How Does Institutional Change Produce Health Disparities?

An Interview with Dr. Mark Hayward
By Dr. Christine Bachrach

In April, Mark Hayward, Centennial Commission Professor in the Liberal Arts at the University of Texas at Austin, gave the prestigious Matilda White Riley Excellence Lecture at the National Institutes of Health.  His talk focused on the need to understand how institutional change shapes the evolution of health disparities and the specific forms they take.  I interviewed Mark, a former member of the IAPHS Board, about how his ideas have evolved.  

Chris: In your talk, you made a case for linking the growing educational gaps in life expectancy in the white U.S. population with the devolution of power to states.  What prompted you to develop these ideas?

Mark: It was an interesting journey.  We started out with classic descriptive demography, tracking and disaggregating the trends.  Others had shown there was a growing educational gradient, but we wanted to know what was driving it.  We knew that life expectancy was increasing for the well-educated,  but declining for those with high school educations or less.  Further, while 20 years ago education effects were tied to completing high school or college, we now were seeing a dose-response effect among those with more than 12 years of education.  Each additional year of education brought reduced mortality above and beyond the effect of earning a postsecondary degree.

Chris:  What did that suggest to you?

Mark: Something happened over that time period, and we reasoned it was an external force affecting how education translates into mortality risk.  Most people assume that individual-level determinants of mortality are constant over time, but they are not – they depend on environmental conditions.  When we started thinking that way, we turned to economic historians –people like Easterlin, Fogel, and Costa – and their ideas about how technological and social trends affect people’s life chances.  Going back to Link and Phelan’s concept of social and economic disadvantage as fundamental causes of health, my colleagues and I recognized that these trends might also affect health disparities.  As the environment changes, the value of education as a resource for garnering health advantages also may change.

Chris: How did you come to think about devolution of power to U.S. states as a key factor?

Mark: Initially, we were looking at technological advances as an environmental driver, but at some point we began to notice that geography mattered a lot. We observed an enormous disparity across states in adult mortality that never existed prior to the 1970s. This was a tectonic shift – the gap in life expectancy that has opened up among our states is much greater than the gap among all advanced European countries.  In work I’ve done with Jennifer Montez and Anna Zajacova, we found that while there were almost no state variations in disability among college graduates, variations were substantial for the less well educated.  

Chris: So what was happening at the state level to produce these differences?

Mark: We are thinking that the New Federalism – the devolution of control over federal programs to the states – is one fruitful handle for exploring state variations.  States became responsible for major social programs in the 1980s, and welfare reforms in the 1990s added to this trend of state control.  This has led to big state differences in program coverage and generosity. In addition, state-level Earned Income Tax Credit (EITC) programs, an important policy option for combating poverty, exist in only half the states and vary a lot in levels of generosity.  State differences in cigarette taxes have also grown rapidly since the 1990s.  And there are many other policy differences to look at: education investments, environmental justice, child welfare, health care, and public health policies.  These policies often don’t arise independently, so it is very hard to parse out which is most important.  We are not making a causal argument at this point.

Chris: A recent Washington Post article pointed to declines in life expectancy among adults aged 25-44 in all but two states and all racial and ethnic groups except Asian Americans.  This suggests that the most recent trends require us to look at minority populations and younger age groups as well as the populations gaining the most attention in recent research.  What are your thoughts on this and will your state-level approach be useful here as well?

Mark: The recent trends are alarming, and researchers absolutely should study them.  The reach of the opioid epidemic is a very likely suspect, but this is probably not acting in isolation.  There are state variations in the mortality increases among young adults that need studying, and there are critical questions to be asked about racial and ethnic differences.  For example, in our data, black-white mortality differences occur mostly among the most highly educated – there are no differences at the less than high school level.  Why is this? It could be discrimination; it may relate to childhood factors, since many highly educated African Americans were born in the Jim Crow South.  These are complex questions and they need to be thoughtfully addressed.

Chris: Mark, you’ve just received a career achievement award named after Matilda White Riley. What advice would she have given to researchers trying to understand how health disparities evolve in a changing context?

Mark:  I can’t say how much OBSSR’s award meant to me. Matilda had an enormous impact on my career.  Her advice was always to look at the macro environment and the changing institutional environment in which people live, work, and play.  That’s where you’ll find the clues to evolving health disparities.