Chair Maloney, Ranking Member Brady, Vice Chairman Schumer and other members of the Committee, thank you for giving me the opportunity to discuss jobs in the United States. We are meeting at a time when the U.S. labor market is beginning to show signs of what I expect will be sustained improvement after moving through the worst downturn since the 1930s by some measures, and since the early 1980s by other measures. The current unemployment rate is 9.7 percent. That is down from a recent high of 10.1 percent last October. This is an improvement, but the rate is still unacceptably high. The economy gained an average of 54,000 jobs per month in the first quarter of 2010, a vast improvement over the 750,000 jobs lost per month in the first quarter of 2009. Yet even with the recent improvement, losses since the start of the recession amount to 8.2 million jobs.
While the collapse in the job market in the wake of the financial crisis has been devastating, job growth in the earlier part of the 2000s was also poor compared to the preceding decade. In my testimony I will focus on two periods: First, I will contrast job growth in the decade of the 1990s with job growth in the 2000s prior to the most recent recession; and second, I will analyze the job losses in the recession that began in December 2007 and the recent stabilization and probable improvement of the job market.
The theme of my remarks is that the U.S. job market is not destined for poor performance because of globalization, technological change, or demographics. Other advanced nations that are subject to these same macro forces have seen stronger job growth than the U.S. in the last decade. I will also elaborate on how the financial crisis affected various segments of the job market, and highlight the lingering effects of the crisis on labor demand, especially among small businesses.
Longer-Term Job Trends Have Not Been Favorable
A look at the jobs picture over the years in the previous decade before the recession indicates that job market performance in the U.S. was poor relative to the 1990s across a number of key metrics. In other words, while the recession has taken a terrible toll on American workers, the job market during the first eight years of the decade of the 2000s--before the recession--was already underperforming.
Table 1 provides several labor market indicators. The number of nonfarm payroll jobs, derived from the monthly establishment survey conducted by the Bureau of Labor Statistics (or BLS), is a standard indicator of employment. Looking over a long stretch of history, despite occasional recessionary periods, the U.S. job market has steadily increased employment to accommodate our growing population until the 2000s. During the 1990s (specifically, from December 1989 through December 1999), the economy gained 21.7 million payroll jobs. By contrast, from December 1999 through December 2009, the economy lost 944,000 jobs. Indeed, as Figure 1 shows, nonfarm payroll employment in the U.S. currently stands at about the same level as it did in September 1999. With no net jobs gained in more than ten years, it is no wonder that many analysts are calling this period the "lost decade." This poor performance is not only due to the recession at the end of the decade. Job gains in the 2000s are weak even if we exclude the losses that occurred during the recession: Over the first 8 years of the 1990s, the economy gained almost 16 million jobs; during the first 8 years of the 2000s, however, payroll employment rose by somewhat less than 7.5 million jobs, a little less than half of the previous decade's 8-year increase.
The lackluster job market performance that is evident in the payroll data also is evident in the BLS's Current Population Survey (CPS), which is a household survey. Consider first the employment-to-population ratio, which is the fraction of the working-age population who report being employed. As Figure 2 shows, the employment-to-population ratio rose 1.3 percentage points from 1989 through 1999, and peaked in April 2000 at a postwar high of 64.7 percent. In contrast, during the decade of the 2000s, the ratio fell nearly 5 percentage points, and is now at a level--58.6 percent--that was last seen immediately following the back-to-back recessions of 1980-82. The decline in the employment-to-population ratio was especially sharp after the recession began, but even prior to the recession the ratio had already given up all of its gains during the 1990s, an indication that job growth failed to keep up with population growth.
More detailed data allow us to compare the experience of individual demographic groups during the most recent decade with their experience in the 1990s. The data show that the trend toward increased labor market participation by women continued throughout the 1990s, but then was partly reversed prior to the 2007 recession; by the end of the decade the share of women working had returned to about the same level as it had been in 1989. The employment-to-population ratio for men edged lower during the 1990s, with an acceleration in this downward trend over the most recent decade. Male employment was then disproportionately affected by the 2007 recession: In the recession, the employment-to-population ratio for men fell more than 5 percentage points, and by 2009 the ratio had fallen to all time lows.
