A Few More Thoughts About Keynes's Victory

-- Posted by Neil H. Buchanan

In my post yesterday, I discussed the views of the winners of this year's faux-Nobel in Economics, Christopher Sims and Thomas Sargent. In particular, I pointed out that the claims of some on the right that Sargent's selection is a slap at Keynesians are simply wrong. Both Sims and Sargent are aggressively apolitical, but even so, at least Sims has said directly that Keynesians are doing the most important work in economics these days. He also approved of further efforts to stimulate the economy through fiscal policy.

The econ nerd in me then made a further argument, which is that Sargent's description of a theory called "rational expectations" was a (perhaps unintentional) endorsement of Keynesianism. Sargent described that theory as merely the assertion that people's expectations (about prices in particular) matter in determining economic outcomes. In other words, the theory of rational expectations is now merely a theory of expectations.

This would not be a problem for the anti-Keynesians if Keynesians had actually built their models on the assumption that people's expectations have no impact on economic outcomes. The fact is, however, that Keynesians have long taken expectations seriously. The only way to get a model that includes expectations in it not to have Keynesian results (most importantly, the prediction that active government policy can improve the economy) is if one assumes that expectations are formed fully rationally, in the very specific and peculiar sense in which rational expectations theorists define that term. (To describe it from a different perspective, the assumption that expectations are formed rationally means that markets are never out of equilibrium, based on the information that is available at any given moment.)

One interesting aspect of this observation is its analog in American politics. In economics, anti-Keynesian models are built on the idea that expectations are formed 100% rationally (by their definition), and then they say that Keynesians build their models on the idea that expectations are formed 0% rationally (i.e., that expectations do not matter). The reality is that Keynesians say that expectations do matter, that people might be partly "rational" in the stylized sense, but that anything less than 100% rationality still leads to Keynesian results.

In American politics, the right (especially the currently ascendant version of the right) takes the attitude that government is always, completely, dangerously, 100% bad and wrong. They then accuse people in the center and on the left of believing that government is 0% wrong. The reality, again, is that the center and the left are fully aware that government is not perfect, and they are looking for ways to improve social outcomes. If a government program can help, and if we can set it up with proper safeguards, then it should be implemented. If a public/private partnership is better, then that is what we should adopt. If tax incentives will do the trick, great. If there is no problem in the first place, then there is no reason even to ask whether there is a role for government.

This is another reason that claims of equivalence between left and right strike me as so absurd. When someone says that, say, Paul Krugman simply has a bias for bigger government that he builds into his analysis, while right-wing economists do the equivalent to support their bias for smaller government, that is simply wrong. Neither Krugman nor anyone I know who counts as non-right has a baseline preference regarding the size of government. The right does. Only pure sophistry can turn "I don't think government is always wrong" into the equivalent of an absolutist statement.

A second issue relating to yesterday's post was raised by the commenter Doug, who pointed out that I had adopted a definition of "helping" the economy that was inappropriately narrow. That is, even if the rational expectations theorists were right that government policy cannot make the economy grow faster, or push the unemployment rate below its "natural" level (as if such a thing exists), there are still plenty of things that government policy can do to make life better, that is, to "help." I completely agree. This is yet another example of the danger of arguing on the other side's terms. I did not mean to concede anything, but only to describe the anti-Keynesian case in its own narrow terms.

Similarly, I noted yesterday the anti-Keynesian claim that austerity can be expansionary. The interesting thing is that this argument, too, is not supported merely by claiming that stimulus is not expansionary. That is, even if stimulus were proven to be non-stimulative, that would not prove that austerity is stimulative. Under the rational expectations theory (or any other theory, for that matter), one needs additional assumptions to turn austerity policies into expansionary policies. Those assumptions, in turn, must withstand empirical scrutiny, and they do not.

Finally, I should note that one can easily believe that the government can do some goods things for the economy in the long run, even if one thinks that it cannot change short-term macro outcomes. To their credit, some prominent anti-Keynesians have argued over the years that the key to long-term prosperity is increased support of education. That is, if the government wants more prosperity, it can raise people's productivity so that everyone's standard of living can rise. This need not be a Keynesian nor an anti-Keynesian point.

Unfortunately, the new move among economists and politicians on the right has been to denigrate government's role across the board. So, the same people who once might have said that the government should not try to "fine-tune" the economy in the short run, but should instead focus on affirmative strategies for long-term growth (like education) are now arguing that teachers are overpaid and underworked, that education (including college) is not worth it (individually or collectively), and that we should simply give up on trying to change the long-term outcome of the economy (and, instead, do unrelated things like shutting down the EPA). That is a coherent argument, in some sense -- the empirical assertion that government always makes things worse -- but it is both defeatist and unsupported by the evidence.

Again, however, I am not saying that government always makes things better. We do not live in a world of absolutes.