For awhile now I’ve been trying to get more journalists to acknowledge that the economy is a powerful driver of political outcomes. So it was interesting to see David Paul Kuhn’s piece at Real Clear Politics Tuesdayy, which opens by declaring, “It’s an old myth that economics determine elections,” and takes aim at the same Paul Krugman column I praised on Monday, as well as a recent piece by Ezra Klein. Kuhn writes:
Klein heavily relied on a graph published along with his story. It illustrated the correlation between income growth and the vote. But one should always look at the outliers. Two jumped out on this graph: 1952 and 1968. Both are classic examples of good economies not saving the party that controlled the White House.
The big factors of the 1952 race: Korean War, a deeply unpopular incumbent and Dwight Eisenhower’s towering personality. Had Ike run as a Democrat, once thought possible, Democrats might have had their sixth consecutive term.
The good economy proved of little weight in the late sixties as well. In the 1966 midterm, Democrats were sunk by the unpopularity of Lyndon Johnson, his Great Society and social issues. Two years later, those social issues (race, urban upheaval) and the Vietnam War particularly took to the fore of the American mind.
Income growth correlates to election outcomes. But the correlation is weaker than Klein contends. It’s also far from determinative.
Real income growth, per capita, was actually larger in the 2000 election year than in 1972. If income growth is destiny, 2000 should have been the landslide. Income growth was significantly larger in 1988 than the greater blowout of 1980. This correlation is even weaker in midterm years.
Kuhn spoke to the right people for this piece—folks like John Sides and Alan Abramowitz, who have informed, directly and indirectly, much of the writing on this topic at Campaign Desk. And he’s right, strictly speaking, that “it’s not always the economy” that drives election outcomes. If you’re trying to predict the midterms, for example, you probably want to look at the generic ballot preference and, to a lesser extent, the president’s approval ratings—both of which are influenced by the state of the economy, but also reflect other inputs.
More broadly, non-economic fundamentals like incumbency and exposure (the number of seats each party has to defend) matter. Campaigns, as Kuhn writes, “matter on the margins”—though some campaign events (conventions) matter more than others (debates). War and other national security threats obviously matter. Ideology matters, in a sense (though it’s largely fixed, and thus mostly unhelpful in explaining short- and medium-term changes). And when new issues emerge that disrupt longstanding political coalitions—well, that matters, too.
So insofar as Kuhn’s piece can correct overbroad generalizations or economic determinism, it’s valuable. As Jonathan Bernstein writes, “If Ezra Klein oversimplified things in his piece, or if there’s more debate than he let on, by all means it’s a good thing to take issue with it.” And as Sides says, in a response post: “Kuhn is certainly right that we should talk about the economy and elections in probabilistic terms like ‘is associated with’ or ‘is strongly related to.’”
But Sides is also right when he says that Kuhn’s reading of Krugman and Klein “is unnecessarily uncharitable.” I wouldn’t argue, as I think Sides does, that for strategic reasons they should feel free to overstate the case; columnists and bloggers (present company included) should be precise with their prose and their analysis, and should be held to account when they aren’t. (What else is the Internet for, anyway?)
But Krugman and Klein shouldn’t be judged only against the Platonic ideal of political commentary (about which there’s of course going to be disagreement), but against political commentary as it actually exists. And it would be hard to claim that their recent writings don’t do far more to advance readers’ understanding than the bulk of political commentary as it actually exists—in which the role of the economy is often absent, and in which things that really don’t matter much, like the latest message war or manufactured scandal, get pride of place. Even if they haven’t got it quite right, they have got us quite a bit closer, and they deserve credit for that.
A final thought: these debates over what “matters” in politics often center around election outcomes. There are good reasons for that: election outcomes are obviously important, and they’re also relatively easy to study. But as we get tied up in discussions about the relative weight of the economy versus presidential approval versus incumbency, etc., etc., in forecasting models, it’s easy to forget that election outcomes aren’t the only thing that matters in politics.