Population Health: Behavioral and Social Science Insights

Income Inequality and Health: A Causal Review

By Kate E. Pickett and Richard G. Wilkinson

Abstract

There is a very large literature examining income inequality in relation to health. Early reviews came to different interpretations of the evidence, though a large majority of studies reported that health tended to be worse in more unequal societies. More recent studies, not included in those reviews, provide substantial new evidence. Our purpose in this chapter is to assess whether or not wider income differences play a causal role leading to worse health. We conducted a literature review within an epidemiological causal framework and inferred the likelihood of a causal relationship between income inequality and health (including violence) by considering the evidence as a whole. The body of evidence strongly suggests that income inequality affects population health and well-being. The major causal criteria of temporality, biological plausibility, consistency, and lack of alternative explanations are well supported. Of the small minority of studies that found no association, most can be explained by income inequality being measured at an inappropriate scale, the inclusion of mediating variables as controls, use of subjective rather than objective measures of health, or followup periods that were too short. The evidence that large income differences have damaging health and social consequences is strong, and in most countries, inequality is increasing. Narrowing the gap will improve the health and well-being of populations.

Introduction

World leaders, including the U.S. President, the U.K. Prime Minister, the Pope, and leaders at the International Monetary Fund, the United Nations, the World Bank, and the World Economic Forum have all described income inequality as one of the most important problems of our time, and several have emphasized its social costs.1-7 Inequality is increasing in most regions of the world, rapidly in most rich countries over the past three decades.8,9 There is a very large literature examining income inequality in relation to health. Early reviews came to different interpretations of the evidence, though a majority of studies reported that health tended to be worse in more unequal societies.10-14 More recent studies, not included in those reviews, provide substantial new evidence.

There is also growing evidence that a wide range of social outcomes, associated with disadvantage within societies, are more common in societies with bigger income differences between rich and poor. Although our objective in this chapter is to assess whether or not wider income differences play a causal role leading to worse health (including the public health issue of violence), we consider studies of other social outcomes where they affect interpretation of the health data.

The first task is to clarify the causal hypothesis and how it has developed as research has progressed. Research was initially focused simply on whether health was worse in more unequal societies, but there is now growing evidence to suggest that this should be seen as part of a wider tendency for a broad range of outcomes with negative social gradients (i.e. more prevalent where social status is lower) to be more common in societies with bigger income differences between rich and poor. Rather than this pattern being confined to physical health, it may apply also to mental health and public health issues such as violence, teenage births, child well-being, obesity, and more.

Whether causality is tested in relation to a hypothesis confined to a relationship between inequality and physical health, or whether the hypothesis extends to problems with social gradients more generally, has important implications for understanding possible causal mechanisms, mediators and confounders. in this chapter, we will focus on the strongest and most important claim underpinning an effect of inequality on health: that large income differences between rich and poor lead to an increasing frequency of most of the problems associated with low social status within societies. Figure 1 provides an illustration of the relationships with which this chapter is concerned. It shows a cross-sectional association between income inequality in developed countries and an index that combines data on life expectancy, mental illness, obesity, infant mortality, teenage births, homicides, imprisonment, educational attainment, distrust, and social mobility. Raw scores for each variable were converted to z-scores, and each country was given its average z-score.15

History

The hypothesis that problems (including poor health) associated with low social status are more common in more unequal societies can be traced back to independent roots in papers on homicide rates and mortality rates. The research literature on homicide and inequality goes back at least 40 years, to a demonstration that they were positively associated among U.S. States.16 The earliest paper on mortality and income inequality—some 35 years ago—showed a cross-sectional association between Gini coefficients of income inequality and both infant mortality and life expectancy at age 5 among a group of 56 developed and developing countries.17 By 1993, a meta-analysis of some 34 studies concluded that there was a robust tendency for violence to be more common where income differences were larger.18 The research on income inequality and health expanded rapidly after the first papers were published in journals of epidemiology and public health.19,20 By 2006, our review of papers on income inequality and health identified 168 analyses, the overwhelming majority of which showed a positive association.13 The two literatures—in criminology and sociology on the one hand, and epidemiology and public health on the other—developed independently and unaware of each other until the late 1990s.21,22

Figure 1. Index of health and social problems in relation to income inequality in rich countries

Chart presents an index of health and social problems in relation to income inequality in rich countries. It shows that when compared with 21 other

Source: Wilkinson R, Pickett K. The spirit level. London: Allen Lane; 2009. Used with permission.

Income inequality is measured by the ratio of incomes among the richest compared with the poorest 20 percent in each country. The index combines data on: life expectancy, mental illness, obesity, infant mortality, teenage births, homicides, imprisonment, educational attainment, distrust, and social mobility. Raw scores for each variable were converted to z-scores, and each country was given its average z-score.15

It was only in 2005 and 2006,23-25 as researchers began to show that the correlates of inequality included teenage birth rates, obesity, and mental illness, that it started to look as if a more general explanatory hypothesis was needed than those which had addressed only physical health and violence. On the assumption that social gradients were often evidence that an outcome was sensitive to social status differentiation, we formed the hypothesis that greater inequality might act to strengthen the effects of socioeconomic status differentiation among outcomes with social gradients.

We tested this hypothesis by analyzing whether or not outcomes with steeper social gradients had stronger associations with societal inequality. We selected 10 different death rates, some with weaker and some with stronger social gradients, as measured by their correlation with county median income, among the 3,139 counties of the United States.26 In a multilevel model controlling for the effects of county income, we then estimated the correlations of these death rates with State income inequality. The results, shown in Figure 2, provided strong confirmation of the hypothesis.

Figure 2. The effect of county-level median household income in relation to contextual effect of State-level income inequality

Chart shows the effect of county-level median household income in relation to income inequality at the State level and the correlation between income inequality and 10 different death rates: prostate cancer, breast cancer, alcoholic liver disease, diabetes, infant mortality, respiratory disease, heart disease, homicide, all-cause mortality-elderly, and all-cause mortality, working age.

Source: Wilkinson RG, Pickett KE. Income inequality and socioeconomic gradients in mortality. Am J Public Health 2008;98(4):699-704. Used with permission.

Notes: Standardized beta coefficient from multilevel model, controlling for county-level income.  r = –0.814; P = .004. a = Standardized parameter estimates (B) from multilevel model after county-level income was controlled. b = Aged 25–64 years. c = Aged ≥65 years.

This work was followed by studies that examined the association among rich developed countries between income inequality and a number of health and well-being outcomes, including the UNICEF (United Nations Children's Fund) Index of Child Wellbeing and the separate components of the Index of Health and Social Problems shown in Figure 1.15,27,28

Popperian Theory Testing

The philosopher of science, Sir Karl Popper, taught that the best evidence of the value of a theory is provided by testing its novel predictions.29,30 A successful theory was "corroborated" (but could never be finally proven true) if it accurately predicted the results of scientific observations that had not previously been expected. The initial evidence of a relation between income inequality and population health using international data was first explicitly tested and confirmed in 1996 by two groups working independently at the universities of Harvard and Michigan, who looked to see if the same relationship could be found among the 50 U.S. States.31,32 There are now very large numbers—hundreds—of replications of these findings in many different settings in societies at all levels of development. Even the more unequal provinces of China have been found to have significantly less good health.33 There now can be no doubt that worse health is at least associated with greater inequality.

The tendency for more unequal societies to have higher homicide rates has also been replicated many times (for one recent review see Rufrancos et al.34). To suggest that a relationship is causal means predicting a subsidiary hypothesis about a mediating mechanism. A testable prediction of a causal mechanism was first suggested on the basis of qualitative impressions only. The hypothesis was that more equal societies were healthier because they were more cohesive and enjoyed better social relations.35 A year later, that prediction was tested quantitatively: path analysis showed that the relationship between greater equality and lower death rates among the U.S. States was mediated by social capital (operationalized as group membership and social trust).36

Another way in which testable predictions have emerged is when new national data have become available that fit previously established relationships between inequality and an outcome measure. This happened when new data on social mobility and on mental illness rates became available for several additional countries, and the data were found to fit previously established relationships between those outcomes and inequality.37

Lastly, the first papers suggesting that mental illness was more common in more unequal societies used general measures of mental illness from the World Health Organization (WHO).25,38 The picture has since been filled out by papers that show more specific forms of mental illness—including depression, schizophrenia, and psychotic symptoms—are all more common in more unequal societies.39-41

There have been a small proportion of negative findings throughout the development of this field. We will discuss some of the reasons for variations in findings at appropriate points later in this chapter.

