Rational expectations isn't the problem; it's the solution.
By Scott Sumner
Edward Luce has a very misleading piece in the Financial Times. Here’s an excerpt:
Why do economists continue to get it so wrong? One answer is that not all of them do. David Blanchflower, who was on the Bank of England’s monetary policy committee during the 2008 crash, insists that evidence of an impending crash was hidden in plain view long before it happened. Blanchflower, whose book Not Working: Where Have all the Good Jobs Gone? is a stinging rebuke to his profession, was consistently outvoted eight to one on the MPC, which sets UK interest rates. Unlike his colleagues, who were using models based on a 1970s-style economy, Blanchflower went out and talked to people. He calls this “the economics of walking about”.
His peers, meanwhile, were relying on “largely untested theoretical models that amounted to little more than mathematical mind games”. Paul Romer, the former World Bank chief economist, calls this “mathiness” — playing with regression to give a false sense of precision. Others might call it alchemy. Lucas, whose Chicago School housed the high priesthood of mathiness, won a Nobel Prize for his rational expectations theory. It demonstrated that the market was always right.
Central bankers do tend to rely on 1970s-style Keynesian models of the economy. But almost everything else is wrong. Paul Romer was criticizing an entirely different set of models, often called “new classical” or “real business cycle” models. These models certainly have flaws, and indeed they may be as useless as Romer claims. But policymakers do not rely upon them. Robert Lucas‘s rational expectations work can be viewed as a critique of 1970s-style Keynesian economics. And by the way, rational expectations most certainly does not suggest that “the market is always right”. The financial system is saturated with moral hazard, a phenomenon long understood and critiqued by “mathiness” economists.
It’s also not clear how “playing with regression to give a false sense of precision” fits in here. That is certainly a problem in economics, but it has nothing to do with the other issues discussed in these two paragraphs. Nor does it have anything to do with the recent policy failures leading up to the 2008 crisis.
In my view, almost all of the top economics programs have too much “mathiness”, not just the University of Chicago. BTW, when I was in grad school, MIT was far more mathematical than Chicago, although that may no longer be true. I went to Chicago precisely because in the 1970s it was more focused on the real world than most other programs.
Most importantly, rational expectations theory is not the problem; it’s the solution. The recession of 2008 occurred precisely because monetary policymakers at the Fed ignored market signals and relied on 1970s-style Keynesian models. The Fed focused on backward looking indicators showing high inflation in 2008, and ignored market signals that growth was slowing dangerously. It was like driving a car while looking in the rear view mirror.
The Fed is now beginning to take market signals more seriously, and that is one reason why monetary policymakers decided to cut rates this year. Older Phillips curve models would have called for a rate increase, and if we had followed them we would already be in recession.