What have microfoundations ever done for us?

For those who do not know the reference, which I think is apposite, see here.

In my previous post on microfoundations I said I disagreed with Paul Krugman’s statement that “So as I see it, the whole microfoundations crusade is based on one predictive success some 35 years ago; there have been no significant payoffs since”. I should say why I disagree. Robert Waldmann has also challenged me along similar lines.

I think the two most important microfoundation led innovations in macro have been intertemporal consumption and rational expectations. I have already talked about the former in an earlier post, which focused on a UK specific puzzle in the late 1980s that I find difficult to address without an intertemporal consumption perspective. However I also think too strong an attachment to a very basic intertemporal view has blinded macro to some critical events in the last decade or so. (See John Muellbauer here, for example.) So let me focus on rational expectations. Again I could look at the UK in 1980/81 and talk about Dornbusch overshooting, but let me try and be less parochial.

Between the rapid inflation of the 1970s and the Great Recession, what events might we look to for rational expectations to help explain? It is not an easy question, because the adoption of rational expectations was not the result of some previous empirical failure. Instead it represented, as Lucas said, a consistency axiom. However between the 1970s and the Great Recession what needs explaining is why nothing very dramatic happened – the Great Moderation. In particular, why did the large rise in oil and other commodity prices around 2005 not lead to the kind of stagflation we saw in the mid-70s and early 80s?

I think an important part of the answer was implicit or explicit inflation targeting by independent central banks. That, in turn, reflected an understanding of the importance of rational expectations. If a central bank had a clear inflation objective, and established a reputation in achieving it, that would anchor expectations and reduce the impact of shocks on the macroeconomy. Just as the Friedman’s expectations augmented, accelerationist Phillips curve helped us understand what went wrong in the 1970s, so the New Keynesian Phillips curve led to better policy around the turn of the century.

Now virtually any empirical claim in macro is contestable. (Indeed, for some this is part of the attraction of the microfoundations approach!) There are other explanations of the weak response to oil price increases, although Blanchard and Gali (2007) do argue that the Great Moderation played an important role. Others might suggest that the Great Recession itself proved that the Great Moderation was an illusion. In a crude sense this does not follow. The Great Moderation was all about the stabilising role that monetary policy can play, and that should always (given Japan) have been conditional on not hitting the zero lower bound. A more challenging argument is that the Great Moderation prepared the ground for the financial crisis, but even if this is correct it does not follow that inflation targeting was not an improvement on what went before – we may just need to do better still. Indeed, if as a result of the Great Recession inflation targets are replaced by price level or nominal GDP targets, I believe rational expectations will be central in making that case.

I think macroeconomics today is much better than it was 40 years ago as a result of the microfoundations approach. I also argued in my previous post that a microfoundations purist position – that this is the only valid way to do macro – is a mistake. The interesting questions are in between. Can the microfoundations approach embrace all kinds of heterogeneity, or will such models lose their attractiveness in their complexity? Does sticking with simple, representative agent macro impart some kind of bias? Does a microfoundations approach discourage investigation of the more ‘difficult’ but more important issues? Might both these questions suggest a link between too simple a micro based view and a failure to understand what was going on before the financial crash? Are alternatives to microfoundations modelling methodologically coherent? Is empirical evidence ever going to be strong and clear enough to trump internal consistency? These are difficult and often quite subtle questions that any simplistic for and against microfoundations debate will just obscure.