Our aim in this review is to go beyond the "counting" methodology of previous major reviews; these mostly divided studies into supportive, mixed, and unsupportive of the income inequality- health relationship and counted them. Among mixed studies, which showed some but not all relationships to be significant, results within a study might vary by geographic scale or by health outcome, measure of inequality, sex or age of subjects, or other variables; merely counting these adds nothing to interpretation, even if we count more. Instead, we conduct a causal review to give a more structured and coherent framework to our examination of the literature. We have aimed to incorporate all new studies that illuminate relevant causal processes.

Epidemiological Criteria for Causality

In observational epidemiology, causality cannot be proven or disproven by any single study—there are no "black swans"—just because income inequality might not affect some health outcomes, or not in some times or places or for some populations, does not mean that it isn't a causal relationship in other contexts. Instead, in epidemiology, a body of evidence needs to be considered, usually including non-epidemiological studies, to judge whether or not an exposure-outcome relationship is causal.

Causal criteria were first proposed in the Surgeon General's Report of 1964, and then refined by Sir Austin Bradford Hill in 1965, in the context of examining the evidence linking cigarette smoking to lung cancer.42,43 The use of causal criteria, indeed even the term "criteria," has been contentious, especially if the criteria are used as a simple checklist or algorithm; however, when used as a framework for thoughtful inference, and used to consider competing causal theories by focusing on crucial observations, they offer a useful organizing structure for critical review.44,45 Bradford Hill's original nine "criteria" have been further refined with modern usage and, as indicated in Table 1, four are considered of major importance.46

Consistency

There are now perhaps 300 peer-reviewed studies of the relation between income inequality and measures of health or homicide. They include both ecological and multilevel studies using cross-sectional, cohort, and time-series designs in many time periods. Looking at bivariate correlations before the use of control variables, the most recent full review found only 6 percent of studies (8 of 128) did not find at least one significant association between greater inequality and worse health.13 After the use of many different kinds of control variables, many of which might be on the causal pathway, 70 percent of the studies reporting either positive or negative (but not "mixed") results found only significant associations between higher inequality and worse outcomes.

These relationships have been found in a wide variety of settings. Much research has focused on the rich, developed, market economies (United Kingdom, United States, Western Europe, Japan, Singapore, Australia, and New Zealand) and analyses of the 50 U.S. States.15,28 Some studies have included developing, emerging, and developed nations, while others have focused on developing countries in particular. Infant mortality is the health outcome most often shown to be positively correlated with income inequality in less developed nations,47-49 although there are also associations with lower life expectancy, higher HIV prevalence, and higher homicide rates.47,50,51 Some studies have focused on particular world regions. For example, Marmot and Bobak found larger declines in life expectancy in the more unequal countries of Eastern Europe following the dissolution of the Soviet Union.52 Biggs and colleagues studied 22 Latin American countries from 1960-2007 and found a substantial relationship between income inequality, life expectancy, infant mortality, and tuberculosis mortality rates; they also reported that when inequality was rising, economic growth was related to only a modest improvement in health, whereas during periods of decreasing inequality, there was a very strong effect of rising gross domestic product (GDP).53 Other studies have shown an association between income inequality and health across states/regions within nations, including, for example, in Argentina,54 Canada,55 Brazil,56 Chile,57 China,33 Ecuador,58 India,59 Italy,60 Japan,61 and Russia.

The geographical scale at which income inequality is measured is, however, an important methodological issue because it points to a distinction between the large majority of supportive studies and the unsupportive minority. In one review, researchers found that after the use of controls, the proportion of analyses classified as wholly supportive of an income inequality effect on health was 83 percent among international studies, but it fell to 73 percent in large subnational areas, and to 45 percent in studies of small areas such as neighborhoods.13 A similar pattern was noted in an earlier meta-analysis of studies of income inequality and violent crime, including homicide18 and a later meta-analysis of multilevel studies.63 Hsieh and Pugh18 concluded that "homogenous estimates of association between income inequality and homicide were reported by studies using states and nations as their sampling units but not by studies using smaller sampling units."

Table 1. Epidemiological framework for causal inference

Criteria Importance
Consistency The association has been replicated in different methodological, geographical, and time settings
Temporality The putative cause must precede the effect, this is an indisputable criterion for causality
Strength of association The stronger an association is the less likely it is that there is some alternative unknown explanation
Specificity There is a high probability that an exposure is causally linked to some outcomes more than to others (many epidemiologists believe this criterion should be dropped
Dose response relationship Increased exposure is related to increased outcomes
Cessation of exposure If exposure changes, positively or negatively, the incidence of the outcome will rise or fall
Consideration of alternative explanations The association is not confounded by one or more other factors
Biological plausibility The association fits with existing biological knowledge
Coherence The association is supported by other scientific knowledge

Source: Adapted from: Gordis I. Epidemiology. 5th ed. Philadelphia: Elsevier Saunders; 2013.
Note: Major criteria appear in bold font.

We have previously suggested that studies of income inequality are more supportive in large areas because in that context income inequality serves as a measure and determinant of the scale of social stratification or how hierarchical a society is.13 Income inequality in small areas is affected by the degree of residential segregation of rich and poor, and the health of people in deprived neighborhoods is likely to be poor, not because of the inequality within each of those small areas, but because they are deprived in relation to the wider society. Studies from the United States and Sweden, which have compared the strength of association at different levels of aggregation, support this interpretation and the need to think carefully about scale before conducting studies.64-66 Another factor that might contribute to the same picture is the possibility that more unequal societies may give rise to greater residential segregation between rich and poor and thereby increase the inequality between areas and diminish the inequality within them.

Together, the studies provide overwhelming evidence that greater inequality is linked to worse health and more violence. Factors such as the size of area64 and the use of conceptually inappropriate controls may provide plausible explanations of the minority of unsupportive studies.

Temporality

The large number of cross-sectional studies, undertaken over several decades, which link income inequality to health and violence, imply that there are relationships over time. As neither income distribution nor health are invariant over time, the fact that cross-sectional associations between them have been reported so many times is in itself an indication that they move together.

The preponderance of cross-sectional studies is partly a reflection of the limited availability of time series estimates of income inequality and, for some countries, for health outcomes. But there are now a growing number of studies of these relationships over time.

A meta-analysis of multi-level studies by Kondo and colleagues included data covering 59,509,857 individuals from nine cohort studies from Denmark, Finland, Norway, New Zealand, Sweden, and the United States, with followup periods ranging from 1-28 years.67 The overall cohort relative risk (95 percent confidence interval) per 0.05 unit increase in the Gini coefficient, a measure of income inequality, was 1.08 (1.06 to 1.10). Studies with baseline data collection after 1990 and a length of followup greater than 7 years had a marginally higher relative risk; these interactions were not modified by the size of the area in which inequality was measured.63

In an international panel study of 21 developed countries over 30 years, controlling for serial correlation and stratifying by age and sex, Torre and Myrskylä, found that high inequality was associated with increased mortality of males and females aged 1–49 and older women but not older men.68

High inequality was associated with increased mortality of males and females aged 1–49 and older women but not older men.

