Inequality and mortality: Long-run evidence from a panel of countries

https://doi.org/10.1016/j.jhealeco.2006.07.003Get rights and content

Abstract

We investigate whether changes in economic inequality affect mortality in rich countries. To answer this question we use a new source of data on income inequality: tax data on the share of pretax income going to the richest 10% of the population in Australia, Canada, France, Germany, Ireland, the Netherlands, New Zealand, Spain, Sweden, Switzerland, the UK, and the US between 1903 and 2003. Although this measure is not a good proxy for inequality within the bottom half of the income distribution, it is a good proxy for changes in the top half of the distribution and for the Gini coefficient. In the absence of country and year fixed effects, the income share of the top decile is negatively related to life expectancy and positively related to infant mortality. However, in our preferred fixed-effects specification these relationships are weak, statistically insignificant, and likely to change their sign. Nor do our data suggest that changes in the income share of the richest 10% affect homicide or suicide rates.

Introduction

Do changes in economic inequality lead to changes in mortality rates? More than 100 articles on this question have been published over the past two decades, but no consensus has emerged (Lynch et al., 2004a). One major reason has been the paucity of reliable historical data on income inequality. As a result, most studies have examined the relationship between inequality and mortality at a single point in time. Because income inequality and mortality are likely to have common causes that cannot all be measured, the cross-sectional relationship between inequality and mortality is unlikely to provide an unbiased estimate of how changes in income inequality affect mortality.

We investigate this issue using a new source of data on economic inequality: the share of personal income received by the richest 10% of adults in Australia, Canada, France, Germany, Ireland, the Netherlands, New Zealand, Spain, Sweden, Switzerland, the UK, and the US. We have annual data covering an average of 62 years per country. As a result, we can control both year and country fixed effects, thereby holding constant both stable country-to-country differences and annual changes in mortality that affect all countries in the same year.

Deaton (2003) and Lynch et al. (2004a) have recently surveyed the literature on economic inequality and mortality. Both reviews conclude that although there are plausible reasons for anticipating a relationship between inequality and mortality, the empirical evidence for such a relationship is weak. We have found only five studies that use time series from developed countries to analyze the inequality–mortality relationship. After examining changes in life expectancy in the UK during the 20th century, both Wilkinson (1989) and Sen (1999) conclude that longevity rose faster when the income gap between the rich and poor narrowed. However, their measures of income inequality are relatively inexact, and they do not try to take account of temporal variation in the effect of technological innovation. Focusing on the last decades of the 20th century in both the US and the UK, Wilkinson (1996) also argues that rising inequality during the 1980s was the main reason why the decline in infant mortality slowed between 1975 and 1985. Deaton and Paxson (2001), in contrast, find no systematic relationship between inequality and health in either the UK or the US from the mid-1970s to the mid-1990s. Likewise, when Lynch et al. (2004b) look at 100-year national trends and 30-year regional trends in the US, they find little evidence of a causal relationship between income inequality and mortality.

This paper extends previous studies by examining long time series for 12 of the world's richest countries rather than one or two. Our findings are consistent with those of Deaton and Paxson (2001) and Lynch et al. (2004b), not with those of Wilkinson, 1989, Wilkinson, 1996 or Sen (1999). In our preferred specifications we find only small and statistically insignificant relationships between income inequality and mortality. This holds true regardless of whether we measure mortality using life expectancy at birth, infant mortality, homicide, or suicide. It also holds true when we introduce lagged measures of inequality, when we focus exclusively on the period since 1960, when we control both the educational level of the population and health expenditures. Our findings suggest that the relationship between income inequality and mortality is either non-existent or too fragile to show up in a robustly estimated panel specification.

The paper is organized as follows. Section 2 presents a simple model of the relationship between inequality and health. Section 3 describes our data. Section 4 presents our results. Section 5 concludes.

Section snippets

A simple model of the relationship between inequality and health

Epidemiologists and social scientists have proposed numerous mechanisms by which income inequality might affect an individual's health. We can group these mechanisms under three broad headings: absolute income, relative income, and society-wide effects of income inequality. We discuss these mechanisms in turn.

