Sabato's Crystal Ball

The Presidential Race: Where Does It Go From Here?

Larry J. Sabato, Kyle Kondik and Geoffrey Skelley, U.Va. Center for Politics October 18th, 2012

If the presidential race seems like it’s hard to get a grip on, that’s because it is — the contest has gone through at least three distinct phases at this point, and where it might go over the final three weeks seems to be anyone’s guess.

In the lead-up to the conventions, President Obama appeared to hold a small lead on Mitt Romney. The national polls would vary slightly, but the president generally held an edge of a few percentage points. This narrowed to an exact tie in the RealClearPolitics average on Sept. 5 — the second day of the Democratic convention — indicating at least something of a post-convention bounce for Romney.

After Obama’s convention, the president got his lead back, and he eventually expanded his national polling edge to 4.3 percentage points in late September. While this was not Obama’s biggest lead of the cycle — he was up 4.7 points as recently as mid-August — it was enough to signal that, barring some big outside development or gaffe at a debate, the president was in a strong position to win reelection.

Obama’s lead was down to 3.1 points by Oct. 3, the day of the Denver debate — and we all know what happened then. Over the course of just a handful of days following Romney’s Mile High rout, Romney took his first national polling lead of the calendar year in the RCP average, reaching a high of 1.5 points on Oct. 10, a week after the debate. And then, over the past week, the race settled into effectively a national tie — as of midday Wednesday, Romney held a tiny 0.4% lead; that includes Gallup’s seven-day tracker, which showed Romney up 51%-45% Wednesday (that does not include any polling following Tuesday night’s debate).

While it has been a topsy-turvy race, it’s also been one without particularly commanding heights or punishing valleys. Since April 10 — when Romney effectively clinched the nomination — he has never topped 48% in the RCP average and never dipped below 43%; Obama has never exceeded 49.5% and never gone below 45.4%. That points to a pretty stable and polarized electorate.

The problem for the president is that in the current average — 47.4% for Romney and 47.0% for Obama — Romney is closer to his peak than Obama is to his. This points to an enthusiasm gap for the president, as does the persistent registered voter/likely voter gaps in public opinion polls, which almost always show the president doing better among the registered voter pool than the smaller likely voter pool.

Perhaps Tuesday night’s debate will help change that. It’s obvious that President Obama did much, much better in the second debate than he did in the first — his answers were crisper and his criticisms of Mitt Romney more pointed. It’s fair to ask where this Barack Obama was two weeks ago, but for Democrats, Obama’s performance was better late than never.

The president won the debate over Romney, although it was not nearly as decisive as the challenger’s win in Denver. Romney got his shots in too, although he also stumbled at times and did not land an effective punch on the president during their exchange on Libya. Of course, Romney will get another shot at pressing Obama on foreign policy at the final debate. Given the questions asked during this town hall debate — they were mainly focused on domestic policy — one has to wonder how much voters actually care about the foreign policy topics that will be discussed in further depth next week.

Ultimately, it’s difficult to say what impact this debate will have on the race. Will it just stanch Obama’s bleeding, will it do more and prompt an Obama uptick, or will it have no effect at all?

Given the race’s ups and downs, we wonder whether we’re still in the third stage of the campaign — the Romney surge — or whether we’re back in the first phase of the campaign: a narrow Obama edge. Or could we be in a fourth, 2000-like stage — meaning a freakishly close election?

Rating changes

The respected Marquette Law School Poll, which had Obama up 11 in Wisconsin before the debate, now shows the president ahead only 49%-48% in Badgerland. This squares with three other surveys conducted after the first debate (but before the second), all showing Obama with small leads inside the margin of error. So Wisconsin is now a toss-up. We also cannot ignore the polls showing tightening margins in both Pennsylvania’s presidential and Senate contest, so we’re moving both to leans Democratic, down from likely. Romney probably only wins Pennsylvania in a decisive national victory — we’ll be stunned if it accounts for his 270th electoral vote — and the challenger’s campaign is wise to downplay his chances both in the Keystone State and Michigan. Meanwhile, it appears that Sen. Bob Casey (D) is running a subpar campaign, and businessman Tom Smith (R) is hammering him with ads. Last month, we ran a list of potential Senate shockersbut none of them would compare to a Casey loss.