Zheng69 reviewed 79 studies of income inequality in relation to mortality: four aggregate and seven multilevel studies examined lagged effects up to 10 years with mixed results. However, all of these studies tested the lagged effect of income inequality in a particular year, treating it as a time-invariant variable and failing to control for a series of previous, subsequent, and contemporaneous income inequalities. Zheng also reviewed eight similar studies of self-rated health; seven of these found a significant effect, with two suggesting the strongest effect persisting through 15 years. Again, however, none of these studies looked at serial measures of inequality. Zheng's reviews included studies published through 2008 and contained in four previously published reviews.11-13,67

We have conducted a further primary systematic search for time-series and panel studies of income inequality and health and identified an additional nine studies, containing 53 analyses, including studies of mortality, life expectancy, infant mortality, under-5 survival rate, and self-rated health.a Of these, 55 percent support a longitudinal effect of income inequality on health (60 percent of within-country studies in the United States, the United Kingdom, and Norway), 37.5 percent do not, and 7.5 percent had mixed results, but all of these studies also suffer from the same methodological problem of not considering time-variant income inequality.

Studies also varied in the inclusion of control variables in the analyses of income inequality and health, including measures of aggregate or individual income or education; ethnic mix; unemployment; alcohol or tobacco consumption; birth, fertility, and divorce rates; benefit payments; health expenditures; and other variables. As some of these may be mediating or moderating factors in a causal pathway leading from income inequality to health, the inclusion of some is questionable, and the estimates of the effect of inequality would be underestimated.

Zheng69 went on to conduct a discrete time-hazard analysis of U.S. national-level income inequality on mortality, controlling for individual income, in 701,179 individuals with a 21-year followup. A detrimental effect of rising inequality began to affect mortality after 3-5 years, and the effect size increased until mortality plateaued at a higher level after 12 years. This finding was robust to different model specifications and different measures of inequality. This study probably provides the best estimates of the average lag time between changes in inequality and mortality. It seems to accord well with other studies, though of course lag times will vary between age groups and causes of death.

A review of time-series and panel studies of income inequality and crime, that included seven studies examining homicide rates (five conducted in the United States or Canada or by international cross-country comparisons), found a significant increase in the murder rate with rising income inequality.34 A study in West Germany found no association between income inequality and homicide rates,70 as did a single international study,71 although the statistical methods of this study were criticized.

Clarkwest analyzed State-level data within the United States from 1970 to 2000 to examine the effects of initial levels and change in income inequality on 10-year changes in life expectancy, finding that States with higher levels of inequality experienced less subsequent improvement in life expectancy.72

It is important to note that, for reasons which are not well understood, health continues to improve over time in most developed countries, with life expectancy rising by approximately 2-3 years with each decade. Against this background rate of improvement, the effect of changes in income inequality is to speed up or slow this background rate. Only when there are catastrophic rises in inequality, as in Russia and Eastern Europe during the transition from communism, does life expectancy actually fall.52

A number of studies have suggested that death rates among the elderly show little or no relation to current inequality, see for example, Torre & Myrskyla.68 However, analyses have shown that the health of older people is independently influenced by socioeconomic status at three different points in the life course, by fetal health and perhaps also by social security and welfare provisions in childhood.73-75 If health in later life is similarly affected by lifetime exposure to inequality, we will need lifelong measures of exposure before we know whether the health of the elderly really is insensitive to their lifetime experience of inequality.

Strength of Association

In general, the stronger an association between putative cause and effect, the less likely it is that the relationship can be explained by other factors. But with health outcomes that are multifactorial, not all causes will have strong effects if they are necessary but not sufficient to cause the outcome alone. Nevertheless, at a population level, even moderate effects can have large impacts.

In international, cross-sectional, unadjusted studies of income inequality in relation to health and social problems in rich countries, the strength of the statistically significant associations vary.15,28,37,76 Correlations with income inequality are higher for mental illness and teenage birth rates (both, r=0.73) and drug use and child well-being (both, r=0.63) than for life expectancy, infant mortality, obesity, and homicide (all, r<0.5). However, when researchers treated income inequality as a common cause of many health and social problems and combined them in one index, which tends to emphasize their common variance, the correlation with an index of problems was so high (r=0.87) that any alternative explanations would need to have extraordinarily strong effects (Figure 1).

When estimated in multilevel models, the size of the effect of inequality usually looks much smaller, as described earlier in this chapter in the section on Consistency. The difference is a matter of what is included as an effect of inequality. Some of the early multilevel studies of the effects of inequality were based on the assumption that the relationship between individual income and health was a reflection of the direct effects of what people's material circumstances did for their health regardless of anyone else in society. The desire was to separate out such effects before looking at the broader contextual effects of inequality, which were assumed to work through quite different pathways involving psychosocial processes hinging on relativities and social comparisons. However, a great deal of research attests to the likelihood that individual income is related to health because it is a marker of individual social status,77,78 and that subjective social status may be more important than objective measures.79 There is also evidence that greater inequality worsens outcomes such as math and literacy scores, social mobility, dropping out of high school, teenage birth rates, and mental illness, all of which might create feedback from higher inequality to increased numbers of people on low incomes.37,39,80 If so, this would mean that multilevel models controlling out the effects of individual income risk seriously underestimate the effects of inequality.

Specificity

In some contexts, specificity is an outmoded causal criterion that dates from when the main health focus was on infectious diseases, which could only be caused by exposure to a specific pathogen. It is less relevant in a context where most health and social problems have multiple, interacting causes, and many outcomes share causes. However, there is an aspect of specificity in the relationship between income inequality and health that is helpful when considering causality and the pathways from one to the other. As we outlined in the History section of this chapter, the adverse effects of income inequality seem to be specific to outcomes that have an inverse social gradient.26 For example, there was no social gradient for breast or prostate cancer mortality and no effect of income inequality, whereas there was a steep social gradient in working age all-cause mortality, and there was a strong association with income inequality. This would explain why the social outcomes included in Figure 1 are more common in unequal societies.

Broadly similar results have been found for child health.81 In 29 Organization for Economic Co-operation and Development (OECD) countries, income inequality was positively related to post- neonatal mortality and teenage overweight, both of which have steep social gradients, but there was no association for suicide, which did not have clear evidence of a social gradient in some countries. One interpretation of this specificity is that income inequality intensifies the health effects of social hierarchy and social comparisons, thus increasing socioeconomic disparities in health. However, there was no association between income inequality and child asthma or adolescent smoking, both of which have some evidence of social gradients.

In international comparative studies, there is also a degree of specificity with regard to income inequality being associated with objective measures of health, rather than subjective measures. This is because, internationally, there is no correlation between life expectancy and the proportion of the population with good self-rated health, although these measures are correlated within countries.82,83

Dose-Response Relationship

A very large number of studies demonstrate statistically significant linear relationships between income inequality and health. The effects on inequality increase step by step from the most unequal of the 50 U.S. States and the most unequal countries to the most equal. However, Kondo and colleagues find a threshold effect, with higher relative risk of mortality in cohort studies with higher levels of income inequality (Gini coefficient >0.30) at baseline.63 In other analyses of health and social problems among developed countries, the relationships tend to appear linear, with no evidence of a step change above some threshold level of inequality (Figure 1). According to the World Bank World Development Indicators database, very few nations have Gini coefficients below 0.30,84 the threshold identified by Kondo and colleagues; among developing nations there is only Afghanistan. Several former Soviet republics (Belarus, Bulgaria, Czech Republic, Kazakhstan, Romania, Slovak Republic, Ukraine) have Gini coefficients between 0.25-0.29, as do the Scandinavian countries (Denmark, Finland, Norway, Sweden), Austria, Germany and Japan. Among OECD countries, only the Netherlands has not experienced a rise in income inequality since the mid-1980s. Thus, most of the world's population is exposed to income inequality above the threshold suggested by Kondo and colleagues, and the proportion of those exposed continues to rise.85

Even within more equal countries, inequality seems to matter. A recent study from Norway86 found an independent effect of regional income inequality on mortality, after adjustment for regional-level social and economic characteristics. A study from Finland87 suggested that widening differences in income inequality account for almost half of the increase in health inequalities, and one from Sweden found a detrimental effect of municipal income inequality on self-rated health.66

Whether each additional increment of inequality above a Gini of 0.30 has a greater effect than it does below that level remains unclear, but there is substantial agreement that there is a dose- response relationship above that level.

