Political Animal


October 25, 2012 11:39 AM Not-So-Big Mo

By Ed Kilgore

Unfortunately, I’ve heard nothing but crickets in response to my request the other day for evidence that media horse-race perceptions of “the lead” or “momentum” in a close presidential race affect actual voters. But Nate Silver does offer some valuable thoughts today on the inherent implausibility of the “momentum” idea, along with the weak support he finds for the idea Mitt Romney’s got it:

The term “momentum” is used very often in political coverage — but reporters and analysts seldom pause to consider what it means.
Let me tell you what I think it ought to mean: that a body in motion tends to stay in motion. That is, it ought to imply that a candidate is gaining ground in the race — and, furthermore, that he is likely to continue to gain ground.
As a thesis or prediction about how polls behave, this notion is a bit dubious, especially in general elections. In races for the United States Senate, for instance, my research suggests that a candidate who gains ground in the polls in one month (say, from August to September) is no more likely to do so during the next one (from September to October). If anything, the candidate who gains ground in the polls in one month may be more likely to lose ground the next time around.

This last insight is based on the mathematical concept of mean reversion, which in politics suggests that the “fundamentals” that produce the “natural” state of any given contest tend over time to override short-term trends. Barring large external events (and those who think the first presidential debate was one of those are basing their judgments on some combination of spin and polls and spinning polls), and particularly at a time when partisan preferences are so fixed and powerful, it’s really implausible to suggest that suddenly one presidential candidate or another has taken flight and is now unstoppably building an ever-greater lead—because of “momentum!” And according to Nate, there’s little or no evidence that’s happening anyway:

Part of the confusion (and part of the reason behind the perception that Mr. Romney is still gaining ground in the race) may be because of the headlines that accompany polls.
We’re still getting some polls trickling in where the most recent comparison is to a poll conducted before the Denver debate. We should expect Mr. Romney to gain ground relative to a poll conducted before Denver. (Mr. Romney may have lost a point or so off his bounce, but he has clearly not lost all of it). But it isn’t news when he does; Mr. Romney’s Denver gains had long ago become apparent, and priced into the various polling averages and forecast models.

This should be part of anyone’s mental furniture by now, but it bears repeating: when looking at polls for “trends,” you have to look at who’s doing the polling, when the poll was taken, and what relationship it has to other polls done in the same place at the same time. Anyone can stack up a list of national or state polls to build a tenuous case for steady Romney gains, but often that involves confusing release dates with polling dates, or simply dismissing contrary data. “Momentum” is usually in the eyes of the spinner, and its power is usually derived from the panicked reactions of the “other side,” which fears (without evidence) that “momentum” turns votes. I don’t see it, but the dynamic explains a lot of what we are hearing from the commentariat this week.

UPDATE: At WaPo, Jonathan Bernstein is even more categorical: “In thinking about elections, the idea of momentum is useless.”

Ed Kilgore is a contributing writer to the Washington Monthly. He is managing editor for The Democratic Strategist and a senior fellow at the Progressive Policy Institute. Find him on Twitter: @ed_kilgore.


  • stormskies on October 25, 2012 11:48 AM:

    Yep, the great CON JOB of the vast amount of the corporate media that promotes this propaganda simply reveals the reality of those that do it: to get pathological liar, buffoon, and corporate automaton Romney installed as president at all costs.

    This is why, to me, one of the greatest dangers to what is left of our country is in fact the vast amount of the corporate media itself, and the 'narratives' and 'story line's' they invent that serve as the vehicles of the propaganda that they want Americans to 'believe'.

    All's one has to do, for example, is observe the likes of David "I am not a used corporate condom" Gregory to see this fact. Right after the last debate in which Obama crushed Romney there was the used condom intoning, while the corporate cum was oozing from it's mouth, nose, and eyes, that this would not stop the momentum of Romney as if the used condom could know such a thing right after the debate.

  • c u n d gulag on October 25, 2012 11:50 AM:

    Though I get nervous from time to time, I've always maintained that women will be the difference in this election.

    And, I feel more confident now than I have in awhile.

