If the exit polls are this unreliable for press’ result-predicting purposes, why aren’t they also unreliable for all the scholarly purposes they are supposedly put to? Garbage is garbage, no?
Not really. My guess is the problem with the exit polls was over-sampling in cities and large suburbs, where the population density makes the data easier to collect. Pollsters clean up the exit poll data by weighting it to the actual vote once it comes in. Unless there’s a sampling problem within a demographic group, the numbers should be pretty reliable.
UPDATE: I see in his BloggingHeads diavlog that Kaus does suspect intra-demographic error — that pro-Hillary blacks, or anti-Obama whites, are reluctant to talk to exit pollsters. Could be, but I’m not sure there’s evidence of that.