Cessation of Exposure

There can be no examples of cessation of exposure to inequality—only of exposure to more or less inequality. Interesting evidence comes from a study by Hamilton and Kawachi88 that assessed whether or not individuals who migrate to the United States from countries with greater income inequality than the United States have better health than those who migrate from countries with less income inequality. Among immigrants who lived in the United States between 6 and 20 years, those for whom moving to the United States was a move towards greater equality had better self-reported health than those for whom it was a move towards greater inequality. Similarly, Auger and colleagues89 found that income inequality was associated with mortality among non-immigrant Canadians but not migrants, although for long-term immigrants the effects tended to approach those of the Canadian-born population.

Also relevant is the striking reversal in international rankings in income inequality and population health between the United States and Japan in the three or four decades following the Second World War.90 In the post-war period, the United States had much lower inequality than it does today and ranked high in the international league table for life expectancy, whereas Japan was highly unequal, with lower life expectancy. But by the end of the 1980s, Japan had become one of the most equal countries and had the highest life expectancy in the world. In contrast, the United States became rapidly more unequal from the late 1960s and is now among the most unequal societies in the developed world. During that period, the U.S. position slipped in the international life expectancy league tables, and it now ranks 40th according to the United Nations.91

Consideration of Alternative Explanations

Given that the epidemiological criteria examined so far support a causal interpretation for the role of inequality, we should ask whether there are any other possible explanations.

A paper by Deaton in 200392 reported that the proportion of black residents in States and Metropolitan Statistical Areas of the United States explained the income inequality-health association. This paper continues to be cited as evidence that income inequality does not affect health, despite the fact that several more recent studies found that ethnic heterogeneity does not confound the income inequality-health association in the United States.12,93-96 International comparisons also show that income inequality is significantly related to health even after adjustment for ethnic heterogeneity.97 Nor does ethnicity explain the income inequality-homicide relationship in U.S. States. To clarify this, one analysis of homicides in the 50 States confined attention to white perpetrators of homicides and showed they were significantly related to income inequality measured only among the white populations of each State.98 It seems likely that ethnic differences attract more attention and seem more important not only when they become markers of social status differences but also when greater inequality makes social status differentiation more powerful, increasing the importance of "downward" social prejudices whether by class or ethnicity.99,100

Another proposed alternative explanation suggests not only that the relationship between individual income and health is curvilinear, such that a rise in income for the poor has a greater impact on health than an equivalent rise in income for the rich, but also that this effect reflects only the direct influence of material living standards on health—not inequality as such. The suggestion is that greater equality would improve average health but only for reasons related to what individual material circumstances do to health, regardless of other income and position in the income hierarchy. The assumption is that someone's health is affected only by their own income and is unaffected by where they are in the income hierarchy. Studies within the United States101 and the United Kingdom,102 as well as international comparisons,47 disprove this explanation, as do the many multilevel studies of income inequality and health reviewed by Kondo et al, which control for individual income and socioeconomic status.67 Such studies show a contextual effect of inequality over and above the effects of individual income.

Income inequality is, and can only be, an ecological variable describing the scale of income differences across a population. Because inferences are usually made only to other ecological variables, such as rates of health or social problems across the same population, the possibility of an ecological fallacy does not arise: inferences are not made from ecological variables to individual risk. However, studies that ask "Whose health is affected by inequality" suggest that although effects are probably strongest among the least well off, they extend to the majority of the population.37,103

It also has been suggested that the income inequality-health association reflects reverse causality—in other words, income inequality is a result of a larger proportion of the population being unhealthy, rather than a cause. The time series studies described above, which show that there are substantial lag periods between changes in inequality and changes in health, disprove this interpretation, as do the findings of cohort studies. In addition, income inequality has been related to many infant and child outcomes, including infant mortality, low birth weight, child well-being, and child mental health problems, which would not be expected to affect inequality.27

The suggestion of reverse causality faces two other difficulties. The first is that more unequal countries appear to do poorly on a wide range of health and social outcomes, while more equal countries do well. If income inequality were a result of worsening outcomes, then it would be necessary to find an alternative explanation for why so many disparate problems—ranging from health to homicides, child well-being, mental illness, and drug abuse—all tend to be worse in some countries than in others. As they are such different outcomes, yet all with similar social gradients, it would be necessary to posit another, very deep-seated explanation closely related to social status differentiation. Lastly, a good deal is known of the economic policies that came in from the late 1970s and led directly to wider income differences.

In terms of pathways from income inequality to health, it has been suggested that more generous welfare regimes, public spending (e.g., on health or transport), more comprehensive social security, and increased investment in human capital development (e.g., education) are all characteristic of more equal societies, and that the relation between income inequality and health may therefore be mediated by these "neo-material" factors. From a neo-material perspective, the association between income inequality and health reflects people's lack of resources, as well as societal underinvestment in such things as "education, health services, transportation, environmental controls, availability of food, quality of housing, [and] occupational health regulations."104

Nevertheless, explicit tests of "neo-material" vs. psychosocial pathways from income inequality to healthy life expectancy, mortality, mental health, and homicide rates conclude that psychosocial factors, such as social capital and trust, mediate the relationship, whereas neo-material factors, such as public expenditure on health or social services, have little or no explanatory role.105-107 Of course, insofar as welfare regimes, social security, and other programs redistribute income, it is difficult to disentangle the independent effects of income inequality and welfare regimes.

Lastly, the tendency is often to imagine that cultural differences lie behind and are the real reason for associations between income inequality and a poorer performance on a wide range of health and social outcomes. Any such hypotheses about the role of culture would of course have to be compatible with the evidence that shows that health changes follow changes in income distribution after a lag of some years. It would also have to be compatible with the evidence of associations in different parts of the world, at different levels of development and with different cultures. So, for instance, Mexico, Russia, and South Africa all have very high levels of income inequality and very high levels of violence, but their cultural identities are very different. Similarly, societies like Japan and the Scandinavian countries have low levels of inequality and low levels of violence despite obvious cultural differences among them. Also interesting is the cultural similarity between Portugal and Spain. Both countries were dictatorships until the mid-1970s, and they share a border. However, their performance on the Index of Health and Social Problems reflects (as shown in Figure 1) their substantial differences in income distribution.

Biological Plausibility

A psychosocial explanation of the effect of income inequality on health and behavioral outcomes is consistent with the biology of chronic stress, new studies of the neuroscience of social sensitivity, and concepts from evolutionary biology. Income inequality is linked to lower levels of social cohesion and generalized trust, suggesting that inequality acts as a social stressor.108-110

Chronic stress impairs memory and increases risk of depression, lowers immune responses, elevates blood pressure and risk of cardiovascular disease, and affects hormonal systems.111 Research shows that the ways in which we relate to one another, such as friendship, social support, and social networks, are as protective for health as smoking is deleterious.112 If we have friends, we are less likely to contract a common cold infection in randomized controlled trials.113 Likewise, if we have a difficult relationship with our spouses or partners, we heal more slowly in trials of experimental wound healing.114 A meta-analysis of 208 laboratory studies of acute psychological stressors and cortisol responses shows that stronger cortisol responses were elicited if tasks were uncontrollable or characterized by "social-evaluative threat" (threats to self-esteem or social status).115 Even low levels of psychological distress were found to be related to mortality in a meta-analysis of 10 large prospective cohort studies.116 Telomere length, a measure of cell aging, was found to be shorter by age 9 among African American boys who lived in highly disadvantaged environments compared to those who were raised in more affluent environments.117

Neuroscience studies also highlight the importance of psychosocial factors for human physiology. A neuroimaging study showed that social pain (exclusion) activated the brain in the same ways as physical pain. The anterior cingulate cortex (ACC) was more active during experiences of social exclusion and was positively correlated with self-reported distress.118 In another study, baseline sensitivity to physical pain predicted sensitivity to social rejection, and social exclusion was associated with more sensitivity to physical pain.119 In two experiments, participants received either acetaminophen (a pain suppressant) or a placebo for 3 weeks. Acetaminophen reduced daily reports of social pain, and functional magnetic resonance imaging showed that acetaminophen reduced neural responses to social pain in areas of the brain previously shown to be related to both social and physical pain.120

Evolutionary explanations of human sensitivity to social relationships and hierarchies stress the importance of belonging and people's need for positive relationships and connectedness. Social exclusion affects cognitive, emotional, and behavioral outcomes, and adaptations to low social rank in both animals and humans include altered levels of hormones and behaviors, such as withdrawal, apathy, or hypervigilance.121 A theory linking submission and subordination to depression suggests that it results from an inability to stop, or escape from, a submissive defeat strategy, and the evidence reviewed by Johnson and colleagues supports this; in more than 20 research studies, people with depression were more likely to report feeling inferior or experiencing shame.122

Coherence

In rich countries, there is no association between average levels of income (e.g. gross national income per capita) and measures of health, such as life expectancy.123-126 Yet within rich countries, there are strong associations between individual income and life expectancy. This pattern suggests that it is relative income within societies that is important for health in rich countries, in turn suggesting that psychosocial mechanisms are relevant.

