January 6, 2009 | Backgrounder on Economy
The American labor market is remarkably dynamic, even during a downturn. Each month, millions of new jobs are created as entrepreneurs start new companies and existing firms hire new workers. Each month, millions of jobs also disappear as uncompetitive firms go out of business and existing companies let workers go. The diverging futures of Detroit, Michigan, and Greensburg, Indiana, illustrate this tension: Honda has opened a new plant in Indiana creating 2,000 new high-paying jobs while General Motors struggles to survive.
Congress should remember this phenomenon when considering how to respond to the sharp increase in the unemployment rate over the past year and a half. In a dynamic economy like America's, unemployment rises when one of two events takes place: when firms increase layoffs, or when firms create fewer new jobs. In the second case, the workers who naturally lose their jobs in the dynamic economy have greater difficulty finding new work, and stay unemployed longer.
In the current economic downturn, newspapers have been filled with reports of layoffs and bleak Christmasses, with most media stories attributing increased unemployment to the growing job losses. Job losses have, indeed, increased over the past year, which naturally and rightly attracts media attention. But increasing layoffs are not the primary reason for the rising unemployment.
The primary reason why unemployment is rising is because employers are creating fewer new jobs. With fewer available jobs, it takes people longer to find work, and thus the unemployed remain out of work longer. Though the turmoil in the financial markets understandably worries many Americans, those with jobs need not be concerned that their employer has become significantly more likely to lay them off than before. Studies show that even during steep recessions, such as the one in 1982, decreased job creation causes most of the increase in unemployment. To reduce unemployment, Congress must take steps to promote -- not hinder -- job creation.
The American economy is highly dynamic. Industries continually expand and contract while entrepreneurs create new companies and uncompetitive firms go out of business. One of the largest companies in America -- Google, Inc. -- is just 10 years old. One of the oldest companies in America -- General Motors -- is struggling to survive. Workers constantly move between jobs. In the average month in 2007, 4.8 million workers were hired at new jobs while 4.6 million workers left their previous job, either voluntarily or involuntarily. In a typical month, roughly 3 percent of all workers leave their job and 3 percent start a new one. In good times and bad the number of jobs created (or lost) each month is the difference between the number of workers starting new jobs and the number of workers leaving old ones.
Two different events can cause unemployment to rise. First, unemployment rises when employees leave their jobs -- either voluntarily or involuntarily -- directly increasing the number of job seekers. Second, unemployment rises when workers are no more likely to leave their jobs but job creation falls. In that case, even if workers are no more likely to lose their jobs, the workers who naturally leave their jobs or enter the labor force have difficulty finding work -- because fewer jobs are available. Consequently, they take longer to find new jobs and remain unemployed longer, and thus more people report being unemployed to the government each month. The total number of workers who lose their jobs can stay constant while the total number of unemployed at any given moment in time increases.
In recent years, the economy has enjoyed historically low unemployment rates. In 2006 and 2007, the unemployment rate stood at 4.6 percent, slightly below what economists in the 1980s considered the natural rate of unemployment. (Since employers constantly create and shed jobs as technology and consumer preferences change, some unemployment will naturally occur as workers transition between jobs.) In March 2007, unemployment hit a cyclical low of 4.4 percent.
Since then, the unemployment rate has risen steadily. Chart 1 illustrates the U.S. unemployment rates since January 2007. The unemployment rate in November 2008 stood at 6.7 percent for workers of all ages, and at 6.1 percent for those ages 20 and above -- both 15-year highs.
The rising unemployment rate has attracted significant political attention. It is being used to promote a second economic stimulus bill following the passage of the Economic Stimulus Act in February 2008.Members of Congress have already extended the length of unemployment insurance benefits to 46 weeks.
Media Focus on Job Losses. Reports on the failures of large employers, such as the problems of General Motors and Chrysler, fill the news. Many workers fear that their jobs are now at greater risk and that they could soon receive a pink slip. Most media coverage has focused on job losses. Behind this coverage is the strong implication that the unemployment rate rises during downturns because firms become more likely to lay off employees, swelling the ranks of the unemployed.
This view is understandable; it was conventional economic wisdom for many years. It also contains an element of truth. Job losses have increased modestly over the past year and a half. The Bureau of Labor Statistics' Job Openings and Labor Turnover Survey (JOLTS) tracks new hires and job separations on a monthly basis. In October 2008, the most recent month for which data are available, employers laid off or discharged 1.4 percent of the workers in the economy, up 0.1 percentage point from a year earlier. Over the past year layoff rates have increased by 8 percent.
These job losses are real and painful for the workers involved. However, they are not enough to explain why the unemployment rate has increased by more than 50 percent over the past year and a half.
Recent research shows that an increased likelihood of workers separating from their jobs is not the main reason unemployment rises during downturns. In the 1990-1991 and 2001 recessions, greater job separations caused very little of the increase in unemployment. In earlier and more severe recessions, such as that in 1982, a rise in job separations explained only one-third of the increased unemployment.
