This is funny. Ubs takes us from “Since 1916, no Democrat has won the White House without winning West Virginia” to “No Democrat has won the White House without winning Missouri since 1824.”
The sad thing is that I’ve seen reputable social scientists do analyses with data over several decades including “state effects,” i.e., coefficients for states that don’t vary over time. Going back to 1916 is sillier than most, but not all, such things I’ve seen. The trouble is that people have been brainwashed into thinking that something called “fixed effects” solves all their problems, so they turn their brains off.
Beyond this, predicting the winner doesn’t make much sense, given that you’re counting all the elections that have been essentially tied; see point 5 here.
What's wrong with including state fixed effects? Sure, there may be things which vary about a state over time, but there are a lot of things which are fixed. Geography comes to mind, but there are plenty of others. If any of these factors is correlated with the regressor of interest and the outcome variable, state fixed effects will solve that omitted variable problem.
FE don't solve every problem, sure, but they solve some.
PLW,
If you're studying politics, you don't want state effects that are constant over a 50-year period. Too much has changed. States that were Republican then go for the Democrats now, and vice-versa. For shorter time periods, maybe. It depends what you're doing. I think we actually put in state offsets for an analysis of redistricting over a 20-year period in our 1994 paper on redistricting, so I don't want to say it can never be done.
How about including state fixed effects as well as state specific linear time-trends. This seems to be the "standard approach" among economists.