What Would an Obama Victory Mean for Political Science?

Noam Scheiber tweets:

This prompted replies from me, “Dave McW,” Brendan Nyhan, and Michael Wagner.  Here’s the thread.  I want to make 2 points:

First, Scheiber is poking at the political science conventional wisdom, which privileges “structural forces like the economy” (his phrase from a later tweet in the thread).  But an Obama victory, as Brendan quickly pointed out, would confirm—or at least not contradict—the influence of structural forces.  I don’t know of a forecasting model that relies only on the economy.  However, if you take Nate Silver’s analysis of various economic indicators, they hardly point to a certain Obama defeat.  Silver’s economic index suggests that as of right now, the economy is better than it was on Election Day in 1980 or 1992, and most similar to 2004—i.e., an election in which a first-term incumbent president who occupied the White House after 2 terms of opposition control won the election amidst a middling economy.  This amplifies a point I’ve made before: it’s important to accurately calibrate expectations about the election outcome to the underlying structural conditions.  I think Scheiber underestimates Obama’s current position vis-a-vis those conditions, which then leads him to assert that an Obama victory must derive from “personality and advertising.”

Second, I’ve been pretty clear that political scientists are not economic determinists who believe that, even in presidential campaigns, things like personality and advertising “don’t matter.”  There is an important difference between Scheiber’s categorical statement and subtler statements that are more in line with the research.  From an earlier post, here is my attempt at a careful statement:

Presidential elections are the elections where we would expect the campaign to persuade relatively few voters.  People tend to have stronger opinions about presidential candidates than candidates at other levels of office.  And, in the age of public financing, at least until 2008, opposing presidential candidates have had roughly equal amounts of money to spend in the general election. Campaigns matter more when the candidates are unknown and when opposing candidates have asymmetric resources so that one can out-spend or out-campaign the other.  This is why campaign events and spending tend to matter more in presidential primaries and congressional elections than in presidential general elections.

Because people continually overestimate the effect of campaigns, this blog holds up the other end of the dialectic by emphasizing the economy and defending those who do.  But plenty of research has identified the effects of campaigns too—research discussed on this blog, e.g., here, here, here—and so it’s time to abandon this whole it’s-either-the-economy-or-the-campaign dichotomy, which seems implicit in M.S.’s portrayal of my position.

And we can do a lot better than just saying “well, it’s both” or saying “campaigns might matter at the margins.”  We have a much better sense of (1) when campaigns matter, as discussed above, and (2) how much they matter.  To give you an example of “how much”: in counties where Obama aired 1,000 more ads than McCain, he received about 0.5% more of the vote than John Kerry did in those same counties (from this article (pdf) by Michael Franz and Travis Ridout).

I’ve certainly tried to knock down categorical statements that things like personality are the key to victory, but I’ve also tried to talk in nuanced ways about when and why campaign advertising might matter. I’ve also talked about how we cannot easily separate the effects of the economy from the effects of the campaign.

In other words, I’ve tried to push against a simple dichotomy between “structure” and the campaign, which seems to be what Scheiber is assuming.  I’m glad that people like Scheiber pay attention to political science, but his thinking seems to reflect more of a caricature than the actual political science literature.

7 Responses to What Would an Obama Victory Mean for Political Science?

  1. JA July 9, 2012 at 3:54 pm #

    “I don’t know of a forecasting model that relies only on the economy.”

    With respect, Professor, isn’t this another case of you trying to have it both ways? You mention Ray Fair’s model quite often, which uses several GDP indicators along with a single incumbency dummy variable.

    While I suppose it’s literally true that Fair doesn’t rely “only” on the economy, it’s can’t be correct to argue that Fair isn’t substantively an economic model. Especially compared to models like Abramowitz, where political indicators like job approval are key components.

    The difference is important, because politically-weighted models like Abramowitz give Obama a much better chance to win than economically-weighted ones like Fair or Hibbs.

    • John Sides July 9, 2012 at 7:52 pm #

      JA: I have mentioned Ray Fair’s model about 5 times in 4+ years of blogging — you can look it up — and always in the context of other forecasting models. It’s not a model that I rely on in any sense so I don’t think I’m trying to have it both ways.

      It’s true that Fair and Hibbs’s models suggest a closer election, but Fair’s model does predict an Obama victory — contra Scheiber’s apparent belief that the economy makes Obama the underdog. Hibbs’s model does predict an Obama loss because it relies on a single economic indicator, change in real disposable income per capita, that is one of the most unfavorable at the moment.

      The broader question is whether “structure” should be construed only to mean “the economy” or whether it should take into account features like (a) incumbency or (b) the length of time a party has controlled the White House — both factors that favor Obama. I don’t know what Scheiber would say, but even if we ignored these factors and focused only on the economy, it’s not at all clear that Obama is the definitive underdog and thus his lead in the polls must derive from Romney’s personality or Obama’s advertising in swing states, etc., etc.

      • Nadia Hassan July 9, 2012 at 11:14 pm #

        I actually think that, and I’ve looked at this for a while, political science findings are *better* suited to explaining Obama’s lead than personality and advertising.

        Solely asserting the power of structural forces and downplaying personality, gaffes, ads, and so forth is not really sufficient to explain what we see, and optimally foreshadow the outcome. ‘Watch the economy’ is useful, but is 8% unemployment a certain defeat? What is the difference between say 1.5% and 2% GDP growth? How does having a 45% versus 50% approval rating matter? All mean a competetive race, which is consequential, but the challenges for the candidates do covary with them in qualitatively meaningful ways.

