Why National Presidential Trial-Heat Polls Vary So Much
Some people have asked us why there is so much variation in the national presidential horserace polls. My guess is that there are a number of reasons for variance from poll to poll:
1. There is volatility because it’s still pretty early in the process and many voters have not started to weigh the pros and cons of the candidates in a meaningful way. So, because they are in flux the polls mirror that.
2. Pollsters are using all sorts of different models to gauge party identification. Using historical models of party ID may be problematic given the swing in the last 12 months toward the Democratic party and the millions of new registered voters (the vast majority of whom are Democrats). There may be unusually elevated sampling bias depending on who has freshly harvested voter registration data and who is using older samples/panels.
3. Some media polls are using Likely Voter screens and some are simply Registered Voter samples. That can hugely impact poll results, as most of the big Obama leads (double-digit) are with Registered Voter samples. 4. Media polls are notoriously cheap and do not have the methodological, sampling or weighting rigor of campaign or academic polls. 5. As a result of all of this, at this stage of the game it’s probably best to simply average all of the polls (as we do).
Our sense is that the Obama lead is somewhere around 7 or 8 points but trending up.
Discussion Area - Leave a Comment
You must be logged in to post a comment.