    First with Akin, and now with Mourdock, I think more and more women, even those in Conservative households, may pull the lever, or fill-in the dot, for President Obama and the Democrats - no matter what they may be telling the pollsters when they're sitting at home, with their husband, brother, or father nearby.

  • Bob on October 25, 2012 12:12 PM:

    I remember four years ago, many conservative pundits pushing the idea in the last two weeks of the election, that McCain was closing in. We know how that turned out.

    For me, this is all about the "main stream media" being snookered by the Repubs, and then fervently hoping for a close race to maintain the interest of readers. In the interest of attracting readers/listeners/watchers, they will continue to push this Romney momentum business against all evidence to the contrary.

    I cannon wait for this election to be over.

  • John Brandon on October 25, 2012 12:41 PM:

    The problem is, you wouldn't be downplaying Romney's momentum unless you thought that the mere mention of it tends to depress the democratic vote or GOTV efforts. Why else would you write this post? You're trying to cheer up democrats who are demoralized and encourage them to make a difference in this election. So obviously you DO think the perception of momentum affects what people do.

  • bigtuna on October 25, 2012 12:57 PM:

    One funny effect of the new reality of billion dollar campaigns - it injects a good slug of cash in certain states, upping employment in a few places, etc. So the repubs and the rovian groups are spending like drunken sailors, thereby amping the economy in at least a few states, where they are trying to convince people that the world will end soon, and maybe cutting the unemployment rate, temporarily, at least a teeny bit.

    tee hee.

    By they way.. maybe I am paranoid, but the big news in Utah a few weeks ago is the change in minimum age for LDS kids to go on missions, and the training sessions for missionaries in training will be shortened. THey said that it was because the LDS leaders feel that the word has to get out to more people faster. Either:

    a. They figure BHO will win, and the end times are closer
    b. WMR will win, and they will dominate more easily with and LDS president [ I am not kidding]

    Hard to read the tea leaves here.

  • LAC on October 25, 2012 12:59 PM:

    Bernstein ought to look at his own paper. The level of delusion and opinion as fact is astounding. And John Brandon, I don't know what or who you are talking about, but the democrats I know are into this, are getting out the vote, and are only dismayed that flip flopper has to be propped up by a media narrative to seem viable.

  • Robert from upstate on October 25, 2012 1:06 PM:

    Hey John B,

    You would not be hovering around here offering your Rethug Big Lie Troll pyscho-babble, if not for the fact that there are large beads of sweat on your very thick eyebrows and hairy palms. Whatever will you all do when you realise that endless piles of corporate and overseas corporate mcash does not guarantee a win?

  • John Brandon on October 25, 2012 1:07 PM:

    LAC, you've just proven my point. If you believe that Romney can be "propped up" by a media narrative, then you're saying that the narrative affects the polls.

  • BJ smith on October 25, 2012 1:12 PM:

    Aree with Gulag & Bob. If there is to be an O victory he will have women to thank, especially if the Mourdock story has legs, also if seniors can shake the truth out of repub nonsense like the lies of R & R that they will save Medicare. Keep hoping women are opening up to each other at their jobs about women's issues. Most women are far to busy to keep up with with all that is being said. At least older people have more time to explore the facts. Mostly as Bob stated, make it go away by being over with a win for the D's.

  • FlipYrWhig on October 25, 2012 1:24 PM:

    "momentum" never made sense to me in sports, either, where it is most commonly used. If a team starts playing well, scoring points, the announcers will say that they have Momentum. Then if the other side starts playing well and scoring points, the announcers will say that _they_ now have Momentum. If you can lose it easily, what's the point of having it? All things considered, does Momentum mean anything more than "doing well" lately? I don't think it does.

  • Tom on October 25, 2012 1:49 PM:

    The "Momentum" meme in itself may be neutral but it is still good strategy. Any time the press is not talking about issues is an advantage to Romney.

  • Jim Bunnell on October 25, 2012 1:54 PM:

    I think the momentum narrative is, in part, to provide cover for rigged election results, a corrupt election apparatus, or politicized court.

  • Sgt. Gym Bunny on October 25, 2012 3:20 PM:

    This may sound like a turd-brain question, but...