Employment-to-population ratios across age groups also display some noteworthy patterns. In particular, for workers age 55 and older, there was a modest uptrend in employment during the 1990s, cumulating in a 1.6 percentage point increase in the employment-to-population rate for the decade. This increase in employment among older workers accelerated during the 2000s, producing a 6.4 percentage point rise in the employment for the decade. The combination of the aging of the baby boom generation which led to an increase in the number of people over age 55 and the increase in their employment rate resulted in a 56 percent increase in the number of employees who were age 55 and over in the first decade of the 2000s. By contrast, the number of workers in each of the other age groups--as well as the share of the age group employed--declined from 1999 to 2009.
These data suggest that, while the effects of the recent recession have dominated changes in employment, the past decade saw considerable underlying softness in employment prior to the recession. BLS data from the Business Employment Dynamics (BED) program provide evidence of a related trend in gross job flows (as opposed to net job creation). The BED data measure the number of jobs added in business establishments that are opening or expanding and the number of jobs lost in business establishments that are closing or contracting. The BED data indicate a trend toward less churning--a reduction in gross labor flows across companies--beginning in the late 1990s and continuing through the 2000s. The BLS's Job Openings and Labor Turnover Survey (JOLTS) likewise shows a decline in worker turnover in the 2000s.
The U.S. labor market is well known for its dynamism, in the sense that reactions to changes in economic conditions occur relatively rapidly, regardless where we are in the economic cycle. Millions of hires take place each month even in the depths of a recession, and millions of separations occur during an expansion. Although such dynamism can be disruptive for workers and companies, it also serves to reallocate workers from declining to expanding sectors, thereby boosting productivity and, ultimately, living standards for Americans. Therefore, the decline in gross labor market flows is a potential concern if it represents a fundamental shift toward a less dynamic U.S. economy.
The decline in churning that is evident in the BED and JOLTS data, however, is at least partly a result of an aging workforce. I make this inference from an examination of job tenure. A decrease in separations and hiring would be expected to result in longer job tenure, all else equal. Indeed, the average worker age 20 to 64 had job tenure of 7.0 years in 1998 and 7.4 years in 2008 according to tabulations of the Current Population Survey (CPS) data by Henry Farber of Princeton University. If age and education are adjusted for, however, Professor Farber finds that job tenure actually fell in this period for both men and women.
Older workers tend to change jobs less frequently than younger workers. It thus appears that that the aging of the Baby Boom generation has led to a more stable workforce, leading to lower separations (and thus less need for external hiring) and less churning within companies. This older, more experienced workforce is likely a source of higher productivity.
U.S. Job Growth Lagged Other Economically Advanced Countries in the 2000s
Available international data suggest that job market performance in the U.S. in the 2000s was poor not only relative to previous decades, but also relative to the experience of foreign countries with advanced economies.
In Canada, for example, payroll employment rose by 2.3 million during the 2000s, a 19 percent increase that largely kept pace with population growth (see Table 2). The overall employment-to-population ratio fell 1.6 percentage points during the 1990s (when the U.S. ratio was rising). During the most recent decade, however, the overall employment-to-population ratio rose 1.1 percentage points in Canada, while the U.S. rate dropped sharply. From 1989 through 2009, the overall Canadian employment-to-population ratio edged down 0.5 percentage point, as the U.S. ratio fell by more than 3.5 percentage points.
Canada's age distribution is very similar to that in the U.S., with a large post-war baby boom cohort. Like the U.S., Canada experienced a sharp increase in employment of older workers in the 2000s. Unlike the U.S., however, Canada also saw a rise in employment for younger workers. Thus, generational crowding – when older workers hold on to jobs longer and crowd out younger workers from the labor market – is an unlikely explanation for the lackluster job growth in the U.S. in the 2000s.
In the U.K., payroll employment also rose during the 2000s, as the U.K. added 1.3 million workers from 1999 to 2009, about a 5 percent increase (see Table 3). The overall employment-to-population ratio rose 0.6 percentage point during the 2000s, in contrast to the sharp decline in the U.S. Across age groups, the U.K. shows a pattern more similar to the U.S., with sharp declines in the ratio for younger workers and a large increase for older workers. In contrast to the U.S., however, the employment-to-population ratio rose for prime-aged workers during the 2000s, in spite of declines associated with the worldwide recession.