Recent studies of income inequality in relation to psychological states and traits and sociological outcomes lend coherence to a psychosocial explanation of the health and social effects of income inequality on health. International comparisons show that status anxiety is higher in more unequal countries, for all socioeconomic groups.127 Status anxiety and trust were found to mediate the association between income inequality and subjective well-being.128 In more unequal countries, people exhibit higher levels of self-enhancement, i.e., believing themselves to be better than average.129 In both ecological and multi-level analyses, people in more unequal U.S. States scored lower on a measure of agreeableness, reflecting less concern for social harmony and getting along with others.130 In more unequal European countries, people show less solidarity; they are less willing to help others.131

Discussion and Conclusions

The body of evidence on income inequality and health points strongly to a causal connection. The major criteria of temporality, biological plausibility, consistency, and lack of alternative explanations are well supported. Of the small minority of studies that find no association, most can be explained by income inequality being measured at an inappropriate scale, the inclusion of mediating variables as controls, the use of subjective rather than objective measures of health, or followup periods that are too short.

Suicides seem to stand as an important exception to the general pattern: they tend to be more common in more equal societies, despite the evidence that depression is more common in more unequal societies.39,132 A possible explanation is that social gradients in suicides are not always consistent internationally.133 Another possibility is that there may be some truth in the view that violence can be directed either outwards or inwards against oneself. If suicide is, like homicide, often a response to adversity, we think it likely that greater equality increases a tendency to blame oneself rather than others for what goes wrong.

Epidemiological causal criteria are not exhaustive. A good test of the validity of a scientific theory is its ability to make successful, testable predictions. The theory that more equal societies are healthier arose from one international study17 and has now been tested in many different contexts. The search for a mechanism led to the discovery that social relationships (social cohesion, trust, involvement in community life, and low levels of violence) are better in more equal societies, suggesting that inequality and health are linked through psychosocial processes related to social differentiation and relative deprivation.61 That inequality does have powerful psychosocial effects is now amply confirmed.

We suggest that the most parsimonious explanation for the effects of income inequality is that larger income differences increase social distances, accentuating social class or status differences. This would explain why income inequality is most closely related to health when measured across whole societies coterminous with social class hierarchies.13,134 Rather than income inequality being a new and independent determinant of health, it is likely to act by strengthening the many causal processes (known and unknown) through which social class imprints itself on people throughout life. This would suggest why, not only health, but a wide range of other outcomes with social gradients are also related to inequality. It also suggests that if class and status are to become a less powerful influence both on individual lives and on whole societies, it will be necessary to reduce the material differences that so often constitute the cultural markers of social differentiation.

As whole populations are exposed to societal income inequality, estimates of the population attributable risk will be high even if, for some outcomes, the causal effect on some outcomes is modest. Kondo and colleagues67 estimated that upwards of 1.5 million deaths (9.6 percent of total adult mortality for the 15-60 age group) could be averted in 30 OECD countries if each country reduced its Gini coefficient below 0.30. If individual income is also related to health partly through psychosocial mechanisms involving relative deprivation, then multilevel models that control out its effects may substantially underestimate the effects of inequality.135 It has been estimated that if the United Kingdom reduced its inequality to the average in other OECD countries, the expenditure savings on physical and mental illness, violence, and imprisonment alone would amount to £39 billion per year.136

Future research should move beyond mere replication of these findings in different samples towards more explicit attempts to clarify the causal relationships, including studies of (1) different measures of income inequality (top- and bottom-sensitive measures, for example) in relation to different health and social outcomes, (2) time lags for different outcomes, (3) further modeling and testing of specific causal pathways, and (4) whether inequalities in wealth are as much a part of the picture as inequalities in income. Comparable measures of wealth inequality are available for only a limited number of countries, but initial explorations of the relationships with life expectancy are interesting. Life expectancy in Denmark, which seems to be an outlier in relation to its more equal distribution of income, appears to fall into place in relation to its large inequalities in wealth.137

The evidence that large income differences have damaging health and social consequences is already far stronger than the evidence supporting policy initiatives in many other areas of social and economic policy, and the message is beginning to reach politicians. The world leaders we mentioned at the start of this chapter have all referred to inequality as a cause of social and economic harm. But to recognize the problem is not the same as tackling it effectively. The gap between the richest and poorest 20 percent of households in countries like the United States and United Kingdom is not only very much wider than it used to be in the 1970s, but it is still twice as large as in some other successful market democracies. The reason why politicians do not do more is almost certainly a reflection of the undemocratic power of money in politics and the media.138 Narrowing the gap will require not only redistributive tax policies but also a reduction in income differences before tax. The halving of top tax rates since the 1970s has led not only to a widening of income differences after tax but, more surprisingly, to an acceleration in pre-tax income differences particularly in the private sector where pay for top executives seems unrelated to company performance.139-141 has written about the risks of policymakers requiring unachievable standards of proof in social epidemiology before they are willing to act, and Popper30 emphasized that scientific theories are never finally proven true. Adopting too high a standard of evidence may mean that it is never considered strong enough. Schrecker quotes Michael Marmot143 as saying "While we should not formulate policies in the absence of evidence to support them, we must not be paralyzed into inaction while we wait for the evidence to be absolutely unimpeachable."

Acknowledgments

We are grateful to Hector Rufrancos for collating the time-series literature. This paper was published previously by Social Science & Medicine (2014 Dec 30; epub); the abstract is available at http://www.ncbi.nlm.nih.gov/pubmed/25577953.

The opinions presented herein are those of the authors and may not necessarily represent the official position of the Agency for Healthcare Research and Quality, the National Institutes of Health, or the U.S. Department of Health and Human Services.

Authors' Affiliations

Kate E. Pickett, Department of Health Sciences, University of York, York, UK; Richard G. Wilkinson, Division of Epidemiology and Public Health, University of Nottingham, Nottingham, UK.

Address correspondence to: Kate E. Pickett, Department of Health Sciences, Seebohm Rowntree Building, Area 3, University of York, Heslington, York, YO10 5DD, UK; email kate.pickett@york.ac.uk.

