Does Income Inequality Cause Poor Health?

Paul Kelleher caught some flack for a blog post last week in which he approvingly cited a 2003 study by Angus Deaton and Darren Lubotsky (DL) that supposedly refutes the idea that income inequality causes poor health. I was curious. Paul sent me the original paper, a 2009 published critique by Ash and Robinson (AR), a reply by DL, and a longer article by Deaton.

Even though the original DL paper is getting old, and the field has progressed substantially since 2003, there is still an important take home message from this debate. Stated simply: income inequality may be one cause of poor health, but income inequality is also likely to be a mediator of other upstream causes of mortality.

The Original Finding and Interpretation

In the original DL paper the authors regress cross-sectional age-adjusted mortality data in U.S. states and Metropolitan Statistical Areas (MSAs) on the region’s measured income inequality. (They focus on the Gini coefficient, but find that the results hold across many summary statistics.) They then take an extra step of introducing a control for the fraction of the MSA that is black during the time period. They summarize their findings in the abstract:

“Across states and Metropolitan Statistical Areas (MSAs), the fraction of the population that is black is positively correlated with average white incomes, and negatively correlated with average black incomes. Between-group income inequality is therefore higher where the fraction black is higher, as is income inequality in general. Conditional on the fraction black, neither city nor state mortality rates are correlated with income inequality. Mortality rates are higher where the fraction black is higher, not only because of the mechanical effect of higher black mortality rates and lower black incomes, but because white mortality rates are higher in places where the fraction black is higher.”

Below is the key table from their study. As the table shows, in the pooled data, the Gini coefficient is a risk factor for mortality in regressions that do not control for fraction black. Once this is controlled for the effect is actually reversed (for example, going from .55 to -.38 for all males. The same basic result holds for the pooled data for females.

One puzzling finding of the DL study is why controlling for the share of blacks would also attenuate the associations between Gini and mortality within whites shown in the righthand panel. If the inequality-mortality relationship were merely an artifact of racial composition, there is no clear reason why the racial composition effect would persist after stratifying on race. Here is what DL say in the conclusion of their paper:

“The fraction of the population that is black is a significant risk-factor for mortality, not only for the population as a whole—which would follow mechanically from the fact that blacks have higher mortality rates than whites—but for both blacks and whites separately. Our empirical results provide no evidence that the association between the fraction black and white mortality is the result of confounding. The effect is robust to conditioning on education, it is present for all age-groups except boys aged 1 to 9, and it is present within geographical regions of the country.”

Is Racial Composition Conceptually Distinct from Income Inequality?

In a reply article Ash and Robinson challenge not only the original analysis but also the substantive interpretation DL offer of their results. First, they argue that the finding that fraction black attenuates the association of the Gini on mortality within whites should cause us to question the underlying social mechanisms in these regions: “Death by human agency is rare; the increased black presence is not ‘‘causing’’ white deaths in any direct way. The presence of blacks in a metropolitan area reflects the presence of social determinants that are detrimental to population health, and it is these factors that warrant identification.”

Second, and relatedly, they argue that “the ratio of blacks to whites in a given locality… captures something about the dynamics of political and economic power in that locality” pointing to historical policies that have contributed to the dominance of blacks by whites, including slavery, Jim Crow segregation, and residential polarization. In the south, where most American blacks live, many political scientists argue that efforts to enact progressive social policies have been stymied by aversion by whites of all income groups against redistribution to blacks, this political dynamic prevents lower-class whites and blacks from banding together to advocate for policies that would jointly promote their wellbeing and improve their health.

I agree with Ash and Robinson, but I think they could have made their point more clearly. I think what we should say is that racial composition is a legacy of particular economic and political processes that are themselves conceptually distinct but related to many of the processes that generate economic inequality. In some settings, and in some regions, they are likely to be more salient than others, and indeed the data from both the DL and AR results shows some variation in the effect of racial composition in different regions of the United States.

DL are absolutely wrong to call racial composition a confounder in their 2003 paper however, and I think their 2009 reply paper is much more careful. To call an omitted variable a confounder is to claim that the explanatory variable (in this case income inequality) is caused by the omitted variable and the omitted variable also causes the outcome (mortality), but that the explanatory variable does not cause the outcome. We should always include confounders in the regression equation because otherwise we will obtain biased estimates of the relationship of the explanatory variable on the outcome.

If Ash and Robinson are right, we should think of income inequality as a mediator of racial composition and mortality, that is racial composition is a causally antecedent variable (at least in some settings), and its effect is transmitted through income inequality. Whether to control directly for either racial composition or income inequality in a regression is therefore a non-obvious question that depends on the type of inference the investigator is interested in drawing. One should only include both of them in the regression when they are both intrinsically of interest, but even so it is generally not possible to identify the independent effect of each with observational data. Better yet, we would hope to find some instrumental variable that allows us to separate the exogenous effect of racial composition from income inequality. But to get to this point we need to have a much deeper and richer understanding of what processes historically have regulated racial politics and resource redistribution in American regions.


