Was Population Health a Predictor of the Presidential Election? Yes, but Not By Itself!
Two months have passed since the US presidential election, and countless analyses have been published that examine the demographic and economic factors that contributed to Donald Trump’s surprising win. Two recent articles from The Economist and The Washington Post add an interesting new consideration to the ongoing post-mortem: the influence of community health on voter preference. These articles provide a compelling way to analyze the election and applying a systems view of community health and political choice can provide an even more nuanced understanding of the trends they describe. We believe the strong association between local health conditions and political choices in 2016 is really a story about entrenched patterns of education, income, and population health that have been traveling together over decades—and playing out differently for demographic groups delineated by race and ethnicity, education levels, and urban proximity.
In the article, “Illness as indicator: Local health outcomes predict Trumpward swings,” The Economist identifies an index of public health statistics that consists of county-level data on life expectancy, and the prevalence of obesity, diabetes, heavy drinking, and physical activity as the strongest predictor of whether the margin of Republican votes increased between 2012 and 2016. The index predicted 43% of the shift in Republican margin, two percentage points more than the next highest, and more publically discussed, an indicator that has emerged: a county’s share of white voters without a college degree.
An article from The Washington Post, “Trump over performed the most in counties with the highest drug, alcohol and suicide mortality rates,” provides another lens on how health may have contributed to the shift in the Republican margin. The shift was greater in counties with the highest overdose and suicide mortality rates. Suicide and overdose deaths are known as “diseases of despair” because they most often result from poor mental health, substance abuse, and alcoholism, and their occurrence in communities is strongly correlated with economic distress.
ReThink Health has a special interest in tracing the effects of education, economics, and equity on health. An initial conclusion from these results may simply be that health is more important than education or income for predicting voter preference. If we view The Economist’s public health index in a broader context, however, that conclusion becomes murkier. The index includes not only life expectancy, but also behaviors that impact health (drinking and physical activity) and the chronic conditions that can develop from negative health behaviors (obesity and diabetes). As Dr. Jack Homer, Senior Modeler at ReThink Health says:
“Both education and health affect employment/income… There is a well-known feedback loop from severe chronic disease to loss of employment/income, which in turn exacerbates unhealthy behaviors and drives even more disease. So, these things are related, and no statistical regression can untangle that fact. It is not *primarily* health or education or income—they’re all connected.”
The interconnection between education, economic stability, and health are reflected in scenarios created by ReThink Health’s System Dynamics model. As described in the August 2016 issue of Health Affairs, the scenarios that showed the greatest impact on population health were ones that combined investments in socioeconomic opportunities, healthier behaviors, and better health care. Rather than viewing public health as a uniquely strong indicator of voter preference, it should be seen as another facet of the same phenomena: The mutually reinforcing and complex social forces related to the quality of life and opportunity that a voter considers when casting their ballot.