Figure 3 illustrates the change in the fraction of the population working in the U.S., Canada, U.K. and Eurozone in the 1990s and various periods of the 2000s. (Comparable data for the Eurozone are not available for the 1990s.) In contrast to the 1990s, it is clear that job growth was dramatically worse in the U.S. than in these other countries in the 2000s, both in the period before the recent recession and in the recent recession. A likely contributing source of the stronger job growth in Canada and the U.K. in the 2000s is that the education levels of their workforces increased more strongly than was the case in the U.S. Interestingly, while the U.S. job market produced fewer jobs (relative to the population) than in these other economically advanced countries in the 2000s, productivity growth was stronger in the U.S. and total GDP growth was roughly comparable over the decade in all three countries.
The international data carry an important implication: The United States' poor labor-market performance in the 2000s was not inevitable. Canada and the U.K. were subject to the same international trends, had access to the same technological advances and faced similar demographic shifts as the U.S., yet they managed to produce significant job increases during the first decade of the 2000s, while the U.S. lost jobs. Based on reviewing other evidence, Council of Economic Advisers Chair Christina Romer has concluded that "structural factors are not central" to the poor performance of the U.S. labor market. Thus, there is little evidence that fundamental structural shifts have taken place that accounted for the weak record of job growth in the last decade.
Job Market Dynamics by Establishment Size over the Business Cycle – New Findings
A variety of comparisons indicate that the U.S. labor market underperformed throughout most of the first decade of the 2000s. But the dominant feature in the jobs picture of the last decade was the acceleration of the pace of job losses during the financial crisis. The aggregate job statistics--a loss of 8.4 million jobs from December 2007 through December 2009--tell only part of the story. Fully 4.2 million private sector jobs were lost in the six months after the fall of Lehman Brothers in September 2008. Job losses in this period exceeded what one would predict from the sharp concurrent contraction in GDP by about 25 percent.[1] The sharp loss in jobs around the time of the financial crisis resulted because the seizure of credit markets caused a sharp drop in economic activity, and because the panic that took hold of financial markets likely spread to employers in other sectors, causing them to react more than normally to a contraction in demand for their goods and services by shedding workers. Lingering uncertainty from the financial crises has also restrained hiring in recent months.
To better understand the dynamics behind the dramatic loss in employment that we have experienced in the past two years, we can examine data on job openings, hires, and separations. These data are collected by the BLS in a survey of business establishments called the Job Opening and Labor Turnover Survey (JOLTS), and published each month by industry group and by region.
Recently the BLS provided the Treasury Department with research data that include an unpublished, unofficial series of job openings, hires, and separations for establishments in multiple size classes. Analyzing employment trends among the different size businesses can help shed additional light on the mechanisms by which the financial crisis induced job losses, and can provide some clues as to policy actions that could be particularly effective in the current environment. Moreover, the research data provided by the BLS are available through February 2010, which makes them by far the most up-to-date data available on employment patterns in small and mid-size businesses.
We aggregated the JOLTS data by establishment size into three categories--establishments with fewer than 50 employees (representing about 40 percent of private sector employment); establishments with 50 to 249 employees (representing about a third of private sector employment); and establishments with at least 250 employees (representing about a quarter of private sector employment). The data on job openings shows that the number of job openings had been falling since early 2007, but openings fell precipitously around the time that the financial crisis moved into high gear, especially for larger businesses (Figure 4). The low job openings rate--defined as job openings as a share of employment plus job openings--reflects the continued difficulty that unemployed persons are having finding work, as there are relatively few job openings for them to apply for. Specifically, in the published February JOLTS data there were 5.5 unemployed persons for every job opening, as compared to an average of two unemployed per opening over the 2001-2007 period. As with the overall employment situation, the job openings rate stabilized last fall and has picked up in the past two months. The increase in job openings, however, is heavily concentrated among larger establishments.
Figures 5 through 7 plot the gross hires and the gross separations for small, mid-size, and large establishments. The difference between hires and separations equals net job gains or losses in the BLS establishment survey[2] Shortly after the financial panic reached its peak in September 2008, a large number of workers were separated from small establishments (Figure 5). Most of the increase in separations was due to layoffs and business closings, as the number of quits was trending down during this period. (The Appendix Figure breaks down separations into layoffs/closings and quits.) The elevated level of layoffs by small establishments continued through February 2009, after which layoffs began to trend down, although they still remain somewhat high in the most recent months compared with the historical average. From the start of the recession to last fall, hiring by small businesses fell at a moderate but steady pace which did not accelerate during the financial crisis. Today the hiring rate by small businesses remains well below its pre-crisis levels.
