Tuesday, October 30, 2012

If Early Voting Numbers are Correct, Obama is a Goner

And it's a big "IF", But here are Gallup’s stats from October 28, 2008: Obama led in early voting by 15 points (55/40).  His election day edge was just three points.  He ended up winning by seven points.  One more note.  According to Gallup, by 10/28/08, the percentage of Americans who said they’d already voted or planned to vote early was 33 percent.  Four years later, that number sits at…33 percent.

Today, Romney currently leads Obama 52% to 45% among voters who say they have already cast their ballots. However, that is comparable to Romney's 51% to 46% lead among all likely voters in Gallup's Oct. 22-28 tracking polling. At the same time, the race is tied at 49% among those who have not yet voted but still intend to vote early, suggesting these voters could cause the race to tighten. However, Romney leads 51% to 45% among the much larger group of voters who plan to vote on Election Day, Nov. 6.

If these numbers are even close to accurate - and Gallup's presidential track record is solid - then Romney will take the national popular vote by roughly 4% - and with it an electoral college majority.

Monday, October 29, 2012

Get Ready for a Crazy 7 Days, and Perhaps Beyond!

I keep saying that 2012 looks like 2004 and that advantages Romney, but there is an important caveat. In 2004 Bush narrowly won OH and narrowly lost many states such as WI and PA. This time around we are seeing similar trends with Romney and Obama close in otherwise Democratic states such as WI, MN, MI and PA. The few polls that have come from solid Democratic states like NJ and CA show Obama with much smaller leads than in 2008 while Romney is well ahead in truly red states.

This does set the stage for that rare event - a split between the electoral college and the popular vote. Though I have no faith in Nate Silvers so-called "election models" I agree with him in one area - Obama has a better chance of winning the electoral college than he does of winning the popular vote (and the electoral college is all that matters - I say that because it's obvious and because I support the electoral college). So it's possible Romney will win the popular vote by 1 or 2% and lose the election by losing OH. In that light, it is possible to see the closing week actions by the president in a different light.

Obama may be expending campaign resources in states like PA and MN because he wants to boost his popular vote numbers, but also because he fears that he may win OH and then lose the race because a state like PA or MN slips away from him.

Now, am I hedging my bets? No, not really. I predicted a Romney win because I believe the fundamentals point to his win. But even with those fundamentals it is not a certainty. Rather than a 50/50 chance Romney wins I see a 55/45 chance. That's hardly overwhelming and I am fully prepared to be wrong - but in the end I've made a prediction based on campaign activity, choices regarding ad buys and campaign stops, on polling numbers among key voting blocs (women and independents), and actual early voting data (not polling data). I do not believe that an election can be predicted the way one might predict a batting average.

Much evidence will come in the next 7 days - that Romney is planning an even in MN and that Biden is going to PA are telling and suggest a race breaking toward Romney. GOP early voting numbers are up in nearly every battleground state and Republican enthusiasm remains higher. But if there is a late break for Obama or for Romney there should be some clear indication by Friday... I know I'll be obsessed with every twist and turn between now and next Tuesday.

Wednesday, October 24, 2012

No, Robo-Polls and Landline Only Polls are not Inaccurate

With the race for the White House so close it seems that whichever side is down in the polls seeks some way to dismiss polling accuracy. At present it is Team Obama that is trying mightily to convince folks that the polls are misleading and their typical line of attack either focuses on so-called Robo-Polls where the survey is conducted via computer instead of a live operator or the lack of cell phone only households in a survey sample. Robo-Polls are not permitted to dial cell phone numbers and the argument is this exclusion of cell phone only households introduces a bias to the detriment of the President.

Unfortunately, there's no support for the anti-Robo-Call or cell phone exclusion argument. A paper written for the American Association for Public Opinion Research (AAPOR) by an ad hoc committee formed to study polling accuracy found no problems with automated polls. They addressed both the cell phone issue and the "robo-poll" question and found: "The exclusion of cell phone only (CPO) individuals from the samples did not seem to have an effect" and "The use of either computerized telephone interviewing (CATI) techniques or interactive voice response (IVR) techniques made no difference to the accuracy of estimates."

Mark Blumenthal, the polling guru from Pollster.com has written often on the issue of robo-polls and concludes there is no evidence they are are any more or less accurate than live operator polls: "automated telephone surveys did at least as well as those from conventional live-interviewer surveys in predicting election outcomes."

Then there's the Pew Foundation, in a study they conducted in 2009 they determined "pre-election polls conducted by telephone did very well in forecasting the outcome of the election in 2008. This was true for polls using live interviewers and those conducted with recorded voices. It was true for those based only on landline interviews and those that included cell phones."

So there you have it - there is simply no evidence to support the argument that robo-polls and the exclusion of cell phone only households introduces any bias into polling results.

Tuesday, October 23, 2012

All Signs Point to a Romney Win

Update: Obviously I was wrong.

The election is a mere two weeks away and I've decided to step away from the twitterverse and provide a more detailed write-up on the state if the race between Barack Obama and Mitt Romney. Anyone who has read my tweets on the campaign will not be surprised when I say that I believe that Romney currently has the momentum and I consider him to be the favorite to win on Election Day. I argue the current electoral college map, national polls, support levels among women and Independents, and the results of the debates all point to a Romney win. I'll consider each in turn.

