I t has been amusing to watch the natural rate of unemployment come down. Two years ago, the community of respectable economists held-though with exceptions including Robert Eisner of Northwestern, Ray Fair at Yale, Harvard's James Medoff, and myself-that 6 percent unemployment was as low as the economy could go without triggering inflation. This meant, in turn, that sustainable economic growth could proceed only at the long-term rate of labor force growth plus the average rate of improvement of the productivity of labor in the recent past, for a growth speed limit of, at the highest, 2.5 percent. Any at tempt to push gross domestic product (GDP) growth any higher would be inflationary, or so we were constantly told.
But since the end of 1995, real GDP growth has averaged 3.6 percent annually. Employment growth and labor force growth were both about double what was expected; productivity growth was pretty much on target. Unemployment, meanwhile, has declined to below 5 percent. And while consumer price inflation rose slightly in 1996-an increase due entirely to rising food and energy prices rather than to any movement of wages-the overall rate of inflation remained negligibly low. Confronted with these facts, the current respectable consensus is that the natural rate of unemployment has fallen to 5.5 percent or 5 percent, or maybe even lower. Those of us who think the very idea of a natural rate should be junked have been encouraged as never before.
But what next? Natural rate devotees remain cautious. For them, the metaphor of a labor market equilibrium, where labor supply just balances labor demand, is too strong; they cannot give it up. True, the equilibrium wasn't 6 percent: We passed that point and disaster didn't happen. And it probably wasn't 5.5 percent: We passed that point too, and again no disaster happened. But it must be somewhere, they think, and we have only 5 percentage points left to go.
Yet the fact that the natural rate turned out to be less than 6 percent-and even less than 5.5 percent-does not make it more probable that the magic number will prove to be 5 percent. Rather, it raises questions about the concept of the natural rate and about the methods used to calculate it. Maybe the natural rate always has been much lower than respected opinion has thought. Or maybe the whole metaphorical notion of a natural rate is misleading.
We don't know the answers to these questions. The real issue, given our ignorance, is: What should we do? Should we pursue falling unemployment and probe for a limit that may or may not be there?
To natural raters, we are like a blind person standing near a cliff by the sea. We may be perfectly safe five feet back from the edge. We may discover by cautious experiment that we can safely take one step forward. This lesson tempts us to change our attitude toward risk, but prudence should hold us back-just because we can take one step safely toward the edge does not mean that we can safely take two more.
This metaphor accurately expresses the natural raters' position on a policy for growth and employment. It is a coherent metaphor, but does it illuminate our actual situation? In my view, it does not. Maybe we need a new metaphor.
Are we really on a cliff by the sea, poised perilously above the waves and the rocks? Or are we in fact down by the beach, on a gentle slope of soft and agreeable sand? What are the risks of another advance-injury and death, or a little wetting of the feet? And if we do get our feet wet, what will it take to get them dry again?
Cliff or beach? This is an issue we can investigate. The way to approach the question is to examine the history of past forays toward the waves. History teaches clearly on this point: Gradually falling unemployment has never resulted in a rapid slide toward inflationary catastrophe. The process is, at worst, extremely gradual. Estimates by Robert Gordon, a pillar of mainstream thinking on the natural rate and a careful student of the past, confirm this. In an article in the Journal of Economic Perspectives earlier this year, Gordon gives a scenario of inflation doubling-to around 6 percent-over eight years, on the assumption that unemployment is kept 1 full percentage point below its estimated equilibrium point. In other words, if the natural rate is in fact 5 percent, then we could enjoy 4 percent unemployment through 2005 without an intolerable increase in the inflation rate. (By that time, of course, the federal budget will have been in balance for more than three years and the Easter Bunny will be President of the United States.)
It is also true that a sudden increase in inflation could happen much sooner-natural rate or no natural rate. A tidal wave, originating in some seismic event in far-off oil fields or elsewhere, could roll in. But in the case of a tidal wave, being five feet back from the water's edge provides no margin of safety.
It therefore makes little difference, from the standpoint of inflation dangers that matter most, whether one pursues full employment or not. The inflation costs of lower unemployment are small, tolerable, and easily reversed if necessary-and that's using pessimistic assumptions. The dangers of an external supply shock, though much greater, are not closely related to the rate of unemployment and cannot be reduced by a slow-growth policy. The lesson to draw is that there is no benefit in failing to pursue full employment.
A second lesson is that the economy is unpredictable in the short run. Those recent high rates of labor force growth and employment growth were not widely predicted and did not result from any policy. But they happened, and had the effect of reconciling higher-than-predicted growth with low rates of wage and price change. Can they continue? Not forever, certainly. But whether they can continue for another year (or three) is simply unknown. To slow the economy now on the assumption that high growth rates cannot continue is to commit a damaging act without justification.
Therefore, at a minimum, policy should do nothing to slow economic growth. Let the economy grow. And if growth slows, policymakers should react quickly by lowering interest rates in an effort to keep progress going. There is certainly no benefit from slower growth and rising unemployment, while the inflationary costs of a stimulative policy in response to evidence of a slowdown are speculative and small.
I happen to think that unemployment can fall another point, to 4 percent, and stay there without disastrous inflation. (Could we go further? Ask me when we get to 4 percent.) I would be particularly confident of this if we had a sensible balanced anti-inflation policy in place-one that did not rely exclusively on raising interest rates and throwing people out of work.
As we think about wading out on that metaphorical beach, some other useful ideas come to mind. To deal with the gentle waves at our feet, a pair of old boots might be helpful. In the 1960s, these were known as "wage-price guideposts" and they helped to keep inflation from accelerating while unemployment fell to below 4 percent. Some new variation on this old-boot theme might be helpful now, preferably one as self-enforcing as possible. And as for the possibility of a tidal shock wave, well, for that you need a boat-or even an ark. But that's another story.
