In the liveblog, Mr. Silver talks about polling, the bailout, the so-called Bradley effect, McCain's difficult road ahead, the upcoming Senate races, and the possibility of Sarah Palin dropping out of the race. For those of you not familiar with Nate Silver, he is very respected and does not hesitate to call it like it is, even when it is bad for Democrats. Fortunately for us, his analysis portends great things for us and Barack Obama!
Transcript
Question:
Your Senate projections have been quite static. Do you expect that to change any time soon? by washington dc
Answer:
The overall partisan picture in the Senate has been fairly static -- we have been projecting the Democrats to pick up between 5-6 seats all year, with an outside chance of a higher number. But, individual races like New Hampshire, Alaska, North Carolina, Kentucky and Oregon have all been moving a lot, sometimes in opposite directions. Senate races tend to break late, so I'm sure there will still be several dramatic finishes. by Nate_Silver
Question:
Dear Mr. Silver: Many of the leading polls -- Gallup, Rasmussen, Marist, among others -- oversample Democrats, while the GW Battleground poll (the only one currently posting a lead for McCain) has a higher Republican sample than the others. Can you comment on how weights are determined and why these polls have such large Democratic samples? Thank you. by Rockville, MD
Answer:
I don't think they oversample Democrats at all. There are more Democrats in those polls because there are more Democrats in the country right now. You can probably make reasonable arguments that the Democratic partisan advantage is anywhere from 5 up to about 11 points, and the different polls tend to reflect figures at different ends of that range. The Battleground poll is crap, though, because they don't weight their sample by age, and because nearly a third of their respondents are age 65 and over. If you're a McCain supporter looking to take solace in the polls, look somewhere else apart from the Battleground poll. by Nate_Silver
Question:
# How come Democrats are getting blamed for the bailout failure when they did what they said they would do? Also, shouldn't Obama press McCain on what he, McCain, intends to cut? McCain has stated many times he will bring down government spending, but has yet to say anything more than earmarks. At least as far as I can tell. Love FiveThirtyEight.com, by the way. by North Platte, NE
Answer:
I'm not sure the public really is blaming the Democrats. ABC had a poll out today suggesting that 40 percent blame the Democrats, 20 percent blame the Republicans, and 40 percent blame everyone (which is probably the correct response). by Nate_Silver [Diarist's note: He later corrected this saying that he got wrong in reverse: 40 percent blame Republicans, 20 percent blame Democrats]
Question:
# In your opinion, what does McCain have to do to win this election? by Washington, DC
Answer:
I'm not sure that it's entirely within John McCain's power at this point. His most powerful differentiating factor from Barack Obama is his experience, but that advantage has not translated to the economy in voters' minds. And even going starkly negative on Obama is dicey, as Obama's favorability ratings are strong. I think he might need Obama to make a significant gaffe, or for there to be some sort of national security crisis. by Nate_Silver
Question:
# Hi, Nate/Poblano. In light of Obama's improving polling numbers in states like WV (down 5-8 points), TN (down 12 points in the MTSU poll) or KY (down 12 in the latest Mason-Dixon poll), do you think that Obama has still a significant problem with "American" voters compared to Kerry's and Gore's performances with those voters? You know, there are a lot of those in southwestern Virginia, southern Ohio or Pennsylvania and Indiana. If Obama could keep the margins small there, nothing would be in his way to score victories in those states. Do those voters still show up as a significant variable with a negative coefficient in your regression models? Best wishes from Germany, Rasmus by Hamburg, Germany
Answer:
Those Appalachian voters are one group where Obama does seem to have made some inroads, no doubt because of the economy. But I'd also look at the number of undecideds in states like West Virginia and Kentucky ... I don't think very many of those undecideds are liable to break for Obama. by Nate_Silver
Question:
# What do you make of the new PPP poll in Florida showing Obama +3? Outlier or predictive of the current trends around BG states? by Rockville, MD
Answer:
I don't think it's really an outlier. Obama is ahead by perhaps 5-6 points nationally, and his numbers in Florida have lagged his national numbers by 2-3 points for most of the past couple of months. So that would point toward a small Obama lead. I think Obama would be the favorite to win Florida if the election were held today -- he'd be the favorite to win a LOT of states if the election were held toady. Fortunately for John McCain, he's still got some time left. by Nate_Silver
Question:
# Nate, you're an avowed Obama supporter. Do you think you would have gotten into the statistical approach to polling and politics if he hadn't run this year? by Rockledge, FL
Answer:
