Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Demographic Change And Economic Transitions inAsia

Md Abu Sayem

Roll no: FH-017Institute of Statistical Research and Training

University of Dhaka

March 13, 2015

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 1 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Outline

1 IntroductionBackgroundObjectives

2 Literature review

3 Data sources and variables

4 MethodologyEconomic modelStatistical Model

5 Analysis and Results

6 Conclusion

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 2 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Background

In recent years, Asian demography is changing quickly.

Young and elderly people require extra health care, prime-ageadults supply labour and savings.

Working age people taking few child, so there might be arelation between age structure and GDP growth rate.

There is huge study to find relation between demography andeconomy in Asia.

All study have found strong relationship.

Some geographical variable also affect the economicalcondition.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 3 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Objectives

To find the relation and pattern among demographic variablesand economic growth.

To show the differences in different regions in terms ofeconomic growth.

To find the effect of income on birth and death rate.

To obtain the causes of increase of life expectancy anddecrease of birth and death rate in different regions of Asia.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 4 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Literature Review

Hussin and Saidin(2012) has examined the impact ofeconomic variables such as foreign direct investment (FDI),openness and gross capital formation to economic growth overthe period 1981- 2008 in ASEAN-4 countries. Using OLSestimation, they found that openness has negative effect onGDP.

Alexandre et. al(2012) investigated if demographic changeplays a role in the dynamic of regional economic growth inBrazil. They used pooled and fixed effect regression model toshow this relation. They found it significant.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 5 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Literature Review

Hussin and Saidin(2012) has examined the impact ofeconomic variables such as foreign direct investment (FDI),openness and gross capital formation to economic growth overthe period 1981- 2008 in ASEAN-4 countries. Using OLSestimation, they found that openness has negative effect onGDP.

Alexandre et. al(2012) investigated if demographic changeplays a role in the dynamic of regional economic growth inBrazil. They used pooled and fixed effect regression model toshow this relation. They found it significant.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 5 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Literature review

Bloom and Finley(2008) examined the role of thedemographic transition in explaining cross-country differencesin economic growth, with a particular focus on East Asia.They showed that with the working-age share beginning todecline in many Asian countries, innovation and flexibility inthe labor market will have required for them to continue toenjoy the high rates of economic growth.

Using cross-country analysis and employing a growthregression for a panel of 78 countries, Bloom and Williamson(1997) examined the effects of demographic transitions oneconomic growth.They found working age people affect GDPgrowth positively and total population growth affectnegatively.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 6 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Literature review

Bloom and Finley(2008) examined the role of thedemographic transition in explaining cross-country differencesin economic growth, with a particular focus on East Asia.They showed that with the working-age share beginning todecline in many Asian countries, innovation and flexibility inthe labor market will have required for them to continue toenjoy the high rates of economic growth.

Using cross-country analysis and employing a growthregression for a panel of 78 countries, Bloom and Williamson(1997) examined the effects of demographic transitions oneconomic growth.They found working age people affect GDPgrowth positively and total population growth affectnegatively.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 6 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Data sources and variables

Table: Data Sources

Variable name Variable Description Source Summary

Growth GDP per capita Annual average change in thelog of real GDP per capita over5 years

WDI 2012

Tropical location Fraction of tropical area PDXLandlocked Landlocked country dummy WikipediaPopulation growth Annual average growth of the

total population over a 5 yearperiod

WDI 2012

log working-age popula-tion to total population

ln(working-age popula-tion/population)

WDI 2012

Life expectancy Life expectancy at birth WDI 2012Population density People per square km. WDI 2012Infant mortality Infant mortality rate WDI 2012

24 countries data has been used from 1971 to 2012.Most of the data has been collected from World Bank(WDI2012).

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 7 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Economic Model

This study has followed the economical model which was derivedby Bloom and Williamson(1997). The model is given as

Economical Model

gy = −λ1y + λ2 ln(L/P) + λ3βX + λ4gpopulation + ε (1)

Wheregy= GDP growth ratey=log GDP per capitaL/P=Ratio of working age population to total populationgpopulation = the growth rate of populationλ′i s are the coefficients

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 8 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical Model

Two types of statistical model was used,

1 Fixed effect model

2 Random effect model

Fixed effect model

We assume that slope coefficients are constant but the interceptvaries over time;

Yit = α1 + α2D2i + α3D3i + · · ·+ β2X2it + β3X3it + · · ·+ uit(2)

where X ’s are the independent variable, Di is the dummy variablesand α, β, λ are the coefficients.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 9 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical Model

Two types of statistical model was used,

1 Fixed effect model

2 Random effect model

Fixed effect model

We assume that slope coefficients are constant but the interceptvaries over time;

Yit = α1 + α2D2i + α3D3i + · · ·+ β2X2it + β3X3it + · · ·+ uit(2)

where X ’s are the independent variable, Di is the dummy variablesand α, β, λ are the coefficients.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 9 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical model

Assumptions of fixed effect model

uit are independent.

