Is the Fed Risking "Dangerous Side Effects"?

Robert Samuelson:

The Fed rolls the dice, by Robert J. Samuelson, Commentary, Washington Post: It was big news last week when the Federal Reserve announced that it wants to maintain its current low-interest rate policy until unemployment, now 7.7 percent, drops to at least 6.5 percent. The Fed was correctly portrayed as favoring job creation over fighting inflation, though it also set an inflation target of 2.5 percent. What was missing from commentary was caution based on history: the Fed has tried this before and failed — with disastrous consequences.

By “this,” I mean a twin targeting of unemployment and inflation. In the 1970s, that’s what the Fed did. Targets weren’t announced but were implicit. The Fed pursed the then-popular goal of “full employment,” defined as a 4 percent unemployment rate; annual inflation of 3 percent to 4 percent was deemed acceptable. The result was economic schizophrenia. Episodes of easy credit to cut unemployment spurred inflation... By 1980, inflation was 13 percent and unemployment, 7 percent. ...

Today’s problem is similar. Although the Fed has learned much since the 1970s ... its economic understanding and powers are still limited. It can’t predictably hit a given mix of unemployment and inflation. Striving to do so risks dangerous side effects, including a future financial crisis. ...

It’s seductive to think the Fed can engineer the desired mix of unemployment and inflation. And the motivation is powerful. About 5 million Americans have been jobless for six months or more. The present job market represents, as Bernanke said, “an enormous waste of human and economic potential.” But the Fed is bumping against the limits of its powers. Are potential short-term benefits worth the long-term risks? It’s a close call.

What does he think a dual mandate means if not the "twin targeting of unemployment and inflation"? That's not unique to the 1970s, it's essentially the Taylor rule (the Taylor principle comes into play as well, but I want to focus on something else). Anyway, he is trying to tell the story about shifting Phillips curve due to rising inflationary expectations, but he misses a key part of the story. A popular explanation for problems in the 1970s, one I think has a lot of veracity, is that the Fed was shooting at the wrong unemployment target (you can find this story in most textbooks, e.g. see Mishkin's text on demand-pull inflation). The Fed was shooting at a 4 percent unemployment target, but because of a large influx of new workers from the baby boom and women entering the workforce, the natural rate of unemployment was actually much higher than 4 percent (new workers tend to have high frictional unemployment rates, and there were also structural changes going on within the economy that led to a higher natural rate of unemployment as well). All told, it's not unreasonable to think of the natural rate had drifted as high as 7 percent, maybe even higher. It eventually came down to closer to 4 percent as the surge of new workers ended and structural change abated somewhat, but for awhile it was elevated above the Fed's 4 percent target. Unfortunately, the Fed didn't not realize this.

Here's how the story goes. The Fed, seeing unemployment drifting toward its natural rate of, say, 7 percent responded to its full employment mandate by using more aggressive policy to create inflation. In the short-run, the policy worked, unemployment did fall due to the inflationary surprise, but as soon as people adjusted their inflationary expectations (and demanded higher wages, etc.), the Phillips curve shifted and we ended up with the inflation we wanted, but the employment gains were lost as the unemployment rate moved toward its natural rate of 7 percent. At that point the Fed says to itself, we must not have been aggressive enough, we need a second round of stimulus and it pumps up the inflation rate even further. Again, this works so long as the inflation is a surprise, unemployment falls in the short-run, but as soon as inflationary expectations adjust once again the employment gains are eliminated, but the inflation remains. As this continues, inflation continues to drift upward until eventually we end up with double-digit inflation and nothing whatsoever to show for it in terms of employment gains.

The fundamental problem here is a miscalculation of the natural rate of the natural rate of unemployment. So the question is, has the Fed made this mistake again? Is the natural rate of unemployment a lot higher than 6.5 percent so that shooting for this target is likely to end up with double-digit, 1970s type inflation?

No for several reasons. First, the Fed is fully aware of this past mistake, and many opposed more stimulus for precisely this reason (e.g. Narayana Kockerlakoata would not support more stimulus until Bernanke convinced him in a series of phone calls that the employment problem was largely cyclical, not structural). If they are shooting at the wrong target, then the policy will not work and they will not continue doing so as they did in the 1970s. They are much more aware of the signs to look for that indicate they've made this mistake. Second, there has been considerable effort to measure the structural/cyclical/frictional unemployment mix for precisely this reason, and the estimates, for the most part, point to a mostly cyclical problem. We didn't have this type of information in the 1970s, in fact we weren't even asking this question. We simply assumed that full employment meant 4 percent and set policy accordingly. Finally, there is an inflation threshold of 2.5 percent, a relatively low level of tolerance for mistakes of this type. If the Fed is wrong about the structural rate, we'll see inflation, and if it the projected inflation rate drifts above 2.5 percent, the program will be reversed. I have no doubt that the Fed is serious abut pulling the plug if inflation rises above 2.5 percent. That's true even if unemployment is still above 6.5 percent.

Samuelson can worry all he wants, he's good at playing the Very Serious Person role (inflation is coming!, the debt will cause interest rates to spike!, there could even be "dangerous side effects, including a future financial crisis"!), but the Fed is not risking a repeat of the 1970s, not even close.

Update: Dean Baker comments on the Samuelson article.