Intended for healthcare professionals


Greater equality and better health

BMJ 2009; 339 doi: (Published 11 November 2009) Cite this as: BMJ 2009;339:b4320

Life expectancy, infant mortality and inequality - once again

From the standpoint of the empirical social scientist, working in the
field of cross-national development research, Dr. Wilkinson’s hypotheses
about the causal trade-off between inequality and low health, manifesting
itself in various social indicators, such as life expectancy and infant
mortality, sound reasonable, but are methodologically and statistically
seriously flawed.

Dr. Jason Beckfield from the Department of Sociology at Chicago
University (2006), like myself a scientist from the empirical, macro-
quantitative current of social science, recently came to the conclusion
that, I quote:

“This study replicates previous work using a larger sample (692
observations from 115 countries over the 1947-1996 period), a wider range
of statistical controls, and fixed-effects models that address
heterogeneity bias. The relationship between health and inequality shrinks
when controls are included. In fixed-effects models that capture
unmeasured hetero- geneity, the association between income inequality and
health disappears. The null findings hold for two measures of income
inequality: the Gini coefficient and the share of income received by the
poorest quintile of the population. Analysis of a sample of wealthy
countries also fails to support the hypothesis”

Much of the research, published by Dr. Wilkinson over the past
decades, looks merely into the bi-variate correlations between inequality
and social performance at the level of OECD democracies. But currently,
global social science now has data about inequality from practically all
over the globe at its disposal.

So do Dr. Wilkinson’s hypotheses hold when we look, for example at a
global sample, which includes the low life expectancies, high infant
mortalities and devastating rates of alcoholism, combined with relatively
low (but rising) inequalities in several former communist countries after
the transformation? Is “equality” associated in a causal way with “good
health” or rather alcoholism? High taxes, low economic growth? These are
just provocative questions, to which quantitative social science has to

Limiting his research to relatively small samples is error number 1
in Dr. Wilkinson’s approach. Error number 2 is that Professor Wilkinson
overlooks the curve-linear trade-off between development levels and
inequality levels. This effect is known to economists and social
scientists ever since the path-breaking article by Professor Simon
Kuznets, Nobel laureate in economics, first published in 1995. A human
being’s intelligence does not depend on the size of the big toe, once we
control for the age factor.

Error number 3 consists in overlooking what colleague Joshua
Goldstein, back in 1985, called the “plateau curve of basic human needs”.
Like inequality, life expectancy depends on development levels, and the
trade-off is again in the shape of an inverted “U”.

I provide the readers of this journal with some of the cross-national
evidence which would be available today on the subject. Table 1 shows the
sometimes powerful effects at the level of the OECD countries, still
vindicating Wilkinson’s approach:

Table 1: Perason Bravais product moment correlations of the difference in incomes between the richest 20% and the poorest 20% (quintile share) in the OECD countries and in the world system

Source: our own calculations with the standard international data, downloadable from; SPSS XV, Innsbruck University

Table 2 now shows the relationship between inequality and the life
quality variables, once we keep the curve-linear effects described by
Kuznets (1955) and Goldstein (1985) constant. What happens is that several
effects are reduced, and explain just around 10% of the variance of the
life quality variables at the level of the world system.

Table 2: partial correlations (keeping constant the nat. logarithm of GDP per capita at purchasing power parity rates and its square) of the difference in incomes between the richest 20% and the poorest 20% (quintile share) in the OECD countries and in the world system

I think a way forward in this debate would be a true dialogue between
medicine, public health and the social sciences. Neo-classical economists
like Barro, Durlauf with associates, and Sala-i-Martin, and world system
scholars, like Herkenrath and Bornschier, who stress the importance of the
globalization factor in blocking long-run socio-economic development, are
correct in showing that nowadays social science has to take a variety of
factors into account to explain the paths of socio-economic development
and decay.


1. Barro R. J. and Sala-i-Martin X. (2003), ‘Economic Growth’.
Cambridge, MA: MIT Press, second edition.

2. Beckfield J. (2006), ‘Does Income Inequality Harm Health? New Cross-
National Evidence’ Journal of Health and Social Behavior, Vol. 45, No. 3
(Sep., 2004), pp. 231-248

3. Goldstein J. S. (1985), Basic Human Needs: The Plateau Curve. World
Development, 13(5), 595 - 609.

4. Durlauf St. N., Kourtellos A., Tan Ch. M. (2008); Are any Growth
Theories Robust? The Economic Journal, 118(1), 329–346.

5. Herkenrath M. and Bornschier V. (2003), “Transnational Corporations in
World Development – Still the Same Harmful Effects in an Increasingly
Globalized World Economy?” Journal of world-systems research, ix, 1,
winter 2003, 105–139,, issn 1076–156x

6. Kuznets S. (1955), 'Economic Growth and Income Inequality' The American
Economic Review, 45, 1: 1 - 28.

7. Kuznets S. (1976), Modern Economic Growth: Rate, Structure and Spread.
New Haven, CT: Yale University Press.

8. Sala-I-Martin X.; Doppelhofer G. and Miller R. I. (2004), ‚Determinants
of Long-Term Growth: A Bayesian Averaging of Classical Estimates (BACE)
Approach.’ American Economic Review, Sep 2004, Vol. 94 Issue 4, p813-835.

9. Tausch A. and Prager F. (1993), ‘Towards a Socio-Liberal Theory of
World Development’. Basingstoke and New York: Macmillan/St. Martin's

10. Wilkinson R. G. (1992), ‘Income Distribution and Life Expectancy’ BMJ,
304, 6820, 165-168

11. Wilkinson R. G. and Picket K. E. (2006), ‘Title: Income inequality and
population health: A review and explanation of the evidence’, 62, 7, 1768-
1784, April

Data Sources:

All the variables, based on such sources as UNDP, are contained in:

Competing interests:
None declared

Competing interests: No competing interests

02 March 2010
Arno Tausch
Adjunct Professor of Political Science
Innsbruck University, Austria