The Impact of Inequality

Richard G. Wilkinson The Impact of Inequality INTRODUCTION R A T H E R T H A N D IS C U S S IN G C O N C E P T S O F F A IR N E S S A S T H E Y M IG H T apply to various aspects of life, I want to draw attention to some of the more important social and health effects that result from different amounts of inequality in modem societies. Most of the evidence comes from comparisons of market democracies or between the 50 states of the United States. As the different amounts of inequality we are compar­ ing are all well within the bounds of what is practically possible, the implications of these comparisons will be relevant to policy. I will start by showing that health and longevity tend to be better in more egalitarian societies. To understand such a relationship means to understand a causal chain that runs all the way from broad issues of social structure right down to the risk factors that affect individual biology. Causality runs from the corrosive effects of inequality on the character of social relations in a society, through the risk factors that modem epidemiology has shown make our health so sensitive to the quality of the social environment, and ends up with the biological effects of social stressors. It is a fascinating journey, showing us how we are all intimately affected by inequality. Essentially, epidemiological research on the determinants of population health has opened up what looks like a new sociology, which will help us to understand ourselves as human beings and what it means to be social. At the same time it takes us back to the political perceptions that have fired radicalism for centuries and made inequality a central theme. Before discussing the evidence, Iwould like first to point out that although we use income distribution as a measure of the amount of social research Vol 73 : No 2 : Summer 2006 711 79 Canada ♦ Sp5 in Australia ♦ ♦ CD Singapore ♦ USA ♦ Denmark 76 ♦ Portugal 75 20 25 30 35 40 45 Gini coefficient (1990-1998) Figure 1 Life Expectancy and Income Distribution in the 21 Richest Countries Data Source: De Vogli et al. (2005). Reproduced with permission from the BMJ Publishing Group. inequality in a society, that does not mean the main effects of inequality are necessarily the direct, socially unmediated effects of different mate­ rial circumstances themselves. Indeed, it seems that the most likely reason income inequality is related to health is because it serves as a proxy for the scale of social class differentiation in a society. It probably reflects not only the scale o f social distances and the accompanying feelings of superiority and inferiority or disrespect, but also, as status differentials increase, how they are likely to become more important than they would be in a more egalitarian society. The effects of social status on health are likely to be exacerbated by the greater insecurities facing those who lose out in the competition for status. A number of theoretical and empirical considerations point clearly in that direction (Wilkinson and Pickett, 2006; Wilkinson, 2005). INCOME INEQUALITY AND HEALTH That health is better in more equal societies is demonstrated by a large body o f evidence. We reviewed some 168 analyses published in peerreviewed journals (Wilkinson and Pickett, 2006). Overall, 78 percent showed at least some statistically significant evidence of a tendency 712 social research for health to be better in m ore equal societies, after the use of w hat­ ever control variables the authors thought appropriate. Comparing just those papers that found no statistically significant supportive evidence w ith those in w hich all associations were significant and supportive, 70 percent were wholly supportive against the rem aining 30 percent, which were unsupportive. If we look at results before the use of control variables and exclude papers in w hich inequality was m easured in areas too small to reflect the overall scale of class inequalities in a soci­ ety, we are left w ith 128 analyses using data on inequality in areas as large as m etropolitan areas, regions, or whole...