That's a fair question. I've always been a politics "fan" but had never followed politics anywhere nearly as closely as I have this year. Having a "hometown" candidate (I'm live in Chicago and went to U of C) whom I'd known about for more than 10 years might well have made the difference. Also, a correction: in that ABC bailout poll that I referred to above, 40 percent blamed the Republicans, and just 20 percent the Democrats (had the numbers reversed). by Nate_Silver
Question:
# Nate, I love your site. Question: Do you think pollsters are currently accurately accounting for the large voter registration on Obama's part, or the anticipated uptick in voter turnouts among youth and African-Americans? And any thoughts on Hispanics voting rates? by madison, wi
Answer:
Ann Selzer, who runs the Des Monies Register's poll and has one of the strongest track records in the country, thinks that some of the polls may be significantly undercounting the youth vote. There is a lot of room for marginal gains among youth and Latino voters, since both groups turned out in very low numbers in 2004, but in significantly higher numbers during the Democratic primaries. I'm less convinced that the polls are undercounting African-Americans, who actually turned out in pretty strong numbers in 2004, but with Obama on the ticket, it would not be surprising if the turnout rate among African-American voters actually exceeded that of whites. by Nate_Silver
Question:
# Do you think there is sufficient groundswell against the "bailout" that nothing will pass prior to the election. by East Hampton, NY
Answer:
I think some of the public sentiment against the bailout was curtailed significantly by the Dow losing 778 points yesterday. I think there's roughly an 80 percent chance that some sort of bill passes before Congress adjourns. by Nate_Silver
Question:
# On Election Day, when the early returns start coming in, will it be an indicator of Obama's strength? Is it North Carolina or Pennsylvania, etc? by Burbank, CA
Answer:
I'd look a lot an Indiana, since that's one of the first states to close its polls (IIRC), and since that's one of the places where Obama is banking on a significant advantage from his ground game. If Indiana gets called for Obama, McCain's chances of winning at that point are probably 20:1 against. On the other hand, if McCain takes the state by 5-6 points or so, it may be a sign that Obama's ground game has fizzled out and that we're in for a long night. by Nate_Silver
Question:
# Did your analysis of the primaries show any sign of the Bradley effect? If not , what difference is there between the 1982 California election and this one? by NY, NY
Answer:
In the primaries, Obama either slightly OVERperformed his polls, or performed about even with them, depending on what metric you look at. There were some regional differences, and differences in the way that undecideds tended to break in different states, but it's not clear that they were related to the Bradley Effect or something else. There is also some evidence if a REVERSE Bradley Effect in which some black voters said they were undecided when they really weren't. The basic difference between 1982 and 2008 is that 1982 was a quarter-century ago. Race relations are different now, the issues facing the country are much different ... how much have you heard this year about crime and welfare? by Nate_Silver
Question:
# Do you think that if a bailout plan passes with mostly Democratic support and the Republicans start a populist uproar about it, they can completely change the dynamics of this election? by Chicago, IL
Answer:
Perhaps they can knock off a couple of Democratic incumbents in the House. But they might have to throw John McCain under the bus in the process, because I think the bill is going to go through the Senate first, and I think McCain's boxed in to voting for it. by Nate_Silver
Question:
# Do you think Obama's recent gains in battleground states such as Florida, Ohio, and Virginia are a product of chance, or do you think the most recent numbers are a harbinger of what's to come? by Ashland, OH
Answer:
Oh, they're certainly not a product of chance. Obama's in good shape right now. I don't necessarily know that we should expect his numbers to continue to improving, however. There seems to be a hard cap at about a 6-point national Obama lead -- he's never been higher than that all year. I think the economic situation dire enough that he could break through that number, but I don't consider it to be the most likely outcome. by Nate_Silver
Question:
# If Palin blows the debate badly, does she drop out because of family reasons? by St Louis, Mo
Answer:
I think it's EXTREMELY unlikely that Palin drops out. George McGovern -- OK, not like he was going to win anyway, but he lost 7 points when he dropped Eagleton from his ticket. And if she does drop out because of a poor debate performance, I think it would look disingenuous to cite family reasons. by Nate_Silver
Question:
# What is your opinion on the impact of "voter caging" in swing states this year, and have you considered attempting to quantify the possibility of disqualified voters in your projections? by Charlottesville, VA
Answer:
Most of the battleground states have Democratic secretaries of states and/or Democratic governors, which is a change from years past. That's not to say that both sides shouldn't be diligent about fighting back against this stuff, but when you get into these sorts of legal intangibles, it's not clear that they favor the Republicans this year. by Nate_Silver