One of the most popular definitions of recessions is that they are periods when real gross national product (GNP) has declined for at least two consecutive quarters. In 1990, real GNP declined between the third and fourth quarters and again between the fourth quarter of 1990 and the first quarter of 1991. Hence, there is general agreement that a recession did occur.

Although the definition worked quite well in this instance, there are several problems with it. One is that it does not provide monthly dates of when recessions began or ended. For this purpose the National Bureau of Economic Research (NBER), whose chronology of recessions is widely accepted, uses monthly measures of production, employment, sales, and income, all expressed in real terms (after allowing for inflation). GNP figures are not available monthly. The NBER found that the latest recession, from business cycle peak to trough, ran from July 1990 to March 1991.

Another problem with the two-consecutive-quarters definition is that there can be serious declines in economic activity even without two consecutive quarters of negative growth. Suppose that in one period, real GNP declines 5 percent in the first quarter, rises 1 percent in the second, and declines 5 percent again in the third. In another period let’s say real GNP declines 1 percent in each quarter. Obviously the first period shows a much more serious drop in GNP, but only the second period qualifies as a recession according to the definition.

These and other considerations have led the NBER to use a broader definition of recessions, which takes into account three dimensions of the decline in aggregate economic activity—its depth, duration, and diffusion across industries. These are known as the three Ds. Measures of this sort for several recent recessions are shown in table 1. This broader concept has also been applied to a much longer period, covering two hundred years of U.S. history (see chart 1).


The Three Ds of Recession: A Brief History




Months Percent Change in Real GNP Unemployment Rate, maximum Percent of Industries with Declining Employment, maximum
Three Depressions
Jan. 1920-July 1921 18 n.a. 11.9 97
Aug. 1929-Mar. 1933 43 -32.6 24.9 100
May 1937-June 1938 13 -18.2 20.0 97
Six Sharp Recessions
May 1923-July 1924 14 -4.1 5.5 94
Nov. 1948-Oct. 1949 11 -1.5 7.9 90
July 1953-May 1954 10 -3.2 6.1 87
Aug. 1957-Apr. 1958 8 -3.3 7.5 88
Nov. 1973-Mar. 1975 16 -4.9 9.0 88
July 1981-Nov. 1982 16 -2.6 10.8 72
Five Mild Recessions
Oct. 1926-Nov. 1927 13 -2.0 4.4 71
Apr. 1960-Feb. 1961 10 -1.2 7.1 80
Dec. 1969-Nov. 1970 11 -1.0 6.1 80
Jan. 1980-July 1980 6 -2.5 7.8 63
July 1990-March 1991 8 -1.2 6.9 73
1920-1938 (5) 20 -14.2 13.3 92
1948-1991 (9) 11 -2.4 7.7 80

SOURCE: Based on table A-2 in G. H. Moore, Business Cycles, Inflation and Forecasting, 2nd ed., 1983. Note that the brief and mild recession of 1945 is omitted here.

Length of Business Recessions and Expansions United States, 1790-1991.
Chart 1. Length of Business Recessions and Expansions
United States, 1790-1991

SOURCE: Compiled from data supplied by the National Bureau of Economic Research Center for International Business Cycle Research, January 1993.
Enlarge in new window
One significant trend seen in the chart is that recessions have been getting shorter and expansions longer. The average recession during the past fifty years lasted eleven months, whereas the average recession was more than twice that long in the nineteenth century. Expansions, on the other hand, now last about twice as long as they did in earlier times. A number of factors account for this trend. One is the growing importance of the service industries, such as trade and transportation, where employment is usually more stable than in manufacturing. As the more stable industries have grown in importance, this has made the whole economy more stable and less susceptible to prolonged and severe recessions. Also, the government has been playing a bigger role in moderating recessions, especially since the thirties. Unemployment insurance has helped to reduce the loss of income during recessions, and monetary policy has been used to reduce interest rates and make credit more accessible.

Recessions have a global as well as domestic dimension. Financial markets in many countries are closely watched in other countries, and many investors are making investments on an international scale. Exports and imports have become more important to business enterprises, and managements must now deal with global competition. As a result periods of recession are likely to encompass many countries at about the same time. During recessions in the United States, a majority of the industrial countries in Europe and the Pacific are apt to show signs of recession also. This, in turn, has depressing effects on the United States by slowing the foreign demand for U.S. exports.

The upshot is that even though recessions are not as severe as they used to be, they have serious consequences in many directions. Some industries, some occupations, and some areas of the country are hit much harder than others. Hence, it pays to keep close watch on them with all the daily, weekly, monthly, and quarterly data that are available for that purpose.


Geoffrey H. Moore is the director emeritus of the Center for International Business Cycle Research, Columbia University, New York.

Further Reading

Burns, Arthur F., and Wesley C. Mitchell. Measuring Business Cycles. 1946.

Center for International Business Cycle Research. Recession-Recovery Watch. Monthly.

Gordon, Robert J., ed. The American Business Cycle: Continuity and Change. 1986.

Moore, Geoffrey H. Business Cycles, Inflation, and Forecasting. 1983.

Moore, Geoffrey H. Leading Indicators for the 1990s. 1989.

U.S. Department of Commerce. Survey of Current Business. Monthly.