“Deaths of Despair” Meets “Black Lives Matter”

How headlines about declining life expectancy distract from the bigger issues.

Posted Jan 16, 2020

Vital statistics are so hot right now. 

No doubt you’ve seen the headlines (like here and here) over the past couple of years describing the declines in overall life expectancy seen in the U.S. since 2014. Those headlines are true: life expectancy (that is, how many years, on average, a person born today can expect to live) declined from 78.8 years in 2014, to 78.7 in 2015 and 2016, to 78.6 in 2017, the most recent year available. That is, about one-tenth of a year each calendar year. This decline comes on the heels of decades of uninterrupted gains in life expectancy.

Kevin Phillips/Pixabay
Yes, there are a lot of numbers in this post. No, this won't be on the test.
Source: Kevin Phillips/Pixabay

I’m not saying two-tenths of a year of life aren’t important—as Pitbull reminds us, “Every day above ground is a great day”—but by the amount of media coverage and alarmist tone you’d think that we were talking about a decline of a decade or two.

What these headlines miss, however, is that this decline in life expectancy was not experienced by all subgroups of the population. In 2015, economists Anne Case and Angus Deaton published an influential paper that examined changes in mortality by race, sex, and socioeconomic status through 2013. This paper was followed by several subsequent publications by these authors and others, which generally report the same set of findings: that the decline in life expectancy seen in the U.S. overall is concentrated among non-Hispanic whites. The causes of death that increased the most in this group during this period were deaths due to unintentional overdoses, suicide, and liver diseases including cirrhosis, which is strongly related to alcohol misuse. For this reason, in policy discussions of their work the authors referred to middle-aged non-Hispanic whites as experiencing “deaths of despair.” You can listen for yourself in this brief video of the authors talking about their findings for the Brookings Institution. 

But what this story fails to address is how life expectancy changed for all other racial/ethnic groups in the U.S. during this same time period, or the inequalities in life expectancy as a function of race that continue to persist. For example, this figure—adapted from briefs produced by the Centers for Disease Control and Prevention (CDC)—shows life expectancy at birth by race and sex.

CDC
Life expectancy by race, ethnicity, and sex: 2006 to 2017
Source: CDC

What this figure shows is that African Americans and Hispanics/Latinos have either seen little or no decline, or actually experienced gains in life expectancy, during this same period as overall life expectancy in the U.S. has fallen. 

But there is a bigger issue that the current framing of the decline in life expectancy is distracting us from acknowledging: the fact that African American men born today have a life expectancy that is, on average, still less than non-Hispanic White men born in 1990. Forget about losses of one-tenth of a year—I’m talking about a difference that is closer to half a decade.

This figure, also adapted from the CDC, summarizes racial, ethnic, and gender differences in mortality in the U.S. It shows in absolute terms both the size of the annual change in mortality from 2016 to 2017 and the racial disparities in mortality when comparing non-Hispanic white men and women and African American men and women.

CDC
Death rates by race and sex: 2016 and 2017
Source: CDC

The increase in the rate of mortality for non-Hispanic white men from 2016 to 2017? That works out to 5.6 deaths per 100,000 people. The difference in the rate of mortality for non-Hispanic white men as compared to non-Hispanic black men in 2017? That works out to 198.2 deaths per 100,000 people.

The final piece of data that I want to address is a bit of good news. This figure illustrates—again, in absolute terms—the white-black mortality gap from 1999 to 2013. This is nearly the same period of time covered by the influential Case & Deaton (2015) paper. It shows that the overall mortality gap in 1999 was nearly six years—that is, at birth, non-Hispanic Whites were expected to live an average of six years longer than non-Hispanic blacks. This gap is now less than four years. 

CDC
Changes in the absolute difference in life expectancy
Source: CDC

What has caused this black-white mortality gap to narrow? Is the gap smaller now because of “diseases of despair” among whites? Not so much.

This last figure shows the contribution of the leading causes of this white-black mortality gap, and how changes in them have contributed to changes in the gap. As you can see, the only “disease of despair” that makes this list is unintentional injuries—which includes accidental overdoses. The majority of the decline in this gap is due to reductions in mortality from heart disease, cancer, HIV, and infant mortality.

CDC
Major contributors to the shrinking black-white mortality gap
Source: CDC

There are still disparities in mortality from these conditions, but it is welcome news that those gaps are narrowing. Yet you wouldn’t know this from the media coverage.

I’m hardly the first person to call for reflection on what's being missed by the news headlines. Ana Diez Roux, Dean of the Dornsife School of Public Health at Drexel University, wrote in the American Journal of Public Health in 2017:

“We should guard against the unintended consequences that the focus on the increase in death rates for some Whites (significant as the are) detract attention from the persistent health inequities by race and social class, which are so large that they dwarf the size of what is a very troublesome increase in some Whites. The act that the despair explanations has been so quickly adopted (with personal responsibility explanation being quickly discarded) to explain the mortality increase in Whites, despite persistent questions of the relevance of factors like despair to Black-White differences, is itself telling. It raises interesting questions about how causal explanations emerge and what influences the speed with which they are adopted.”

She then suggests what some of these questions are: “Why is it that Blacks and Hispanics are somehow protected from the adverse health effects of the adverse economic trends creating despair in Whites? Have they not experienced similar adversity and frustrated expectations? Is frustrated expectation really the driving factor, and is it so differentially distributed by race that recent economic circumstances would cause increases in death rates in Whites but not other groups?”

Those aren’t rhetorical questions, but they are challenging ones. Ones that should encourage us to reflect on what those vital statistics are telling us about life and death in America today. 

And even if we learn the answers to them, the fact will remain that data does not tell a story. Rather, people tell stories with data. Researchers, clinicians, and the media interpret numbers, and that interpretation comes with a lens and perspective that for simplicity generally emphasizes some pieces of data at the expense of others. That emphasis in turn shapes the discussions that policymakers have when it comes to public health, including how they prioritize issues and allocate resources. 

**PS: Want to explore life expectancy in your neck of the woods? Check out this neat data visualization tool of mortality from the wiz kids at 538. 

References

Anne Case and Angus Deaton. (2015). Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. Proceedings of the National Academy of Sciences. 112 (49) 15078-15083;

Ana V. Diez Roux (2017). Despair as a Cause of Death: More Complex Than It First Appears. American Journal of Public Health 107, 1566_1567, https://doi-org.proxy.lib.umich.edu/10.2105/AJPH.2017.304041

CDC. National vital statistics report. June 24, 2019. Available here: https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_07-508.pdf

National Center for Health Statistics. NCHS Data Brief No. 328, November 2018. Available at: https://www.cdc.gov/nchs/products/databriefs/db328.htm

National Center for Health Statistics. NCHS Data Brief No. 218 November 201. Available at: https://www.cdc.gov/nchs/data/databriefs/db218.pdf