References

  1. Obama B. State of the Union address, 2014. Available at http://www.whitehouse.gov/the-press-office/2014/01/28/president-barack-obamas-state- union-address; Accessed December 5, 2014.
  2. Cameron D. Hugo Young lecture, November 20, 2009. Available at http://www.theguardian.com/commentisfree/2009/nov/10/big-society-government-poverty-inequality.. Accessed December 5, 2014.
  3. Pope Francis. Evangelii Gaudium. Vatican City: Vatican Press; 2013.
  4. Lagarde C. Speech at World Economic Forum, January 23, 2013. Davos, Switzerland. Available at https://www.imf.org/external/np/speeches/2013/012313.htm. Accessed December 7, 2014.
  5. Moon B-K. Remarks at informal General Assembly Thematic Debate on Inequality, July 8, 2013. Available at http://www.un.org/apps/news/infocus/sgspeeches/statments_full.asp?statID=1918#.UfDb943EPO4. Accessed January 27, 2015.
  6. Elliott L. Climate change will 'lead to battles for food', says head of World Bank. The Guardian; April 3. 2014. Available at http://www.theguardian.com/environment/2014/apr/03/climate-change-battle-food-head-world-bank. Accessed January 27, 2015.
  7. Global risks 2013. Geneva: World Economic Forum; 2014. Available at http://www.weforum.org/reports/global-risks-2013-eighth-edition. Accessed January 27, 2015.
  8. Ortiz I, Cummins M. Global inequality: beyond the bottom billion—a rapid review of income distribution in 141 countries. New York, NY: UNICEF; 2011.
  9. Divided we stand: why inequality keeps rising. Available at http://tinyurl.com/pz28olv. Accessed January 27, 2015. Paris: OECD Publishing; 2011.
  10. Wagstaff A, van Doorslaer E. Income inequality and health: what does the literature tell us? Annu Rev Public Health 2000;21:543-67.
  11. Lynch J, Smith GD, Harper S, et al. Is income inequality a determinant of population health? Part 1. A systematic review. Milbank Q 2004;82(1):5-99.
  12. Subramanian SV, Kawachi I. Income inequality and health: what have we learned so far? Epidemiol Rev 2004;26:78-91.
  13. Wilkinson RG, Pickett KE. Income inequality and population health: a review and explanation of the evidence. Soc Sci Med 2006;62(7):1768-84.
  14. Macinko JA, Shi L, Starfield B, et al. Income inequality and health: a critical review of the literature. Med Care Res Rev 2003;60(4):407-52.
  15. Wilkinson RG, Pickett KE. Income inequality and social dysfunction. Annu Rev Sociol 2009;35:493-512.
  16. Loftin C, Hill RH. Regional subculture and homicide: an examination of the Gastil-Hackney thesis. Am Sociol Rev 1974;39:714-24.
  17. Rodgers GB. Income and inequality as determinants of mortality: an international cross-section analysis. Popul Stud 1979;33:343-51.
  18. Hsieh C-C, Pugh MD. Poverty, income inequality, and violent crime: a meta-analysis of recent aggregate data studies. Criminal Justice Rev 1993;18:182-202.
  19. Wilkinson RG. National mortality rates: the impact of inequality? Am J Public Health 1992;82(8):1082-4.
  20. Wilkinson RG. Income distribution and life expectancy. Br Med J 1992;304(6820):165-8.
  21. Wilkinson RG, Kawachi I, Kennedy B. Mortality, the social environment, crime and violence. Soc Health Illness 1998;20(5):578-97.
  22. Wilson M, Daly M. Life expectancy, economic inequality, homicide, and reproductive timing in Chicago neighbourhoods. Br Med J 1997;314(7089):1271-4.
  23. Pickett KE, Kelly S, Brunner E, et al. Wider income gaps, wider waistbands? An ecological study of obesity and income inequality. J Epidemiol Community Health 2005;59(8):670-4.
  24. Pickett KE, Mookherjee J, Wilkinson RG. Adolescent birth rates, total homicides, and income inequality in rich countries. Am J Public Health. 2005;95(7):1181-3.
  25. Pickett KE, James OW, Wilkinson RG. Income inequality and the prevalence of mental illness: a preliminary international analysis. J Epidemiol Community Health 2006;60(7):646-7.46
  26. Wilkinson RG, Pickett KE. Income inequality and socioeconomic gradients in mortality. Am J Public Health 2008;98(4):699-704.
  27. Pickett KE, Wilkinson RG. Child wellbeing and income inequality in rich societies: ecological cross sectional study. Br Med J 2007;335:1080.
  28. Wilkinson RG, Pickett KE. The problems of relative deprivation: why some societies do better than others. Soc Sci Med 2007;65(9):1965-78.
  29. Popper K. The logic of scientific discovery. Abingdon: Routledge; 2002.
  30. Popper K. Conjectures and refutations: the growth of scientific knowledge. Abingdon: Routledge; 2014.
  31. Kaplan GA, Pamuk ER, Lynch JW, et al. Inequality in income and mortality in the United States: analysis of mortality and potential pathways. Br Med J 1996;312(7037):999-1003.
  32. Kennedy BP, Kawachi I, Prothrow-Stith D. Income distribution and mortality: cross sectional ecological study of the Robin Hood index in the United States. Br Med J 1996;312(7037):1004-7.
  33. Pei X, Rodriguez E. Provincial income inequality and self-reported health status in China during 1991-7. J Epidemiol Community Health 2006;60:1065-9.
  34. Rufrancos H, Power M, Pickett KE, et al. Income inequality and crime: a review and explanation of the time–series evidence. Sociol Criminol 2013;1:103.
  35. Wilkinson RG. Unhealthy societies: the afflictions of inequality. London: Routledge; 1996.
  36. Kawachi I, Kennedy BP, Lochner K, et al. Social capital, income inequality, and mortality. Am J Public Health 1997;87(9):1491-8.
  37. Wilkinson RG, Pickett K. The spirit level: why equality is better for everyone. London: Penguin; 2010.
  38. Pickett KE, Wilkinson RG. Inequality: an underacknowledged source of mental illness and distress. Br J Psychiatry 2010;197:426-8.
  39. Messias E, Eaton WW, Grooms AN. Economic grand rounds: income inequality and depression prevalence across the United States: an ecological study. Psychiatr Serv 2011;62(7):710-2.
  40. Burns JK, Tomita A, Kapadia AS. Income inequality and schizophrenia: increased schizophrenia incidence in countries with high levels of income inequality. Int J Soc Psychiatry 2014;60(2):185-96.
  41. Johnson SL, Wibbels E, Wilkinson R. Economic inequality is related to cross national prevalence of psychotic symptoms. Forthcoming.
  42. Smoking and health: report of the Advisory Committee to the Surgeon General of the United States. Washington, DC: Department of Health, Education, and Welfare, 1964.
  43. Hill AB. The environment and disease: association or causation? Proc Royal Soc Med 1965;58(5):295-300.
  44. Bhopal RS. Concepts of epidemiology: an integrated introduction to the ideas, theories, principles, and methods of epidemiology. Oxford: Oxford University Press; 2002.
  45. Rothman KJ, Greenland S, Lash TL. Modern epidemiology. Baltimore, MD: Lippincott Williams & Wilkins; 2008.
  46. Gordis L. Epidemiology. 5th ed. Philadelphia: Elsevier Saunders; 2013.
  47. Babones SJ. Income inequality and population health: correlation and causality. Soc Sci Med 2008;66(7):1614-26.
  48. Flegg AT. Inequality of income, illiteracy and medical-care as determinants of infant-mortality in underdeveloped-countries. Popul Stud 1982;36(3):441-58.
  49. Hales S, Howden-Chapman P, Salmond C, et al. National infant mortality rates in relation to gross national product and distribution of income. Lancet 1999;354(9195):2047.
  50. Drain PK, Smith JS, Hughes JP, et al. Correlates of national HIV seroprevalence: an ecologic analysis of 122 developing countries. J Acquir Immune Defic Syndr 2004;35(4):407-20.
  51. Fajnzylber P, Lederman D, Loayza N. Inequality and violent crime. J Law Econ 2002;45:1-40.
  52. Marmot M, Bobak M. International comparators and poverty and health in Europe. Br Med J 2000;321(7269):1124-8.
  53. Biggs B, King L, Basu S, et al. Is wealthier always healthier? The impact of national income level, inequality, and poverty on public health in Latin America. Soc Sci Med 2013;71(2):266-73.
  54. De Maio FG, Linetzky B, Ferrante D, et al. Extending the income inequality hypothesis: ecological results from the 2005 and 2009 Argentine National Risk Factor Surveys. Glob Public Health 2012;7(6):635-47.
  55. Daly M, Wilson M, Vasdev S. Income inequality and homicide rates in Canada and the United States. Can J Public Health 2001;43(2):219-36.
  56. Rasella D, Aquino R, Barreto ML. Impact of income inequality on life expectancy in a highly unequal developing country: the case of Brazil. J Epidemiol Community Health 2013;67(8):661-6.
  57. Subramanian SV, Delgado I, Jadue L, et al. Income inequality and health: multilevel analysis of Chilean communities. J Epidemiol Community Health 2003;57(11):844-8.
  58. Larrea C, Kawachi I. Does economic inequality affect child malnutrition? The case of Ecuador. Soc Sci Med 2005;60(1):165-78.
  59. Rajan K, Kennedy J, King L. Is wealthier always healthier in poor countries? The health implications of income, inequality, poverty, and literacy in India. Soc Sci Med 2013;88:98-107.
  60. De Vogli R, Mistry R, Gnesotto R, et al. Has the relation between income inequality and life expectancy disappeared? Evidence from Italy and top industrialised countries. J Epidemiol Community Health 2005;59(2):158-62.
  61. Kondo N, Kawachi I, Subramanian SV, et al. Do social comparisons explain the association between income inequality and health?: Relative deprivation and perceived health among male and female Japanese individuals. Soc Sci Med 2008;67(6):982-7.
  62. Walberg P, McKee M, Shkolnikov V, Chenet L, Leon DA. Economic change, crime, and mortality crisis in Russia: regional analysis. Br Med J 1998;317(7154):312-8.
  63. Kondo N, van Dam RM, Sembajwe G, et al. Income inequality and health: the role of population size, inequality threshold, period effects and lag effects. J Epidemiol Community Health 2012;66(6):e11.82
  64. Chen Z, Gotway Crawford CA. The role of geographic scale in testing the income inequality hypothesis as an explanation of health disparities. Soc Sci Med. 2012;75(6):1022-31.
  65. Franzini L, Ribble J, Spears W. The effects of income inequality and income level on mortality vary by population size in Texas counties. J Health Soc Behav 2001;42(4):373-87.
  66. Rostila M, Kolegard ML, Fritzell J. Income inequality and self-rated health in Stockholm, Sweden: a test of the 'income inequality hypothesis' on two levels of aggregation. Soc Sci Med 2012;74(7):1091-8.
  67. Kondo N, Sembajwe G, Kawachi I, et al. Income inequality, mortality, and self rated health: meta-analysis of multilevel studies. Br Med J 2009;339:b4471.
  68. Torre R, Myrskylä M. Income inequality and population health: an analysis of panel data for 21 developed countries, 1975-2006. Popul Stud 2014;68(1):1-13.
  69. Zheng H. Do people die from income inequality of a decade ago? Soc Sci Med 2012;75(1):36-45.
  70. Entorf H, Spengler H. Socioeconomic and demographic factors of crime in German: evidence frompanel data of the German states. Int Rev Law Econ 2000;20(1):75-106.
  71. Messner SF, Raffalovich LE, Chrock P. Reassessing the cross-national relationship between income inequality and homicide rates: implications of data quality control in the measurement of income distribution. J Quant Criminol 2002;18(4):377-95.
  72. Clarkwest A. Neo-materialist theory and the temporal relationship between income inequality and longevity change. Soc Sci Med 2008;66(9):1871-81.
