THE RECENT RECESSION, THE CURRENT RECOVERY, AND STOCK PRICES BY WILLIAM D. NORDHAUS COWLES FOUNDATION PAPER NO. 1045 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Box 208281 New Haven, Connecticut 06520-8281 2002 http://cowles.econ.yale.edu/WILLIAM D. NORDHAUS Yale University The Recent Recession, the Current Recovery, and Stock Prices As OF APRIL 2002 THE signs of economic recovery are sprouting like spring flowers. The United States has emerged from the short but painful winter that followed the burst of the asset bubble in early 2000 and the anxiety caused by the events of September 11 and the anthrax scare in the fall of 2001. This report addresses the current state of the economy. What were the salient characteristics of the 2001 recession? What is the state of the econ- omy today? How does the latest business cycle compare with earlier cycles? What are the initial conditions of and therefore the prospects for the recovery? What are the conditions of profits and equity markets? Although it is clearly too early to write the definitive economic history of the recent downturn, it is nonetheless useful to describe the terrain as it appears in early 2002. 2001: The Mildest Recession To begin with, it appears that the 2001 recession was extremely mild. Real GDP growth was close to zero in the second half of 2001, and unem- ployment rose relatively little. Was it really a recession? For that matter, what is a recession? From the point of view of economic welfare, business downturns are undesirable, and therefore the subject of concern and study, because they I am grateful for helpful suggestions from Ray Fair, George Hall, Ken Petrick, and members of the Brookings Panel.200 Brookings Papers on Economic Activity, 1:2002 reduce the nation's output below its potential, reduce people's real incomes, and cause the economic pain of involuntary unemployment. The National Bureau of Economic Research (NBER) defines a recession as a period with "a significant decline in activity spread across the econ- omy, lasting more than a few months, visible in industrial production, employment, real income, and wholesale-retail sales."' Technically, a recession or contraction is a period of declining economic activity because the term refers to the slope of the economy's trajectory rather than its level relative to a high-employment benchmark. From an eco- nomic point of view, however, the levels of output, unemployment, and employment relative to a high-employment baseline are probably more important than their derivatives. To gauge the severity of the 2001 recession compared with earlier ones, I first examine some of the fundamental variables involved in busi- ness cycles. Figure 1 shows a measure of business conditions that is designed to capture substantial declines in output. It shows an index that is zero if the economy is growing and equals the two-quarter growth rate of real GDP if that term is negative. The shaded areas are the NBER con- tractions (recessions). This measure is an accurate predictor of NBER recessions in that it calls every NBER recession and has no false posi- tives. The latest recession barely sneaks in the door, however. The nega- tive growth for 2001 is barely visible and is by a wide margin the smallest downward spike of any postwar recession. A closely related indicator of cyclical conditions is total nonfarm employment. If we construct a series based on two-quarter employment growth rates (not shown) like that in figure 1, periods in which employ- ment declined are also good predictors of NBER recessions. Job loss was substantial in the 2001 recession, but the series has a pattern very similar to that in figure 1, with the 2001 recession virtually tying with the 1970 recession for the smallest postwar decline. Figure 2 shows an analogous measure for the unemployment rate. Here the line shows increases in the four-quarter unemployment rate when unemployment is rising and is at zero when it is falling; recessions are again shown as shaded areas. This measure produces several false posi- tives, indicating that the unemployment rate is a less reliable indicator of 1. NBER Business Cycle Dating Committee, March 11,2002, at www.nber.org/cycles/ recessions.htm1.WilliamD. Nordhaus Figure 1. Recessions and Declines in Real GDP Growth, 1948-2001" Percent a year Source: Author's calculations based on data from Bureau of Economic Analysis. National lacome and Product Accounts (NIPA). a. Two-quarter growth rate of real GDP when negative and zero otherwise. Shaded area< indicate NBER-dated recessions. NBER recessions than real GDP growth. A slight modification would hold that when the four-quarter change in the unemployment rate is greater than 1 percentage point, we have a reliable indicator of recession. By the unemployment rate standard, the last two recessions have been the mildest in postwar history. A final indicator of business cycles is the output gap, which measures the percentage difference between potential and actual GDP. The analy- sis of gaps is controversial because of the difficulties in defining and measuring potential output. Some classical approaches hold that output is essentially always equal to its potential, so that output gaps either are def- initionally zero or quickly disappear. Among schools that believe in gaps, some have measured the gap by interpolating between output peaks. Oth- ers use Okun's law and calculate the difference between actual output and the output that would be consistent with a benchmark unemployment rate such as the NAIRU; this in turn gives rise to other ambiguities aboutBrookings Papers on Economic Activity, 1:2002202 Figure 2. Recessions and Increases in the Unemployment Rate, 1948-2001" Percentage points Source: Author's calculat~ons hased on data from Bureau of Labor Stat~stlcs. a. Four-quarter change In the civilian unemployment rate when positive and zero othenvise. Shaded areas indicate NBER-dated recessions. whether the NAIRU is a wage or a price NAIRU and to empirical diffi- culties in measuring the NAIRU. For present purposes I use a NAIRU-based measure of potential output derived from estimates of potential GDP from the Congressional Budget Office (CB0).2 Figure 3 shows the output gap (again, only if positive) over the postwar