The Electoral College
There can be no question but that Romney has made significant gain in the battle for 270 electoral votes in the past 3 weeks. Romney has a solid hold on every state John McCain won in 2008 and has reclaimed two Obama states - Indiana and North Carolina. But that tells only part of the story. Florida and Virginia have also moved significantly toward Romney and if you look at the top ten advertising market ad buys Virgina is nowhere to be found and Florida is at the bottom of the list. Democratic operative and CNN contributor Paul Begala told CNN just yesterday that Florida has been lost to Romney. That Virginia is not even in the top ten speaks to the reality of race in that state. Add to that the fact that the scuttlebutt is that Obama for America has ceased polling in Virginia and team Obama may have pulled upwards of half of their staff out of the state to be used elsewhere.

With Florida and Virginia in the Romney column I see a current electoral map with 237 electoral votes for Obama and 248 for Romney. Six states remain as toss-ups - Nevada, Colorado, Iowa, Wisconsin, Ohio, and New Hampshire.

Among those six states, Colorado has been trending to Romney and Nevada has been trending to Obama - that would create 243 votes for Obama and 257 for Romney. There has been much talk of everything coming down to Ohio, but that is now only true for Obama. In the map I'm seeing, Obama must win Ohio to win, but he'd only have 267 electoral votes and would need New Hampshire, Iowa, or Wisconsin to cross the victory line.  Romney can hit 270+ by winning Wisconsin and New Hampshire or Wisconsin an Iowa. Romney doesn't need Ohio, but if he does win it, he wins it all. And the trend in Ohio is toward Romney. Since the first debate Obama's average lead in OH has fallen from 5% to below 2% and multiple polls show Obama receiving 47% support or below - that's a dangerously low level of support for an incumbent. If one looks to the last 5 presidential elections, in 4 of the 5, late deciders broke strongly for the challenger. In Ohio specifically, late deciders broke for John Kerry by a 22 point margin in 2004. Kerry still lost, but a key difference between George W. Bush and Barack Obama is the fact that Bush was closer to the 50% mark already and there were fewer undecided voters in 2004.

But the electoral story goes deeper. Recent polling data and campaign actions suggest that once safe Obama states may now be in play - Pennsylvania and Minnesota. Paul Ryan attended a rally in Pittsburgh last weekend and in Minnesota the Obama campaign has made an ad buy. Worse, I'm hearing rumblings today that one of two campaigns is conducting polls in Oregon. Obama won all three states by comfortable margins in 2008. Romney is unlikely to win any of the three, but the simple fact that they are in play demonstrate how Obama's fortunes have changed. I see an end result that puts Romney at 285 electoral votes as a minimum.

The National Polls
The national polls provide the best evidence of Romney's momentum. Prior to the first debate on October 3rd, Obama held a clear advantage.

The table above, from Realclearpolitics.com makes clear the strength of Obama's position just before the first debate. Obama was ahead by an average of 3 to 4 points and more important he was regularly hitting 49% and 50% in the polls.

Then came the debate....

Obama's advantage was gone and worse his 49% and 50% marks were now accompanied by 45% and 46% levels of support - deadly for any incumbent.  Some on the American Left have tried to argue that Romney gained temporary momentum after the first debate and that momentum stalled after the 2nd debate - that is clearly wishful thinking.


Romney is in a better position today than after the first debate. Romney is hitting the 49%, 50% and 51% mark and appears to be establishing a lead. The one poll that offered Obama hope was the IBD/TIPP tracking poll that had Obama with a 6 point lead on Sunday, but within two days the lead fell to 2% suggesting the lead was just the result of a bad day of polling. Obama is regularly receiving support in the 45%-47% range and it's hard to see how he climbs back to 50.1%. 

Support Among Key Voting Blocs
When you take a glance deeper inside the polls you see greater trouble signs for Obama. Obama is counting on a large gender gap to win him a second term - namely he hopes to win women by a large margin. That was the basis for the "war against women" meme and the constant talk of contraception access. But the approach isn't working. The Monmouth/SurveyUSA survey shows what has been a common element most national polls - Obama is not winning women by a margin sufficient to win. Romney has a double digit lead among men, but Obama has only a mid single digit lead among women. Team Obama has failed to recognize that women are more concerned with the economy than any other issue. The focus on "gender issues" came at the expense of presenting a plan for improving the economy in his second term.

Independent voters are the other key group. Obama won Independents by 7-8 percentage points in 2008, yet in most national polls and in most swing states - including Ohio - Romney has a clear advantage among Independent voters - usually in the 5-7% range. Absent the gender gap and the support of Independent voters, Obama would need Democratic turn-out to match or exceed 2008 levels. But 2008 was an election marked by tremendous enthusiasm among Democrats and the party's 7 percentage point turn-out advantage likely was a high water mark that will not be repeated. If the party's turn-out advantage winds up at the 4% range (the average of 2004 and 2008) Obama will fall short of victory.

