Campaigns and Counterfactuals

Carlisle Rainey discusses a potential reason political scientists and political reporters have different views of campaign effects: they use different underlying counterfactuals, in two senses:

First, political scientists tend to discuss the effects of small changes in campaigns, while journalists tend to imagine big changes. Second, political scientists construct counterfactuals in which campaigns are responding to each other and cancelling out, while journalists tend to hold one campaign constant and vary the other.

The first means that political scientists imagine a world in which, say, a candidate did not commit a gaffe or air a particular ad, but journalists imagine a world in which that candidate did not campaign at all.  The latter counterfactual leads journalists to infer big effects but the former leads political scientists to infer small effects.

I disagree with this characterization, because I don’t think it accurately represents the thinking of journalists.  I think journalists do have an implied counterfactual that is similar to that of political scientists.  Journalists are interested in measuring the effects of specific new events, new messages, and so on because that is what they are writing about everyday.  It’s not at all newsworthy to muse on what the world would be like if Mitt Romney, the RNC, and GOP-affiliated super-PACs and 501c’s raised no money. It is newsworthy to speculate on what the attacks on Romney’s time at Bain Capital will mean in November.

Similarly, I think journalists are highly attuned to the ways in which campaigns respond to each other.  Indeed, the back-and-forth between campaigns—Obama attacks Romney on Bain, news outlets report on whether Romney really retired in 1999, Romney surrogates go on Sunday shows and talk about “retroactive” retirement, Romney surrogates bring up Teresa Heinz Kerry’s tax returns, etc., etc. —is also what reporters cover everyday.

The difference between many political scientists and many journalists—I say “many” to mitigate the tendency to over-generalize in these sorts of posts—is not the counterfactual, but everyone’s priors about the causal effect implied by the counterfactual.  I think the prior of many political scientists is that many discrete campaign events or activities don’t have a net effect on the outcome—because the events are not well-known to persuadable voters, because the events are not persuasive for whatever reason, because events “cancel” out as Rainey suggests, and so on.  I think the prior of many journalists is different.  Most wouldn’t assume implausibly large effects of events in the average presidential campaign; few would say, I imagine, that Romney’s record at Bain threatens to turn the election into a 1964-style landslide.  But I think they do imagine that such events could make the difference between winning and losing, whereas political scientists are more skeptical or at least more cautious.  And I think journalists imagine that events cancel out less frequently than political scientists believe because journalists are constantly judging whether one candidate’s message is better than the others, or whether one candidate is winning the news cycle but the other isn’t—even though persuadable voters may not arrive at the same judgment or indeed make not make any judgment at all.

To be clear, I agree with Rainey that political scientists could be clearer about the counterfactuals underlying our claims about campaign effects, but I am not sure that those counterfactuals distinguish us from political reporters all that much.

[Cross-posted at The Monkey Cage]

John Sides

John Sides is an associate professor of political science at George Washington University.