Life expectancy and human capital: long-run evidence from a panel of countries

ABSTRACT

This article examines the long-run impact of life expectancy on human capital investment for a panel of 14 countries over the period 1870–2010. Using recently developed panel time series techniques, we find (i) that life expectancy at birth has a statistically significant long-run effect on schooling and (ii) that long-run causality is unidirectional from life expectancy to schooling.

KEYWORDS:

• Life expectancy
• human capital
• coefficient heterogeneity
• cross-sectional dependence

JEL CLASSIFICATION:

• I15
• O11
• N30
• C23

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ One of the theoretical rationales for the hypothesized positive relationship between life expectancy and education is that gains in adult life expectancy increase the horizon over which investments in schooling will be paid off; an increase in the probability of surviving in the future therefore raises the incentive for individuals to invest in their own education (e.g. Kalemli-Ozcan, Ryder, and Weil 2000). Another rationale is that parents make trade-offs between the quantity and the quality of their children within their given budget constraints; thus, by reducing parental costs of educating each surviving child, a decline in child mortality may lead to greater investments in children’s human capital (e.g. Kalemli-Ozcan 2002).

² If the variables are I(1) and cointegrated, then an ARDL can be written as a conditional error correction model (ECM). By construction, the ECM version and the original ARDL model yield identical estimates of the long-run parameters and their standard errors.