Kinder, Gentler Recessions

Commentary, Disruption, David Henderson

Future recessions will be less severe and shorter than past recessions and the information technology revolution deserves some of the credit for the shift.

The optimists say that the IT revolution allows annual economic growth to average 4 to 5 percent, rather than the accustomed 2 to 3 percent. There is a plausible case for this higher growth since current government measures understate the actual growth of the economy. An apparent growth of 2.5 percent in the gross domestic product may really be 1 or 2 points higher.

Even if you discount the increase in long-term economic growth, the IT revolution has made recessions less severe. It has done so in two ways. First, it facilitated financial derivatives which most firms use today to reduce risk and hedge themselves against adverse changes in monetary and fiscal policy.

The financial newspapers in the ’70s often carried stories of major firms whose earnings were zapped by movements in exchange rates, interest rates, and prices of key raw materials. Such stories have virtually disappeared. What caused the change? The vast majority of large companies and even many small ones now use financial derivatives to insulate themselves from such unfavorable movements. As a result, the whole economy has become much less vulnerable to changes in interest rates, one of the traditional causes of recessions.

The second way the IT revolution has made recessions less severe is by allowing for the reduction of inventories. For decades, macroeconomists have noted the crucial role of inventories in the business cycle. In 1976, for example, Harvard economists Martin Feldstein and Alan Auerbach wrote that 75 percent of the cyclical downturn in GDP from the peak of a business cycle to its trough could be accounted for by the reduction in business inventories. The explanation: when demand for the goods fell, companies, still sitting on a pile of goods, would lay off workers. Because firms often detected decreases in final demand only after a long lag, the drop in production due to, say, a 10 percent drop in final demand could easily have been 20 percent or more. Firms would then have had to lay off even more workers.

But now, databases produced by Oracle, Informix, and Sybase allow companies to relate salespeople’s orders to data on products ready to ship. The correspondence between final demand and production is therefore much tighter, enabling companies to work with lower ratios of inventories to sales. Although a decline in demand for a company with lean inventories could still lead to layoffs, the layoffs would be less severe than when inventories were high. To the extent that companies throughout the economy are holding smaller inventories, recessions caused by declines in demand will be shorter and shallower.

Another factor, the government deregulation of transportation in the late ’70s and early ’80s, made it easier for companies to use just-in-time methods of production, an element of so-called "lean thinking," by making the shipping of goods by truck and rail cheaper, more reliable, and more responsive to customers’ demands. Between 1981, when deregulation had just begun, and 1987, inventories fell from 14 percent of the GDP to 10.8 percent. Of course, some of this reduction was explained by the fact that 1981 was a recession year and 1987 a boom year. But transportation economists also believe that deregulation allowed companies to pare inventories substantially.

Although IT isn’t solely responsible for lean thinking and transportation improvements, it has played a role in implementing both. Sophisticated IT now lets companies reroute trucks while they’re on the road so that they can deliver shipments to a higher-valued customer and let a lower-valued customer wait a little longer. And improvements in manufacturing often involve the use of computers. Lumber mills, for example, have changed dramatically. Almost no one works there anymore. Computers instead of human sawyers size up trees and estimate the best cuts to make to minimize waste, then guide lasers in making the cuts.

Recessions now are more likely to be part of what economists call "real" business cycles. Such recessions occur when some factor that’s unrelated to monetary policy changes. Examples include the OPEC-engineered increase in the price of oil and, more recently, the decline in U.S. defense spending, increasing competition resulting from newly liberated economies in the Third World and in Eastern Europe, and the computer revolution.

In fact, the 29% decline in annual defense spending between 1987 and 1996 (from 6.2 to 3.5 percent of GDP) was a major factor in the last U.S. recession in 1990-91. But such a recession is the healthy outcome of a free economy, not one to be avoided. However painful the transition is for the workers involved, it is a necessary part of a dynamic, growing economy. The government could have prevented a recession only by maintaining the level of defense spending. But then resources spent on defense would not have been available for other uses, uses in which these resources are now more valued. Keeping defense workers in make-work jobs would have been no more justified than subsidizing buggy manufacturers in the face of Henry Ford’s awesome productivity. Just as a person is sometimes better off taking a salary cut to retool for higher-paying work later, so too is the economy better off now because of the 1990­91 recession.

Another factor that could cause a real, healthy recession is, ironically, the computer revolution itself. That revolution has allowed banks and insurance companies to merge their operations and fire many white-collar workers who have been made redundant by database software. Ford used computers to reduce the number of employees in its accounts payable department by 75 percent. For a while, people put out of work by this improvement in productivity did not find jobs, which drove up the unemployment rate, one of the traditional indicators of a recession. But it had to be higher–until these people found new work, which they did.

Will there ever be recessions in the future? There will, but they are less likely to be as severe as those in the past. Moreover, those that do occur are more likely to be healthy adjustments to changes inevitable in a dynamic economy. And part of the credit for this belongs to the digital revolution.