Figure 1 College Graduates and Vote Share for Medicaid Expansion. Notes : Each circle represents a locality, and its size reflects the total number of votes cast. The x-axis shows the share of the adult population in the locality with at least a bachelor's degree, and the y-axis shows More

Figure 7 Vote for and Support for Democratic Candidates in the 2014 House Midterm Elections by Insurance Type. Notes : Data are recycled predictions and their 95% Huber-White confidence intervals from regressions of insurance type. The dashed lines indicate predicted outcomes among individuals More

Figure 3 County-level changes in gubernatorial Democratic two-party vote share (2002–6) and TennCare enrollment (2004–6), weighted by population. Sources : Tennessee secretary of state, TennCare annual reports. Notes : Larger circles represent more populous counties. Moore County omitted More

Figure 1 Average ACA individual-market premiums by Republican 2012 vote share. Circles represent the average price in 2014 for a silver-level plan in counties binned in 2% bandwidths by Republican vote share in 2012, while triangles represent the same measure in 2017. The size of the circles More

Figure 3 Association between Republican vote share and premiums: 2014–17. Estimated effect of a 10-point difference in Republican vote share on average monthly silver-level premiums in the first four years of the ACA marketplaces. Estimates derived from covariate-adjusted model analogous More

Figure 1 Gains in insurance and Democratic vote by county. More

Figure 5 Coefficient of Regressing Legislator Vote on H.R. 3590 onto Voter Preference by Fifteen-Year Age Groups Source : CCES 2008, 2010, and 2011 (Ansolabehere 2012 , 2013a , 2013b ) More

Figure 6 State-level correlation of Trump vote share with mask wearing and COVID-19 infection rate. Note : Pearson correlation coefficient printed on each plot. Data source: COVIDcast, Delphi Research Group ( https://delphi.cmu.edu/covidcast/ ). More

Figure 6 Medicaid Density and County Voting Rates More

Figure 1A Differences in the Predicted Probability of Voting in 2008 by Chronic Health Conditions and Race More

Figure 1B Differences in the Predicted Probability of Voting in 2008 by Chronic Health Conditions and Education Source : Author's analysis of 2009 Behavioral Risk Factor Surveillance System data (CDC 2009 ) Note : Predicted probabilities are based on stratified models displayed in table 4 . More

Figure 2 Reasons for not voting (by health and income). Source : Multinomial logit model in SI. Predicted probability of reasons for not voting by income and health. All other variables fixed at median values. More

Figure 2a Positions on House Roll Call Votes, Left-to-Right DW-NOMINATE Ideology Scale: 93rd Congress (1973–74)—Nixon/Ford Source : https://legacy.voteview.com/dwnomin.htm More