  73. Smith GD, Hart C, Blane D, et al. Lifetime socioeconomic position and mortality: prospective observational study. Br Med J 1997;314(7080):547-52.
  74. Hoynes HW, Schanzenbach DW, Almond D. Long run impacts of childhood access to the safety net. Working paper 18535. Cambridge, MA: National Bureau of Economic Research; 2012. Available at http://www.nber.org/papers/w18535. Accessed January 27, 2015.
  75. Johnson RC, Schoeni RF. Early-life origins of adult disease: national longitudinal population-based study of the United States. Am J Public Health 2011;101(12):2317-24.
  76. Pickett KE, Wilkinson RG. Child wellbeing and income inequality in rich societies: ecological cross sectional study. Br Med J 2007;335(7629):1080.
  77. Adler NE, Epel ES, Castellazzo G, et al. Relationship of subjective and objective social status with psychological and physiological functioning: preliminary data in healthy white women. Health Psychol 2000;19(6):586-92.
  78. Marmot M. Status syndrome: how your social standing directly affects your health and life expectancy. London: Bloomsbury; 2004.
  79. Singh-Manoux A, Martikainen P, Ferrie J, et al. What does self-rated health measure? Results from the British Whitehall II and French Gazel cohort studies. J Epidemiol Community Health 2006;60(4):364-72.
  80. Doing better for children. Paris: OECD Publishing; 2009.
  81. Bird PK. Social gradients in child health and development in relation to income inequality. Who benefits from greater income equality? York: University of York; 2014.
  82. Barford A, Dorling D, Pickett K. Re-evaluating self-evaluation. A commentary on Jen, Jones, and Johnston (68:4, 2009). Soc Sci Med 2010;70(4):496-7; discussion 8-500.
  83. Dorling D, Barford A. The inequality hypothesis: thesis, antithesis, and a synthesis? Health Place 2009;15(4):1166-9; discussion 3-5.
  84. World development indicators: distribution of income or consumption. Washington, DC: World Bank; 2014. Available at http://wdi.worldbank.org/table/2.9. Accessed January 27, 2015.
  85. Smeeding T, Morelli S, Thompson J. Recent trends in income inequality in the developed countries. Progressive Economy 2014;March(2):24-33.
  86. Elstad J. Does the socioeconomic context explain both mortality and income inequality? Prospective register- based study of Norwegian regions. Int J Equity Health 2011;10(1):1-11.
  87. Aittomaki A, Martikainen P, Rahkonen O, et al. Household income and health problems during a period of labour-market change and widening income inequalities - a study among the Finnish population between 1987 and 2007. Soc Sci Med 2014;100:84-92.
  88. Hamilton TG, Kawachi I. Changes in income inequality and the health of immigrants. Soc Sci Med 2013;80:57-66.
  89. Auger N, Hamel D, Martinez J, et al. Mitigating effect of immigration on the relation between income inequality and mortality: a prospective study of 2 million Canadians. J Epidemiol Community Health 2012;66(6):e5.
  90. Bezruchka S, Namekata T, Sistrom MG. Improving economic equality and health: the case of postwar Japan. Am J Public Health 2008;98:216-21.
  91. United Nations Department of Economic and Social Affairs (UN DESA) Population Division. World population prospects, the 2010 revision. New York, NY: United Nations, 2011.
  92. Deaton A, Lubotsky D. Mortality, inequality and race in American cities and states. Soc Sci Med 2003;56(6):1139-53.
  93. Ash M, Robinson DE. Inequality, race, and mortality in U.S. cities: a political and econometric review of Deaton and Lubotsky (56:6, 1139-1153, 2003). Soc Sci Med 2009;68(11):1909-13.
  94. Ram R. Income inequality, poverty, and population health: evidence from recent data for the United States. Soc Sci Med 2005;61(12):2568-76.
  95. Subramanian SV, Kawachi I. The association between state income inequality and worse health is not confounded by race. Int J Epidemiol 2003;32(6):1022-8.
  96. Subramanian S, Kawachi I. Response: in defence of the income inequality hypothesis. Int J Epidemiol 2003;32(6):1037-40.
  97. Ram R. Further examination of the cross-country association between income inequality and population health. Soc Sci Med 2006;62(3):779-91.
  98. Daly M, Wilson M. Cultural inertia, economic incentives, and the persistence of "southern violence". In: Schaller M, Norenzayan A, Heine S, et al (Eds): Evolution, culture, and the human mind. New York, NY: Psychology Press; 2010.
  99. Kennedy BP, Kawachi I, Lochner K, et al. (Dis)respect and black mortality. Ethn Dis 1997;7(3):207-14.
  100. Andersen R, Curtis J. The polarizing effect of economic inequality on class identification: evidence from 44 countries. Res Soc Stratif Mobil 2012;30(1):129-41.
  101. Wolfson M, Kaplan G, Lynch J, et al. Relation between income inequality and mortality: empirical demonstration. Br Med J 1999;319(7215):953-5.
  102. Wood AM, Boyce CJ, Moore SC, et al. An evolutionary based social rank explanation of why low income predicts mental distress: a 17 year cohort study of 30,000 people. J Affect Disord 2012;136(3):882-8.
  103. Subramanian SV, Kawachi I. Whose health is affected by income inequality? A multilevel interaction analysis of contemporaneous and lagged effects of state income inequality on individual self-rated health in the United States. Health Place 2006;12(2):141-56.
  104. Lynch JW, Smith GD, Kaplan GA, et al. Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. Br Med J 2000;320(7243):1200-4.
  105. Elgar FJ. Income inequality, trust, and population health in 33 countries. Am J Public Health 2010;100(11):2311-5.
  106. Elgar FJ, Aitken N. Income inequality, trust and homicide in 33 countries. Eur J Public Health 2011;21(2):241-6.
  107. Layte R. The association between income inequality and mental health: testing status anxiety, socialcapital, and neo-materialist explanations. Eur Soc Rev 2012;28(4):498-511.
  108. Delhey J, Dragolov G. Why inequality makes Europeans less happy: the role of distrust, status anxiety, and perceived conflict. Eur Sociol Rev 2014;30(2):151-65.
  109. Lancee B, Van de Werfhorst HG. Income inequality and participation: a comparison of 24 European countries. Soc Sci Res 2012;41:1166-78.
  110. Uslaner EM, Brown M. Inequality, trust, and civic engagement. Am Politics Res 2005;33(6):868-94.
  111. Sapolsky R. Sick of poverty. Sci Am 2005;293(6):92- 9.
  112. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Med 2010;7(7):e1000316.
  113. Cohen S, Doyle WJ, Skoner DP, et al. Social ties and susceptibility to the common cold. JAMA 1997;277:1940-44.
  114. Kiecolt-Glaser JK, Loving TJ, Stowell JR, et al. Hostile marital interactions, proinflammatory cytokine production, and wound healing. Arch Gen Psychiatry 2005;62(12):1377-84.
  115. Dickerson SS, Kemeny ME. Acute stressors and cortisol responses: a theoretical integration and synthesis of laboratory research. Psychol Bull 2004;130(3):355-91.
  116. Russ TC, Stamatakis E, Hamer M, et al. Association between psychological distress and mortality: individual participant pooled analysis of 10 prospective cohort studies. Br Med J 2012;345.
  117. Mitchell C, Hobcraft J, McLanahan SS, et al. Social disadvantage, genetic sensitivity, and children's telomere length. Proc Natl Acad Sci U S A 2014;111(16):5944-9.
  118. Eisenberger NI, Lieberman MD, Williams KD. Does rejection hurt? An FMRI study of social exclusion. Science 2003;302(5643):290-2.
  119. Eisenberger NI, Jarcho JM, Lieberman MD, et al. An experimental study of shared sensitivity to physical pain and social rejection. Pain 2006;126(1-3):132-8.
  120. DeWall CN, MacDonald G, Webster GD, et al. Acetaminophen reduces social pain behavioral and neural evidence. Psychol Sci 2010;21(7):931-7.
  121. DeWall CN, Deckman T, Pond RS, Jr., et al. Belongingness as a core personality trait: how social exclusion influences social functioning and personality expression. J Pers 2011;79(6):1281-314.
  122. Johnson SL, Leedom LJ, Muhtadie L. The dominance behavioral system and psychopathology: evidence from self-report, observational, and biological studies. Psychol Bull 2012;138(4):692-743.
  123. Preston SH. The changing relation between mortality and level of economic development. Popul Stud 1975;29(2). Reprinted in: Int J Epidemiol 2007;36(3):484-90.
  124. Cutler D, Deaton A, Lleras-Muney A. The determinants of mortality. J Econ Perspect 2006;20(3):97-120.
  125. Kenny C. Getting better: why global development is succeeding—and how we can improve the world even more. New York: Basic Books; 2011.
  126. Wilkinson R, Pickett K. The spirit level: why more equal societies almost always do better. London: Penguin; 2010.
  127. Layte R, Whelan C. GINI DP 78: Who feels inferior? A test of the status anxiety hypothesis of social inequalities in health. Amsterdam: Amsterdam Institute for Advanced Labour Studies; 2013.
  128. Delhey J, Dragolov G. Why inequality makes Europeans less happy: the role of distrust, status anxiety, and perceived conflict. Eur Sociol Rev 2014;30(2):151-65.
  129. Loughnan S, Kuppens P, Allik J, et al. Economic inequality is linked to biased self-perception. Psychol Sci 2011;22(10):1254-8.
  130. de Vries R, Gosling S, Potter J. Income inequality and personality: are less equal U.S. states less agreeable? Soc Sci Med 2011;72(12):1978-85.
  131. Paskov M, Dewilde C. Income inequality and solidarity in Europe. Res Soc Strat Mob 2012;30(4):415-32.
  132. Daly MC, Oswald AJ, Wilson D, et al. Dark contrasts: the paradox of high rates of suicide in happy places. J Econ Behav Org 2011;80(3):435-42.
  133. Burrows S, Laflamme L. Socioeconomic disparities and attempted suicide: state of knowledge and implications for research and prevention. Int J Inj Contr Saf Promot 2010;17(1):23-40.
  134. Kondo N, van Dam RM, Sembajwe G, et al. Income inequality and health: the role of population size, inequality threshold, period effects and lag effects. J Epidemiol Community Health 2011;66(6):e11.145.
  135. Pickett KE, Wilkinson RG. Greater equality and better health. Br Med J 2009;339:b4320.
  136. The cost of inequality. London: The Equality Trust; 2014.
  137. Nowatzki NR. Wealth inequality and health: a political economy perspective. Int J Health Serv 2012;42(3):403-24.
  138. Gilens M, Page BI. Testing theories of American politics: elites, interest groups, and average citizens. Perspect on Politics 2014;12(3):564-81.
  139. Chemi EG, Giorgi A. The pay-for-performance myth. Bloomberg Business Week, July 222, 2014. Available at http://www.businessweek.com/articles/2014-07-22/for-ceos-correlation-between-pay-and-stock-performance-is-pretty-random. Accessed January 27, 2015.
  140. Piketty T, Saez E, Stantcheva S. Optimal taxation of top labor incomes: a tale of three elasticities. Cambridge, MA: National Bureau of Economic Research; 2011.
  141. Tosi HL, Werner S, Katz JP, et al. How much does performance matter? A meta-analysis of CEO pay studies. J Manage 2000;26(2):301-39.
  142. Schrecker T. Can health equity survive epidemiology? Standards of proof and social determinants of health. Prev Med 2013;57(6):741-4.
  143. Marmot M. Inequalities in health: causes and policy implications. In: Tarlov A, St. Peter R (Eds): The society and population health reader, vol 2. A state and community perspective. New York: New Press; 2000.

