What Went Wrong in Macro - Overview
Paul Krugman and I agree that something went very wrong in macroeconomics. This post (and a separate one that gives more of the historical details) continues a discussion (see here and here) about how things went so far off track and what we should learn from this episode.
My claim is that the problems cannot be attributed solely to obstinacy on the part of Lucas and his supporters. They were making two points: (i) The existing multi-equation Keynesian models were fatally flawed. (ii) It is possible to formulate a rigorous model of the labor market that generated a version of what we would now call an expectations augmented Phillips curve. In 1978, they would still have been open to extensions to the preliminary model that Lucas had proposed to make point (ii). Unfortunately, by this time, both of their points were being treated with scorn, and this did lead to obstinacy.
Their open-mindedness can be inferred from a paper that Lucas and Sargent presented in the summer of 1978. In it, they describe a range of extensions or alternatives to Lucas’s model of an expectations augmented Phillips curve. They acknowledge the value of the sticky wage, long-term contract models that Stan Fischer and John Taylor were pursuing, although they did note that because the terms of the contract would change after a policy induced shift to a high inflation rate, these models were unlikely to recover a traditional Phillips curve. They endorse search models of the labor market. They support extensions of Lucas’s initial model of imperfect information, some of which could have taken the form of recent models of sticky information. So in 1978, there was still room for the rest of the profession to engage seriously in an evaluation of the issues that Lucas and his supporters were raising and for a broadly shared consensus to emerge.
Yet by 1978, from some vocal corners of the profession, the response had already taken on a tone that ranged between dismissive and derisive. As I describe in the more detailed post, I observed this type of response from Robert Solow when I was a student at MIT.
No doubt, some of the hostility was a defensive reaction by people who had devoted their careers to building large macro models. For them, Lucas and Sargent were the bearers of bad news. Support for this style of work was about to collapse.
For Lucas and Sargent what seems to have rankled most was a response which took existing policy as given and judged they by asking whether it implied support for this type of theory, without considering the theory on its own merits. From people who cared deeply about good policy, a better response would have been to take the new research agenda seriously, see what policies better models actually supported, and trust that if their policy beliefs were correct, the science would support it.
Within a few years, Lucas and his supporters seem to have concluded that the right way to respond to dismissive derision was to return fire indiscriminately. They seem to have stopped engaging with any macroeconomists outside of their well defined circle of supporters, even though some of these outsiders were taking them seriously. The ensuing self-imposed isolation let real business cycle models take root and persist for far too long. Their group-think allowed a coordinated move away from the use of data to evaluate or test a theory, and toward the use of calibration instead. Loyalty seems to have precluded any internal criticism as theory became opaque and misleading.
I am not saying that the wrong turn in macro was “Solow’s fault.” But neither do I agree that it was “Lucas’s fault.” In fact, I do not think that the wrong turn can be understood as the result of a character flaw in any of the economists who were involved.
Nor am I saying that the dismissive attacks that I associate with Solow–attacks that failed to engage on the underlying scientific issues that Lucas and his supporters were raising–justify the behavior that emerged later from the isolation and group-think of the rebel camp. On the contrary, I believe that that they should be held to account for losing their commitment to Feynman integrity which is one of the cornerstones of science.
The general conclusion I draw is that all economists would be better off if we trust science. Solow, and Lucas both seem to have convinced themselves that to make progress on the path each was pursuing, they needed to take a shortcut that departed from the norms of science.
In retrospect, in each case, the shortcut backfired. Dismissive criticism ended up fueling the worst type of new classical macro, which did interfere with good policy. Isolation undermined the commitment to science that was central to Lucas’s work when he started out. So it might be possible to get everyone to agree that it is better to commit to the norms of science.
But because the social dynamic of science is unstable, this commitment requires active support. A little acrimony or a bit of politics (and certainly the two together) are all it takes to for science to fall apart.
The people who achieve prominence in science are generally reasonable people with good intensions, so all it might take to keep things on track might be a few small corrections–a reminder that to criticize something, you have to take it seriously; or that anger from a nursed grievance takes away IQ points.
But the response has to come quickly. Over time, grievances amplify; positions harden. When they do, our only recourse is to count on Bohr’s observation that sometimes, science progresses funeral by funeral.