With the CBS/YouGov poll showing little change in the race, I'd like to return to this thread and ask a somewhat different question: who was helped by the Selzer/DMR/CNN poll not showing up?
I will posit that the likeliest answer is 'Biden.'
He was in fourth w 15 in their last two polls and for a reason: their method--no education weights, super tight screen, whole RV universe--hurts him at every point. No other pollster has been so tough on him
He was in fourth w 15 in their last two polls and for a reason: their method--no education weights, super tight screen, whole RV universe--hurts him at every point. No other pollster has been so tough on him
Other Iowa polls give the sense of a Biden-Sanders race, and that has the potential to help Biden under a number of circumstances on caucus night. A fourth place finish in Selzer, forgetting the risk of a gaining Klobuchar, would have undermined that advantage, right or wrong
And as I just alluded to but separately, the chance that a final poll from Selzer would have shown a gaining Klobuchar in double digits (like 7 straight polls pre CBS YouGov), and potentially close to Biden, would have been a serious issue as well
If Klobuchar didn't surge, like CBS, I'd wonder whether Pete lost out by not getting this poll. In our poll, at least, he would have been 2nd and closed to within 4 points of Bernie if we had used Selzer's screen and weights. He could use a late show of strength v Biden
Anyway, we'll never know. But polls have been very
consistent on their House Effects in Iowa, since the methodological differences are unusually big and highly correlated with vote choice. If the race is stable, I think we have pretty strong priors here
consistent on their House Effects in Iowa, since the methodological differences are unusually big and highly correlated with vote choice. If the race is stable, I think we have pretty strong priors here