Decline in Job Creation at Fault
The main reason unemployment rises during economic downturns is that job creation falls while the labor force continues to grow, and new jobs are more difficult to find. Those without work stay unemployed longer, driving up the unemployment rate. This may seem counterintuitive, and this is not most people's impression of the media's message. It is also cold comfort to any breadwinner who has recently received a pink slip. But it is true, nonetheless, and implies distinct policy strategies to reduce unemployment.
Studies and the data show that it is a drop in job creation, not a rise in job losses, that explains the majority of the increase in the unemployment rate during recessions.
Longer Duration of Unemployment
Not only the unemployment rate rises in recessions. So does the average length of time workers spend unemployed. Chart 2 shows the year-on-year percentage changes in the average duration of unemployment and in the number of unemployed workers.
The number of workers unemployed at any given point in time and the average length of time workers stay unemployed rise in almost perfect unison. At the peak of the 2001 recession, unemployment rose 47 percent and the length of an unemployment spell by 39 percent. At the height of the 1990-1991 recession unemployment rose 32 percent and the duration of unemployment rose 33 percent.
If workers were quitting or being laid off from their jobs at higher rates, but could find new jobs at the same rates as usual, then the number of unemployed would rise, but not the average length of unemployment. The fact that the number of unemployed and the time workers spend unemployed rise together shows that lower job creation is the main force increasing unemployment. If it takes workers a third longer to find new jobs, a third more workers will remain unemployed when the Bureau of Labor Statistics interviews them each month.
Detailed Look at Job Losses
A detailed look at job losses confirms this finding. The Heritage Foundation recently used data from the March 1976-March 2007 supplements to the monthly Current Population Survey (CPS)to estimate the probability that any given worker lost his or her job over the past year. This study classified two types of workers as job losers: workers who had a job in the previous year but were unemployed at the time they answered the March survey, and workers who changed jobs over the past year but spent more than two weeks unemployed between jobs.
The first measure of job losses includes the effects of workers staying unemployed longer during recessions, since more workers will report that they are currently unemployed when interviewed because they have not found work. The second measure is not affected by how long workers stayed unemployed between jobs; it only measures the total number of workers that become unemployed and change jobs. An eight-week and a 28-week spell of unemployment are treated identically. Chart 3 shows the yearly job-loss rate between 1976 and 2007 for both types of workers.
The two measures of job losses show very different results. The measure including the currently unemployed is affected by the duration of unemployment and fluctuates wildly with the business cycle. The measure that records the total number of workers who lose their jobs and spend an extended period of time unemployed does not respond strongly to the business cycle. Even during a deep recession, such as in 1982, it barely rises.
The principal reason unemployment rises during recessions -- even deep ones -- is not that employers lay off larger numbers of employees. It is because new jobs become harder to find and, therefore, the unemployed stay without work longer. In a better economic climate, they likely would have found new work much sooner.
The Current Downturn
This is why unemployment has increased over the past year and a half. JOLTS survey data reveals this clearly. Chart 4 displays the job hire and separation rates since January 2007.
Job-separation rates have not increased. In fact, they have fallen. In March 2007, 3.4 percent of workers left their job; last October, 3.1 percent did. This occurred despite the slight increase in layoffs and discharges because workers have become less likely to quit their jobs.
Why has unemployment risen when fewer workers are leaving their jobs each month? New-hire rates have fallen sharply since the unemployment rate hit its low in March 2007. Then, 3.5 percent of workers started work with a new employer; today that number is only 3.0 percent.
Thus, far into the current downturn it has been the fall in the hiring rate, not large drops in job security, which has done the most to increase unemployment. The experience of past downturns suggests that, should the downturn worsen, this phenomenon will continue to be the case.
The Congressional Response
In order to address rising unemployment, Congress must understand why unemployment has increased in the first place. Businesses are tightening spending and loathe taking risks in the current economic climate. Firms are curtailing discretionary investments while trying to weather the economic storm. To reduce unemployment Congress needs to encourage firms to innovate, invest, and take risks and remove policies that discourage them from doing so.
One policy proposed by President-elect Barack Obama would encourage exactly this innovation and job creation. During the campaign, Obama proposed eliminating the capital gains tax on start-up companies. Repealing the capital gains tax on start-ups would encourage more venture capital investment in new businesses. Most new businesses fail. Venture capital funds invest in many new companies and make up the losses in the many failures with large profits on the handful of start-ups that succeed.
If Congress repealed the capital gains tax on start-up businesses, investors would earn greater after-tax profits on the few successful start-ups. The higher returns would encourage venture capital funds to invest in more new companies, including some riskier ventures they will not now invest in. The greater profits from successful companies would compensate for the risk of failure from others. The net result would be more investment, more start-up businesses, and more jobs. This would increase job creation and lower unemployment.
Congress should also remember to do no harm. Labor unions are pressing Congress to respond to the weakened economy by passing the misnamed Employee Free Choice Act (EFCA), which effectively deprives workers of their right to vote in a secret-ballot election on whether or not to join a union. Allowing union organizers to pressure workers into joining a union would greatly expand union membership.
Allowing unions to pressure millions of Americans into joining would further reduce job creation -- the driving force behind unemployment. Academic research shows that employment growth slows dramatically once unions organize a company. Union contracts make business expansion much less profitable, so unionized companies expand much more slowly, if at all. Passing EFCA would increase unemployment.
