The Right Perspective on Debates

by John Sides on September 29, 2012 · 15 comments

in Campaigns and elections

Gwen Ifill:

Gallup polls going back decades show precious little shift in established voter trends before and after debates.

Chris Cillizza:

…there are relatively few examples of times in which the general election presidential debates fundamentally altered the course of a race.

Miranda Green:

Presidential debates rarely have much effect on election outcomes.

It’s nice to see these, rather than hype about the debates as “game-changers.”  Maybe political science is “killing the campaign narrative” after all?

UPDATE: Add John Harwood to this list.

UPDATE #2: And Mark Barabak.


sarang September 29, 2012 at 8:38 pm

For once I have a mild preference for the conservative counter-narrative, viz. debates-not-mattering is what the Obama partisans would prefer to be the case, so they’ll go with it.

sarang September 29, 2012 at 8:40 pm

PS I think the best-case scenario for political science in public discourse is to become like evolutionary psychology, i.e., something that is trotted out in support of views that people are already predisposed to hold.

Guilherme September 30, 2012 at 4:51 pm

And where do you think these predispositions come from, chump?

Zachary Alain September 29, 2012 at 10:52 pm

Kill the narratives! Too long have we suffered them! I want do bathe in the blood of mediocrity! Let boring reasonableness smother the writhing sensationalist in the womb!

Tobler September 30, 2012 at 12:27 pm

“It’s nice to see these, rather than hype… Maybe political science is “killing the campaign narrative” after all?


Ya think ?

Why not just wait for the actual election ?

Political-Science adds almost nothing but hype to the election run-up.

Obsession with polls & horse-race is foolish.

The debates themselves add little substance– since neither Romney nor Obama will speak honestly. That’s the standard for professional-politicians (dishonesty/spin)… and seems quite acceptable to the general political-science community as well.

John Sides September 30, 2012 at 12:54 pm

Tobler: If you think political science is synonymous with “obsession with polls & horse-race [sic],” then I encourage you to spend some more time on this blog. Or maybe read some political science.

JS September 30, 2012 at 9:13 pm

Agreed, and speaking as not a political scientist*, just a politically-interested follower of blogs such as this one and Plain Blog About Politics.. Poly Sci is about issues like explaining how the “Invisible Primary” works, and batting down all of the scenarios I could come up with where Newt Ging… I mean Rick Santor… I mean any of the not-Romneys might have put together a coalition to get through the weirdest primary season of my lifetime.

It may take a couple more election cycles for this to sink in with “the media”, I’ll grant you. But that’s more because the media types have a strong financial incentive to keep the horse race frame alive in the future, not because the actual political science is wrong.

*one elective “Intro to Political Science” class circa 1980

Nadia Hassan September 30, 2012 at 10:27 pm

Professor Sides, Professor Shaw brings up the 2000 experience, but I am wondering if the 2000 debates tended to matter for character traits or ideology. Bartels and Zaller 2001 and Fiorina et. al 2002 both found that Gore was disadvantaged because he was perceived as more ideologically distant from the median voter. Is there any research that tracked perceptions of candidate ideology in 2000 to figure out when those perceptions may have developed?

Troy Smith October 1, 2012 at 3:00 am

To summarize: Presidential debates don’t make a difference, except sometimes they do.

In other words, the following statement is just as accurate: the upcoming debates may make a significant amount of difference.

Until pundits or political scientists can accurately predict beforehand that these specific debates will (not) influence the election, then it is irresponsible to imply to the public that these debates are likely meaningless.

Is it not irresponsible for political scientists to think they are doing their science in a vacuum and proceed heedless of how their actions might affect actual events, especially when their conclusions would be overturned if political scientists did a better job teaching and respecting civics and civic responsibility? If more citizens considered knowing the candidates, their character and positions as a duty commensurate with voting, then the debates would probably have greater influence.

Nadia Hassan October 1, 2012 at 6:00 am

Well, it is often asserted that they’re a big deal, and nobody is saying not to watch them per se.

John Sides October 1, 2012 at 7:53 am

Troy: That’s wrong in several respects. The best summary is not your cheeky version, but this: “Debates sometimes create shifts in the polls, but they rarely decide the winner of the election.” That’s just the historical record.

Furthermore, you suggest that if political scientists say that debates rarely matter, then somehow we are failing to teach voters civic responsibility — as if political science is somehow powerful enough to lower the Nielsen ratings for the debates. But it’s the voters who generally don’t react to debates. Political scientists are just reporting that fact.

