There's no such thing as a good recession

fedgazette Editorial

Arthur J. Rolnick | Senior Vice President and Director of Research, 1985-2010

Published September 1, 1989  | September 1989 issue

Maybe you saw it. The cover of the July 1 issue of the "Economist" shows a sketch of the earth, a globe suspended in space, which has a sick-looking human face with a thermometer in its mouth. The caption reads, "What we need is a good recession." The idea seems to be that the world economy is overheated and the best medicine for its inflationary fever is a policy-induced recession. A clever analogy, and possibly the right diagnosis, but the wrong prescription. For recent experience and research by some economists have shown that there's no such thing as a good recession. Their suggested prescription—and mine—is a good monetary policy.

The "Economist's" cover story does a fair job of describing the old, wrong medicine, which is still widely prescribed. Titled "A good recession," the story suggests that the world's major industrialized countries, especially the United States, should consider engineering a slowdown in economic activity—declines in employment, production, income and spending—for the overall health of the world economy. And it claims that the sooner this is done, the better. According to the "Economist," the choice is "to aim for a victory over inflation at the risk of a shallow economic recession now; or to accommodate a further spell of rapid growth and upward-creeping inflation, followed by a much deeper recession later." A good recession, then, is a recession that policymakers cause at just the right time to produce a sharp decline in inflation with only a mild decline in growth.

Why do the policy options, whether timely or not, boil down to a tradeoff between inflation and economic growth? The argument is based largely on two assumptions. One is that there is a precise relationship between measures of price level changes and measures of economic activity. This assumption seems firmly founded in history, in a well-documented positive correlation between movements in these measures. Since World War II, most of the data for most industrialized countries reveals that when rates of inflation are high, so are rates of economic growth.

The other assumption is that this correlation can be exploited by policymakers, that they can actually trade some economic growth for more stable prices. Behind this assumption is the notion of "sticky" wages, or the idea that dollar wages are slow to adjust to market forces. If wages are sticky, then they don't increase when inflation does. Thus, when inflation increases, the real value of these wages—what they cost employers or what they can buy workers—drops. And less expensive labor results in increased employment and output. Conversely, if wages are sticky, then when inflation decreases, real wages rise. And more expensive labor results in decreased employment and output. The clear implication is that policymakers can raise employment and output if they are willing to accept more inflation. Or they can lower inflation if they are willing to sacrifice employment and output.

Given these assumptions, the call for a good recession when inflation heats up seems reasonable. But, in fact, it's not. For its two underlying assumptions have recently been challenged by experience and by relatively new and persuasive economic research. Together they suggest that a recession is an ineffective way to fight inflation.

The U.S. economic experience of the 1970s seems to have undermined the first assumption behind the good-recession prescription. In that decade, the positive correlation between inflation and economic growth rates that had been so stable for so many years changed dramatically. As inflation rose to double digits, employment and output not only failed to rise as expected; at times, they dropped. Indeed, in 1975 the unemployment rate reached 9 percent, its highest level since the 1930s. Rather than stimulating economic activity, inflation seemed to be stifling it. The positive correlation had turned negative.

While this was happening, economists were attacking the theory that supported the other good-recession assumption. The research of what we now call new classical economists extended standard economic principles to the analysis of expectations. It found that, even if inflation and economic growth were correlated, there could be no exploitable tradeoff between them. It found, that is, that the second assumption behind the good-recession prescription is false.

According to the new classical view, the old inflation-growth correlation cannot be explained simply by sticky wages; it must also involve unexpected monetary policy changes. For when policy changes are expected, so are inflation changes, and then wages will not stick. Both employers and workers care about real wages, not just dollar wages. Thus, they will make sure that any expected price level changes are quickly incorporated into wage agreements. Since an expected shift from, say, a 10 percent inflation policy to a 5 percent inflation policy won't affect real wages, it won't affect employment and output either. It follows that only unexpected changes in inflation—changes that employers and workers have been surprised by and so have not protected themselves against—will affect real wages and produce a correlation between inflation and growth. Clearly however, this correlation does not represent a tradeoff, because people learn. Any systematic attempt to change inflation and growth quickly becomes predictable.

The new classical economists see policymakers' options quite differently than the old medicine prescribers do. Since, the new classicals say, there is no tradeoff between inflation and economic growth, there is no such thing as a good recession. In fact, to combat inflation, they say, we don't need any kind of recession. What we need is a well-designed monetary policy aimed at reducing inflation, one that is both predictable and credible so that it has little effect on real wages. I agree.

The way to make a policy predictable is to keep it as simple as possible and communicate it well in advance of policy actions. Credibility comes with experience. Thus, if the Federal Reserve announces that its objective is to reduce inflation from 5 percent to 2.5 percent over the next several years by gradually reducing money supply growth, then its credibility depends on its demonstrated willingness to follow through.

With a predictable and credible policy to reduce inflation, real economic activity does not have to be affected. As we have seen, theoretically, the rate of change in the overall price level can decline without reducing employment and output. This theory is supported by several recent studies of countries that devised successful policies to control inflation. The most dramatic success stories come from Europe shortly after World War I. In the early 1920s; many European countries suffered from severe inflation. Yet by announcing and following a credible price stabilization policy, they were able to end that inflation within a relatively short time and without disrupting real economic activity. Other research on countries that suffered from more moderate inflation has found the same thing. Historically, that is, if the public has understood and believed in the new price stabilization policies, then these policies have appeared to work without affecting employment or output much, if at all.

Lowering the world's inflationary fever today thus does not require a good recession, as the Economist and others would have us believe. Instead, it requires a firm and open commitment by all industrialized countries to modest money growth. With a dose of that sort of medicine, the world would have a good chance of moving toward a healthy balance of both economic growth and price stability.