Polarization leads to nationalization of elections
Do swing states really swing? Are the presidential campaigns right to focus so much time and money on a small set of swing states?
Brad Jones, a graduate student in political science at U. Wisconsin, has produced some extraordinary graphs in a blog post at CivilPolitics.org, showing that states used to swing widely from election to election, particularly in the decades after WWII. Knowing how a state voted in one presidential election didn’t usually give you a strong basis for predicting its vote in the next few elections. So it would make sense for candidates to pour huge amounts of money into the few states that could plausibly be shifted.
But as political polarization has increased since the 1980s, the states have begun to lose their individual personalities and assume their place in a single ranked list, based (I assume) on the percentage of the population that is liberal or conservative. In other words, if you know how liberal or conservative a state is, you can predict with high accuracy how it’s presidential vote will turn out. As Jones puts it:
politics has become increasingly nationalized as it has polarized. This nationalization would explain the stable rankings and uniform shifts that have characterized recent elections. The shifts in election results are not concentrated among the handful of states that receive endless barrages of campaign advertising. Rather, all of the states have tended to move toward the candidate who ultimately wins the election.
You gotta see the graphs to believe it. One really interesting finding: this same pattern of extreme predictability is not new. Jones shows that it also held during the last period of extreme political polarization, in the late 19th Century. Polarization does weird things to our democracy. It makes one moral fault line become highly stable and salient, rather than having multiple possible fault lines and shifting coalitions, which I think is a healthier situation, less prone to demonization.