a Refer to Appendix: Summary of Studies of Income Inequality and Health. Available at http://www.york.ac.uk/healthsciences/our-staff/kate-pickett/#publications


Kate Pickett Kate Pickett, FRSA, FFPH, is Professor of Epidemiology in the Department of Health Sciences at the University of York and previously was a National Institute for Health Research Career Scientist. She co-authored (with Richard Wilkinson) The Spirit Level: Why More Equal Societies Almost Always Do Better and is a co-founder of The Equality Trust. She was awarded a 2013 Silver Rose Award from Solidar for championing equality and the 2014 Charles Cully Memorial Medal by the Irish Cancer Society.
Richard Wilkinson Richard Wilkinson, FRSA is Emeritus Professor of Social Epidemiology at the University of Nottingham, and Honorary Professor at UCL and the University of York. He co-authored (with Kate Pickett) The Spirit Level: Why More Equal Societies Almost Always Do Better and is a co-founder of The Equality Trust. He was awarded a 2013 Silver Rose Award from Solidar for championing equality and the 2014 Charles Cully Memorial Medal by the Irish Cancer Society.
Page last reviewed July 2015
Page originally created August 2015
Internet Citation: Income Inequality and Health: A Causal Review. Content last reviewed July 2015. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/professionals/education/curriculum-tools/population-health/pickett.html
Back To Top