Reëlection Watch: August 11, 2012
How are things going for the President now? Let’s dive in.
The trend I mentioned two weeks ago of President Barack Obama’s approval/disapproval rating in the Real Clear Politics average appears to have reversed. We’re back to the same old story of oscillation right around zero, as it has been since the beginning of spring. In other words, it’s reverted to the mean, after a brief dip into negative territory. The peak of negativity I mentioned last time in the Right Track/Wrong Track polls remains steady.
The improvement in Congress’s approval spread continues, mostly through increases in approval rather than decreases in disapproval. The generic Congressional ballot moved back into Republican territory, but we’ve seen oscillations there around zero in the same way that we’ve seen oscillations in Obama’s approval/disapproval gap. I expected the noise to diminish as we got more frequent polls, but that has yet to materialize. At this point, I don’t expect the generic ballot to tell us much of use anyway. It’s time to put it up in the top shelf of the closet, in the back.
In the national popular vote matchup of Obama versus presumed Republican nominee Mitt Romney, the President’s lead has grown sharply in the past two weeks, manifesting itself in a combination of growth for Obama and shrinkage for Romney. It’s a large enough and sudden enough shift that it leaves me wondering if it’s going to be a blip that will be followed by a reversion to the mean. If it’s a trend, it’s one that’s significant enough to put Obama in the territory of a confident national popular vote win. We’ll see in a couple of weeks how sticky this shift really is. Interestingly, Obama’s lead over Romney today is about the same as his lead over Senator John McCain on this date four years ago. On the other hand, Senator John Kerry had a lead, albeit a bit smaller, over President George W. Bush on this date in 2004.
Obama’s favorability polls remain pretty consistently over 50 percent favorable among “Americans” or “registered voters”. Oddly, we haven’t had favorability polls of “likely voters” recently. Still, whether you look at recent numbers or longer history, the President has positive favorability, even while his job approval has been lower. Romney’s favorability polls haven’t changed in a statistically significant manner, and the higher favorable numbers I mentioned two weeks ago appear to have been statistical noise. Obama continues to maintain the edge here.
While these polls still indicate a possible Obama victory, they are typically farther removed from the key signal of electoral votes than are many other indicators. We’ll hit the others down below.
As of yesterday, Intrade had Obama at a 58.7, up a couple of points from last time. This is the fourth increase in a row, and big enough to indicate a trend, reversing his April-through-June decline.
Things remain in a pro-Obama state on the national scene, and the cloudy view last time appears to have cleared.
In this edition, the section on Mitt Romney is being replaced by a little on the economy.
This week, in Richmond, California, a crude oil separator exploded at the largest Chevron refinery in the state, stopping its ability to convert crude oil into its primary components. The portion of the plant that further refines those primary components is still functional, and it will continue to produce retail product from its onsite supply of those primary components. However, until the separator is repaired, no crude oil will be processed at that plant.
The national refining infrastructure has very little excess capacity, a situation that oil companies like to blame on excess regulation. In reality, there is little excess capacity because excess capacity translates to wasted capacity, and thus wasted money. There is no incentive for any oil company to build a plant whose sole purpose is to remain idle until something tragic happens at the others.
As a result, the national supply of refined petroleum products will necessarily decline over the next few months. This has already led to a spike in gasoline prices on the wholesale markets, and will most likely result in an increase of about 30 cents per gallon in the near future.
It’s unclear what impact, if any, this will have on either the overall economy or Obama’s prospects for reëlection. This is something we will need to follow closely over the next month.
The Electoral College
This is the first August Reëlection Watch. We’re now using a further narrowed set of bands for tossups and leaning states.
Here’s what the Electoral College looks like, based on current polling data:
Here are the states with new data since last time, covering only those discussed around the Internet as “leans” or “tossups”, from reddest to bluest:
- Arizona was polled by Public Policy Polling, who found an 11-point lead for Romney. This is within the range of expected values from previous polls. In light of the narrower bands, plus a consistent double-digit margin for Romney, I’m moving Arizona into the “Likely Romney” column. I have always been skeptical that Arizona was in play, but at this point I’m convinced that it’s not; it’s the first state to make a move into the “Likely Romney” column since I began tracking the electoral votes.
- Georgia had two new polls, one from SurveyUSA and one from InsiderAdvantage. Both used likely voter models; the former showed Romney with an eight-point lead, while the latter has him up by nine. There’s no change here, since these are well within the historical margins.
- Indiana was polled by Rasmussen. Indiana is a difficult state to poll, because state law prohibits robo-polling. This means that pollsters’ techniques in Indiana are inherently different from their techniques elsewhere, which means that we can’t assume that anything we know about a particular pollster can be used in Indiana. So, while Rasmussen’s 16-point Romney lead would ordinarily push the Hoosiers into “Likely Romney”, I can’t safely make that assumption here. The data are simply too sparse, and the error bars too wide. Indiana’s a “Leans Romney” state in terms of data, though probably a “Likely Romney” state in reality.
