Re: The Inflation/Unemployment Equation That Failed But Keeps On Winning

by Patrick Brennan

Regarding criticism of the Fed’s implicit belief in the Phillips Curve, a few things: Clearly, the experience of the 1970s did call into serious question the way Keynesians had used and understood the Phillips Curve (which posits a trade-off between inflation and unemployment).

But it’s odd to argue that today’s experience does too, as Tim and his reader claim. In Tim’s piece last night, he points out that there has been inflation since, say, 2006, and thus people earning the same nominal wages are indeed seeing them eroded. But rates of inflation and inflation expectations have been low in historical and absolute terms, and are nothing like what we saw in the 1970s, so it’s not clear how today’s experience is supposed to be another nail in the coffin of the Phillips Curve.

Tim also scorns the idea that the long-term inflation–unemployment relationship has been replaced with a short-term one. Recognizing the Phillips Curve only works in the short run is, roughly, what Milton Friedman and Edmund Phelps introduced in the late 1960s — to their great credit, before the empirical case became crystal clear.

When it did, Keynesians conceded the point and adjusted their view of inflation and unemployment too, though they ended up in a somewhat different place than Friedman and Phelps. But both views roughly agree that there is some kind of interaction between inflation, inflation expectations, and employment levels that is important in the short term, but not the long term. Tim suggests this is logically impossible: “To believe this you’d have to believe modeling tools that are not precise on a large scale somehow become more precise at the small.”

That’s begging the question, especially given that economics (and the contemporary understanding of the Phillips Curve) involves expectations — which are more important in the short run than the long.