About Brendan Saloner

I am a postdoctoral fellow at the University of Pennsylvania in the Robert Wood Johnson Health and Society Scholars Program. I completed a PhD in health policy at Harvard in 2012. My current research focuses on children's health, public programs, racial/ethnic disparities, and mental health. I am also interested in justice and health care.
This entry was posted in Blog posts and tagged , , , . Bookmark the permalink.

14 Responses to Does Income Inequality Cause Poor Health?

  1. Paul Kelleher says:

    Thanks, B. On vaca writing this from a mobile device so sorry for typos.

    I just want to note that the “approving” reference of mine was more to the Leigh, Jencks, Smeeding survey. I wanted to highlight DL as a study bearing on what LJS say is the only “equivocal” case of income inequality potentially causing mortality.

    Finally, I can’t tell from your post (bc I don’t know all the terminology). Are you in the end disputing DL’s conclusion: “We regard this result as showing that there is no direct effect of income inequality on health”?

  2. Mel Bartley says:

    for a very lucid account of many of the issues raised here

  3. Mel Bartley says:

    see also
    N Kondo et al. “Income inequality, mortality and self rated health: meta-analysis of multilevel studies” BMJ 2009; 339 doi:10.1136/bmj.b4471

  4. Hi Paul,

    Sorry for misquoting you on that bit — I need to read that survey article, but I’m sure it’s good.

    Yeah, thanks I should have been clearer. I do think there is a direct effect of income inequality on poor health, but as I was trying to say it’s hard to quantify the size of that effect because some of it is transmitted from downstream causes like racial politics and social norms. I do instinctively resist the relative deprivation hypothesis, at least I don’t think inequality per se is toxic (like some chemical exposure), but depends on a very particular set of institutions and historical context.


  5. Hi Mel,

    Many thanks for the helpful references. I look forward to reading some of the comprehensive reviews on this topic (it’s definitely not my area of expertise).


  6. Jim Dunn says:

    The Deaton finding is a curious one, but I have always been deeply troubled about two aspects of it, both relating to the fact that the analysis is severely ‘under-socialized’ and quite superficial (although statistically impressive). The first problem is this: %Black at the state or metro level is clearly implied to be a marker for _something_, but what? Is a greater proportion Black indicative of less discrimination or more? In the absence of other evidence about the _actual_ causal mechanisms, I could make the case for both. But my guess is that it’s a whole lot more complex than that, and it could even be spurious, or attributable to some third factor associated with all of: %Black, income inequality and mortality. So without evidence of the actual causal mechanisms, then I’m not giving this paper or debate too much credibility. The second issue is American exceptionalism, and I can best illustrate it with an example. In 2005, we published a paper (citation below) examining the relationship between metropolitan income and working-age mortality in 5 countries – Canada, the US, the UK, Sweden and Australia. While under review with another journal, a reviewer, leaning on the Deaton paper, argued that our findings were flawed because we had failed to control for ‘proportion black’. We argued, unsuccessfully(at least to that journal editor), that such a variable was _only_ relevant to the US. I still stand by that position for reasons related to my first point: we don’t know what %Black is a marker for, but I can guarantee you that it is a marker for a _different causal process_ in each of these countries (and may not be a marker for anything at all in some of them). So unless and until someone comes forward with something more credible than the speculation that it may have to do with some vaguely conceived causal mechanism, let’s allow this debate to lie dormant for longer. Racial differences in health care are not a convincing answer and there is no evidence that they can be untangled from income differences in health care in the US. Moreover, to dismiss the income inequality ~ health relationship on the basis of this one analysis is clearly premature – indeed, rather than answering any one question, this analysis raises many more questions, some of them quite important for public health.

    Jim Dunn, CIHR-PHAC Chair in Applied Public Health, McMaster University and William Lyon Mackenzie King Visiting Chair in Canadian Studies, Harvard School of Public Health 2011-12

    Ross, N.A., Dorling, D., Dunn, J.R., Henriksson, G., Glover, J., Lynch, J.W. and Ringback Weitoft, G. (2005) Metropolitan income inequality and working age mortality: A cross-sectional analysis using comparable data from five countries. Journal of Urban Health, 82(1):101-110.

    See also:
    Dunn, J.R., Schaub, P. and Ross, N.A. (2007) Unpacking income inequality and population health in North American Cities: The peculiar absence of geography. Canadian Journal of Public Health, 98(S1): S10-S17.
    Dunn, J.R., Burgess, B. and Ross, N.A. (2005). Income distribution, public services expenditures, and all-cause mortality in U.S. states. Journal of Epidemiology & Community Health, 59(9): 768-774.

    • Hi Jim,

      Thanks so much for your terrific comments. I look forward to checking out your papers. Given all the lit that people are sending my way, perhaps I’ll write a follow-up post.

      I think Ash and Robinson provide some helpful thoughts about how to think about %Black, although I see your point “it could even be spurious, or attributable to some third factor.” I have a few ideas for some other contextual variables related to race that people might look at (perhaps they already have): a dissimilarity index measuring the residential segregation in the MSA, a measure of political participation/representation including voter turnout by race, a measure of racial segregation in hospitals or health care (to directly test the medical apartheid hypothesis), or some composite indicator of racial attitudes (for example, the percent of individuals in the MSA that express feelings of racial distrust or strong aversion to other races). I think these would be more conceptually important than %Black. What do you think?