- Missouri was polled by Rasmussen and Post-Dispatch/Mason-Dixon. Rasmussen found a six-point unadjusted lead for Romney (five adjusted), while Post-Dispatch came away with a nine-point Romney lead, a result that matched the WeAskAmerica poll from last time. Missouri is still a “Leans Romney” state.
- North Carolina was polled by Public Policy Polling, who saw an unadjusted three-point Obama lead (which adjusts to a tie), and Rasmussen, who found Romney up by five unadjusted points (four adjusted). It’s just on the “Tossup” side of the border with “Leans Romney”.
- Iowa was polled by Rasmussen, and the two-point unadjusted margin for Romney (corresponding to an adjusted one-point margin) is rather close to Public Policy Polling’s half-point adjusted margin for Obama from two weeks ago. No change here; Iowa’s still an up-the-middle “Tossup”.
- Florida had two new polls this time. CBS/New York Times/Quinnipiac saw a six-point unadjusted lead for Obama (eight adjusted), while Public Policy Polling found him up by one, unadjusted (this translates to two points for Romney). The noise is still of larger magnitude than the margin, making it impossible to determine a leading candidate in the Sunshine State from polls alone. Florida remains a total enigma, and thus an up-the-middle “Tossup”.
- Colorado was polled by three firms in the past two weeks. CBS/New York Times/Quinnipiac saw a five-point unadjusted lead for Romney (four adjusted), while Rasmussen found an unadjusted tie (which adjusts to a little over a point for Obama) and Public Policy Polling had Obama up by six unadjusted points (three adjusted). With the overall trend, it seems that Quinnipiac had an outlier this time. I still think Colorado is on the blue side of the middle, but within Tossup range.
- Virginia was polled by both Rasmussen, who saw Obama up by two unadjusted points (three adjusted), and CBS/New York Times/Quinnipiac, who found Obama up by four unadjusted points (six adjusted). Virginia is still on the blue edge of a Tossup, but another round of polls like this will push it over the edge into “Leans Obama”.
- Ohio got a new poll from CBS/New York Times/Quinnipiac, who saw a six-point unadjusted lead for Obama (seven adjusted). Ohio’s polling is still noisy, but the broader trend continues to suggest that we’re beyond the point where the noise is a larger amplitude than the margin. Ohio is a “Leans Obama” state, contrary to what Real Clear Politics believes.
- Michigan had two new polls in the past two weeks. Both EPIC-MRA and Rasmussen came away with six-point Obama margins, though Rasmussen’s adjusts to seven. The Mitchell Research poll I mentioned last time does, indeed, appear to have been an outlier. Michigan is still a “Leans Obama” state.
- Wisconsin was polled by two firms in the past fortnight. Marquette University found a five-point Obama lead, while CBS/New York Times/Quinnipiac saw a six-point unadjusted (eight point adjusted) lead for the President. These are both within the bounds of the long-term trend. Wisconsin remains in the “Leans Obama” column.
- Pennsylvania was polled by CBS/New York Times/Quinnipiac, who saw an astonishing 11-point unadjusted lead for Obama. This adjusts to 13 points, but it’s far enough from the trend that I suspect we’re looking at an outlier here. Barring confirmation of this sort of lead, Pennsylvania remains a “Leans Obama” state. With confirmation, however, it would move to “Likely Obama”.
- Connecticut was polled by Public Policy Polling, who saw an eight-point unadjusted Obama lead, corresponding to a five-point adjusted lead. Connecticut has always been polled as a closer race by PPP than by Quinnipiac, the only other pollster to publish results for the Constitution State. Could it be a “Leans Obama” state instead of a “Likely Obama” state? I don’t think so, but we should at least keep an eye out.
In the past two weeks, most states of interest were polled, most by multiple companies. Arizona was the lone mover this time. Obama still ends up with a likely 275 electoral votes. That’s two times in a row where Obama could lose all tossups and still stay in the White House.
Romney has been returning to the mean, which doesn’t bode well for the challenger. The Electoral College remains in Obama’s court.
If I had to predict an Electoral College result, I’d keep things exactly where they have been for the past month, and I have increasing confidence. That prediction would give Obama 303, and Romney 235. In that scenario, Obama would be 62 votes shy of his 2008 tally.
How do you feel about these predictions? Do you differ on them? If so, how, where, and why?
- Reëlection Watch: July 28, 2012 (logarchism.com)
- CNN Poll: Obama holds 7-point lead over Romney (politicalticker.blogs.cnn.com)
- How Romney’s Pick of a Running Mate Could Sway the Outcome (fivethirtyeight.blogs.nytimes.com)
- Shouldn’t Romney be ahead by now? (politico.com)
- Obama’s Lead Grows as Romney’s Support Slips (foxnews.com)
- Gallup: CancerGate Ad Flap Hurting Obama (breitbart.com)
- Ten Swing States Could Decide the 2012 Election; Obama Leads in Nine of Them (salem-news.com)
- Polls show Obama ahead in presidential race (capitalfm.co.ke)
- Aug. 1: Obama Extends Electoral College Advantage (fivethirtyeight.blogs.nytimes.com)
- The disappearing undecided voter (politico.com)