The Debates
The prior paragraphs make clear the gains Romney made from the first debate, but his gains continued through the 2nd and 3rd - debates he "lost." After the 2nd debate, a CNN poll of viewers found Obama scored a narrow victory, but on the most important question - who would be the better leader on the economy - Romney won 58%-40%. After the final debate on foreign policy, many of Obama's supporters were convinced of the president's clear victory. But CNN again found viewers gave the president a narrow win and on the only question that mattered - is the candidate ready to be commander-in-chief - Obama and Romney tied. A different poll of  viewers by PPP (a Democratic firm) asked if the debate made viewers more or less likely to support a candidate. Among Independent voters, 32% said they were more likely to vote for the President and 48% answered less likely. For Romney, the numbers were reversed, 47% more likely and 35% less likely. 

The debates also allowed Romney to undo the negative effects of $100 million in swing state advertising by the Obama campaign to define him as an unacceptable candidate. Going into the fall Obama enjoyed a tremendous favorability advantage over Romney. Romney was considered to be one of the least liked candidates ever. As of today, Romney enjoys a 6% advantage over Obama on favorability! It marks a stunning turnaround.

For all of the reasons discussed above, and many more, I see a race that has shifted dramatically in Romney's favor and at present I consider him the most likely victor. Obama's Intrade advantage has collapsed and one of his biggest supporters, Nate Silver of the 538blog, has begun to hedge his bets and discuss the limits of his probability models (which are actually variants of the rather commonplace Monte Carlo method). If you follow Silver religiously you should really move on, because he's not doing anything special and the modifications he's made to the Monte Carlo model all favor Obama - for instance he tends to add greater weight to polls favorable to the President and will even weigh such polls more heavily when they are older and clearly contradicted by newer surveys. Those who follow the futures markets have seen in the past 24 hours how quickly they can change.

All indications are that the results of the election will be close, and it may well come down to a single state that makes the difference, but right now Romney is the candidate with the momentum and the likely victor. I will add, however, that there is the possibility - if national polls are truly capturing the dynamics of the race - that Romney will win by a more comfortable margin.

I may well be wrong, but a week before the the 2010 midterms when most political science models and analysts (including Nate Silver) were predicting GOP House gains in the 50 seat and below range - with many models predicting the GOP would fail to even capture the majority in the House - I predicted a 65 seat gain (the actual was 63) and I correctly called all of the Senate races (also something Silver failed to do) - so I'm just sayin'...

Sunday, October 14, 2012

Is Obama in Trouble in Minnesota?

Update - 10/25: Mitt Romney has purchased TV ads in Minnesota. Recent polls suggest a very close race there, but Romney may be trying to force President Obama to divert resources to this assumed to be safe state. Romney may as well be seeking the positive coverage such an ad buy would generate as it shows him playing offense on traditionally Democratic turf. We'll know better on November 6th...

Update II - 10/26: The Obama campaign has now started running TV ads as well. Minnesota is in play.

There were whispers that team Obama sensed some danger in Minnesota - a state most have considered to be a safe Obama state. Then came word that the campaign was sending Jill Biden to the state. If the campaign was not concerned they wouldn't be sending anyone. They can't risk sending Obama or Joe Biden because that would make the campaign's Minnesota worries front page news. I was hesitant to accept the Minnesota news, but then I received the following e-mail from my colleague and co-author Steve Schier, the H. and Edward C. Congdon Professor of Political Science at Carleton College in Minnesota. Schier is an expert on Minnesota politics:
"Here is a link to a memo from GOP pollster Glen Bolger revealing a new poll indicating Obama ahead by only 47-43 in Minnesota. A competitive race?

The two most recent SurveyUSA polls of the sixth and eighth Congressional districts provide supportive evidence for Bolger's poll result. In the eighth, a D +3 district, Romney led in the Survey USA poll 47-45. Romney led by 17 points -- 53-36 -- in the sixth district, which has a partisan index of R +7.

Add to that the fact the a major surrogate, Jill Biden, will campaign in the Twin Cities and Duluth next week.

If Minnesota is competitive, Obama is in trouble."
Make no mistake, if Minnesota is in play then Obama has already lost the election. It's now clear that the Romney debate bounce was more than just a bounce. It has remained and translated to crucial swing states. Romney is surging in Florida and Virginia and appears to be ahead in Colorado, New Hampshire, and perhaps Nevada. In Ohio, Romney has eliminated Obama's once considerable lead and now the race is a statistical tie. Even Pennsylvania has become competitive.

The Gallup and Rasmussen tracking polls have the race at 49%-47% - advantage Romney. At present, the 2012 race is starting to look like 2004 with Romney as Bush and Obama as Kerry. Assuming 2/3 of the undecideds break for the challenger (common in an election featuring an incumbent) the 49-47 race would end up 51-48 - just as in 2008. In a race that close Romney would win, but receive 300 electoral votes or less. He'd likely lose Minnesota and Pennsylvania - but both would be close.

Beyond what's happening in Minnesota, one need only look at Obama's average in the national polls. Obama is averaging 46% - that's deadly for an incumbent. Obama needs to be polling above 47% to have a chance at re-election.