Using econometric models to predict recessions
Since the focus of this article is on how,
and how well, econometric models forecast
recessions, I begin by defining a recession. A
statistical model requires a precise definition,
and, as we will see, different definitions lead to
different econometric approaches. After I describe
these econometric methods, I evaluate
the recent forecasting performance of one
econometric method: the NBER's Experimental
Recession Index, which I developed jointly
with James Stock of Harvard University. It
turns out that this index did not perform well
over the current recession. Unraveling the
reasons for its poor performance says much
about the causes of the current recession.