Vo Tan Thanh Diep
Vo Tan Thanh Diep
Vo Tan Thanh Diep
Paper submitted to
The ninth Vietnam Economists Annual Meeting
Da Nang City, 11 – 12th August 2015
Abstract
This study has demonstrated new evidence sustaining the idea that variation in
demographic factors is an important determinant of growth in per capita income. Using an
panel dataset from 1990 to 2013 at the country-level in the Southeast Asia, this study is
conducted to analyze the following two key areas. First, the prevalent of a number of the
demographic factors impact on the economic growth. Second, the bi-directional causality is
revealed among: (i) the population growth and the economic growth; and (ii) the life
expectancy and the economic growth. Furthermore, the two new econometric techniques,
Driscoll and Kraay estimation, and structural equation model are applied. Finding from this
study manifests the substantial effects on the economic growth from: the positive effect of
population growth, the significant boost of the life expectancy, and the pressure of dependent
population. In addition, it is a worth noting in its kind of literature due to the confirmation of
the significant inverse effects of the economic growth on the two demographic factors.
2. Literature review
2.1. Demographic factors
There has been a long time since most of economists focused on population growth and
population size to evaluate the effects of demography. It can be clarified by the analysis of
the demographic transition. In general, it is exhibited the change over the birth rate and death
rate, and the trend of population growth. Nevertheless, in recent times, it is a concentration of
the age structure, beside the population growth, and the population size. It is not only about
the contribution of each stage of life, but also about the different behavior and contribution
each age group that could lead to the difference in economic performance. The high
proportion of young people will create a high economic burden due to their small savings,
and a great budget being spent for in many other life aspects such as health, education and so
on. However, this could be a potential labor source for country in the future, when population
growth falls based on the drop of fertility rate, the nation will get benefit of high proportion
of working-age and low dependence age (the young and the elderly). Nevertheless, when the
countries keep improving the living standard, life expectancy will increase and the ratio of
elder in population will be higher. The resources for elderly will be greater than the product
they create, and then restrict the income per capita of the nation (Bloom et al. 2003). In such
explanation, the age structure could be a reasonable proxy to distinguish the countries that
have same population size and growth.
Another examined factor is life expectancy. As the definition of United Nations (2013),
life expectancy at specific age is the expected years that a person at that age could live base
on the assumption that age-specific mortality levels is constant. In this paper, life expectancy
is the short word for life expectancy at birth which was the number of years that one newborn
could live if the age-specific mortality rates was not change. Due to the fact that longevity
reflects the mortality as well as the health of country, life expectancy could be an important
demographic proxy in the process of economic development.
2.2. Demographic factors and economic growth
The characterization of modeling the relationship between demographic factors and
economic development has been challenged and debated since the time of Malthus (1798). It
has portrayed the influential dimension of the demographic factor on the economic growth by
defining the Malthusian stagnation mechanism. Nevertheless, this has been changed in recent
year due to the changeable idiosyncratic of demographic factors, the economic situation, or
even the changeable modern conception. Based on this way of analyzing, a number of
researches has been constructed as a summarized framework for the related theories which
includes: (i) the traditional “Malthusian regime” that focused on the stagnation from the
relationship between the controlling process of demographic transition (population growth)
and the food issue; (ii) the “post-Malthusian” state which implied the slight impact of greater
population on the growth; (iii) the “modern approach” postulates the steady growth
mechanism with a group of number of factors that impact on growth.
2.2.1. The perspective of Malthusian Regime
The main idea of this view is population without control increases geometrically but
food supply only rise arithmetically because of limit of some factor of production such as
land and technology. As stated by Malthus (1798), the economic performance will be higher
with small population, on the one hand, high income per capita have the supportive function
to the expansion of population. On the other hand, the higher population with circumscribed
resource will lead to the increase in price, it will have a preventable effect to the standard of
living. The Malthusian theory implies that population size will be self-adjusted to adapt with
the supply of production. The raise in product will counterbalance by increasing in population
size, thus the technologically superior countries only had higher population density and the
economic performance did not reflect technological progress.
Although the recent paper of Acemoglu and Johnson (2007) provides an opposite
finding to almost previous studies: the increase in life expectancy contributes to the reduction
in economic growth based on cross-country database over the 1940-2000 periods. The
positive relationship between life expectancy and economic growth has been confirmed in a
number of empirical studies with different econometric method such as Barro and Lee
(1994), Caselli et al. (1996), Gallup and Sachs (2000), de la Croix et al. (2009). Moreover,
the study of Cervellati and Sunde (2011) which use framework of Acemoglu and Johnson
(2007) as benchmark revealed that the improvement in life expectancy would leads to the
increase in income per capita. From these finding, life expectancy is expected to positively
affect the economic growth.
