Report | Wages, Incomes, and Wealth

The Jobless Recovery: Suffering from the recession’s aftershocks, labor market conditions continue to worsen

Issue Brief #186

Download PDF

Though the recession that began in March 2001 has not yet been declared officially over, most economists believe it ended early in 2002. However, the labor market downturn is far from behind us. Today’s labor market is much weaker than it was one or even two years ago, and the “jobless recovery” grinds on.

The jobs picture is so serious that steps to stimulate the economy and generate job growth are urgently needed. Any stimulus proposal should be evaluated primarily on its impact on job creation and its ability to reverse the current trend of weakening wage growth.

An examination of the trends in key labor market indicators between the end of 2000 and the end of 2002 shows widespread and persistent weakening in the labor market. Key findings are:

  • Despite the fact that the recession is widely believed to have ended last year, unemployment rose throughout 2002, ending the year at 6.0% in December. Since the most recent economic peak, the jobless rolls have expanded by 2.8 million.
  • Compared with the end of 2000, there are now 2.1 million fewer private sector jobs. Payrolls contracted not only over the recessionary year of 2001, but also over the alleged recovery year of 2002.
  • The decline in private sector jobs at this point in the recovery is greater than in any of the past three recessions/recoveries.
  • The lack of job creation also has led to long spells of unemployment. The average unemployment spell has increased by more than five weeks compared with its level in the fourth quarter of 2000. There are now 1.7 million people who have been jobless for more than half a year.
  • The lack of employment growth has led to slower growth of the labor force, as fewer people choose to compete for scarce jobs. The labor force is now growing half as fast as it was two years ago, a sure sign of a weak labor market.
  • The rise in unemployment has led to slower wage growth, real income losses, and higher poverty rates.
  • In 2001, when the unemployment rate climbed to 4.8% from 4.0% in 2000, real household income fell by 3% for the poorest households and 2% for middle-income households, while poverty increased by 0.4 percentage points. Although income and poverty data for 2002 are not yet available, given that unemployment was another point higher in 2002 (5.8%), the incomes of low- and middle-income households very likely fell even more last year.
  • According to a new report from the Bureau of Labor Statistics, median weekly earnings fell 0.5% in real terms over the past year (from fourth quarter 2001 to fourth quarter 2002).

The jobless recovery

The tables below compare the recent trends in key labor market indicators over three time periods: the last quarters of 2000, 2001, and 2002. These periods span the time from the labor market peak of the last business cycle through the most recently completed quarter. If the alleged recovery had reached the labor market, we would expect to see trends worsening during 2001 but then improving in 2002. Instead, we see a steady worsening. Some negative trends, such as long-term unemployment, even accelerated.

Table 1 shows one of the primary indicators of labor market weakness: the unemployment rate, which is the share of the workforce unsuccessfully looking for work. At the end of 2000, the unemployment rate stood at 4.0%. With the onset of recession in early 2001, the rate began to climb, reaching 5.6% by the last quarter of that year. Since then, joblessness has continued to rise, reaching 5.9% in the most recent quarter. As the second panel shows, the ranks of the unemployed grew over this two-year period from 5.6 million to 8.4 million, an increase of 2.8 million unemployed people.

Table 1

Table 1 also shows unemployment rates by gender and race. Note that the African American rate has risen the most, reaching double digits by the last quarter of 2002, at 10.8%. This represents an increase of 582,000 in the number of black unemployed persons.

Table 2 shows one reason why unemployment keeps rising: the absence of job growth. After growing by 1.6% in 2000, the nation’s payrolls contracted by 0.8% in 2001 (1.4% in the private sector), and continued to decline, though less quickly, through 2002. Job losses over this period amounted to 1.4 million overall and 2.1 million in the private sector (since government hiring is less sensitive to the business cycle, private sector employment growth is more indicative of economic conditions).

Table 2

The manufacturing sector has been particularly damaged, falling by 6.7% (1.2 million jobs) in 2001 and 3.7% (another 642,000) last year. Job losses in retail trade actually accelerated in 2002, driven partly by particularly weak holiday hiring.

Figure A compares private sector employment losses over four different recessionary periods. In each case, we measure employment growth from the low point of unemployment through the subsequent 27 months (since we are now 27 months beyond the unemployment low of 3.9% in October 2000). Even though other recessions were deeper and lasted longer than the most recent one, payrolls have fallen persistently throughout this 27-month period. Thus, as shown in the figure, employment losses in the most recent recession and subsequent jobless recovery have been greater than in any of the other three recessions and recoveries. Other recessions may have led to greater losses initially, but by this point in those recoveries, the economy had bounced back and payrolls were again expanding.

