Here’s a jarring thought: Most political analysts, and most political strategists for our two political parties, have been operating off flawed data and flawed assumptions. The result has been one political surprise after another, and the election of the two most unsatisfactory presidents, in the minds of many voters, since Franklin Pierce and James Buchanan in the 1850s.
The flawed data has been apparent for some years, since a June 9, 2016, column by New York Times election analyst Nate Cohn. As he recently reprised in a Twitter thread, his argument was that there are more noncollege white voters in the electorate than most analysts, relying on exit polls, believed.
Cohn pointed to other data sources — the Census Bureau’s Current Population Survey, and the voter file data compiled by the Democratic firm Catalist — that showed more noncollege white voters than the exit poll.
Cohn’s column, as I noted at the time, was buttressed by an earlier series of articles by Sean Trende in RealClearPolitics. Together, they undermined the theory that the Obama Democrats created an ascendant and growing majority coalition of college-educated white people and racial minorities. And the theory that Republicans, by relying on the declining number of