The experiences of mid-size and large establishments around the time of the financial crisis were notably different. As mentioned, small establishments responded by quickly laying off a large number of workers. Mid-size establishments (Figure 6) and large establishments (Figure 7) responded by sharply cutting back on hiring in the months immediately after the crisis, and while they also increased layoffs, the increase was not as large as that seen by the small establishments. Of course, the net effect is that total employment contracted severely across establishments of all sizes in the months following the crisis.
The JOLTS data can be used to construct a rough summary measure of notional net labor demand, which is a measure of companies' desired change in employment. Specifically, I define notional net demand as the net job change (total hires minus total separations) plus the total number of job openings, relative to total employment. Results for each size category are displayed in Figure 8. It appears that notional labor demand increased steadily for large establishments throughout 2009. Notional labor demand is more volatile for mid-size and small establishments, but it appears to have increased at a more moderate pace than it has for large establishments.
The analysis of the JOLTS data highlights how the improvement in the labor market seen to date has been unevenly distributed across establishments of different sizes. On the positive side, labor demand has generally trended up at large private sector establishments since reaching a trough in February 2008. Moreover, large establishments have apparently increased employment in five of the six months since September 2009--a possible early sign of durable job growth. At the lower end of the size distribution, however, labor demand by small establishments has continued to be weak, with notably low rates of new hires. The challenges small businesses are facing remains a significant concern to policymakers within the Administration. The Administration has consistently supported efforts to assist small businesses through both numerous provisions in the Recovery Act as well as more recent proposals.
Consequences of a Low-Pressure Labor Market
The JOLTS data are consistent with a story in which many small businesses responded to the shock of the financial crisis by quickly laying off workers and shutting down operations, while the first line of response for larger companies was to freeze hiring. Large companies also increased layoffs over the ensuing months. This pattern is consistent with small employers having lower fixed costs associated with hiring and employment than large employers. It is also consistent with small companies being unable to access credit to maintain employment when demand for their products collapsed in late 2008. Larger companies, which also faced frozen credit markets and declining product market demand in the fall of 2008, eventually had access to corporate debt markets, which enabled them to reduce layoffs and expand employment as the financial markets improved in 2009. Small businesses, which are more dependent on bank financing which remains tight, however, are still facing severe challenges. The Administration's small business proposals, such as the proposals to create a $30 billion small business lending fund and raise the cap on SBA 7(a) loans to $5 million, are particularly well timed given the difficulties that small businesses continue to face in the aftermath of the financial crisis.
Arthur Okun characterized the 1960s as a high-pressure labor market. Lawrence Katz and I similarly described the 1990s as a high-pressure labor market in a 1999 Brookings Paper.[3] I think it is fair to say that we have had what could be characterized as a low-pressure labor market so far in the 2000s, punctuated by a deep recession at the end of the decade that in turn featured excess job losses as the financial crisis infected the rest of the economy. We don't know definitively what the causes were for the low-pressure labor market so far in the 2000s. The deep recession that began in 2007 obviously didn't help job performance. Nevertheless, it is clear that the tax cuts that were intended to boost the economy in 2001 and 2003 did not result in better performance in the labor market than what was achieved in the 1990s, a period when government revenue increased and the deficit was reduced and eventually eliminated.
The consequences of a low-pressure labor market are obvious. Job growth that is not strong enough to accommodate a growing labor force results in higher unemployment. Unemployment carries severe personal and social costs, and can also reduce future economic performance as out-of-work individuals see their skills atrophy and their attachment to the labor market erode. But there are additional, more subtle consequences of a low-pressure labor market. When times are bad, workers are more likely to be forced to take dead-end employment, as opposed to having the opportunity to work more hours in better jobs with on-the-job training, career ladders and fringe benefits. A chronically weak labor market has also been found to raise income inequality and prevent families from leaving poverty. For all these reasons and more, the Administration is steadfastedly committed to working with Congress to enact policies that promote sustainable job growth and that lay the foundation for every American to enjoy the opportunity to share in the tremendous prosperity that our nation is capable of producing.