E (uit) ∼ N(0, σ2)

Slope coefficients are constant

Random effect model

The random effect model is;

Yit = β1 + β2X2it + β3X3it · · ·+ ωit (3)

Where

ωit = εi+uit

{εi = individual specefic error componentuit = combined time series and cross-section error component

(4)

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 10 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical model

Assumptions of fixed effect model

uit are independent.

E (uit) ∼ N(0, σ2)

Slope coefficients are constant

Random effect model

The random effect model is;

Yit = β1 + β2X2it + β3X3it · · ·+ ωit (3)

Where

ωit = εi+uit

{εi = individual specefic error componentuit = combined time series and cross-section error component

(4)

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 10 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical model

Assumptions for εi and uit

εi ∼ N(0, σ2)

uit ∼ N(0, σ2)

E (εi , uit) = 0 E (εiεj) = 0 i 6= j

E (uituis) = E (uitujt) = E (uitujs) = 0 (i 6= j ; t 6= s).

Assumptions for ωit

E (ωit) = 0

var(ωit) = σ2ε + σ2

u

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 11 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Statistical model

Assumptions for εi and uit

εi ∼ N(0, σ2)

uit ∼ N(0, σ2)

E (εi , uit) = 0 E (εiεj) = 0 i 6= j

E (uituis) = E (uitujt) = E (uitujs) = 0 (i 6= j ; t 6= s).

Assumptions for ωit

E (ωit) = 0

var(ωit) = σ2ε + σ2

u

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 11 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Analysis and FindingsGraphical representation: GDP growth in 1971 and 2012

-5 0 5 10 15 20mean of gdp_growth

United Arab EmiratesThailand

SyriaSri Lanka

South KoreaSingapore

Saudi ArabiaPhilippines

PakistanOman

North KoreaNepal

MalaysiaKuwaitJordanJapanIsrael

IranIndonesia

IndiaHong Kong

ChinaBrunei

Bangladesh

Figure: GDP growth in 1971

-2 0 2 4 6 8Mean of GDP growth

United Arab EmiratesThailand

SyriaSri Lanka

South KoreaSingapore

Saudi ArabiaPhilippines

PakistanOman

North KoreaNepal

MalaysiaKuwaitJordanJapanIsrael

IranIndonesia

IndiaHong Kong

ChinaBrunei

Bangladesh

Figure: GDP growth in 2012

Overall GDP growth of Asia has increased

GDP growth of Saudi Arabia and Iran has decreased

GDP growth of Bangladesh was negative in 1971, but it is now in positive.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 12 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Analysis and ResultsGraphical representation: Realationship of fertility rate and life expectancy with GDP growth

68

1012

14Lo

g G

DP

per

wor

ker

40 50 60 70 80 90Life expectency

Figure: Relation between GDPper worker and Life expectancy

68

1012

14Lo

g G

DP

per

wor

ker

0 2 4 6 8Total Fertility Rate

Log GDP vs Total fertility rate

Figure: Relation between GDPper worker and Fertility rate

GDP growth per worker and Life expectancy is positively related

GDP growth per worker and fertility rate is negatively related

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 13 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Regression ResultsGDP growth rate is independent variable

Table: Explaining cross country economic growth, 1971-2012

(1) (2) (3)Independent variables OLS OLS OLS

log GDP per capita −.381∗∗∗ −0.759∗∗∗ −0.749∗∗∗

(-3.71) (-4.16) (-6.06)Proportion land area in tropics .502 0.00498 1.242∗∗

(1.33 ) (0.01) (2.36)Access to port(landlocked) −1.927∗∗ −2.11∗∗ -0.840

(-2.19) (−2.31) (-0.94)Population growth rate −0.252∗∗ 0.220∗∗

(2.42) (2.03)Log ratio of working age to total population 5.061∗∗

(2.39)Life expectancy -0.587

(-0.95)Infant mortality −0.0196∗

(-1.76)Eastern Asia 0.999∗

(1.76)Southern Asia −2.112∗∗∗

(-3.50)South-eastern Asia -0.968

(-1.51)Constant 7.915∗∗∗ −5.577 10.446∗∗∗

(9.28 ) (-0.65) ( 8.82)

Wald χ2 17.79∗∗∗ 29.11∗∗∗ 49.86∗∗∗

AIC 6026.173 5653.57 5992.815

BIC 6045.731 5692.165 6031.916

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 14 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Regression ResultsInterpretation

OLS-1 represents the relation between GDP growth per capitaand geographic variable. Which shows that the landlockedcountry has a significant effect on GDP growth.

