Romney's Swing State Problem

I'm going to sound like a broken record but here are the "swing-state" poll numbers released in the last 12 hours according to Taegan Goddard:

Colorado: Romney 48%, Obama 47% (CBS/NYT/Quinnipiac)
Florida: Obama 47%, Romney 46% (NBC/WSJ/Marist)
Michigan: Obama 49%, Romney 42% (Detroit News)
Michigan: Obama 46%, Romney 44% (Gravis)
Nevada: Obama 47%, Romney 45% (Suffolk)
Ohio: Obama 51%, Romney 45% (NBC/WSJ/Marist)
Ohio: Obama 47%, Romney 46% (Pulse Opinion Research)
Pennsylvania: Obama 47%, Romney 45% (Pulse Opinion Research)
Virginia: Romney 48%, Obama 47% (NBC/WSJ/Marist)
Virginia: Obama 51%, Romney 46% (CBS/NYT/Quinnipiac)
Virginia: Obama 48%, Romney 48% (Pulse Opinion Research)
Wisconsin: Obama 50%, Romney 47% (CBS/NYT/Quinnipiac)
Wisconsin: Obama 50%, Romney 46% (Pulse Opinion Research)

I put "swing-state" in air quotes because Michigan and Pennsylvania are not swing states but I won't get started on that rant again. These are just the states that Goddard lists.

Some of these pollsters are better than others and I won't bother with that discussion right now other than to say, look at where Obama is in polls that call cell-phones. For purposes of our discussion here, let's assume these polls are all equally valid.

Who is likely to win the election today if these polls are accurate? Mitt Romney is ahead in exactly 2 of these 13 polls and he's tied in one more. If we average the polls in states where there is more than one poll, Obama wins every state listed here except Colorado. Now, many of these states (like Florida) would be close. But guess what? Florida has always been close. And guess what else? Let's give Florida to Romney along with Colorado. Now who wins? The President.

Everyone is talking about how the first debate "wiped out" everything that happened between the convention and the debates when Obama built a bit of a cushion. I don't believe that's exactly correct but let's say it is. Now who wins? The President.

Yesterday, the President said, "What’s important is the fundamentals of what this race is about haven’t changed." He's absolutely right.

UPDATE: Greg Sargent snagged an interview with Geoff Garin, arguably the top Democratic pollster in the country and got Garin to outline why Obama is doing so much better in swing states like Ohio than he's doing nationally:

Geoff Garin, the pollster for the Obama-allied Priorities USA, tells me that his polling shows that views of Romney are more fixed in the battlegrounds than nationally. “In the swing states, voters are much more apt and able to quote back the main case against Romney,” he tells me.

Garin adds that his polling has tested voter reaction to various arguments against Romney, such as the idea that his economic policies would favor the wealthy or burden the middle class. He says voter agreement with those suggestions is “higher where the advertising has occurred,” and adds: “All the swing state advertising has had a measurable and lasting impact.”

Not Much to Say - Obama Still Winning

A few people have asked me why I haven't been blogging much in the past few days. The main reason is that there isn't much to say ... and that's good news for the President.

There have been lots of polls out but they don't tell us very much that is new ... at least it isn't new if you've been reading here for the last couple of months. Here's a new poll. The President is winning in Maine. Here's one that shows Romney winning in Georgia. And here's one that shows Obama winning in Michigan. There are no surprises here.

Obama is winning right now. If you look at the pollster.com trend estimates by state, you see that Obama is leading by smaller or larger margins in all the "battleground" states. Here's a listing of Obama's lead (or deficit) in pollster.com's current trend estimates nationally and for all the states for which they have multiple polls:

National - 1.2
AZ - (0.5)
FL - 1.2
IA - 10.0
MI - 4.0
MO - (3.0)
MT - (9.0)
NV - 8.2
NH - 8.8
NC - 2.6
OH - 3.8
PA - 7.6
VA - 4.5
WI - 5.7

The only states Obama is losing in this group are states that Obama lost in 2008. But let's look at this from a more conservative angle. Sean Trende recently pointed out that most of these state polls are looking at registered voters rather than likely voters. There is very good reason for pollsters to do this because it is exceedingly difficult to get an accurate read this far out from the election on what the likely voter pool will be. But Trende correctly points out that likely voter pools are usually more favorable to Republicans than registered voter pools and that, on average, the shift from one to the other means about a 3-4-point shift in the Republican direction. So, let's take state polls above (the national polling does include more likely voter polls) and shift all of them 3.5 points in Romney's direction:

AZ - (4.0)
FL - (2.7)
IA - 6.5
MI - 0.5
MO - (6.5)
MT - (12.5)
NV - 4.7
NH - 5.3
NC - (0.9)
OH - 0.3
PA - 4.1
VA - 1.0
WI - 2.2

So who wins in this scenario? Obama. And he's still got a little bit of room to spare. "Really?" you say. Yes, really. In fact, he's at 303 electoral votes. "Well that Ohio result is so close it scares me. What if we give Ohio to Romney? Does Obama still win?" you say. Yes. Obama still wins. "Well, I've noticed Colorado isn't listed there" you say. "You're just cooking the books by giving Colorado to Obama!" Well, I would predict Obama wins Colorado right now (and I don't think it'd be that close) but okay, let's give Colorado to Romney. What happens now? Obama still wins ... with 276 electoral votes. You can play around with the math on this here but I'm telling you ... Romney's path to 270 is VERY narrow right now.

