Lynch and others reviewed 98 studies examining the association between income equality and health. They found little support for the idea that income inequality plays a major role in population health differences between or within most countries. Support was strongest for a link in the United States, however, so they completed a review of 100-year U.S. national and 30-year U.S. regional trends in mortality and compared those trends with U.S. income inequality. The analysis showed little support for a link between income inequality in the United States and population health over time, except for mortality from suicide and homicide. The analyses appear in The Milbank Quarterly (2004;82(1):5-99 and 2004;82(2):355-400). Lynch and co-author George Davey Smith are recipients of Robert Wood Johnson Investigator Awards in Health Policy Research.
Q: First, a few definitions. In your abstract you mention "rich" countries. How do you define "rich" and "poor"?
A: If you look at per capita GDP [gross domestic product], 20 or 30 countries stand out as being much richer than the rest. There's a well-known association between per capita GDP and life expectancy in poorer countries. In richer countries, variation in population health is not strongly linked to income. Trying to explain that variation is part of what led to this area of research.
Q: Can you define "health" as you used the term in your research?
A: Up front in our paper, we say that we need to think specifically about health, and not in terms of generalities. The tendency is to talk about health as this generic thing.
When you're studying the relationship between income inequality and health, you need to be specific in terms of what health outcome you're talking about: Is it suicide, breast cancer, prostate cancer, heart disease, depression? The way we structure societies might have different effects on alcoholism than it will on breast cancer, for example, and the relationship between income inequalities may be different for each disease or condition. Some processes linked to income inequality have more plausible explanations for certain outcomes than others. So if lower-income people get stressed because they're not doing as well as others, that might show up in the rates of violence or suicide, but not so much in the rate of breast cancer, for example.
Q: What is the income inequality-health hypothesis?
A: We all know that being rich or poor affects your health as an individual: we know the rich do better. The income-inequality hypothesis says that it's the size of the gap between the rich or poor that makes a difference, regardless of any one person being rich or poor. Think of it as the gap between a ceiling and a floor. Maybe in country A there's seven feet between the ceiling and the floor; in country B, the floor is at the same level but the ceiling is higher. The income inequality hypothesis says that country B will have poorer health overall.
Q: What are the mechanisms behind that relationship?
A: There are two schools of thought. One is a psychosocial hypothesis: people make social comparisons, and poorer people see richer people doing better and that essentially makes poorer people feel bad. It causes stress and anxiety that could lead to increased depression, violence, poorer nutrition and other negative results.
The other hypothesis says that the income inequality gap is a marker for other social policies. For example, take two countries—the United States and Sweden. The income gap is smaller in Sweden, but there are also lots of other differences between the countries: social, economic and health policy differences, to name a few. So when you see better health in Sweden, it may not be due to income inequality but may indicate that income inequality is a marker for other social investments the Swedes make to improve health.
Q: Can you give us an overview of the findings?
A: Part 1 was the first systematic review in this area. We were motivated to do it because no one had gone through every single study and reviewed them all. While we may have missed one or two, the omission of those would not seriously affect the overall pattern of the findings.
The review shows that the evidence is limited for international differences, across countries, for income inequality being related to health. When you look at countries other than the United States—Canada, Spain, Australia, Sweden, and others—you don't see a relationship. It may be that an inequality threshold must be reached before you see an effect, because among the richest countries, the United States is the country with the most income inequality.
Then in Part 2 we say well, if this relationship holds in the United States, what does it look like over time? If income inequality really is a major determinant of population health, then when you change income inequality, health should change. We looked at that idea over a century, and it was difficult to generate much support for a relationship. We looked at many different causes of death, and except for suicide and homicide, it was difficult to generate support for a relationship between income inequality and mortality.
For example: the largest decline in income inequality was from the end of the Depression to World War II, and we did not see a resulting increase in population health outcomes. However, at that same time, smoking underwent a huge surge in popularity. I've heard people make the argument that reducing the income equality helped made cigarettes more available to more people and so declining income inequality may have had a perverse effect on population health because it helped fuel the rise in smoking.
Q: So in the end, there's no solid link between the size of the gap between the rich and the poor in a country and the health of that country's population?
A: Back to that ceiling and floor analogy: I think we are saying that as long as the floor is high enough and extensive enough—as long as there is a minimum standard of wages, health care, job conditions—then overall the population's health will be OK. If the social floor is high enough, the ceiling probably doesn't matter too much. That's the big story for population health.
The other theme is that sometimes when we look at things in one point in time, the relationship looks very promising and we can make up all kinds of plausible stories about how it works. But when we look at the same things over time, it's hard to see how there's a plausible relationship. That's the case with the relationship between income inequality and health. To link income inequality changes to changes in health, income inequality has to be linked somehow to risk factors for major disease outcomes over time, and we've seen that it just isn't, except for maybe suicide and homicide.
Q: Why measure income inequality, rather than education or some other measure?
A: Income inequality is only one measure of the amount of inequality in a society. It's a convenient thing to study because income data are easily available. You could easily talk about education inequality in the same way, but that's trickier to measure and compare, both within and across countries. At the individual level, education is an important precursor to income. At a social level, when you're talking about income inequality in a society, income is a convenient marker for education.
I think education is likely to be very important, however. There's one study in the United States that shows that the effects of income inequality on mortality disappear when you adjust for educational differences.
Q: Why is it important to do these kinds of studies?
A: If you're interested in maximizing population health, you want to know the kinds of social arrangements that will help you do so as well as what societal characteristics are conducive to better health. People talk about democratic structures and access to health care—are these systems really conducive to better health? One element of answering that question involves finding out what size gap between the rich and poor is a country willing to tolerate, and how does it impact the health of the people there. The United States has high income inequality amongof all industrialized nations, and this research was motivated by that fact—we were trying to find out how income inequality is associated with health.