The unemployment rate has risen sharply over the past year and a half. Many American workers fear that their jobs are at risk, and Congress is trying to find policies that will reduce unemployment. Congress and the American public should understand that job losses have only slightly increased and are not the main reason unemployment has risen. Unemployment is up because employers are creating fewer jobs than in the recent past. In order to promote job creation and reduce the unemployment rate, Congress should adopt President-elect Obama's proposal to eliminate the capital gains tax on start-up companies. Congress should also remember to do no harm and reject policies, such as effectively eliminating secret-ballot organizing elections, that cause employers to create fewer jobs.
James Sherk is Bradley Fellow in Labor
Heritage Foundation calculations based on data from the Department of Labor, Bureau of Labor Statistics, "Job Openings and Labor Turnover Survey," 2007 / Haver Analytics.
Adults are workers 20 years of age and older.
for example, Louis Uchitelle, "Jobless Rate Climbs to 5.7% as
51,000 Jobs Are Lost in July," The New York Times, August 2,
2008, at http://www.nytimes.com/2008/08/02/business/02econ.html (December
16, 2008). See also Neil Irwin and Michael S. Rosenwald, "Job
Losses Accelerate, Signaling Deeper Distress," The Washington
Post, p. A1, October 23, 2008, at http://www.washingtonpost.com/wp-dyn/content/article/200
8/10/22/AR2008102203709.html?hpid=topnews (December 16, 2008).
Department of Labor, Bureau of Labor Statistics, "Job Openings and Labor Turnover Survey"/ Haver Analytics. The publicly released monthly separations figures are seasonally adjusted, but the layoff and discharge rates are not, and display substantial seasonal variability. Consequently the year-on-year change is reported to avoid conflating seasonal changes with a true change in layoff rates.
Robert Hall, "Job Loss, Job Finding, and Unemployment in the U.S. Economy over the Past Fifty Years," National Bureau of Economic Research Macroeconomics Annual 2005 (Cambridge, Mass.: MIT Press, 2005), at http://www.stanford.edu/ ~rehall/nberjobloss.pdf (December 16, 2008); Robert Shimer, "Reassessing the Ins and Outs of Unemployment," NBER Working Paper No. W13421, September 2007; Michael Elsby, Ryan Michaels, and Gary Solon, "The Ins and Outs of Cyclical Unemployment," January 2007, NBER Working Paper No. W12853, at http://ssrn.com/abstract=959129 (December 16, 2008).
Elsby, Michaels, and Solon, "The Ins and Outs of Cyclical Unemployment," p. 11.
"The job-finding rate is the key variable in understanding the large fluctuations in unemployment over the past 50 years. The separation rate, the other determinant of unemployment, has been stable, by all the available evidence." Hall, "Job Loss, Job Finding, and Unemployment," p. 135.
Note that the average duration of unemployment, as measured by the Bureau of Labor Statistics, is the average duration of unemployment at the time a worker is interviewed, not the length of the total spell of unemployment, which usually concludes some time after the worker is interviewed.
James Sherk, "Job-to-Job Transitions: More Mobility and Security in the Workforce," Heritage Foundation Center for Data Analysis Report No. 08-06, September 2, 2008, at http://www.heritage.org/research/Labor/cda08-06.cfm. The March CPS asks workers who they worked for in the previous year, as well as for whom they currently work.
JOLTS provides data on all job separations on a seasonally adjusted basis. The JOLTS survey also provides data on job separations by cause of separation: quits, layoffs and discharges, and other separations. However, the layoff and discharge separation rates are not seasonally adjusted, and, therefore, all job separations were used in this figure.
The figure displays hires and separations because both series are seasonally adjusted. The figures for layoffs and discharges, a subset of the overall separation rate, are not seasonally adjusted. The layoffs and discharge rates have increased slightly on a year-on-year basis, as discussed above.
James Park, "TV Ads Promote Employee Free
Choice," AFL-CIO News Blog, September 3, 2008, at http://blog.aflcio.org/2008/09/03/tv
-ads-promote-employee-free-choice-act/ (December 16, 2008); "Hot Topics: The Employee Free Choice Act," American Rights at Work, undated, at http://www.americanrightsatwork.org/employee-free-choice
-act/home/the-employee-free-choice-act-20080418-555-102.html (December 16, 2008).
Barry T. Hirsch, "What Do Unions Do for Economic Performance?" in James Bennett and Bruce Kaufman, eds., What Do Unions Do? A Twenty-Year Perspective (Edison, N.J.: Transaction Publishers, 2007), pp. 214-218; Barry T. Hirsch, "Unionization and Economic Performance: Evidence on Productivity, Profits, Investment, and Growth," in Fazil Mihlar, ed., Unions and Right-to-Work Laws (Vancouver, B.C.: The Fraser Institute, 1997), pp. 35-70; S. Bronars, D. Deere, and J. Tracy, "The Effects of Unions on Firm Behavior: An Empirical Analysis Using Firm-Level Data," Industrial Relations, 1994, pp. 33, 426-451.