Sources:
2016 Election Results from the New York Times »
Population data from the American Community Survey 2014-2018 »
The four county characteristics presented here are population density against higher education, with the color reflecting which party won a majority of votes, and the size of each county point reflecting the total population.
In the past Presidential election, it was often observed that voters were divided along regional and educational lines. The so-called "urban-rural divide" is somewhat apparent in the two most populous states in the country: as we approach population densities of 500 and higher the counties that vote Democratic represent a higher proportion of counties. This proportion grows significantly after 1000 people per square mile. Conversely, counties with population densities of 50-100 and below vote Republican for the majority. Interestingly, the remaining cities do not seem to be predictably for either party based on population density. There is a positive correlation with the proportion of the population with a college degree or higher in California, but this correlation is much lower in Texas. This education divide has been noted in election results, but whether the level of education or other variable correlated to the regional divide is a causal factor is unexplained by the data present here. Some other things to note about both charts include the overall state favorability towards one party or another, and the positive correlations between population density and population size, and between population density and proportion of people with higher education.