Is there a life expectancy Preston Hypothesis for Africa?

Abstract

Background: The inconclusive evidence surrounding the Preston hypothesis and scarcity of findings in Africa motivated the focus of this study. This paper examined the Preston hypothesis for all countries in Africa, over the period 1960-2018. Findings are shown for the effect of per capita income and income-proxy variables on life expectancy.

Methods: Using a two-way fixed effect model, bivariate and multivariate regression were used to determine the strength of income and its proxy variables in explaining health. The direction of causality between income and health was also examined using the Dumitrescu & Hurlin (DH) panel granger causality test. The data used were sourced from the World Development Indicators (WDI) provided by the World Bank.

Findings: The study revealed the existence of the Preston hypothesis but only in the bivariate model. Findings showed stronger effect of the proportion of Investment to GDP ratio than per capita income and other income proxy variables in explaining life expectancy. The multivariate result also showed stronger effect of investment to GDP ratio, immunization rate and fertility rate on life expectancy than per capita income and proxy variables for income. The DH test revealed reverse causality between per capita income and life expectancy.

Conclusion: There is weak evidence of the existence of the Preston Hypothesis in Africa. The gains in life expectancy are largely attributed to investment, immunization and fertility rate than per capita income. Efforts to improve health in Africa should give top priority to raising investment, increase in immunization rate and reduction in fertility rate. This focus should rank high in policy maker’s agenda.