        Time for A Change, which did quite well in the 538 review a few months ago sees a tight race with an Obama advantage and polls concur. If Obama was ahead 1.5-3 points, whereas his approval was in the low 40s, growth was sub 2%, more bad jobs reports, and so forth, then it would be time to consider things like personality and advertising (though, I suspect the GOP agenda would be a more likely culprit for a situation of this sort).

  2. Chris Lawrence July 9, 2012 at 4:51 pm #

    JA: the obvious rejoinder to that comment is that Ray Fair is an economist, not a political scientist.

    In any event, most of the forecasting models seem to suggest Obama will win. In fact, I think they’re overly optimistic on that score; I’m not at all certain the subjective economy perceived by Americans is quite congruent with the economic figures that, for example, say the recession has been over for over three(!) years now.

  3. Nadia Hassan July 9, 2012 at 6:07 pm #

    (1) I think Lynn Vavreck’s theory will be a winner of the 2012 election.

    (2) To be fair, some like “Bread and Peace” rely primarily or solely on the economy. You also supplied Ezra with the graph and he went out and talked about how political scientists can predict the vote to within 2 points without focusing on *only* the economy. Though, it’s not entirely clear this is entirely true You talked about how you were glad to see him and others make the case cleanly and boldly. Although many of you think of it as an antidote or corrective, I wonder if it actually perpetuates the divide.

    (3) I wish press coverage was infused by the models core ideas’, findings, and especially The Message Matters’ framework. Consider how ABC News covered the latest polling in swing states.


    “It’s clear what some perceive as Romney’s passive strategy is starting to frustrate his supporters who want to see a more robust push-back.

    Recall the Wall Street Journal’s editorial board, which last week, wrote: “Team Obama is now opening up a new assault on Mr. Romney as a job outsourcer with foreign bank accounts, and if the Boston boys let that one go unanswered, they ought to be fired for malpractice.”

    But, getting into the fight that Obama wants to pick is a dangerous trap. It does Romney no good to try and defend bank accounts in the Caymans or the value of private equity.

    The real question is when will we see Romney start to define himself on his terms?

    Where is the People Magazine profile? Or the Good Housekeeping interview?”

    I think what you told me earlier this year better captures the situation. I asked, “Given various economic scenarios–is it straightforward that Republicans should run a clarifying campaign focused sharply on the economy? ”

    (I think clarifying campaign is an intuitive phrase, but if anyone needs definition: http://newsroom.ucla.edu/portal/ucla/ucla-political-scientist-discovers-98567.aspx)

    “Everything hinges on how the economy does between now and November. If growth slows or unemployment increases — perhaps because of Eurozone events — then I think the GOP should run a clarifying campaign. But if the current trajectory continues, then the GOP confronts a harder task. Could it successfully emphasize the economy even though it’s growing? Would it have to change the subject? If so, to what? The deficit? The size of government? It’s not obvious to me — mainly because Lynn’s work shows how hard it is to find such a message when the economy benefits the incumbent party. To me, the size of government would be a possible message, but it really depends on whether a majority of voters sides with the GOP over Obama on this issue. Obviously, Obama seems to feel he can campaign successfully by defending a role for government. And many voters support some government programs. Ultimately, I don’t know whether this message would work for Republicans.”

    There might be situations where campaigns can embarass political scientists, or surmount challenging structural factors.

    Check out one forecaster’s post-mortem in 2000: ” All that forecasting models can
    offer well in advance of an election is some sense of the advantage or disadvantage candidates have as the campaign begins. They offer a starting point of sorts,
    not the final resting place. Campaigns matter. They always have and they always will. If this wasn’t already clear before the 2000 election, it is now.”

    But, it’s not a function of the kinds of tactics that pundits often imagine.

    And so far, the evidence suggests that 2012 is not 2000. Mixed economic growth and approval ratings in the 40s are consistent with a tight race with a small Obama lead, and that’s what the polling so far shows.

  4. Nadia Hassan July 9, 2012 at 6:20 pm #

    Chris, the forecasting models are mixed.


    There’s a post on the AEI blog that plugs in four models, and Romney wins the 2-party vote in three of them.

    Income growth has remained totally anemic. Check out this report on focus groups for more:

  5. paul gronke September 18, 2012 at 5:04 pm #

    First, Hibbs’s model is called bread and PEACE. Peace is not the economy.

    Second, campaigns “matter” in so far as they focus on the fundamentals, i.e. they seldom have an independent effect above and beyond. This is one of those counter factual puzzles that John blogged about earlier but the insight is pretty intuitive: in what world would an incumbent NOT talk about a well performing economy?

    Ipso facto, the campaign and advertising don’t “matter” because you can’t find an independent contribution, but that doesn’t mean that if you could somehow magically turn off the advertising, it would not hurt the incumbent.

    Finally, even Hibbs admits

    The only postwar presidential election results not well explained by the Bread and
    Peace model are 1996 and 2000. In 1996 the vote received by the incumbent Democrat
    Clinton was 4% higher than expected from political-economic fundamentals, whereas in
    2000 the vote for the incumbent Democratic Party candidate Gore was 4.5% less than
    expected from fundamentals. I am tempted to argue that idiosyncratic influence of
    candidate personalities took especially strong form in those elections, with the ever
    charming Bill Clinton looking especially attractive when pitted against the darkly
    foreboding Bob Dole in 1996, and the unfailingly wooden Al Gore paling by comparison to
    an affable George W. Bush in 2000. Alas, this line of reasoning is entirely ad hoc and
    without scientific merit.

    Personally, I disagree with him on the last point. There is nothing unscientific about trying to explore the reasons for large outliers. There remains an idiosyncratic element to campaigns, it’s just not that large or that important.