Income Inequality and “The Spirit Level”

The issue of income inequality is a hot topic. Stories are appearing everywhere, with arguments and counterarguments being thrown around, as people try to decide whether the issue is a red herring or a harbinger of revolutions to come. The Economist provided this graphic to illustrate the concern.

The Gini coefficient is a recognized measure of income inequality with the higher the numbers indicating that more income is collected at the higher levels. Zero would indicate a uniform distribution. Inequality is rising in the countries indicated. Of particular interest is the fact that it is increasing significantly in China and India. There is certainly a story to be told there, but I am sure it is not a simple one. One might suspect that each country is a unique case with its own set of unique conditions.


The most interesting contribution to the debate, although perhaps not ultimately the most enlightening, comes from Richard G. Wilkinson and Kate Pickett who published a book in 2009 titled The Spirit Level: Why More Equal Societies Almost Always Do Better. These authors attempt to come up with a grand theory which seems to couple most of the ills of society to the existence of significant income inequality, or perhaps just inequality in general. Presumably inequality generates a physical stress with deleterious health effects. The book apparently caused a stir in the UK when it was published. It is not readily available in the USA. The reception to their thesis seems to depend on political viewpoint as much as anything. Liberals seem to love the notion. Unfortunately, social scientists have been active in attacking the data sets and methods used, as well as the conclusions.


Consider one of the charts used to make their case. This one involves infant mortality rates by country plotted against a measure of income inequality.

If one glances quickly at the chart one sees a definite correlation between higher income inequality and higher infant mortality. However, if one looks closely there are some significant outliers. Consider Singapore and the USA. They define the low and high points of infant mortality, but they are also the two highest points in terms of income inequality. These two countries clearly fall out of the authors’ grand synthesis. Presumably, the results can be explained by the difference in health care systems operative in the two countries and have nothing directly to do with income inequality. One could make the same healthcare system argument to explain the Japan and USA results. One begins to suspect that each point on that curve has its own national narrative that brings in complicating factors. If one looks at the charts the authors provide at their website, there always seem to be complicating factors which could lead to different interpretations. Many of the trends are those that you would expect from the incidence of poverty. But does that have anything to do with income inequality per se?


Although my progressive leanings originally drew interest in this idea, the more I have thought about it and the more I have stared at the data, the more dubious I have become. Interested people can find out more from the authors’ website, which has more the appearance of a social movement than a scientific data center. References to arguments against their thesis can be found here.