This makes the Electoral College count 267 for Obama, 235 for Romney and 36 toss-ups (Colorado, New Hampshire, Virginia and Wisconsin). We do not have the data to move Iowa, Nevada or Ohio onto the list of toss-ups. Keep in mind that as of this writing we have little indication as to what impact the second debate will have on this utterly wild race.

Map 1: Updated Crystal Ball Electoral College ratings



Bypassing the Buckeye State?

If there is one state that everyone seems to be focusing on right now in the Electoral College, it’s Ohio. And why not — as we’ve mentioned many times, Ohio has been the most consistent presidential bellwether for more than a century, selecting the winner 27 of the past 29 elections.

President Obama has consistently held a slightly bigger lead in Ohio than he has nationally, and that still appears to be the case. That might have something to do with his support of the auto bailout, and Romney’s opposition to it. It might have something to do with lingering suspicion of Republicans after last year’s punishing fight to overturn restrictions on public sector unions. It also might have something to do with the state’s unemployment rate, which is below the national average (and which appears to also be boosting Republican Gov. John Kasich, who struggled last year during the labor fight). Or maybe it’s because Romney is a Massachusetts presidential candidate — Ohioans voted against John F. Kennedy in 1960, Michael Dukakis in 1988 and John Kerry in 2004. Put all these things together, and maybe that gives Obama a little bit more of an edge in Ohio than he has in some other places.

Ohio could still end up going either way, although we’re keeping it in the “lean Obama” category for now. And even though it’s likely that Ohio will end up voting for the winner — as it usually does — both candidates have plausible paths to victory without the Buckeye State.

For Obama, the path is to win all the states he won with a greater share of the vote in 2008 than he won nationally (as in, by more than 52.9%). That would get him exactly to 272 electoral votes — just barely over the finish line:

Scenario No. 1: Obama wins without Ohio




The really challenging state here for Obama is Colorado, where Romney holds a tiny edge in public polling.  Indeed, Colorado and Virginia — seen six months ago as potential firewall states for Obama because of their growing minority and professional populations — now both look like toss-ups that might even be tilting to Romney, while whiter, more blue collar states where Democratic presidential candidates have traditionally fared better, such as Ohio and Wisconsin, might be better bets for the president.

In the end, it’s hard to see how Obama wins Colorado but loses Ohio, but it’s also not impossible.

Scenario No. 2: Romney wins without Ohio




For Romney, a Buckeye-less victory means not only wresting away Obama’s three 2008 Southern victories — Florida, North Carolina and Virginia — but also picking up a few more states that appear more challenging. Iowa has been sparsely polled in the post-debate phase of the campaign, but perhaps it, like so many other states, has tightened up; Romney would need it if he didn’t win Ohio. He also would need Wisconsin, a toss-up that nonetheless will be a heavy lift for Romney given its Democratic history.

Scenario 3: The tie




The simplest way to explain this scenario is that Obama retains all the states John Kerry won in 2004, while adding New Mexico and Ohio. Under the 2004 map, that would make Obama president. But the 2010 reapportionment sent a handful of electoral votes from the Blue states to the Red states: just enough to make this scenario an exact 269-269 tie. But 269 would make Romney president, because in the event of a tie, the new House of Representatives would pick the president: each of the 50 delegations would get a single vote. Given that Republicans are almost assuredly going to control a solid majority of House delegations in the next Congress (see below), Romney would win.

Of course, this scenario involves Romney winning Nevada, but that might be very difficult. The Democratic machine in the Silver State that carried Sen. Harry Reid (D) over the finish line in 2010 appears to still be operating at full efficiency; Democrats continue to have a wide registration advantage in Nevada, and Jon Ralston, the prominent Nevada politics expert, believes that Obama has a lead there and that polls understate his advantage, much like they did for the president in 2008 and for Reid in 2010.

We don’t have a dog in this race, but we are rooting for one thing: no tie! A 269-269 Electoral College outcome would inevitably be a national crisis on par — or worse — with the 2000 Florida cliffhanger, especially if Romney lost the popular vote.