Data on inequality and health

Data quality has been a major problem in studies of the relationship between income inequality and health. As Judge, Mulligan and Benzeval (1998:569) note in their review of the literature:

“Many of the studies use multiple sources of income distribution data and/or data from a wide range of years, which makes comparability between countries questionable. Only five of the studies use data based on a measure of equivalent disposable income. In fact, we believe it is the generally poor quality of

Empirical strategy and results

Most of the existing literature relies on comparisons across countries at a single point in time or on changes over time within one or two countries. We begin by estimating an equation similar to the one often reported in this literature:mjt=α+β(Share10)jt+γZjt+εjtwhere m is the measure of mortality (life expectancy or infant mortality) for country j in year t, Share 10 the income share of the richest 10% of the population, Z the real GDP per capita, and ɛ is an error term. Standard errors are

Conclusion

While there is a strong consensus in the literature that the correlation between income and health is positive, there is much less agreement about the relationship between income inequality and health. This paper has used a new measure of inequality – the income share of the richest 10% of the population – to test the relationship between inequality and mortality. Because we have longer time series for more countries than past studies, we have tried to circumvent some of the problems that

Acknowledgements

Thanks for valuable suggestions that helped improve earlier drafts to Philip Clarke, Andrew Clarkwest, David Cutler, Christian Dustmann, Richard Frank, Campbell Murray, Joseph Newhouse, Betsey Stevenson, S.V. Subramanian, and an anonymous reviewer, as well as seminar participants at the Australian National University, the University of Adelaide, the 2005 SOLE/EALE annual meetings, and the 2005 NBER Summer Institute.

References (32)

  • A. Deaton

    Health, inequality, and economic development

    Journal of Economic Literature

    (2003)
  • Deaton, A., Paxson, C., 2001. Mortality, income, and income inequality over time in Britain and the United States. NBER...
  • P. Fajnzylber et al.

    Inequality and violent crime

    Journal of Law and Economics

    (2002)
  • I. Kawachi et al.

    Social capital, income inequality, and mortality

    American Journal of Public Health

    (1997)
  • S. Knack et al.

    Does social capital have an economic payoff? A crosscountry investigation

    Quarterly Journal of Economics

    (1997)
  • Leigh, A., in press. Does equality lead to fraternity? Economics...

    • Cited by (87)

    • The effect of housing price inequality on mental health

      2023, Labour Economics

    • Relative deprivation and health revisited: New evidence from middle-aged and older adults in rural China

      2023, China Economic Review

    • Rising childhood income inequality and declining Americans’ health

      2022, Social Science and Medicine
      Citation Excerpt :

      The relationship between income inequality and health has been extensively studied, but current literature has provided mixed findings. Many scholars now are skeptical of the adverse impact of income inequality on health (e.g., Leigh and Jencks, 2007; Mellor and Milyo, 2003) while others believe the initial theoretical arguments and maintain that the effects may still exist in certain situations (e.g., Lillard et al., 2015; Tibber et al., 2022; Zheng, 2012). This uncertainty has high stakes as most developed and developing countries have experienced dramatic increases in income inequality in the past several decades.

    • Income inequality and non-communicable disease mortality and morbidity in Brazil States: a longitudinal analysis 2002-2017

      2021, The Lancet Regional Health - Americas
      Citation Excerpt :

      However, this association is often inconsistent in studies that use data from multiple countries, as often social, cultural and political particularities that are not accounted for [14,15]. Indeed, several studies have often pooled men and women together [16,17] or performed cross-country studies without fully accounting for potential national differences [15,18,19]. The use of controls for unobserved time and geographical characteristics (i.e. fixed-effects) are a methodoligcal advancement for analysing health data.

    • Iconic buildings’ contribution toward urbanism

      2020, Alexandria Engineering Journal

    • Relative deprivation and suicide risk in South Korea

      2020, Social Science and Medicine
      Citation Excerpt :

      For instance, Andres (2005) examined the determinants of suicide rates in 15 European countries from 1970 to 1998 but found no changes in suicide rates in response to an increase in the Gini coefficient. Leigh and Jencks (2007) examined mortality data over a decade (from 1903 to 2003) for 12 developed countries (Australia, Canada, France, Germany, Ireland, Netherlands, New Zealand, Spain, Sweden, Switzerland, the UK, and the US) and estimated a weak correlation between suicide rates and the income share of the top decile. The null effect of income inequality was also reported in Germany for the period of 1980–2000, where the association between income inequality and suicide rates at the state-level was estimated using a dynamic panel regression specification (Neumayer, 2004).

    View all citing articles on Scopus
    View full text
    Copyright © 2006 Elsevier B.V. All rights reserved.