      I also take your point about American exceptionalism. I had not intended to speak to this question in other countries, but I want to emphasize again my contention that particular institutions and history matter arguably more than the mere presence of income inequality.

      Anyway, those are some thoughts,


  7. Brendan,

    In that case, are you sure you are disagreeing with Deaton/DL? I took them to be arguing that income inequality is not per se bad for health. You say you agree with this, but then you also believe that there is a direct effect from income inequality to health. Is it crucial to characterize that which you say has a direct effect as “income inequality”? Note that DL do NOT dispute (at least not here) the claim that DIFFERENCES in income predict and/or have a direct effect on DIFFERENCES in health. That claim could be true but follow mechanically (to use a phrase of DL’s) from some version of the absolute income hypothesis, right? 

    So it would help me tremendously if you could restate the point on which you differ with DL. I don’t yet quite see the disagreement. 


    • Hi,
      Sorry let me restate/better state some of my points:
      1. I believe that income inequality causes bad health, but probably mainly in interaction with other institutions and social contexts. I can’t prove that, it’s a hunch.
      2. DL say things which seem to commit them to different claims. Sometimes they seem to only be saying that there is no direct effect of income inequality on health, which leaves the door open to mediation, but other times they talk about income inequality as a confounder. The two diagrams I drew try to highlight how their picture would differ from my own. Note that on my picture, there is an arrow going both from composition to health and also an arrow that runs through the income inequality pathway. I have no way of saying which of these relationships is more substantively important, and empirically they are challenging to empirically disentangle. (My comments above to Jim might provide better ways of measuring the racial composition pathway).

      Is this still unclear? Shouldn’t you be sipping a tropical drink instead of blogging??

  8. This is a rich discussion, and I’ve already learned much. I guess one of my difficulties with studies like DL from a methodological standpoint is how to suss out all the different variables. I am no methodologist, but I can’t understand how one would truly be able to disentangle the effects of racial inequalities on health from those of income inequalities, especially because of the life course hypothesis, which suggests that the pathways between inequalities and poor health are embodied from extremely early in the life span (even prenatally).

    My point is not that one can’t operationalize the variables to say something useful about the relationship between various inequalities and health, but rather, and I argue this in a forthcoming piece, the attempt to atomize different demographic characteristics when they are connected in what Powers and Faden term “densely woven patterns of disadvantage” seems quixotic. The conclusion, if this is right, is hardly that social inequalities don’t exist, but that the pathways through which any particular set of such inequalities operate to produce poor health and its distribution are multidimensional, multifactorial, and are intricately bound up with other social disadvantages (because, of course, they cluster).

    My concerns here go beyond the specific DL piece we are discussing, and become rather more important when one tries to control for the SDOH to prove something else has a significant effect on health, whether that something else is “health care services” or the black box epidemiology often used to think about genetic causal contributions to disease. I think it is methodologically difficult if not impossible to control for the SDOH, and more to the point, given the evidence in tot that social and economic conditions have on health and its distribution, I have great difficulty understanding why ever we would wish to do so.

  9. Hi Daniel,

    Thanks for this interesting comment. I think I mostly agree with what you say, both conceptually and methodologically, but I want to make a few comments:
    1. Let’s agree that there are densely woven patterns of disadvantage, and that these disadvantages have a tendency to reinforce one another in terms of health, social stratification, and many other processes we are interested in. That’s an important discovery. Still there are further questions that we might want to know. Historically and sociologically, we might be interested in knowing whether there are certain beliefs and political processes that caused those particular disadvantages, and not others, to become densely woven. Attending to places in which those disadvantages reinforce each other and where they come apart can help us to understand this process better. Policymakers might also want to know how we can unravel certain threads out of the pile. For example, many policies involve single systems like healthcare and from that vantage point, it makes a lot of sense to try to develop interventions that deal with (say) racial discrimination, independent of socioeconomic disadvantage or income inequality. So I don’t like the idea of only saying disadvantages are bounded up together, full stop, I want to know why and what we can do about particular disadvantages.
    2. You also say that it’s methodologically difficult if not impossible to control for the SDOH. For many questions, I agree with that assessment, especially if we’re interested in what economists would call “general equilibrium” effects, the effects of different social or economic forces averaged over the entire population. But sometimes we are interested in something else, we want to know how to address specific disadvantages in specific contexts, and there it is feasible to examine controlled or even natural experiments that exogenously change the influence of a particular factor or set of factors on a group. This is one place where I think policy researchers and SDOH people might sometimes be talking past each other.


  10. Pingback: Health Promotion Headlines from Robyn and Meghan – August 15, 2011 | Health Nexus Today / Nexus Santé aujourd'hui

  11. Pingback: Mind the (Income) Gap « creatingreciprocity

  12. Pingback: Mind the (Income) Gap - creatingreciprocity

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.