FiveThirtyEight looks at the polling data for the upcoming 2018 midterm elections and imagines the results for the U.S. House of Representatives if only women, men, nonwhite voters and white voters by education level voted. It’s a thought exercise they’ve indulged in before, with the presidential race in 2016, and it serves to indicate the demographic divide in voting intentions. (Cartographically, the maps suffer from the usual problem of U.S. election maps of congressional districts—large, sparsely populated districts in the middle of the country dominate the map.)

Previously: Trump, Clinton and the Gender Gap; What If Only … Voted?

Based on data from GovTrack, this map displays the ideological leanings of current (at the time) members of the U.S. House of Representatives by their district. “The data is based of numbers from 0-1. If the congressman is a 0 he is the most liberal in the House. If a congressman is at 1 then he is the most conservative. If the congressman is a 0.5 they are centrist. […] The most conservative congressman is Jeff Duncan, a Republican from South Carolina’s 3rd District. The most liberal congressman is Barbara Lee, a Democrat from California’s 13th District.”