Figure A

Table 3, meanwhile, shows that the decline in employment has meant not only more unemployment, but also much longer spells of joblessness. The table shows the trend both in average weeks spent unemployed and in the share of the unemployed who have been out of work for at least half a year. Both series show that the problem of long-term joblessness worsened in 2002, as the jobless recovery took hold. Since the final quarter of 2000, the average time spent unemployed has grown from about three months (12.4 weeks) to about four and a half months (17.9 weeks), an increase of well over a month. At the same time, the share of the unemployed who have been without work for at least 27 weeks has almost doubled, from 11.1% to 20.9%. Thus, by the last quarter of 2002, 1.7 million job seekers had been without work for at least half a year.

Table 3

Some commentators have argued that, since unemployment has yet to surpass 6%, this period of recession and slow recovery has been more benign than previous ones. Such analysis overlooks two countervailing factors, however. First, the pain of recession is experienced not only in the level of unemployment but in its trend. In this regard, the two percentage-point increase in unemployment over this downturn is not far behind the increase over a comparable time period in the last recession (which was 2.3 points). Second, in the current recession, the lack of job creation has led to a marked decline in the growth of the labor force. Since fewer workers are competing for scarce jobs, this has the effect of preventing the unemployment rate from rising more quickly, masking the full extent of the underlying hardship.

Table 4 presents the percentage changes in the labor force over the fourth quarters of the past three years, showing a clear deceleration from 0.9% in 2000 to 0.4% in 2002. The second column of the table shows the net growth in the number of persons in the labor force over the past few years. Between the fourth quarters of 1999 and 2000, the labor force grew by 1.2 million. It grew slightly less in 2001. Then, in 2002, growth dropped dramatically, with only 508,000 people joining the labor force that year: a clear symptom of weak job creation.

Table 4

Table 5 shows that the loosening of what had been a very tight labor market has meant slower growth in wages. It shows the annual changes in hourly wages over three years (comparing data from the third quarters of each year, since data are not yet available for the most recent fourth quarter). Wage changes are shown by gender for low-, middle-, and high-wage workers, and are based on nominal wages – not adjusted for inflation – with the rate of inflation shown in the last column.

Table 5

Nominal wage growth has slowed considerably. By the third quarter of 2002, nominal hourly wages for low-wage workers grew 1.4%, compared to 5.3% in 2000. Middle and high wages have similarly decelerated. Inflation also slowed over this period, so these smaller nominal wage increases do yield more buying power than they would have a few years ago. However, a new report by the Bureau of Labor Statistics with data through the end of last year finds that the nominal median weekly earnings of full-time workers grew 1.7% between the fourth quarters of 2001 and 2002, which is below the rate of inflation for that period (2.2%).1

Even if the labor market were to improve soon, nominal wage growth would likely be slow to revive. This is because there is a considerable lag time between rising unemployment and its negative impact on wage growth (wage growth this year may be just beginning to suffer from last year’s rise in unemployment).

Table 6 shows the resulting impact on living standards as reflected in the Census Bureau’s annual income data. Although such data only go up to 2001, the impact of the recession is quite clear. The contracting economy, particularly the weakening labor market, helped drive real household incomes lower, with the largest losses occurring at the bottom of the income scale. Income fell 2.9% in real terms for the poorest households, and 1.8% for middle-income households. It grew slightly only for those at the top of the income scale. (Poverty also rose by 0.4 percentage points in 2001.)

Table 6

It is important to note that these losses occurred simultaneously with the rise in the average annual unemployment rate from 4.0% in 2000 to 4.8% in 2001. Unemployment rose another point in 2002, on average, to 5.8%, and it is highly likely that the incomes of middle- and low-income households contracted further last year.

These wage and income trends reveal that the weak economy is not simply a problem for the unemployed. Many of those who have kept their jobs are facing less job security and are considerably less likely to see the type of wage increases that prevailed a few years ago. In 2000, when unemployment was hovering around 4.0%, the tight labor market ensured that the gains of economic growth were broadly shared and that wages and incomes rose for most working families. These conditions are now unlikely to return until growth accelerates and unemployment falls below 5%.

In short, these tables help reveal the current weakness in the labor market, and its consequences. High unemployment rates relative to a few years ago are beginning to lower the living standards of many working families. The persistent lack of job growth is leading to lengthening spells of unemployment and slower wage and income growth. The primary goal of domestic economic policy at this point should be to target and reverse these negative trends.


1. See “Usual Weekly Earnings of Wage and Salary Workers: Fourth Quarter 2002.” U.S. Bureau of Labor Statistics.

Issue Brief #186, 6 pages, 8 1/2″ x 11″, paper

FREE (shipping charges apply)

See related work on Income and wages | Wages, Incomes, and Wealth

See more work by Jared Bernstein