امید به زندگی و سرمایه انسانی: شواهدی از انتقال اپیدمیولوژیک بین المللی

Exploiting preintervention variation in mortality from various infectious diseases, together with the time variation arising from medical breakthroughs in the late 1940s and the 1950s, this study examines how a large positive shock to life expectancy influenced the formation of human capital within countries during the second half of the 20th century. The results establish that the rise in life expectancy was behind a significant part of the increase in human capital over this period. According to the baseline estimate, for one additional year of life expectancy, years of schooling increase by 0.17 year. Moreover, the evidence suggests that declines in pneumonia mortality are the underlying cause of this finding, indicating that improved childhood health increases human capital investments.

What causes some countries in the world to remain underdeveloped and what can be done to help those countries escape economic stagnation and poverty? In an attempt to answer such challenging but relevant questions, one part of the literature has focused on the relationship between the health (life expectancy) and the wealth (GDP per capita) of countries. This literature is typically motivated by a strong positive cross-country correlation between health and wealth – the so-called Preston curve (Preston, 1975) – that is to say, healthier countries are also wealthier countries. The current paper continues this line of inquiry. A seemingly important mechanism between health and wealth is the human capital channel: healthier individuals, who expect longer lives, have stronger incentives (and are more able) to acquire human capital skills. While a range of micro studies confirm the relevance of the argument, the conclusions in recent macro-empirical studies are less clear: Acemoglu and Johnson (2007) and Lorentzen et al. (2008).1 Motivated by this puzzle, the paper revisits the relationship between life expectancy and human capital. To investigate the causal effect of life expectancy on human capital, the analysis exploits a large positive shock to life expectancy, which was a part of the international epidemiological transition,2 caused by the breakthrough of antibiotics and new intervention techniques (e.g., malaria eradication). More specifically, the identification strategy relies on the interaction between preintervention variation in the mortality from different infectious diseases and time variation occurring from the new medical technologies in the mid-20th century. Utilizing this interaction as a plausible exogenous variation in life expectancy, the strategy compares years of schooling for cohorts of individuals in countries with a higher burden of infectious diseases to cohorts of individuals in countries with a lower burden of infectious diseases before and after the medical advancements. The analysis finds that countries with higher levels of preintervention infectious disease mortality experienced larger increases in schooling years, suggesting that life expectancy had a positive effect on the accumulation of human capital.3 According to the baseline estimate then for every extra year of life expectancy, years of schooling increase by 0.17 year. In addition, the first-stage results reveal that the new medical technologies were accountable for an increase in life expectancy of about 6.3 years.4 The implication is that the health shock increased years of schooling by 1.1 years. The analysis also establishes that this conclusion is robust to the potential convergence dynamics of schooling, functional form specifications, and so on. The evidence suggests that the decline in pneumonia mortality was the main force contributing to the rise in human capital. As the mortality from that disease is concentrated at earlier ages of life, this instigates the argument that childhood health is a particular important factor in the process of human capital accumulation. However, it should be stressed that the baseline finding seems not to be attributable to the large decline in infant mortality observed over the 20th century. By demonstrating that the effect of health on human capital is positive, the paper delivers new insights to the macro-empirical literature. Deploying the same empirical strategy, Acemoglu and Johnson, 2006 and Acemoglu and Johnson, 2007, henceforth AJ, find no relationship between average years of schooling in the adult population and life expectancy. The fact that their human-capital variable is only marginally affected by the entry of the new and better educated cohorts of individuals could, however, account for this discrepancy.5 Moreover, while data availability on schooling restricts AJ to study the relationship in the 1960–2000 period, this paper exploits newer schooling data at the cohort level based on the 5-year age groups from Barro and Lee (2013), allowing a rigorous empirical analysis with data on years of schooling throughout the 20th century.6 An assessment of other empirical evidence on the quantitative importance of health improvements on human capital at the country level is mixed. On the one hand, some studies have demonstrated positive correlations (e.g., Zhang and Zhang, 2005, Tamura, 2006 and Murphy et al., 2008), and the current paper contributes by establishing a positive link running from health to schooling and hereby supports the conclusions made in that research. On the other hand, the study by Lorentzen et al. (2008), which exploits geographical variables to identify the effect of adult mortality on economic outcomes, finds that mortality has no effect on human capital. Relying on an alternative identification strategy, exploiting the panel data structure to eliminate country fixed effects, and using cohort data on schooling, the analysis here reaches the opposite conclusion: the human capital channel is important in the understanding of how health is related to economic development. This work also relates to the research of Cervellati and Sunde, 2011a and Cervellati and Sunde, 2011b. Their analysis reveals that the impact of life expectancy on GDP per capita is negative and statistical insignificant before the onset of the demographic transition, whereas after its onset the effect is positive and significant. In Cervellati and Sunde (2009), they argue that the relationship between life expectancy and average years of schooling in the population follows a similar pattern. When considering cohort based data on schooling instead, the evidence presented here indicates that health has a positive effect on human capital for all countries in the AJ sample.7 The paper is organized as follows. Section 2 provides a short overview of theories of the human capital channel. Section 3 describes the data. Section 4 outlines the empirical strategy and presents the main findings. Section 5 reports the robustness analysis. Section 6 offers a concluding discussion.

This article has provided evidence on the relevance and importance of the human capital channel: healthier countries with longer-lived populations also spend more (effective) time on schooling and hereby acquire better human capital skills. Using the same empirical strategy as proposed by Acemoglu and Johnson (2007) to identify the effect of life expectancy on human capital, the analysis finds that for one extra year of life expectancy, years of schooling increase by 0.17 year, which implies that the medical advances of this period raised schooling by about 1.1 years. Thus, the bulk of evidence indicates that life expectancy has a positive effect on human capital. Moreover, the mechanism underlying the result seems to be the mortality declines in pneumonia, pointing towards early life health being an important determinant of human capital. Nevertheless, this is far from conclusive evidence on the mechanism and future research should dig deeper into this issue. At this point, however, a discussion of the limitations of the result is warranted. In particular, potential concerns about the empirical approach put forward in other papers should be discussed (Aghion et al., 2010 and Bloom et al., 2013). One concern is that the regressions should have included initial life expectancy.21 However, as shown by Aghion et al. (2010), if the model is in fact misspecified then because of mean reversion in life expectancy this would bias the estimate on life expectancy downwards, which actually strengthens the results presented in the current paper.22 Another important concern, put forward in Bloom et al. (2013), is that the disease-specific mortality rates component in the health shock variable is endogenous to economic development. However, the current type of empirical strategy has the flavor of a difference-in-difference model and other studies following this approach also use preintervention health conditions interacted with the intervention date (e.g., Bleakley, 2007, Bleakley and Lange, 2009, Lucas, 2010 and Bhalotra and Venkataramani, 2012). Furthermore, the similarity in empirical strategy to various micro studies makes it easier to compare the macro and micro-estimates qualitatively, which was the starting point of the current study.