Troy Smith October 1, 2012 at 1:22 pm

John: I’ll go with your explanation over my “cheeky version,” but your version “[the debates] rarely decide the winner of the election” still allows for the possibility that these 2012 presidential debates could influence the winner of the election. Until political science can actually predict specific events rather than prognosticate based on means and averages, then such claims as made in the above statements are irresponsible in that the claim exceeds the data and conclusions and the claim may actually influence the outcome.

Your characterization of my argument about failing to teach civic responsibility mischaracterizes my argument by confusing causality. Teaching the claim “the debates rarely matter” is not failing to teach civic responsibility. Rather failing to teach civic responsibility may lead to public apathy or cynicism and the debates not mattering.

I’ll grant you that political scientists do not always influence political events, but they do often enough, especially when their research is used by pundits or government officials, to warrant responsible considerations regarding how the data and conclusions are being used. As two examples of many that could be cited consider Arthur Schlesinger’s comment in _The Imperial Presidency_ conceding that “historians and political scientists, this writer among them, contributed to the rise of the presidential mystique” (Schlesinger 1973, ix). Similarly, Grodzins’ and Elazars’ claims of cooperative federalism was an exaggeration but helped justify a more active federal government. Responsible science and responsible citizenship, I think, requires recognizing when political scientists’ writings may influence events. As professionals, I believe political scientists have a responsibility to try to correct the partisan use and policy misuse of political science’s findings.

Do I think the statements above are partisan efforts to influence public perceptions? Consider, if Mitt Romney were leading in the polls, would the above claims have been written the same way?

I think responsible science requires that when political scientists’ research does make it to such audiences, political scientists have a responsibility to ensure the science is correctly interpreted.

John Sides October 1, 2012 at 1:44 pm

Troy: How about this? “Debates have created shifts in the polls, but they rarely have decided the winner of the election.” Past tense, describes the historical pattern, doesn’t predict the future. Which is what the journalists I quoted above were describing. They are correct.

But what the polisci can further tell us is *why* the historical pattern is true, which is what I elaborated on in my piece on debates. And that underlying logic can inform predictions about the overall effect of this election’s debates.

So, no, I don’t think anyone is being irresponsible here, nor do I think the scholarly evidence is being misused. And while I can’t speak for Ifill, Cillizza, and company, I would certainly say the same thing if Obama, not Romney, were trailing in the polls by 4 points.

Troy Smith October 1, 2012 at 4:20 pm

John: I enjoyed your article. As a corrective to the claim “debates matter most of the time,” your article is insightful and valuable. I accept your conclusion. I will note the “rarely” in it and enough wiggle room in your analysis to allow that we cannot conclusively deny that some debates may have influenced voter behavior (beyond which your measuring tools could identify) and future debates might make a difference in electoral outcomes.

A question I am trying to focus attention on is what do/should the audience take away from the article(s). Speaking only to political scientists who understand the limitations of your research, your emphasis is OK. When such research goes to the public, however, it will, in my opinion, lead many readers to jump to the conclusion “don’t pay attention to the debates, because they don’t matter.” Which is why I think it would be wise to give greater acknowledgement to the research’s limitations.

If we consider what is the “responsible” message to send the public, recognizing that the “message” may shape public behavior, would it be: (1) the debates rarely matter in electoral outcomes, so there is no need to pay attention; or (2) the debates’ influence is less than we once thought, although they still have the potential to influence the electoral outcome; or (3) the debates are an important element of civic responsibility for members of a democratic society; they matter for civic awareness and responsible civic behavior, so, please, pay attention to them.

I do not think you are intentionally implying #1, although I do think that is the conclusion many will reach from the above stories. I find #2 more thoughtful and civicly responsible. I think I could make a good case that #3 is something that should be taught and emphasized in political scientists’ professional engagements with those who are not political scientists.

John Sides October 1, 2012 at 4:37 pm

Troy: I don’t disagree with that at all. It’s true that a horse-race perspective on the debates is just one perspective, and there are civic goods that can come from debates even if they do not alter the dynamics of the race. In my piece, I noted, for example, that people do learn things from debates — with the research of Tom Holbrook and others in mind. And so I would certainly urge people to watch them, even as I would urge reporters not to exaggerate their role in determining elections.

Comments on this entry are closed.

Previous post:

Next post: