Judge K, Mulligan J-A, Benzeval M. Income inequality and population health. Soc Sci Med 1998; 46:567-79.

Judge, concerned that the data on the income inequality health relationship has been influential and has been cited with confidence, continues his questioning of the relationship. The paper reviews published studies of the relation between income distribution and measures of health among countries, where the income data sets are not collected uniformly. It neglects all the studies of comparable data sets done within countries. The reviewed studies are faulted for using poor quality data, not adjusting for national incomes, and not adjusting incomes for taxes, and transfer to yield measures of equivalent disposable income. The authors then do their own analysis using data for 16 OECD countries, adjusting for household composition, year of data and including the effects of taxes and benefits. They find the expected relationship, looking at partial correlations, though not as high as in some of the reviewed studies. They conclude: "we cannot sate with confidence that income inequality is not associated with national health. ......results in this paper provide only very modest support for the view that income inequality is associated with variations in average levels of national health among rich industrial nations." Wilkinson's letter in response (see below) points out their exclusion of studies within countries where comparable data are available that demonstrate the relationship. He points out that their own analyses show statistically significant correlations, even with the small number of cases. His study published with McIsaac cited above deals with the more recent data from the Luxembourg Income Study, and points out the difficulties of non-response rates, concentrated among the poor and rich. Only two studies out of 21 available, fail to confirm the hypothesis, and those are based on data which can explain the reasons. Then Judge's reply (see below) to Wilkinson's letter questions further why the relationship didn't show up on some cross-national data sets which, they state, was the purpose of their study. They go on to consider the US data, and wonder if the findings could be the result of more general societal characteristics. They wish to emphasize the inter-relatedness of different dimensions of poverty and social inequality per se rather than focus particular attention on the distribution of income.

Abstract 

A number of studies have suggested that inequalities in the distribution of income may be an important cause of variations in the average level of population health among rich industrial nations. However, what is missing from the debate so far is any systematic review of evidence about the relationship between different measures of income distribution and indicators of population health. This paper aims to bridge that gap. First, it summarizes the recent English language literature on this topic and illustrates the methodological problems that weaken the inferences that can be derived from it. Secondly, it presents new empirical estimates of the relationship between different measures of income distribution, infant mortality and life expectancy based on the most authoritative data published to date. In contrast to most earlier studies, we find very little support for the view that income inequality is associated with variations in average levels of national health in rich industrial countries. Some possible explanations for these differences are outlined.

Keywords

• health inequalities
• income
• income distribution
• income inequality
• inequality
• infant mortality
• international
• life expectancy
• mortality