Skip to main content
. Author manuscript; available in PMC: 2018 Oct 1.
Published in final edited form as: J Rural Stud. 2017 Oct 1;55:227–236. doi: 10.1016/j.jrurstud.2017.08.010

Figure 5. Model-Estimated Difference in Mean Trump Over-performance by County Characteristic, Industrial Midwest.

Figure 5

Note: N=504 counties

Model-estimated differences are from bivariate linear regression models. The bars represent the difference in mean Trump over-performance between counties in the bottom 25th percentile (Q1) vs. top 25th percentile (Q4) for all characteristics except nonmetro county and persistent population loss (both dichotomous). All estimates are from unadjusted models, but all models use clustered standard errors to account for nesting of counties within states.