No.

jdb: We're releasing the presidential and Senate numbers tomorrow. Dave: Less than half a percent said they'd not vote on the measures, which is why they show up as 0% when rounded. It's hard to capture undervotes in a poll since we only take completed interviews, and people don't usually volunteer they won't be voting on something in a poll even if they don't end up voting on it. But it's fair to say there will probably be fewer undervotes on a charged issue like same-sex marriage than for past ballot measures. Either way, it's close, and enough undervotes could tip the results against the amendment.

George, we did the CT poll for the League of Conservation Voters. We tweeted about it yesterday, with a link to the results.

1.) It's party self-ID, not based on registration records. We don't weight for party ID. 2.) Turnout by party never matches registration. Unaffiliated voters are almost always less likely to turn out than partisans, so the proportion of Ds and Rs will almost always be higher than their proportion in the overall registered electorate.

Mike Hall: According to exit polls, 20% of the VA electorate in 2008 was AA. We have 17%. There's no indication there has been much of a drop in enthusiasm from black voters for voting for Obama or voting in general. In fact, in our national polls, black voters remain one of the most enthused groups about turning out to vote--way more than whites and almost as much as Tea Partiers and Hispanics. Also, the party breakdown in 2008 was D+6. No room to complain here, especially considering the continually growing and blue-ifying NoVa.

Jeff, you comment here all the time (unless it's a different Jeff). As always, demographic info is in the link to the full results at the bottom of the post.

Cas: We didn't say anything about white men, just white voters of both genders. TallDave: We have it as D+4, but we never weight for party ID, which shifts and isn't necessarily correlated with party registration.

We take a few suggestions a week, but we can't take everything. Suggestions also have to be in by Thursday morning every week.

Iowa has been more of a battleground state than MI, WI, and certainly MN in the last several cycles, going back to '96 or '92 at least.

Tom didn't write that or the above post; I did. Note that I said: "Of course, this is all contingent upon Romney not winning in November."

Jay: For a GOP primary, that's typical. What would you expect?

2010 was a midterm election and a Republican wave year. Of course there was an R turnout advantage. But that's a different beast than a presidential election. 2012 will be much closer to 2008, which was D+8, according to exit polls.

Yeah, I didn't realize that until after I posted this. A number of people have alerted us to that. Oh well. We asked the question to provide a little levity, and it's funnier if you assume respondents were thinking about the fictional town, as we were.

Jay, you can't "make adjustments" in how many Democrats or Republicans or independents are voting for particular candidates. That's not polling. That's called fabrication.

We have the sample at GOP +2, and in '08 it was GOP +8 according to exit polls. So you're complaining about what? Also, women always are a larger share of the electorate as men. Women +4 is identical to '08.

Yes, but he won't exceed that by much if at all in 2012 with Barack Obama on the ticket campaigning for Tim Kaine. So expect Kaine to get almost all of the 11% undecided black voters to move up to about 78-80%. Still lower than Obama will pull, but you won't see Kaine in the 60s, just like you didn't see Webb there.

That makes no sense. Paul is a libertarian with a little L, as is Johnson, and they're both Republicans. Paul does stray from some libertarian doctrine, but he's what passes for libertarianism in the GOP these days.

We polled Goode in Virginia, but he wasn't a known quantity even there, so there's not much point in polling him elsewhere at this point. Third-party candidates' shares of the vote are always inflated at this point; they inevitably get less than the polls say they will, especially this far out. We may include third-party options in our polls as the election gets closer, especially if it looks like Johnson or others could have a spoiler effect.

Paul leads Obama by 22 points with independents, not with everyone. He does 3 points better than Romney with all voters because of that large lead with independents. He does worse than Romney does with Republicans. Also, we're not media.

Xerxes: You obviously know nothing about NC. The state has about a D+13 registration advantage. '08 was also D+11.

Elizabeth, if you'd actually read the results, you'd have seen we did poll Johnson.

Matt D: Bev wasn't this unpopular when she was elected. Her popularity tanked within a few months of taking office.

Mark: You obviously don't know anything about NC, which is one of the fastest-growing states in the country, and has been for 20 years. Some estimate that a third of the state's population did not live here four years ago. Most of the voters who move here are more liberal than the natives and more educated. We asked in a poll one month last year when respondents first lived in NC, and those who moved here or were born 20 years ago or after were going for Obama by 20 points over Romney. (Tom Jensen and I are among this group.) That was a huge chunk of the electorate, and this was when Obama was in worse shape than he is now. So Obama's decline with independents and and moderate/conservative Dems is buffered by the sheer demographic shifts taking place in this state, plus his continuing popularity with black and Hispanic voters (and growing popularity with the latter because of Republican rhetoric on immigration). That's why you see him holding his ground or even improving in other growing, diverse states like VA, CO, NM, and AZ relative to states which are declining in population and which are older and whiter, like OH, PA, NH, and most of the Midwest.

Our polls in '09 on Maine's referendum were dead on, and there's less social desirability bias with automated polling than live interviewer polling.

Geoff: Not sure why you'd guess that. The numbers would be even worse if Cali WEREN'T represented proportionally.