If you’re like me, you’ve been glued to the poll numbers lately. Obama has a lead in every national poll I’ve seen and he is more or less ahead in all the battleground states considered crucial to get past 270. However, as some of the pollsters are noting, the poll numbers haven’t been reflecting relatively large leads in early voter polling. There could be several reasons for this and one possible explanation is that many of the polls are based on a false premise: that calling landlines can get you a representative sample.
Take the Pew numbers. They have Obama ahead by 15 while most other national polls are in the 5-10 range. What’s going on here, and who is correct? The genius site five-thirty-eight has this little gem buried in the text of this post.
A brief aside about the Pew poll, which pegs the race at Obama +15. Pew has a very good reputation. Their polls, however, have also had about a 3-point Democratic lean this cycle. There may be good reasons why a poll leans a certain way; for instance, Pew calls cellphones, whereas most other pollsters do not. But a +15 from Pew isn’t quite as meaningful as if, say, that result were coming from Gallup, which has been very neutral overall this year.
ChrisH over a Denialism Blog notes the same thing. What we don’t know is the balance of calls, landlines vs. cellphones. Surely there are some national numbers on the proportion of people that have both, just a landline (can such a person still exist) or just cell phones. Who is getting this right? I’d be willing to bet that Pew is working hard to be accurate on this (Gallup? maybe. the others, not so confident — especially if they already ignore cellphones). We know that the younger crowd leans toward Obama in large numbers and my guess would be that this demographic is heavily weighted toward the no landline part of the population. Honestly, I’m not sure I know one person my age or younger that has a landline. I haven’t had one for almost ten years!
Whatever the answer is, the early voting numbers look truly incredible and are hard to reconcile with the poll numbers. Here’s to hoping that a paradigm shift is in order for polling in this country and that a landslide is on its way!
UPDATE: Chuck Todd proves my point:
A brand-new NBC/WSJ/MySpace poll illustrates not only McCain’s challenge come Election Day, but also the challenge the Republican Party could face in future elections. In the poll, Obama enjoys a more than 2-to-1 advantage over McCain among first-time (read: 18-21 year olds) and lapsed voters, 69%-27%. These voters have a much more positive view of Obama (64%-27% fav/unfav rating) than average voters do (56%-33% fav/unfav in last week’s NBC/WSJ survey). What’s more, they have a much more negative view of McCain (29%-59%) and Palin (23%-54%) than average voters do. All of this suggests that a big turnout among these new and lapsed voters would benefit Obama on Election Day. The only question is: Will they turn out? In the poll, 66% say they are certain to vote — but that’s far less than the 90% of all voters who said that in last week’s NBC/WSJ poll. Dem pollster Peter Hart compares this (potential) Obama advantage among young voters with the evangelical advantage Bush built in ’04. Yet unlike Bush, Obama can count on this advantage in every state, not just in the handful of areas where evangelicals are concentrated. What does this mean? The young vote/new voter demographic could provide Obama a 3-5 point buffer with the rest of the electorate.
That’s what myspace can do for you… wait till FaceBook gets involved!