OLS-2 represents the relation between growth in GDP percapita and geographic and demographic variables. Here, wecan see that Landlocked and among 4 demographic variablesthree of them are significant. From demographic variables,growth rate in total population has a negative effect ongrowth in GDP per capita where ratio of working agepopulation to total population has a significant positive effect.

OLS-3 shows the geographic and regional effect on GDPgrowth per capita. From results we see that proportion landarea in tropics has a positive effect. Eastern Asia andsouthern Asia has a significant effect on change in GDPgrowth pattern relative to western Asia.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 15 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Effect of income on life expectancyRegressions results: Life Expectancy is independent variable

Table: The effect of income on life expectancy

(1) (2) (3)Independent variables OLS(Fixed effect) OLS OLS

log GDP per worker −0.0038 0.0527∗∗∗ 0.0359∗∗∗

(0.75[1]) (26.87) (16.89)Dummy 1981-1990 0.052∗∗∗ 0.0562∗∗∗ 0.062∗∗∗

(14.12[1]) (7.82) (9.95)Dummy 1991-2000 0.094∗∗∗ 0.094∗∗∗ 0.104∗∗∗

(23.63[1]) (13.13) (16.76)Dummy 2001-2010 −0.073∗∗∗ 0.126∗∗∗ 0.139∗∗∗

(19.1[1]) (17.46) (22.19)Eastern Asia 0.0517∗∗∗

(8.22)Southern Asia −0.069∗∗∗

(-9.44)South-eastern Asia 0.011∗

(1.72)Western Asia 0

(omitted)Constant 4.233∗∗∗ 3.639∗∗∗ 3.796∗∗∗

(88.42[1]) (194.89) (170.91)

Wald χ2 1363.9∗∗∗ 2147.58∗∗∗

AIC -3411.483 -2139.016 -2402.877

BIC -3387.405 -2114.938 -2364.352

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 16 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

The effect of income on life expectancyInterpretation

1 The period 1971–1980 was taken as base period.

2 There is no effect of income on life expectancy in OLS-1(Fixed effect). Life expectancy increased in period 1981–1990and 1991–2000 but decreased in 2001–2010 relative to baseyear.

3 Log GDP per worker has a significant positive effect on lifeexpectancy in OLS-2. This regression also shows that lifeexpectancy has significantly increased in all period.

4 OLS-3 represents the effect of income on life expectancy fordifferent periods and different regions. It also shows thesignificant positive effect of income on life expectancy.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 17 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Relationship between fertility and incomeRegressions results: Log fertility is independent variable

Table: The effect of income on fertility rate, 1971-2012

(1) (2) (3)Independent variables OLS(Fixed effect) OLS OLSConstant 1.712∗∗∗ 2.379∗∗∗ 3.107∗∗∗

(9.0[1]) (22.87) (25.22)log GDP per worker −0.02 −0.095∗∗∗ −0.159∗∗∗

(0.98[1]) (-8.71) (-13.85)Dummy 1981-1990 −0.231∗∗∗ −0.188∗∗∗

(14.82[1]) (-4.79)Dummy 1991-2000 −0.475∗∗∗ −0.396∗∗∗

(27.59[1]) (-9.92)Dummy 2001-2010 −0.706 −0.613∗∗∗

(36.47[1]) (15.27)Eastern Asia −0.881∗∗∗

(-25.05)Southern Asia −0.369∗∗∗

(-9.04)South-eastern Asia −0.479∗∗∗

(-14.00)Western Asia 0

(omitted)

Wald χ2 425.28∗∗∗ 870.37∗∗∗

AIC -849.5339 -490.999 730.0886

BIC -824.995 1016.071 754.1613

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 18 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Relationship between fertility and incomeRegressions results: Interpretation

1 The above table explains the effect of income on fertilitytaking log fertility as dependent variable.

2 Column–2(OLS-1) shows the fixed effect relation betweenfertility and income per head and four different periods. Herelog GDP per worker is not significant and period 2001–2010 isalso not. But period 1981–1990 and 1991–2000 has a negitiveeffect on fertility. Which means fertility has decreased in theseperiods relative to period 1971–1980.

3 Column–3 (OLS-2) shows the random effect relation betweenlog fertility and income for different periods. This declaresthat income has a significant negative effect on fertility.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 19 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Conclusion

Bloom et. al.(1999) showed that western Asia has achievedHuge change in GDP growth using the data up to 1990.

GDP growth in western Asia has decreased, but increased inother area.

Land area in tropics affected GDP growth negative in previous1990, but in this study it has found positive.

Increase in total population growth showed a negativerelationship with GDP per capita growth.

Increase in working age population growth have a positiveeffect on GDP growth.

Southern Asia is still poorer than other regions.

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 20 / 21

Introduction Literature review Data sources and variables Methodology Analysis and Results Conclusion

Thank You

Md Abu Sayem March 13, 2015 Demographic Change And Economic Transitions in Asia 21 / 21