August
12, 2003, 7:00 a.m.
Not All Recessions Are Alike
The
economic dip of 2001 was a lot different from its predecessor.
n length, the 2001 recession (dated April 2001 to November 2001) was relatively
short, a lot like the 1990-1991 recession (dated August 1990 to March
1991). Based on official recession-dating by the National Bureau of Economic
Research, each was eight months in duration, compared with the 11.8 month
average for the five previous recessions.
In many ways, however,
the two most recent recessions were quite different.
Leading into the 2001
recession, the dollar was very strong. Commodity prices had suffered a multi-
year deflation, and CPI inflation was falling to very low levels. In sharp
contrast, the 1990-91 recession was preceded by a relatively weak dollar
and rising inflation.
In terms of government
policies, the two recessions were almost opposites. The government cut tax
rates during and following the 2001 recession, whereas it raised rates substantially
in 1990. And the Federal Reserve began cutting the fed funds interest rate
in January 2001, months before the official beginning of the 2001 recession.
In contrast, it was reluctant, and hence late, to ease in 1990-91.
The widely used rule
of thumb is that two consecutive quarters of declining real gross domestic
product may indicate a recession. So using GDP as a guide, how do the two
recessions stack up? The comparison is interesting.
Real GDP fell in each
of the first three quarters of 2001 and in the three quarters surrounding
the 1990-91 recession. This fulfills the two-consecutive-quarters rule.
From its peak in the first quarter of 2001, GDP fell 0.6 percent through
the third-quarter GDP trough. The 1990-91 recession was also modest from
a GDP perspective, with GDP falling 1.3 percent from peak to trough.
However, for quarterly
data, the NBER dates the recession as Q2 2001 to Q4 2001. Using that definition,
GDP actually rose 0.2 percent in the 2001 recession, illustrating
the shallowness of this economic dip.
Here are some other
key peak-to-trough economic statistics for the two recessions (from NBER):
Recession
Dating 1990-91: August 1990 to March 1991 (peak was July 1990) 2001:
April to November (peak was March 2001)
Real Personal
Consumption 1990-91: -0.4 percent 2001: 2.1 percent
Real Disposable
Income 1990-91: -0.8 percent 2001: 0.1 percent
Unemployment Rate 1990-91: Peak = 5.5 percent Trough = 6.8 percent 2001: Peak
= 4.2 percent Trough = 5.6 percent
Nonfarm Payroll
Employment 1990-91: -1.23 million (-1.1 percent) 2001: -1.63 million
(-1.2 percent)
Real Fixed Nonresidential
Investment 1990-91: -4.5 percent 2001: -8.0 percent
Real Residential
Investment 1990-91 -13.1 percent 2001: -0.9 percent
Contribution of
Inventories to Real GDP Growth 1990-91: -0.6 percentage point 2001: -0.8 percentage point
Change in S&P
500 Stock Market Index 1990-91: +8.1 percent 2001: -8.0 percent
Change in Federal
Funds Rate 1990-1992: -2.0 percent 2001: -2.5 percent
Weakness in business-investment
spending was clearly the prominent factor behind the 2001 recession, while
weak consumption spending and residential investment played their usual
important roles in the 1990-91 recession.
The weakness in job
growth was similar in the two recessions. However, this weakness has extended
far past the end of the 2001 recession, in contrast to the 1990-91 episode.
The trough in nonfarm
payroll employment was May 1991, just two months after the recession trough.
In contrast, nonfarm payrolls have continued to decline through the June
2003 figure. But the rapid job growth and very low 3.8 percent unemployment
rate achieved prior to the 2001 recession explain the sluggishness in employment
in the current post-recession period.
Not all recessions
are alike. And yet, they all have their political ramifications — which
brings us to a final interesting difference between the 2001 and 1990-91
economic periods: their dating relative to presidential elections.
The 2001 recession
began just 2 months after the presidential term began, and it ended 36 months
before the next presidential election. In contrast, the 1990-91 recession
occurred in the second half of the presidential term and ended 20 months
before the November 1992 election. In fact, the NBER did not announce its
end date until December 22, 1992 — over a month after the presidential
election.
Consequently, the 2001
recession should be a less important factor in the 2004 election than the
1990-91 recession was in 1992.
<% dim printurl
printurl = Request.ServerVariables("URL")%>
n length, the 2001 recession (dated April 2001 to November 2001) was relatively
short, a lot like the 1990-1991 recession (dated August 1990 to March
1991). Based on official recession-dating by the National Bureau of Economic
Research, each was eight months in duration, compared with the 11.8 month
average for the five previous recessions.
