The Federal Reserve must lower interest rates now to avoid a recession, rising unemployment
by Robert E. Scott and Christian Weller
Despite its half-percentage-point interest rate cut on January 3, 2001, the Federal Reserve must quickly make even deeper cuts to lessen the damage it has done to the economy. In each of the past two recessions, the Fed cut real interest rates by five to six percentage points before unemployment began to decline. Given this history with past recessions, the Federal Reserve should reduce interest rates by at least two to three percentage points within the next year. Without swift action by the Fed, unemployment is likely to rise by at least one to two percentage points, if the economy follows the pattern of previous recessions.
The Fed’s sharp 1.75 percentage-point hike in interest rates in 1999 and 2000 has slowed the economy and may even tip it into a recession. Rising energy prices, stock market instability, and several other important factors suggest that the risk of a U.S. recession is great. Since there are no signs of a significant, economy-wide increase in inflation, prices and wages are unlikely to surge if the Fed cuts interest rates this year. Even if the U.S. avoids a recession, a relaxation of Fed policy now will not damage the economy.
Sharply rising interest rates usually cause recessions
Although each recession is unique, they seem to share some patterns. First, interest rate increases by the Federal Reserve’s Open Market Committee (FOMC) have been one important cause of recessions, with sharp increases in real (inflation-adjusted) short-term interest rates heralding a slowdown.1 Second, months or years after interest rates peak, growth slows, real output actually falls, and then unemployment begins to rise. These patterns are apparent in both of the recessions that took place between 1979 and 1994.
The relationship between interest rates and output (gross domestic product) during the last two recessions is illustrated in Figure 1.2 (Since quarterly growth rates are volatile, GDP growth is measured here using a three-quarter moving average.)
The 1979-82 recession: The first of these – the so-called “double dip” recession of 1979-82 – is unusual for several reasons. First, interest rates and real GDP growth were both extremely volatile in this period. Real interest rates ranged from 0.5% to 9.5%, and real GDP growth from -3.1% to 4.9% per year, with both varying wildly. But since unemployment rose steadily throughout, this entire period can be viewed as one continuous recession.3
Furthermore, sharp interest rate spikes preceded the periods of negative output growth. Real interest rates rose by more than two percentage points in the fourth quarter of 1979 (1979:4) and by an additional three percentage points in 1980:1. GDP fell for the first three quarters of 1980, representing a delayed response, or lag, of about one quarter. The lag between the first interest rate spike and the first of the two dips in the 1979-82 recession is highlighted with a shaded bar in Figure 1 (also see Table 1 and below for further discussion of GDP and unemployment lags).
Real interest rates spiked three more consecutive quarters beginning in 1980:4. GDP subsequently declined for four consecutive quarters, from 1981:3 to 1982:2. Lags between interest rate hikes and output declines are hard to measure in this period due to their volatility, but appear to lie somewhere between the first and second quarters.
The 1989-92 recession: This recession followed a two percentage-point increase in real interest rates that began in 1988:1 and peaked in 1989:1-2. GDP fell for three quarters, starting in 1990:3, representing a lag of six to seven quarters from the peak in real interest rates in 1989:2 (highlighted by the vertical bars in Figure 1). This pattern indicates that the lag between higher interest rates and falling output grew from one recession to the next.
More recently, between 1999:2 and 2000:3 (not shown), the Fed increased real interest rates by about 1.2 percentage points,4 and, as anticipated, real GDP growth has since declined. Past trends would indicate that there is a significant risk of a recession in 2001.
Falling growth rates eventually lead to rising unemployment
The 1979-82 recession: Unemployment tends to lag behind rising interest rates and declining GDP growth, peaking one to two years after GDP has bottomed out (Figure 2). In the case of the 1979-82 recession, GDP growth began an erratic decline in 1979:2. Although unemployment declined slightly in that same quarter, it began to rise the next quarter (1979:3) and thereafter, eventually reaching its peak of 10.7% in 1982:4. In fact, the unemployment rate increased in every quarter between 1979:2 and 1982:4 for a total increase of more than five percentage points. Unemployment did not recover to pre-recession levels until late 1988, nearly five years after GDP reached its lowest point. The peak level of unemployment in 1982:4 occurred 10 quarters (30 months) after GDP reached its initial low point in 1980:2 (indicated by the first set of vertical bars in Figure 2). Unemployment lags behind changes in GDP growth for a variety of reasons, including the tendency of many firms to resist firing workers because of the potential loss of key, employee-specific skills.
The 1989-92 recession: The recession of 1989-92 was much deeper and longer than expected, and it took much longer to develop than the recession of 1979-82. GDP growth began to decline rapidly in 1989:2. Unemployment remained stable for some time, and then began to rise in 1989:4 (a lag of two quarters). GDP growth reached a low point in 1990:4, and the level of GDP hit a trough in 1991:1. Unemployment rose steadily and peaked in 1992:3, six to seven quarters after GDP growth and level reached their nadir (second set of vertical bars in Figure 2). This pattern demonstrates that peak unemployment lagged by 18 to 21 months behind the trough of output in this cycle.