    With all these polls a-going on, why haven't I ever been polled? I don't even think I know anybody who has been polled this election season...

    Just askin'...

  • bos'n on October 25, 2012 3:52 PM:

    John Brandon, LAC means the media is 'propping up' MittMo the way a cardboard cutout is propped up by the unseen supports behind it. The cardboard sign is still only a crude facsimile of a living thing, much like MittMo. I think that Republicans and their corporate media allies are boosting the MittMo story in hopes that if their voter suppression and vote fixing schemes work enough to produce uncertainty and chaos on election day, they can jump in and plausibly declare victory for Mitt, much like they did for GWB in 2000. I'm saying BO wins this; if he doesn't look for dubious chicanery in places like Florida, Ohio, & Pennsylvania, states where GOP election officials have already shown their hands in many, many ways.

  • bos'n on October 25, 2012 3:56 PM:

    I meant to put a comma between "if he doesn't, look for dubious..." Hope that makes the meaning clearer.

  • Doug on October 25, 2012 6:26 PM:

    Actual momentum may still be present in some primariy campaigns and I know it USED to occur occasionally during national nominating campaigns - back when states held conventions. Nowadays? Nope.
    Merely yet another attempt by the MSM to justify the facile "horse race" meme.

  • Keith M Ellis on October 25, 2012 7:41 PM:

    There's several things going on in these ideas of momentum.

    First and foremost, it's simply that we have a very strong instinct to identify patterns, whether there's really a signal there, or it's mostly just random noise.

    The purest example of this, in both the mathematical and psychological senses, is the Gambler's Fallacy: that a string of independent events represents a "trend" which predicts both short-term events (the next bet should follow the trend, i.e., "I'm on a streak!") and long-term events (good luck or bad luck will have to end and its opposite will take hold). The Gambler's Fallacy both sees momentum and an opposing force, depending upon a person's state of mind. ("Reversion to the mean" isn't the same thing, as it doesn't actually predict individual events or otherwise claim that subsequent events are somehow influenced by the previous aberrant ones. It's just a different view of the law of large numbers, saying that there will be really unusual events now and then, but the overall pattern will follow the probabilities, so those occasional events will skew the average, the mean, only temporarily.)

    So this is partly just about the essential human cognitive mechanism of pattern-finding.

    However, in both politics and sports there an entirely plausible, even undeniable, social psychological effects involved.

    In competitions, where there is rough parity in skills and preparation, motivation and optimism/pessimism can be a strong factor and it's pretty sensitive to the perception of such trends. However, I'd argue that in both individual and team sports, such examples of psychological "momentum" are only truly real at the level of individual games, not sequences of games.

    In both competitions and politics, there's also just the consensus forming nature of social psychology. This is also undeniable, and it can interact with the previous motivation/optimism/enthusiasm factor where there's a social consensus formed about performance. And in the more general social sphere, clearly there's momentum in how a consensus starts to form and then rushes like a tide over a whole group.

    So in these respects, there's some valid reasons for people to expect there to be "momentum" in political campaigns.

    The problem with both these factors as applied to national or otherwise high-profile political campaigns is that they each rely on things not present in these contexts.

    The psychological motivation stuff, and in a group context, is real only in a short-term time-frame and these quick bursts of enthusiasm or discouragement is not what people are talking about when they're talking about "momentum". They're talking about it being much more long-term. But it's not, because it would take a truly unusually large impetus to affect peoples' enthusiasm in such a long-standing respect. Which does happen from time to time, but not from a bad debate performance. More from the revelation of a bad scandal that disqualifies a candidate.

    And the consensus-forming stuff, where momentum is quite evident in all of our own personal experience, also doesn't apply in these contexts because, well, there's too many people and, crucially, consensus momentum occurs mostly where there were no identifiable initial affiliations/positions. But Presidential elections begin with almost everyone always sorted, having affiliations, and also this is a relatively information-rich environment. There's not going to be a national consensus forming on whether one prefers hamburgers or hot-dogs because our preferences were mostly established prior to the context of the question being asked and there's very little information offered in the form of what other people think relative to what we already know about our own experience. Picking Presidents is much more like this than it is in an emerging consensus in a medium-sized group situation.