Recently, my aunt attended a lecture and the lecturer (a political scientist) said the election will come down to "FLOHPA," the key states of Florida, Ohio, and Pennsylvania, and whichever candidate wins 2 of the 3 will win. My aunt was quite alarmed because she thinks Obama is in trouble in Florida. But guess what? The old "FLOHPA is decisive" argument is not necessarily true. 1) I just showed you how Obama can lose Ohio and Florida AND lose Colorado and still win. 2) For the love of all that is holy, let's take Pennsylvania out of the "swing" state column (see my earlier rant on this). Obama is either winning Pennsylvania or he's losing VERY badly. But the election is NOT turning on Pennsylvania.

My broader point for today is this. Take the state polling into account and then shift everything 3.5 points in Romney's direction ... and Obama is still in the driver's seat. I'm not saying the election is over and I'm not saying Obama is running away with it. But Obama is winning right now and nothing happened this week (no, not even the gay marriage stuff) to change that.

Swing State Polls

I said the other day that Obama seems to be doing better in Ohio than he is doing nationally. Today, a few swing state polls seem to confirm that Obama is indeed doing quite well right now.

The Purple Poll shows Obama up by 5 in Ohio, up by 2 in Virginia, tied in Colorado, and down by 2 in Florida. If those are the actual results on Election Day, Romney loses the election for sure. A new Rocky Mountain Poll in Arizona has Obama up by 2 there. Arizona is not an absolute must-win state for Romney but it would be very bad if he lost there.

It is a long, long way to go. But early on, this race is resembling the 1996 race in one key respect. It "feels" stable to me. It doesn't seem like we're likely to see wild swings back and forth in the absence of some big external shock. Intrade has Obama as about a 60% bet to win. I think he's in a little bit better position than that.

Enthusiasm

The lack of enthusiasm for Romney is what hurt him in Colorado. Some very interesting graphs from Seth Masket on this point.

The summmary is that Republicans voters who are "okay with" Romney are not turning out for him in very good numbers. So even the numbers he's getting in polls are a bit soft.

We're Number 1 ... or Number 2!!!

Mitt Romney got absolutely shellacked in Missouri, is getting shellacked in Minnesota, and he may even lose Colorado. From TPM:

Romney thanked all his supporters, including those who have gone out caucusing tonight in Colorado. "The race is too close to call in Colorado at this point, but I'm pretty confident we'll come in Number 1 or Number 2."

That's right ... he's "pretty confident" that he'll come in "Number 1 or Number 2." As Andrew Sullivan pointed out tonight, it appears that if he doesn't have a week and millions of dollars of negative ads to pound his opponent with negative ads, he can't seem to win.

We've now had 7 contests (IA, NH, SC, FL, NV, MN, MO) with one more pending. Romney has won 3 with CO pending. Santorum has won 3 ... with CO pending. Rick Santorum is pushing Mitt Romney. Really.

UPDATE: Yeah, Romney lost Colorado too. Something really bad has happened with Mitt Romney tonight. The bottom really dropped out. In talking about the low turnout in Florida, I had said Michael McDonald was actually understating what a major potential problem this was for Romney. Nate Silver points out that this pattern has finally cost Romney ... big-time:

We have repeatedly noted the pattern in which Mr. Romney's stronger states and counties have been associated with lower Republican turnout.

So far, it was not clear that this had lost Mr. Romney a state -- save perhaps Iowa, when virtually anything might have altered the result.

But in Colorado, where the demographics were reasonably favorable to Mr. Romney -- he won 60 percent of the vote there in 2008 -- it may have made the difference. Mr. Romney's stronger areas in the state were associated with turnout declines of about 20 percent. But turnout was steady or slightly up in places where Rick Santorum did well.

Among other problems for Mr. Romney, this suggests that suggests that the caucus states could be problematic rather than advantageous to Mr. Romney, with his superior organization being outmatched by very conservative voters who have low levels of enthusiasm for him.

So, we've now had 8 contests, Santorum has won 4 and Romney has won just 3. That's a pretty poor resume for Mr. Inevitable.

Minnesota, Missouri, Colorado

Well, I got one thing right last week when I said Santorum is the biggest threat to Romney. Santorum has a really good shot to win two of tonight's contests (Minnesota and Missouri) and he is even in good position to come in second in Colorado.

My reasoning was only slightly off. I had said that Santorum dropping out would boost Gingrich. It seems Santorum is a threat to beat Romney in key places. But the reasoning is not as far off as you might think. Santorum doesn't just have a chance to win in Missouri ... he could get close to 50%. This matters because my logic had been that there are some states where Santorum + Gingrich > Romney. Missouri provides a real test of this proposition because Gingrich is not on the ballot. Even with Paul on the ballot, Santorum is inching towards 50% (he got 45% in the PPP poll).

Meanwhile, Romney is downplaying the importance of tonight's contests. Like Iowa and South Carolina, they don't really matter apparently. My favorite part of this memo is their invoking the fact that John McCain lost 19 contests in 2008 so we're really doing just fine. What ever happened with that President McCain anyway?