H1b. Life expectancy has positive effect on economic growth in Southeast Asia
Age structure was considered as the new aspect in the investigation of effect of
demography on development. The finding from studies of Crenshaw et al. (1997),
Azomahou and Mishra (2008) using cross-national dataset suggested the distress of
dependence age group to economic growth. Moreover, these findings was maintained
in the case of Asia countries since both the young and old dependence fraction. On the
other hand, the increase in portion of working age population seemed to stimulus the
economic growth in recent empirical studies (Azomahou and Mishra, 2008; Bloom
and Finlay, 2009). In conclusion, the economic growth could get more benefit from
both the raise in proportion of working age population and the reduction in proportion
of dependent population.
H1c. Age dependency ratio has negative effect on economic growth in Southeast Asia
Two-way relationship
The previous research has been examined the simultaneous relationship between
population growth and economic growth to confirm the impact of the former on the
latter. Particularly, the instrument variable technique was observed in the studies of
Bloom and Williamson (1998), Bloom and Finlay (2009), Bloom et al. (2010) to deal
with this problem. Similarly, the endogenous problem which results from two-way
relationship between life expectancy and economic growth was considered in a
number of empirical review such as Acemoglu and Johnson (2007), Ashraf et al.
(2008), and Cervellati and Sunde (2011). However, most of papers has just handled
the simultaneous problem to robust the effect of demographic factors on economic and
disregard of the possible impact from the opposite direction. Therefore, the objective
of this study is not only examining the impact of demographic factors on economic
growth but also clarifying the bi-directional causality of this relationship.
H2a. There is an existence of the bi-directional causality between population growth
and economic growth in Southeast Asia
H2b. There is an existence of the bi-directional causality between life expectancy and
economic growth in Southeast Asia
Figure 1. The conceptual framework
In this case, the Driscoll and Kraay estimation obtains the covariance matrix that deals
with the heteroskedasticity, the autocorrelation and the cross-sectional dependence issues.
(Hoechle, 2007)
3.2.2. The bi-directional causality
The current literature has just concerned the presence of the potential
endogenous problem in the relationship between demographic factors and economic
growth. Nevertheless, these papers have not clarified the possibly inverse effect of
economic growth on several factors of demographic factors (population growth, or life
expectancy). It means that this study has been concerned the analysis in both two way
of the relationship – bi-directional causality – formulated as following equations:
Model 5:
Ln (GDP per capita) it = 𝛽 0 + 𝛽 1 (Population growth) it + 𝛽 2 Ln (Life expectancy) it
+ 𝛽 3 Ln (Life expectancy 1990) it + 𝛽 4 (FE_MA) it + 𝛽 5 (Trade) it
+ 𝛽 6 (GCF) it + 𝛽 7 (Transition) it + 𝛽 8 (Civil liberties) it
+ 𝛽 9 Ln (Density 1990) it + 𝜀 it
Model 6:
Population growth it = 𝛽 0 + 𝛽 1 Ln (GDP per capita) it + 𝛽 2 Ln (Density) it
+ 𝛽 3 (Labor female) it + 𝛽 4 (Urban) it + 𝛽 5 (Food) it + 𝜀 it
Model 7:
Ln (Life expectancy) it = 𝛽 0 + 𝛽 1 Ln (GDP per capita) it + 𝛽 2 (Population 65) it
+ 𝛽 3 (Death) it + 𝛽 4 (Immunization) + 𝛽 5 (Fertility) it
+ 𝛽 6 (Human capital) it + 𝜀 it
Where,
(i) Dependence variable: log of GDP per capita, the population growth, and
life expectancy are presented respectively in the model 5, 6, and 7.
This study has been analyzed the problem of bi-directional causality by applying
the structural equation model (SEM). This modeling equations has been considered
widely as a platform of the econometric technique to quantify the theoretical issue, the
causal relationship, or the mechanism of transmission. Its procedure may provide the
crucial framework which specifies the influential trait based on the view of the causal
mediation mechanism, the simultaneous or non-recursive structural equation.
(Pugesek et al., 2003; Markus, 2012; and StataCorp., 2013)
The idiosyncrasy of this method is the joint determination of the dependent
variables, and a simultaneous equation. Usually, it can be dealt with the biasness of
the endogenous problem, the structural error of the equation, and the correlation
between the explanatory variables and the residual may cause the inconsistent
estimation in the traditional OLS model (Wooldridge, 2012).
