It is very hard to find things to worry about. The polling has looked really good, the stories are being written that the pundits are lining up behind an Obama re-election. POTUS standing next to Chris Christie in a rare moment of bipartisanship makes a stark contrast to the countless editorials in Ohio calling Romney out for his lies on Chrysler shipping jobs to China. Romney has had to resort to all-out lying to try to win, and has had to go into PA because he knows OH and WI are quite probably out of reach.
Still - some of us like to worry. We need to worry. If there were nothing for us to worry about we might sleep easy at night, and ceilings need to be stared at. So, if instead of a horror movie, you want to see a few truly cautious signs... if you have nails that are screaming to be chewed on... read below the fold.
Edit: Nate re-ran 538. He had a run in the 6pm hour (I drafted this diary while dkos was down) and another run in the 8pm hour Eastern time. Now he is at 83.7% up from 82.9% yesterday in the Now-cast (up 0.8%) and 81.4% up from 80.9% yesterday in the forecast (up 0.5%). So strike one reason to worry. Makes it even harder to be a worry wart.
Reasons that worry warts should be worried about Election Day:
1) Gallup's November Surprise. Sandy may have helped the president win the election but not exactly how you might think. Every day for the preceding two weeks, Gallup's LV model had been driving a ton of media narrative. Peaking at Romney +7, the poll showed Romney +5 on the last day of the tracker before the storm. Without Gallup, the national poll average has switched to a narrow Obama lead. Even Rasmussen pulled into a tie. However, Monday afternoon at 1pm, right before voters go to the polls, Gallup will put out another poll. There is no reason to believe that with their existing methodology they will get anything other than similar results to their previous poll. If the national race has indeed moved a fraction of a point towards the president in the last few days, then Gallup should have a 4-5 point Romney lead on Monday. Immediately upon the release, the Romney camp would use this as a rallying cry and the news casts on election eve will be about #mittmentum.
When you go to fivethirtyeight remember - Nate uses the Gallup tracker as a data point. Gallup not releasing polls has caused their poll to age and be less impactful. Had Gallup been pumping out 5-7 point Romney leads all week, Nate would have the president at lower chance to win. How much lower? Don't know.
2) RAND showing voters flipping Obama to Romney. Sam Wang says the "Shifts-Between-Candidates" graph is "the only graph at RAND I follow." For 18 days after the first debate, voters were moving from Obama to Romney. Note the graph is a trailing 7 day average. The day after the 2nd presidential debate, that 7 day average shifted towards the president. For a full 7 days, voters showed movement to the president, undoing (by my rough eye-balling) about 1/3 of the shift to Romney. However there has been a spike in the last 4 days for Romney and now he is peeling off Obama voters. If people changing their opinions are changing their opinions to Romney in the period AFTER the storm, that runs totally counter to the meme we want to be hearing. Which also gets to the next point.
3) While Obama is ahead in all the swing state polls, some are tightening. For example yesterday's Grove poll showing Obama up 48-47 in Florida followed a poll from a week before from Grove showing Obama up 47-45, so Romney gained one. The 10/27 poll of Florida from SurveyUSA showing a 47-47 tie came 9 days after SurveyUSA had Obama up 47-46. In Iowa, NBC/Marist Had Obama up 51-43 on 10/17 and that was cut to 50-44 on 10/29. So some polls, if you pick and choose and look at trends within the same pollster, are converging down to very slim leads in many swing states. Additionally a bunch of the favorable polls are from Ipsos which is an online poll, so untrustworthy.
4) Surveymonkey has it real close. Too close for comfort. Yes, it's an on-line poll, and I just bashed those. However it had 500,000 participants from 10/3 to 10/30, usually people at the site to perform a survey for some other commercial purpose (for example, my company uses them to do internal surveys about departments or events all the time). This isn't a poll that is likely to have been freeped. What does their poll say? Well, some obvious things, like Romney would have been really glad to have had the election in October as he would have crushed Obama. In fact, between 10/10 and 10/19 he had a solid electoral college lead. As of right before Sandy, Obama had a lead for sure, but only 273-265 in the electoral college. If Surveymonkey is right, then Romney has to flip only one single state in order to prevail, and the Gallup world view and the world view of the internal polls of the Republican candidates is more real than the public polling released.
http://www.slideshare.net/...
5) Gallup is explaining their world view - and it's scary. Voter turnout among Registered voters has been on a steady upward trend over the last 5-6 elections. However Gallup's asking of how likely people are to vote and are they giving quite a lot of thought or some thought to the election shows that participation (in their opinion) would be less than 2004 or 2008 levels, but exceed 2000 levels. This assumes of course that there is a correlation between answers to these questions and voter participation, which Gallup seems to think they have shown. Now in 2004, we thought we got out our vote, but the Republicans got out more of theirs. If turnout fell to less than 2004 levels, that is likely our demographics not showing up. While early voting is promising in some instances, huge swaths of the electorate are going to need to vote on Tuesday and if they don't show up, it doesn't matter who they may have told PPP that they were voting for. http://www.gallup.com/...
Original Content is below - remember my edit that Nate did a revised run, I just felt I should leave the original so everyone could laugh.
So - Nate's up to 81.1% - up a meager 0.2% from yesterday even with Obama leading in 15 of the 17 Swing State polls today and the other two being ties (one tie from Ras in Ohio and one from on-line Ipsos in CO), and where he led in 3 of the 5 trackers and tied in two others. With a polling day of 18 wins, 0 losses, and 4 ties he could only inch up 0.2%. This the 11th all-time highest day in Nate's forecast, which is great. From a high of 87.1 on 10/4, the President fell to a low of 61.1 on Oct 12, a drop of 26.0 points. The President has made up almost 77% of the drop, showing Romney's "bounce" clearly fading. But... Nate's "Nowcast" fell from 82.9 to 82.5 today - so I guess today's polling may really have been good for John McCain.
So all you "glass half full" types can remember that most phone polls don't offer Spanish or call cell-phones, or talk about how we're allegedly getting our infrequent voters out in FL and NC and the Republicans are not, or how Oregonian voting officials filling in Republican ticket on absentee ballots that left races blank really would only have a small impact. Those of us worry warts need to reach a bit to find something to be concerned about. I guess that's good news. When you have to rely on surveymonkey to make you stressed, or a poll that isn't coming out for almost 3 days that might (might) show bad numbers, then it's likely a pretty good situation. So I'll just do the worrying for you while you're getting all psyched up for GoTV :).