Retesting the income inequality hypothesis: Pooled time-series-cross-section regression with a new statistical case selection method

Abstract

The Wilkinson hypothesis, which posits that it is income inequality, rather than economic growth, that determines population health in wealthy societies has been widely debated over decades. This article, with an original case selection process, retests the controversial hypothesis with the time-series-cross-section (TSCS) regression model. The main findings are that disposable income inequality is not statistically significantly associated with life expectancy, and its influences on the health of the elderly population are counterintuitively beneficial. Infant mortality rate is the only health indicator that both has the statistically significant association with income inequality and corresponds with the Wilkinson hypothesis. Finally, the relationship between income inequality and population health is not as simple as the hypothesis suggests and found to be dynamic and complex depending on what health indicator is used for the test.