Concerning the following equation of the simultaneous equation:
y1 12 y2 11x1 12 x2 1 (1)
y2 21 y1 22 x2 x3 2 (2)
In this case, the system of equation has presented the joint determination of the
endogenous variables y1 and y2 by a group of exogenous variables - x1, x2 , and x3 . The
determination of the endogenous variables has been demonstrated by the following
mechanism of the path analysis (StataCorp., 2013):
Figure 2. The non-recursive mechanism.
In this model, the “loop” is represented for the non-recursive relationship (the bi-
directional causality) of the two endogenous variables. (StataCorp., 2013).
4. Empirical Results
4.1. Summary statistic
Our sample in the period of 1990-2013 covers the period when most of the Southeast
Asian countries have experienced the demographic transition. According to Table 1, the mean
of GDP per capita based on purchasing power parity, which is our response variable, is
18725.690 dollars with a standard deviation of 25472.700 dollars. This quite high standard
deviation presents the fact of the relatively high income inequality in this area. The
descriptive statistics of demographic factors are reported in Panel B, there is a relative
variation in population growth and life expectancy. Besides, the components of population
provides general picture about demography in Southeast Asia over the 1990-2013 period. As
could be seen from Panel B, the working-age obtains a high proportion from total population
- 62.3% - which could be a good signal for motivating the growth process. Furthermore, the
portion of young people is fairly high - 33.005% - in comparison with total population. The
other proxy of young population - Dependence young shows that two adults will have to be
responsible for one child. Although this age group could be an encouragement for the labor
force in the future, a large burden in present time could damage the economy in short run.
Last but not least, two variables which illustrate the fraction of old group is considerably
small, the maximum value of it is only 10.671%. This age structure may reflect the fact that
most of the countries in sample still have been obtaining the demographic bonus according to
the definition of United Nations (2004).
Table 1. Summary statistics. The table provides summary statistics for the variables in the full
sample.
Mean Std. dev. Min Max No. of obs.
Panel A - Dependence variable
GDP per capita 18725.690 25472.700 1011.094 78958.090 238
Pane B – The demographic factors
Population growth 1.865 0.891 -1.476 5.322 250
Life expectancy 69.282 6.885 48.480 82.346 240
Population 0_14 33.005 8.411 15.680 50.032 250
Population 65 4.690 1.848 1.915 10.671 250
Population 15_64 62.305 6.994 47.521 73.783 250
Dependence old 7.389 2.282 3.280 14.490 250
Dependence young 55.246 20.470 21.290 105.283 250
Panel C - Other control variables
Transition 0.360 0.481 0.000 1.000 250
FE_MA 71.781 15.786 47.826 97.494 240
Death rate 6.424 2.453 2.758 15.709 240
Fertility 3.122 1.468 1.150 7.122 239
Labor female 57.843 14.481 24.600 80.100 240
Urban 3.378 1.437 -1.476 7.164 240
Human capital 2.219 0.400 1.590 2.965 198
Civil liberties 4.614 1.153 3.000 7.000 241
Immunization 82.513 18.311 18.000 99.000 238
Density 720.682 1838.379 18.390 7813.857 250
Food 98.665 48.117 27.310 495.070 250
Source: Author’s analysis
Moreover, the components of population provides more interesting results. The fraction
of working-age population (Population 15_64) from model (4) shows a statistically
significant and positive effect to per capita output. This finding is compatible with the
argument of Crenshaw et al. (1997) and Azomahou and Mishra (2008) that the improvement
in participation rate of labor force would directly lead to a raise in economic growth. In
contrast, it is apparent from model (2) and (3) that the unfavorable impact (significant at 1 per
cent) of the dependent population in comparison with total population (Population 0_14,
Population 65_above) and the working-age population (Dependence young, Dependence
old). This result confirms the results of Bloom et al. (2010) about negative effect of both
young and old dependence on short-term growth. Combining with the data from overview
section, it is apparent that the raise in working-age population by 8% has undoubtedly
improved economic performance in Southeast Asia during the period of 1990-2013.
Moreover, the improvement could be resulted from the reduction in young dependency by
more than 10%. Despite the slight increase in the old dependency ratio, its impact may results
in a huge burden for the economy due to the significantly large marginal effect. In particular,
the magnitudes of these coefficient are noticeable, the distress of older people which is
illustrated by coefficient of Population 65_above (-0.533) and Dependence old (-0.322) is
considerably greater than the one from youth which could be shown in coefficient of
Population 0_14 (-0.106) and Dependence young (-0.022). This finding could demonstrate
that the old population is becoming a pressure in Southeast Asia in future unless the
government could make a suitable pension policy and health care system.
