Just over five months ago, when Wiesman and I published our initial predictions for the outcome of this election, the only thing we differed on was whether the President would win Florida. Today, we still differ only on whether the President would win Florida. My official map:
Five months ago, I gave the following rationale for giving Florida to Obama:
1. I predict Obama’s popularity will continue to improve as election season wears on. The economy will continue to improve (barring some natural or man-made catastrophe), and it will become increasingly clear that there is only one adult in this race and he’s not from Massachusetts. In particular, I think the debates are going to be particularly kind to President Obama. Furthermore, if the Ron Paul devotees have anything to say about it, Mitt Romney won’t be enjoying the post-convention bump that Republicans typically experience. Add all this together, and I think Obama can pencil in an extra 1-3% on top of his current poll numbers in all the swing states.
2. Florida is not Mitt Romney’s sort of state. If you’ll remember, Romney won the incredibly important Florida primary by flooding the state with a torrent of negative attacks on Newt Gingrich. Romney enjoyed a 5:1 spending advantage over Gingrich in that race and though he won by 15 points, turnout was down by over 300,000 voters, a 15% decrease in turnout in a state whose population has grown by nearly 3% in the same period. As the exit poll data showed that night, this is not a state in which the conservative base is particularly enamored of Mitt Romney and presidential elections are all about turning out your base.
3. Demographics. I suppose this could be folded into “Florida is not Mitt Romney’s sort of state,” but it’s important enough to get its own header. Florida has the largest population (17.3%) of elderly citizens in the United States, which is significant both because older voters have higher turnout than other age cohorts and because Republicans in Congress have been lining up behind Paul Ryan’s budget plan, which would slash Social Security and Medicare. Obama has already been hammering this plan as “social Darwinism,” and we can expect his attacks on the Ryan budget to intensify as Election Day draws nearer. I expect senior citizens to be highly mobilized in this election, which can only bode poorly for Romney in Florida. It is also worth noting that Florida has large Hispanic, African-American, and Jewish populations that should also be in Obama’s camp, although Wiesman was correct to note that GOP voter suppression acts of late may be successful in limiting the effect of those groups.
4. I’m just a homer. Goes without saying.
Certain of the predictions underlying my call of Florida have gone better than others. As I expected, economy has continued to improve, Romney did not get very much of a post-convention bounce (although this may have had more to do with Clint Eastwood’s chair than disruption by the Paulites), and the President is generally more popular than he was back in May, to the tune of an extra 1.8% in the RealClearPolitics average of his approval rating.
What I did not count on, however, was the dominating performance Romney put on in the first debate, which really set back a lot of the progress I anticipated. On October 3rd, the night of the first debate, Obama was riding high: 538 was giving him an 86.1% chance of winning and giving him an average of 319.3 electoral votes. I don’t see a way to look at old projections for individual states, but I distinctly remember mocking Wiesman as Obama took the lead in the Florida polls (a 2 point lead, according to RCP) and even started to trend toward taking back North Carolina. Nine days later, that lead had almost completely disappeared, Romney had his biggest polling average lead of the campaign in Florida, and Obama has been scrambling back ever since.
So, that’s where we are. As Wiesman said in his post, 538 is putting Obama’s chances in Florida at 45%. I am less confident now than I was in May, but I still believe that 45% shot will come through and Obama will prevail in Florida. Here are my reasons:
1. Florida is trending Obama’s direction. Obama’s peak in Florida came in the wake of the Democratic convention, with the president capturing a lead of 3.2 points in the RCP average. Nine days after the first debate, he hit his nadir, with Romney ahead by 3.2 points in the same measurement. Since then, Obama has closed the gap to 1.8 points, with increasingly favorable polls coming in all the time. I think this reflects Florida’s fundamental identity as a 50/50 state that shifts with the political winds. The winds are pushing in Obama’s direction, and I believe that bodes well for him there.
2. The demographics in Florida remain in Obama’s favor. Just as I said back in May, Florida has large populations of elderly and minority voters that should be highly motivated to turn out for Obama. In May, I reasoned that the Ryan budget would be a big factor in this motivation. I think that still holds, but the leaked video of Romney’s comments about “the 47%” is even more powerful with those groups. Obama has been running ads invoking those remarks incessantly, and it has proven a powerful political weapon.
3. I’m a homer. As true as ever.
Despite the above, I’ll not be surprised if Florida does go the other way. As Wiesman said, there’s no telling what effect all the voter suppression tactics pushed by the Republican leadership in Tallahassee will have, and it’s certainly possible that Romney could win “legitimately” too.
Where I disagree with Wiesman is in his assessment of the “worst-case scenario.” If Wiesman’s read of the situation is to be believed, the worst Obama can do tomorrow night is winning with 281 electoral votes. Though I applaud Wiesman’s optimism – and would invite him to extend that optimism to the president’s chances in Florida – the fact of the matter is that the worst-case scenario is much worse than that. Here is the worst-case scenario, as I see it.
Here’s what this map represents: Romney wins Florida, wins the close races in Virginia, Colorado, and Ohio, rides a recent uptick in the polls in New Hampshire, and pulls out an almost unthinkable comeback in Pennsylvania. In order for this map to come to fruition, two things would have to be true: the polls would have to be severely missing in their likely voter models, and the GOP’s voter suppression tactics would have to be more effective than most people are giving credit for.
I don’t think that this is at all a likely state of affairs, but I don’t think Wiesman has the right of it in saying that Obama cannot lose. The polls and expectations of neutral observers are in Obama’s favor, but we don’t have to go far to find examples of the polls and expectations being wrong: Harry Reid keeping his Senate race in 2010, Obama’s victory in Indiana in the 2008 general election, Hillary Clinton’s shocking comeback in the 2008 New Hampshire primary…the list goes on. As Nate Silver said in a post this weekend,
We’ve about reached the point where if Mr. Romney wins, it can only be because the polls have been biased against him. Almost all of the chance that Mr. Romney has in the FiveThirtyEight forecast, about 16 percent to win the Electoral College, reflects this possibility.
The state polls may not be right. They could be biased. Based on the historical reliability of polls, we put the chance that they will be biased enough to elect Mr. Romney at 16 percent.
Now, 16 percent – or 13.7%, as the 538 model now pegs Romney’s chances – is not much. To analogize to poker, it’s about the chance that a pair of Aces will be beaten by King-Queen offsuit if those hands go all-in before the flop. But as a poker player himself, Wiesman should understand that King-Queen will sometimes crack those Aces, and sometimes it will happen at the worst possible time.