Thomas Ogorzaleka, Spencer Piston, and Luisa Godinez Puig write:
When social scientists study relationships between earnings and voting choices, their measures implicitly examine folks to others within the nationwide financial distribution. But an absolute earnings degree . . . doesn’t have the identical that means in Clay County, Georgia, the place the 2016 median earnings was $22,100, because it does in Previous Greenwich, Connecticut, the place the median earnings was $224,000. We tackle this limitation by incorporating a measure of 1’s place in her ZIP code’s earnings distribution. We apply this strategy to the query of the connection between earnings and whites’ voting choices within the 2016 presidential election, and check for generalizability in elections since 2000. The outcomes present that Trump’s help was concentrated amongst nationally poor whites but in addition amongst regionally prosperous whites, complicating claims concerning the function of earnings in that election. This sample means that social scientists would do effectively to conceive of earnings in relative phrases: relative to at least one’s neighbors.
Good to see that individuals are persevering with to work on this Crimson State Blue State stuff.
P.S. Concerning the graph above: They need to’ve included the information too. It will’ve been straightforward to place in factors for binned information simply on prime of the plots they already made. Clear profit requring near zero effort.