Similarly, the estimates also confirm the hypothesis about bi-directional causality in the
relationship between life expectancy and GDP per capita. The results show that GDP per
capita, as expected, affect life expectancy directly, although the magnitude of this impact is
fairly small when the one percent raise in GDP per capita only contribute to 0.002%
improvement in life expectancy. Contrariwise, the positive impact of life expectancy on GDP
per capita is considerably high when one percent increase in life expectancy would be
associated with an 8.682% increase in income per capita. This finding confirms the positive
influence of life expectancy on economic growth in previous studies (Barro & Lee, 1994;
Gallup & Sachs, 2000; Ashraf et al., 2008).
5. CONCLUSIONS
Using a panel dataset from 1990 to 2013 at the country-level in the Southeast Asia
(classified by United Nations, 2015), findings from this study provides the empirical evidence
of significant effect on the economic growth from the three key following areas. First, there
is the positive effect of the population growth. Second, the life expectancy presents a
substantial boost to the growth in income per capita. Third, the significant results of age
structure are presented, including: (i) the negative effect dependence population (proxied by
population 0_14, and population 65) and the dependence working-age structure (proxied by
dependence young, and dependence old); and (ii) the positive effect of working age
population (proxied by population 15_64). These findings could infer that the old population
is becoming a pressure in Southeast Asia in future unless the government don’t make a
suitable pension policy and health care system.
In addition, this study has demonstrated the significant inverse effects of the
economic growth on the two following demographic factors. First, the negative effect
of GDP per capita on the population growth is present which is reflected by the impact
of different population policy control in the Southeast Asia. Second, the GDP per
capita affects positively on the life expectancy. In general, these findings have filled
the gap of current literature by confirms the significant presence of the bi-directional
causality between the demographic factors (population growth, and life expectancy)
and the economic growth.
From these results, the governments should observe the experience of population policy
of other countries and concern different policies due to the stage of demographic transition in
each countries. Moreover, there could be certain tradeoffs between the cost for current
population burden and future benefit from huge working age group, and policymakers should
consider carefully these costs and benefits of in order to maintain the sustainable growth and
ensure the benefit for the next generation. In addition, the sustainable pension system and the
improvement in statutory retirement age were some considered decisions that Southeast
Asian government should endeavor to prepare for the potential risk from aging population.
Finally, the pronounced contribution of health to economic growth process was confirmed by
the significantly positive impact of life expectancy on GDP per capita in Southeast Asia. On
the other hand, the significant relatively small effect of output per capita on life expectancy
may reflect the incipient health services in this area. In order to preserve the positive impact
of life expectancy on income per capita and prevent the pressure from aging population, the
measures to improve life expectancy should more focus on the child and working-age health.
Reference
AbouZahr, C., & Royston, E. (1991). Maternal mortality: a global factbook.
Acemoglu, D., & Johnson, S. (2007). Disease and Development: The Effect of Life
Expectancy on Economic Growth. Journal of Political Economy, 115(6).
Ashraf, Q. H., Lester, A., & Weil, D. N. (2008). When does improving health raise
GDP? (No. w14449). National Bureau of Economic Research.
Azomahou, T., & Mishra, T. (2008). Age dynamics and economic growth: Revisiting the
nexus in a nonparametric setting. Economics Letters, 99(1), 67-71.
Barlow, R. (1994). Population growth and economic growth: Some more
correlations. Population and Development Review, 153-165.
Barro, R. J., & Lee, J. W. (1994, June). Sources of economic growth. InCarnegie-Rochester
conference series on public policy (Vol. 40, pp. 1-46). North-Holland.
Barro, R. (1996). Health and economic growth. Harvard University, Cambridge, MA.
Bloom, D. E., & Finlay, J. E. (2009). Demographic change and economic growth in
Asia. Asian Economic Policy Review, 4(1), 45-64.
Bloom, D. E., Canning, D., & Finlay, J. E. (2010). Population aging and economic growth in
Asia. In The Economic Consequences of Demographic Change in East Asia, NBER-
EASE Volume 19 (pp. 61-89). University of Chicago Press.
Bloom, D. E., & Williamson, J. G. (1998). Demographic transitions and economic miracles
in emerging Asia. The World Bank Economic Review, 12(3), 419-455.
Bloom, D. E., Canning, D., & Malaney, P. N. (2000). Population dynamics and economic
growth in Asia. Population and Development Review, 257-290.
Bloom, D. E., Canning, D., & Sevilla, J. (2004). The effect of health on economic growth: a
production function approach. World development, 32(1), 1-13.
Bloom, D., Canning, D., & Sevilla, J. (2003). The demographic dividend: a new perspective
on the economic consequences of population change. Rand Corporation.