The next recession
Real GDP growth declined sharply from 5.6% in 2000:2 to 2.2% in 2000:3.5 There are already signs that the economy will slow down even further, with many leading economic indicators in decline since February 2000. For example, housing fell 13.6% between February and December 2000.6 The motor vehicle sector was also hard hit in 2000 by energy price increases and falling consumer confidence. In fact, the sales of domestic autos declined 24% between February and December 2000, and domestic lig
ht truck sales declined 21% in the same period.7 Furthermore, the purchasing manager’s index has fallen steadily throughout this period and has been below 50 since August, suggesting that a downturn is a distinct possibility (any value below 50 in this indicator is taken as a sign that the surveyed purchasing managers expect a slowdown, whereas a value above 50 indicates an expected upswing). Finally, the Bureau of Economic Analysis’ composite index of leading indicators was flat or declining in every month between March and December 2000.
As for the immediate future, the U.S. economy also faces a number of threats in early 2001. First, the doubling and quadrupling of some energy prices (especially for petroleum products and natural gas) is a supply shock that will further depress output. Second, corporate earnings are likely to decline with output, thereby depressing stock prices. The recent U.S. economy was built on stock market speculation in the 1990s, and a rapid stock market decline could result in financial instabilities because the household and business sectors are mired in debt. Rising loan defaults could make lenders wary enough that a slowdown in credit expansion could hamper a recovery, even if interest rates are lowered. Another threat comes in the form of the rapid and continuous growth of the U.S. trade deficit. The resulting surge in net foreign debt increases the financial fragility of the U.S. economy. As more overseas capital is needed to finance future trade deficits, an economic slowdown may make it increasingly difficult to attract foreign capital or to avoid the outflow of capital to other financial areas, particularly the 11 member countries of the European Union. Finally, recent planned and unplanned power outages and brownouts in California pose a threat to that state’s economy. Since California generates about 13% of total U.S. GDP, any disruption to the California economy will have ripple effects throughout the country.8
Since real interest rates probably peaked in the third quarter of 2000, this suggests that the United States may now be headed for a period of very slow growth or recession. If the economy does enter a recession, output could decline until at least early 2002 if the recession follows the patterns established between 1989 and 1992. The downturn could last much longer in a worst-case scenario, which could involve a stock market collapse, international trade and financial crisis, or both. If this occurs, unemployment will begin to increase in the near future but not peak for at least two to three years, long after GDP has begun to “recover.”
The Fed must rapidly cut interest rates now if it wishes to minimize the damage done to the economy, and any future damage to unemployment in particular, as a result of its 1.2 percentage-point hike in real short-run interest rates between 1999 and 2000.
In the stubborn recession of 1989-92, the Fed was forced to cut real interest rates by more than five percentage points between 1989:2 and 1994:1 (Figure 1) to lower unemployment from its peak of 7.6%. If the last recession is any indication, the Fed must soon cut interest rates by at least 1.75 to 3.5 percentage points to reduce or eliminate the risks of a recession. In particular, a rate cut by at least 0.5 percentage points is necessary at the Fed meeting in January. Short of that, the Fed will signal to workers and markets that it is willing to tolerate a recession and growing unemployment.
In addition to these steps, a fiscal policy response is probably required as well to help those who would be most vulnerable in a downturn and to stimulate the economy. The most important steps would include: (1) improvements in social safety nets, such as extended unemployment insurance, benefits coverage, and aid to those not eligible for welfare or who are unemployed through no fault of their own, (2) acceleration of public investment in schools, transportation, and other needed infrastructure, and (3) design of any tax cuts to be temporary and heavily weighted to low- and middle-income consumers, who are the most likely to spend the additional income.
It’s clear that the Fed’s sharp interest rate hikes in 1999 and 2000 have slowed the economy and may even have placed it in jeopardy. Rising energy prices, stock market instability, and several other important factors suggest that the U.S. economy is in great risk of tipping into recession. With no signs of a significant, economy-wide increase in inflation, prices and wages are unlikely to surge if the Fed cuts interest rates this year. And even if the U.S. avoids a recession through quick action, a relaxation of Fed policy at this point will not damage the economy.
1. The FOMC sets rules to control the federal funds rate (FFR), the interest rate that banks charge each other for overnight loans. All short-term loans, as well as longer-term loans to a lesser degree, are influenced by the FFR.
2. In 1973 the U.S. moved to a flexible exchange rate system, and liberalized capital flows into and out of the United States, which profoundly changed its economic structure. The impact of monetary policy was greatly increased and that of fiscal policy (taxes and government spending) was reduced. Thus, we analyze only the recessions following 1973.
3. Although the NBER considers two recessions to have occurred in this period, this paper sees it as one, based on the continual rise in unemployment.
4. Figures 1 and 2 include information through 2000:2 only because we report GDP growth as a three-quarter moving average.
5. Actual quarterly data are not averaged.
6. See “Economic Indicators Summary” data file from .
7. See “Auto and Truck Sales and Production” data file from . All data reported at seasonally adjusted annual rates. Heavy truck sales (domestic and foreign) declined 23% in the same period.
8. See “Gross State Product 1992-98” data file from .
Bureau of Economic Analysis. 2001. “National Accounts Data.” http://www.bea.doc.gov/bea/dn1.htm
Bureau of Labor Statistics. 2001. “Employment and Unemployment.” http://stats.bls.gov/datahome.htm
Federal Reserve Board. 2001. “Statistics and Releases.” http://www.federalreserve.gov/releases/
National Bureau of Economic Research. 2001. U.S. Business Cycle Expansions and Contractions. Cambridge, Mass.: NBER.