The Washington PostDemocracy Dies in Darkness

Economists aren’t politically driven, report economists

Bush CEA chair Greg Mankiw, left,  and Obama CEA chair Christina Romer. (Dominick Reuter for MIT)

People — journalists in particular — have a tendency to label economists as liberal, or conservative, or libertarian, much as they would elected officials or political pundits. In some limited sense, this is fair enough. Economists are humans with political opinions just like anybody else. But people typically mean something more than this — that economists' actual research is indicative of political values they hold.

Some economists are pretty upfront about this. In his article "The Politics of Political Economists," the future Nobel laureate George Stigler argued* that economists tended toward conservatism because the findings of their research straightforwardly supported conservative policy prescriptions. On the other hand, radical economists on the left have disputed that the idea of economics as an objective science even makes sense, arguing that one's research will inevitability reflect one's values, and there's no use trying to avoid it. "Opposing truth and ideology is in my view a methodological error," the Marxist economist Stephen Marglin once wrote. "What is ideology, after all, but the unproved assumptions, beliefs, and values that must underlie any intellectual inquiry, or for that matter, any form of contemplation or action?"

But hold up — do economists actually disagree about major issues in the discipline, and do those disagreements map along ideological lines? A new white paper by the University of California San Diego's Roger Gordon and Gordon Dahl takes a stab at answering that. As their data, they use a set of surveys conducted by the University of Chicago's Booth School of Business, in which dozens of economists are asked their view on everything from trade policy to the debt ceiling to capital gains taxes.

Perhaps surprisingly, Gordon and Dahl find widespread consensus among the respondents. Out of 80 analyzed questions, 32 showed absolute consensus: All responses were either uncertain or all on one side. For example, the question "Will the Fed’s new policies in 2011 increase GDP growth by at least 1% in 2012?" received only responses of "uncertain," "disagree" and "strongly disagree." A question where all responses were uncertain, agree or strongly agree would similar count as an area of consensus. Even in the other 48 questions, where there was some actual disagreement, it tended to be lopsided. All told, Gordon and Dahl found that just 6 percent of responses went against prevailing opinion.

So what explains this disagreement? Funnily enough, not party. Pairs of economists who served in Washington for different parties were no more likely to disagree than a random pair. Indeed, just about nothing — from the school where the economists got their PhDs, the departments they're part of, and their age — has a major, significant effect on the level of disagreement.

The one thing that does have a big effect is the size of the research literature. When an area is new and relatively understudied, there are many more "uncertain" responses as well as responses that dissent from the prevailing view. A question on fracking, a fairly recent technological development in extracting natural gas without much economic literature around it, provokes more disagreement than any, for example. In this way, economics works like any other science. The more an area is studied, the fewer people are uncertain and the more convergence there is around a single explanation.

The other area where Gordon and Dahl found effects was on confidence levels. They found that people who'd worked in Washington were significantly less likely to say they were uncertain about an issue. Women, and people without policy experience, were more reticent, perhaps reflecting industry pressures against self-confident women (or in favor of self-confident men). And some schools, in particular Chicago and MIT, were full of know-it-alls:

Respondents who got their degrees at Chicago are far more confident than the other respondents, with almost as strong an effect for respondents with Ph.D.’s from MIT and to a lesser extent from Harvard. Respondents now employed at Yale and to a lesser degree Princeton, MIT, and Stanford seem to be more confident.

One shouldn't read too much into this. The Booth surveys, while surveying both apparent liberals and apparent conservatives, omit people on the true fringes of both sides. There aren't any Marxists or post-Keynesians on the left, and there aren't any Austrians or ultra-libertarians on the right. The survey's left flank is Emmanuel Saez (who supports very high taxes on the rich but is basically a social democrat) and its right flank is Robert Hall (who invented the flat tax, which, while radical in its way, has a lot of real supporters in American politics). But the study does confirm that the economics profession has a coherent mainstream consensus, with many shared beliefs about the subject, a consensus that holds across party and ideological lines.

Update: It turns out that University of Michigan economist and friend-of-Wonkblog Justin Wolfers has a smart response to this paper, arguing that a re-analysis suggests an ideological element to economists' views. Do check it out.

* No non-paywalled version seems to be available, sadly.