Caselli, F., Esquivel, G., & Lefort, F. (1996). Reopening the convergence debate: a new look
at cross-country growth empirics. Journal of economic growth, 1(3), 363-389.
Cervellati, M., & Sunde, U. (2011). Life expectancy and economic growth: the role of the
demographic transition. Journal of Economic Growth, 16(2), 99-133.
Coale, A. J., & Hoover, E. M. (1958). Population growth and economic development in low-
income countries: a case study of Indias prospects.
Crenshaw, E. M., Ameen, A. Z., & Christenson, M. (1997). Population dynamics and
economic development: Age-specific population growth rates and economic growth
in developing countries, 1965 to 1990. American Sociological Review, 974-984.
Cuberes, D., & Teignier, M. (2014). Gender inequality and economic growth: a critical
review. Journal of International Development, 26(2), 260-276.
De la Croix, D., & Licandro, O. (1999). Life expectancy and endogenous growth. Economics
Letters, 65(2), 255-263.
De la Croix, D., Lindh, T., & Malmberg, B. (2009). Demographic change and economic
growth in Sweden: 1750–2050. Journal of Macroeconomics, 31(1), 132-148.
Fogel, R. W. (1994). Economic growth, population theory, and physiology: the bearing of
long-term processes on the making of economic policy (No. w4638). National Bureau
of Economic Research.
Frankel, J. A., & Romer, D. (1999). Does trade cause growth?. American economic review,
379-399.
Gallup, J. L., & Sachs, J. D. (2001). The economic burden of malaria. The American journal
of tropical medicine and hygiene, 64(1 suppl), 85-96.
Galor, O. (2011). Unified growth theory. Princeton University Press.
Hoechle, D. (2007). Robust standard errors for panel regressions with cross-sectional
dependence. Stata Journal, 7(3), 281.
Kelley, A. C., & Schmidt, R. M. (1995). Aggregate population and economic growth
correlations: the role of the components of demographic change.Demography, 32(4),
543-555.
Kormendi, R. C., & Meguire, P. G. (1985). Macroeconomic determinants of growth: cross-
country evidence. Journal of Monetary economics, 16(2), 141-163.
Kuznets, S. (1960). Population change and aggregate output. In Demographic and economic
change in developed countries (pp. 324-351). Columbia University Press.
Lutz, W., & Qiang, R. (2002). Determinants of human population growth.Philosophical
Transactions of the Royal Society B: Biological Sciences,357(1425), 1197-1210.
Malthus, T. R. (1798). An essay on the principle of population. Oxford University Press
(reprinted in 1993), Oxford.
Markus, K. A. (2012). Principles and Practice of Structural Equation Modeling by Rex B.
Kline. Structural Equation Modeling: A Multidisciplinary Journal, 19(3), 509-512.
McDonald, S., & Roberts, J. (2002). Growth and multiple forms of human capital in an
augmented Solow model: a panel data investigation. Economics letters,74(2), 271-
276.
Poland, G. A., Tosh, P., & Jacobson, R. M. (2005). Requiring influenza vaccination for
health care workers: seven truths we must accept. Vaccine,23(17), 2251-2255.
Preston, S. H. (1975). The changing relation between mortality and level of economic
development. Population studies, 29(2), 231-248.
Pugesek, B. H., Tomer, A., & Von Eye, A. (Eds.). (2003). Structural equation modeling:
applications in ecological and evolutionary biology. Cambridge University Press.
Ross, C. E., & Wu, C. L. (1995). The links between education and health.American
sociological review, 719-745.
Sanderson, W. C., Lutz, W., & Scherbov, S. (2013). The End of World Population Growth in
the 21st Century: New Challenges for Human Capital Formation and Sustainable
Development. Routledge.
Sarafidis, V., & Wansbeek, T. (2012). Cross-sectional dependence in panel data
analysis. Econometric Reviews, 31(5), 483-531.
Shaw, J. W., Horrace, W. C., & Vogel, R. J. (2005). The determinants of life expectancy: an
analysis of the OECD health data. Southern Economic Journal, 768-783.
Soares, R. R. (2007). On the determinants of mortality reductions in the developing
world. Population and Development Review, 33(2), 247-287.
StataCorp. 2013. Stata 13 Base Reference Manual. College Station, TX: Stata Press.
United Nations (2013). World Population Ageing 2013. New York: United Nations.
United Nations (2015). World Population Prospects: The 2015 Revision. New York: United
Nations.
Wei, Z., & Hao, R. (2010). Demographic structure and economic growth: Evidence from
China. Journal of Comparative Economics, 38(4), 472-491.
Wooldridge, J. (2012). Introductory econometrics: A modern approach. Cengage Learning.