n ⇓

A rise in men’s wages generate only an income effect

wF τn + c ≤ [wM ] ⇑ +[wF ]Oded Galor The Demographic Transition

EvidenceTheories


.The Decline in the Gender Wage Gap

If women work and raise children, an increase in wF increases theopportunity cost of raising children more than family incomei.e.,

wF ⇑ =⇒ |Income effect| < |Substitution effect|

=⇒ n ⇓ (even if preferences are homothetic)

A rise in men’s wages generate only an income effect

wM ⇑ =⇒ n ⇑


EvidenceTheories


Women’s Relative Wages and Fertility - US


EvidenceTheories


Women’s Relative Wages and Fertility - Evidence

US (1970s): wF ⇑ =⇒ n ⇓ & wM ⇑ =⇒ n ⇑ (Heckman and

Walker (ECT 79))

Sweden’s demographic transition: (wF /wM ) ⇑ =⇒ n ⇓Schultz (1985)

France (1876—1896): reduction in the gender literacy gap had anadverse effect on fertility, accounting for income per capita,educational attainment, and mortality rates (Murphy (2009))


EvidenceTheories


The Rise in the Demand for Human Capital - Main Thesis

The acceleration in the rate of technological progress in the 2nd phaseof industrialization increased the demand for human capital

education enabled individuals to cope with a rapidlychanging technological environment

The rise in the demand for human capital induced a substitution ofquality for quantity of children triggering a demographic transition

=⇒ reduction in fertility


EvidenceTheories


The Model - Preferecnes

u = (1− γ) ln c + γ[ln n + β ln h]

c ≡ consumption

n ≡ (surviving) children

h ≡ quality (human capital) of each child

β ≡ degree of preference for child quality; β < 1


EvidenceTheories


The Model - Budget Constraint

yn(τq + τ ee) + c ≤ y

y ≡ household potential income

τq ≡ fraction of the household’s unit-time endowment required toraise a child, regardless of quality

τ e ≡ fraction of the household’s unit-time endowment required foreach unit of education per child

(τq + τ ee) ≡ time cost of raising a child with a level of education(quality) e

y(τq + τ ee) ≡ opportunity cost of raising a child with quality e


EvidenceTheories


The Model - Human Capital Formation

h = h(e, g)

he (e, g) > 0 & hee(e, g) < 0HC is increasing (in decreasing rates) in the parental timeinvestment in the education of the child

hg (e, g) < 0 & hgg (e, g) > 0HC is decreasing in the rate of technological progress(obsolescence of HC in a changing technological environment)

heg (e, g) > 0Education lessens the obsolescence of HC in a changingtechnological environment

h(0, g) > 0 & lime→0 he (e, g) =∞; lime→∞ he (e, g) = 0Basic level of human capital & interior solution


EvidenceTheories


The Model - Optimization

{n, e, c} = argmax γ[ln n + β ln h(e, g)] + (1− γ) ln cs.t. yn(τq + τ ee) + c ≤ y

since c = y [1− n(τq + τ ee)] ⇐⇒

{n, e} = argmax γ[ln n+β ln h(e, g)]+(1−γ) ln y [1−n(τq + τ ee)]


EvidenceTheories


Optimization


with respect to n:

γ

n=(1− γ)y(τq + τ ee)y [1− n(τq + τ ee)]

γ[1− n(τq + τ ee)] = (1− γ)(τq + τ ee)n

n(τq + τ ee) = γ


EvidenceTheories


Optimization


with respect to e :

γβhe (e, g)h(e, g)

=(1− γ)ynτ e

y [1− n(τq + τ ee)]

since n(τq + τ ee) = γ

γβhe (e, g)h(e, g)

= nτ e =⇒ βhe (e, g)h(e, g)

=τ e

(τq + τ ee)

βhe (e, g)(τq + τ ee) = τ eh(e, g)


EvidenceTheories


Optimization

n = γ/(τq + τ ee)

τ eh(e, g) = βhe (e, g)(τq + τ ee)=⇒

e = e(g , β, τ e , τq),

n = γ/[τq + τ ee(g , β, τ e , τq)]


EvidenceTheories


Testable Predictions - Investment in Quality

The optimal level of investment in child quality increases if:

The technological environment changes more rapidly

∂e(g , β, τ e , τq)/∂g > 0

Preferences for child quality are higher

∂e(g , β, τ e , τq)/∂β > 0

The cost of raising a child (regardless of quality) increases

∂e(g , β, τ e , τq)/∂τq > 0

The cost of educating a child decreases

∂e(g , β, τ e , τq)/∂τ e < 0


EvidenceTheories


Testable Predictions - Investment in Quantity

The optimal number of children decreases if:

The technological environment changes more rapidly

∂n/∂g < 0

Preferences for child quality are higher

∂n/∂β < 0

The cost of raising a child (regardless of quality) increases

∂n/∂τq < 0

The cost of educating a child increases and the elasticity of childquality with respect to the cost of child quality is smaller than onein absolute value

∂n/∂τ e < 0 if [∂e/∂τ e ][τ e/e] > −1


EvidenceTheories


Human Capital Fomration and the DT - England


EvidenceTheories


Growth Rates 1870-1913 and DT


EvidenceTheories


Supporting Evidence: Cross-Country

Cross Section of Countries (1870-2000) - educationalattainment has been negatively associated with fertility, accountingfor income per worker and mortality rates (Murtin (2009).

Cross Section of Countries (1960-1999): adverse effect on netfertility of an increase in productivity in advanced stages ofdevelopment, when education demand dominates (Lehr 2009)


EvidenceTheories


Supporting Evidence: Cross-Country

US (1910s): Eradication of hookworm —a positive shock to thereturn to child quality - had an adverse effect on fertility(Bleakley-Lange-2009)

Prussia (19th century): the rise in human capital formation hashad an adverse effect on fertility (IV: Land concentration &Distance from the birthplace of Protestantism - Wittenberg)(Becker-Cinnirella-Woessmann (2010)

France (1876—96): the level of education attainment had anadverse effect on fertility rates during France’s demographictransition, accounting for income per capita, the gender literacygap, and mortality rates. (Murphy 2009)

England (1580-1871) Adverse effect of family size on individualliteracy (IV parental fecundity). (Klemp-Weisdorf (2010))


Date post:	22-Aug-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

The Demographic Transitionmedia.virbcdn.com/files/dd/a3b6b1e276a56f69-DTBeamer-Dec...The Demographic...

Documents