"...So how do you estimate potential? There are two main methods. One looks at past levels of output and/or unemployment, and basically uses some weighted average of the past as an estimate of what’s normal and presumably appropriate. The other looks at inflation, and tries to back out the unemployment rate consistent with price stability.
And both methods break down completely under depression conditions, which is what we have right now."
"Here’s a rough illustration from US data. I show unemployment rates versus the change in the unemployment rate over the next three years (three years because there is some tendency of things to get worse in recessions before they get better).
You can see that while unemployment rates jump around — hey, stuff happens — there has historically been a tendency for high rates to be followed by rapid declines; V-shaped recoveries. But not this time — because the Fed found itself pushing on a string. And a Hodrick-Prescott filter, or whatever, would interpret this as a rise in the natural rate of unemployment. Back in my 1998 Japan paper, I pointed out that the methods then being used to estimate output gaps would have concluded that 1930s America was back at full employment by 1935; it’s the same thing now."
"What about inflation? Again, historically inflation has tended to rise when unemployment is low, fall when it’s high. Here’s the US unemployment rate versus the change in core PCE inflation over the next year:
But right now we have high unemployment combined with more or less stable core inflation. Typical models would interpret this as a sharp rise in the natural rate, from maybe 5.5 to 8 percent. But what it almost surely reflects instead is the stickiness of inflation at low levels; the long-run Phillips curve is
not vertical thanks mainly to downward nominal wage rigidity,and that reality is central to what’s happening now.
I wish that these were narrow technical issues, of no importance for real-world policy. Unfortunately, they’re not. Understating output gaps leads to excessive demands for austerity and excessive complacency at central banks; this perpetuates the depression; and the longer the depression goes on, the more misleading the standard estimates become."
