Intended for healthcare professionals


Politics and mortality in the United States

BMJ 2022; 377 doi: (Published 07 June 2022) Cite this as: BMJ 2022;377:o1308

Linked Research

Political environment and mortality rates in the United States, 2001-19

  1. Steven Woolf, professor
  1. Department of Family Medicine and Population Health, Virginia Commonwealth University, Richmond, VA, USA
  1. steven.woolf{at}

Mounting evidence suggests a link between republican politics and higher death rates

For decades, people in the US have experienced lower life expectancy and poorer health outcomes than populations in other high income countries, and the problem has worsened over time.1 This pervasive disadvantage has been attributed to unhealthy behaviors, a dysfunctional healthcare system, and adverse socioeconomic and environmental conditions, but these downstream conditions arise from upstream policies and social values, many of them quintessentially American.1234 These include the US Constitution’s protections of states’ rights and gun ownership, resistance to social welfare programs or to restrictions on personal freedoms, and systemic racism.5

Such attitudes are not uniform across the country. US policy reflects the choices not only of the federal government but also of 50 states and their incorporated counties and cities. Stark geographic inequalities in both policy and health outcomes have widened over time.6 Life expectancy began to diverge dramatically across US states in the 1990s, increasing in states such as New York where the Democratic party and progressive policies dominated, and stagnating or decreasing in states with more conservative governments and Republican majorities.1 Southern and Appalachian states, generally led by Republican governments, have the nation’s lowest life expectancy and highest poverty rates.

Political polarization in state governments intensified over the past decade.7 Republican governors and legislatures adopted more conservative policies that affected population health, such as opposing Medicaid expansion, minimum wage legislation, and tobacco and gun controls.89 Studies of the 2016 and 2020 presidential elections reported that counties voting for the Republican candidate Donald Trump had higher—and larger increases in—mortality rates than counties favoring his Democratic opponent.10111213

Such studies hint at an association between elections and mortality but cannot establish the existence or direction of a causal relationship. The association could represent the harmful effects of Republican policies, but it could also reflect a preference for Republican candidates among disadvantaged voters14 or characteristics outside the control of politicians—such as changes in the economy, migration, or demographics—that happen to predominate in places that elect Republicans.

A linked study by Warraich and colleagues (doi:10.1136/bmj-2021-069308) advances the evidence by examining the temporal relationship between US presidential voting and county mortality rates.15 Between 2001 and 2019, counties that favored Republican candidates experienced half the reduction in mortality rate (11%) observed in Democratic leaning counties (22%). Republican leaning counties also experienced smaller decreases in mortality in white populations and rural areas, and no decreases after 2009.

Between 2001 and 2019, the absolute difference in mortality rates between Republican leaning and Democratic leaning counties jumped by 541%, from 16.7 to 107.1 deaths per 100 000 population. Votes for governor followed the same pattern. Study weaknesses included the binary classification of counties as Republican or Democratic on the basis of the favored presidential candidate in the preceding election, rather than a continuous measure such as vote share, or margin. Votes for local officials, governors, and legislators in statehouses and Congress—which may bear more on county mortality rates—were not examined.

These limitations aside, corroborating evidence about the potential health consequences of conservative policies is building. For example, Montez and colleagues reported that states adopting more conservative policies between 1970 and 2014 experienced smaller improvements in life expectancy. The authors estimated that the increase in US life expectancy that occurred during 2010-14 would have been 25% steeper in women and 13% steeper in men had this transition to conservative policies not occurred.16

Political influence on US mortality rates became overt during the covid-19 pandemic, when public health policies, controlled by states, were heavily influenced by party affiliation. Republican politicians, often seeking to appeal to President Trump and his supporters, challenged scientific evidence and opposed enforcement of vaccinations and safety measures such as masking.17 A macabre natural experiment occurred in 2021, a year marked by the convergence of vaccine availability and contagious variants that threatened unvaccinated populations: states led by governors who promoted vaccination and mandated pandemic control measures experienced much lower death rates than the “control” group, consisting of conservative states with lax policies and large unvaccinated populations.18 This behavior could explain why US mortality rates associated with covid-19 were so catastrophic, vastly exceeding losses in other high income countries.19

Observers of health trends in the US should keep their eye on state governments, where tectonic shifts in policy are occurring. While gridlock in Washington, DC incapacitates the federal government,7 Republican leaders in dozens of state capitols are passing laws to undermine health and safety regulations, ban abortion, limit LGBT+ rights, and implement more conservative policies on voting, school curriculums, and climate policy.20 To understand the implications for population health, researchers must break with custom; although scientific literature has traditionally avoided discussing politics, the growing influence of partisan affiliation on policies affecting health makes this covariate an increasingly important subject of study.


  • Research, doi: 10.1136/bmj-2021-069308
  • Competing interests: I have read and understood the BMJ Group policy on declaration of interests and declare the following interests: none.

  • Provenance and peer review: Commissioned; not externally peer reviewed.