Truths and Myths about the 2008 Election, Part I

by John Sides on November 5, 2008 · 10 comments

in Campaigns and elections

I’ve been nattering on about this, as have Brendan Nyhan and others. Even the Washington Post finally noticed yesterday. If you look at the forecasting models by political scientists (summarized here) and also include the models of Ray Fair and Doug Hibbs, Obama’s predicted share of the two-party vote was 53.8%. He won 53%.

What this means: first, he did no better or worse than expected. His victory is historic because he is our first African-American president. But his victory was not a surprise. Moreover, the 53% margin means that he did not squeak by, but neither did he win some landslide.

Second, beware interpretations of the election that fixate on the campaign. Journalistic accounts of this election will undoubtedly emphasize how good Obama’s campaign was and how bad McCain’s campaign was. This New York Times story—headlined “Near-Flawless Run From Start to Finish Is Credited in Victory”—gives you a taste of that kind of interpretation. (See Larry Bartels in the Post story on why journalists and politicians have a “fundamental conflict of interest” on this issue.)

But the fundamentals are arguably more important. I’ll paraphrase something I heard Tom Mann say last week. As Obama, would you rather have a bad economy and an unpopular president, or a 2-to-1 spending advantage? You’d take the bad economy and the unpopular president.

Also note that there was a great deal of stability in the polls in the last month. Jim Stimson’s graph and commentary is worth reading as a blow-by-blow narrative of what wasn’t changing. Mark Blumenthal at Pollster and Nate Silver at 538 have also diligently refused to overinterpret every wiggle in the polls.

Of course, this is not to say that the campaign didn’t matter. It may be that the campaign helped move voters in line with the outcome that the fundamentals predict; see Andy’s work here and here, as well as Tom Holbrook’s book.

It may also be that some dynamics in attitudes helped produce Obama’s victory. See this report (pdf) by Samuel Popkin and Doug Rivers, which identifies some relevant trends, such as the growing perception that Obama “understands people like me” and promotes policies that would benefit the middle class. Further analysis will tell us whether those trends were truly consequential to the outcome.

It may also be that the campaign helped mobilize some voters to go to the polls, in line with Don Green and Alan Gerber’s research on get-out-the-vote campaigns.

Ultimately, we won’t know how much the campaign mattered until much later, when we can analyze some systematic data. In the meantime, expect impressionistic renderings to dominate, but read them skeptically. I also hope that journalists will note this success of political science theories of elections and take them more seriously in the future. This piece at Inside Higher Ed is a nice start.


johnt November 5, 2008 at 11:51 am

Where did you get the final popular vote percentage from? CN, 538, and the NYT are still showing some precincts uncounted.

I agree with your main point. I just happen to have a bet on the popular vote total (with 53 as the guess!), so I’m looking for the official final number.

John Sides November 5, 2008 at 12:32 pm

John, my numbers are preliminary — whatever’s on the usual websites. I should have noted that in the post.

Mike3550 November 5, 2008 at 4:33 pm

I am sympathetic to the argument that fundamentals help decide elections and that campaigns are blown out of proportion by journalists because they are a) exciting and b) pay their bills.

That being said, I also think that the reaction against the pendulum can swing too far the other way to basically say campaigns don’t matter at all. I should note that part of perspective might be due to my perspective as a sociologist in contrast to political science. However, I think that there are two ways Obama’s campaign mattered:

1. Without the incredible organization, strategy, and campaign skills there is no way that Obama would have gotten out of the primary. Now, I know that the “campaigns don’t matter” really is talking about the two-party general election. However, by simply saying “campaigns don’t matter” I think that it covers up how spectacularly Obama played all of his cards to even end up in the general.

2. Second, the fact that Obama–as a black man–was able to pull within the margin of a usual victory speaks to the ability of his campaign skills. Again, coming from a sociological perspective, race is so incredibly salient in so many aspects of our lives as Americans, it is astounding that so many Americans were willing to put those sentiments aside and vote for a black man. I suspect that this is what you might mean by “It may be that the campaign helped move voters in line with the outcome that the fundamentals predict” — but I think that understates how amazing Obama’s accomplishment to be the first African American President really is.

James November 5, 2008 at 9:34 pm

Mike, I sympathize with your arguments, particularly #1, with which I’m in agreement.

But on point #2, if models based on the fundamentals predicted the outcome we observed yesterday, then isn’t the most plausible explanation that the models work reasonably well, and that race isn’t as salient in national elections today as we might have guessed? The alternative position, which you adopt, seems to me to assume that the general salience of race in American life translates into a political effect at the polls, and that the black candidate’s campaigning was superior, and that these two effects happen to have roughly balanced each other out.

I’m not trying to deny the magnitude of Obama’s achievement, or the role of race or good campaigning. But these results suggest to me that race may have been of lesser importance than we’d assumed, and that estimating the magnitude of any effect due to race would be particularly hard with this information, as we’re assuming that effect was roughly balanced by good campaigning.

Mike3550 November 6, 2008 at 10:47 am

James, I’m not sure that the most plausible explanation is that race doesn’t matter that much in national politics. Unfortunately, there is no counterfactual to compare this campaign to: a black candidate for President who didn’t run the campaign that Obama ran. Since I study racial influences on racial attitudes in other areas of life (mainly housing, but also racial attitudes in other arenas as well), I would just be very surprised if racial attitudes, either stated or hidden (e.g. “I would vote for Obama, but I’m not sure my best friend would…), didn’t factor into the election.

On the other hand, for the sake of our country and moving forward on race issues, I hope that your argument is the correct one!

James November 7, 2008 at 9:24 am

Mike, I don’t mean to suggest that racial attitudes in our country, which are certainly common in other aspects of life, don’t play out in voting behavior, too.

I just don’t see how the most *plausible* explanation in this case could be said to be that racial attitudes skewed the vote, and superior campaigning skewed it right back.

We have fairly good models of presidential elections, and they predicted this result–and suggest that campaigning may play a very minor role.

So your argument is that there is racism in the vote, and that it was masked by an equal, but opposite, effect that the models say shouldn’t amount to such a large factor.

I think the more *plausible* explanation, on the evidence, really should be that neither race nor campaigning played much of a role here. It’s the simpler explanation, and fits both the models and the evidence. Perhaps my willingness to believe this is because I think it’s not unreasonable that our society’s racial attitudes didn’t translate into voter behavior in this election: those who might have been reluctant to vote for a black candidate may have been voting for the Republican anyway, or may have felt that the economy trumped their feelings about the particular candidate (as the models predict).

If race did play a factor, the question becomes, how much? By your logic, we don’t know the magnitude of the effect, because it was offset by strong campaigning. This suggests to me that the effect of race could be anywhere from tiny to quite large, and we simply don’t know.

Finally, I believe we do have counterfactuals to compare this election to: not black candidates who ran weaker campaigns, but white candidates who ran strong campaigns. White candidates, regardless of their campaigning prowess, run as well as Obama did in these circumstances–not better or worse. At least according to the statistical work done by many elections specialists.

Comments on this entry are closed.

{ 4 trackbacks }

Previous post:

Next post: