Congratulations for Not Forecasting Recessions

Mark Thoma at Economist’s View cites an article by Daniel Gross about the inability of economic forecasters to forecast recessions. A maintained assumption of everyone who discusses this seems to be that this failure of economic forecasting reflects badly on economists if reflects on them at all. I would suggest, however, that the inability to forecast recessions in practice is, when properly understood, actually a point in economists’ favor.

Suppose that we can represent the consensus forecast in one number that indicates the consensus probability of recession. A consensus recession forecast would then refer to a situation is which this number exceeds 50%. But we know that the consensus usually changes slowly, so a 51% recession forecast is likely to have been preceded, with some lag, by a 40% forecast, and that by a 35% forecast. But what happens when the consensus calls for a 35% or 40% chance of recession? What happens is that policymakers – usually at the Fed – start aggressively trying to prevent a recession. (That is if they didn’t already start when the consensus was at 30% or lower.) If they do a good job, the consensus never gets to 50%. So if the Fed is doing a good job, it’s unlikely that one will ever see a consensus recession forecast, successful or otherwise.

And the Fed’s ability to do a good job of preventing recessions depends in part on its economists – as well as on those various economists over the years that have studied the implementation of monetary policy and helped the Fed learn how to use it effectively. If it were the case that forecasters could successfully forecast recessions, it would indicate that these economists had failed in their task.