Abstract
A number of industries underwent large and permanent reductions in employment growth at the beginning of this decade. We investigate the sources of these permanent changes in employment growth and what the consequences were for the U.S. economy. In particular, we find that relative declines in demand rather than technological innovations were the key drivers of the elevated levels of job destruction and permanent layoffs in the affected industries. In addition, most workers that were displaced in downsizing industries relocated to other sectors. While this process of reallocation led to large increases in productivity (and a reduction in labor's share of aggregate income) in industries shedding workers, it also resulted in prolonged periods of unemployment for many displaced workers, along with sizable reductions in earnings that were consistent with substantial losses in their specific human capital. Putting these pieces together, we estimate the costs to those adversely affected by these events to have been 1/2 percent to 1 percent of aggregate income per year.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See, for example, the exchange between Lilien [1982] and Abraham and Katz [1986].
See, for example, Kirkegaard [2009], Phelps [2008], and Williams [2009].
In addition, Oliner, Sichel, and Stiroh [2007] argue that the large gains in productivity during this period were partly the result of the reallocation of labor and materials across industries.
We adopt the industry breakdown used by the Bureau of Economic Analysis (BEA) in its publication of gross output by industry. For the private nonfarm sector, the BEA publishes annual data on output and employment for 59 industries.
We chose these periods because average annual employment peaked in 2000 and troughed in 2003.
Our results are not qualitatively different if we exclude this industry from the group of downsizing industries.
For a discussion of this issue, see Mann [2003].
See, for example, the discussion in Chapter 2 of the 2004 Economic Report of the President.
To measure these co-movements in the data, we first calculate the change in the growth rate of each variable from 1990–2000 to 2000–03 by industry. We then aggregate the relevant industry-level cross products using industry employment shares as weights.
In particular, several industries—including ambulatory health care services and credit intermediation—appear to have benefited from positive supply innovations. Ambulatory health care services (a trend increasing industry) may have benefited from the weak labor market in 2000–03, making it easier and less expensive to recruit new workers. Credit intermediation may have benefited from the sharp reduction in short-term interest rates over this period.
These correlations could also be consistent with technology changes that were biased against labor if the elasticity of substitution between labor and other factors of production was greater than one. For example, the co-movements for information and data processing services are consistent with advances in telecommunications technology making possible the substitution of lower-priced labor overseas for domestic labor.
The industry classification scheme in the Current Population Survey (CPS) switched from an Standard Industry Classification basis to a North American Industry Classification System basis in 2002. We develop a concordance between the two classifications that yields a fairly consistent definition of industry pre- and post-2002. This requires a less disaggregated classification than the BEA's (33 industries instead of 59). We label any CPS industry containing a significant share of employment from BEA downsizing industries as downsizing. 13 of the 33 industries are classified as downsizing, and they account for about 23 percent of employment.
We use the 1994–2000 period rather than 1990–2000 because the 1994 CPS introduced a break in the classification of permanent layoffs.
Reductions in industry employment do not necessarily imply increases in permanent layoffs. Given that gross hiring is well above zero even in industries experiencing sharp declines in employment, it is possible—in an accounting sense—for reductions in employment to be accomplished with a reduction in hires rather than an increase in layoffs.
In particular, Ramey and Shapiro [2001] find that the time required for reallocation is proportional to the specificity of capital, and Eisfeldt and Rampini [2003] find that liquidity in secondary physical capital markets is procyclical. This latter result suggests that the cost of moving resources out of downsizing industries will be higher if downsizing is concentrated in downturns.
One caveat to this result is that our estimates of the recovery of productivity in downsizing industries may be biased upward. By identifying downsizing industries as those with relatively larger and more persistent employment declines, we have introduced a selection bias against finding a sluggish response of productivity to a drop in relative demand.
In support of this hypothesis, Loungani and Rogerson [1989] show that unemployment spells for workers who permanently change industries are longer lasting than other unemployment spells.
Because displacement is defined as a job loss that occurred as a result of a plant closing, insufficient work, or a position being abolished, these job losses were likely involuntary.
To compute displacement rates we use CPS measured employment in the denominator.
Displacement rates in those industries that did not downsize rose noticeably in 2001, but they subsequently dropped off, leaving the average 2001–03 displacement rate little different from its 1997–99 average.
More specifically, we estimate the completed duration of nonemployment spells for the censored observations as the number of weeks of observed nonemployment plus 1/μ, where μ is a constant hazard rate estimated from the observed completed spells. In fact, hazard rates appear to be slightly U shaped, decreasing from 0 to 50–75 weeks and then increasing thereafter, though the limited number of completed long-duration spells increases the uncertainty of the slope as duration increases. We also computed nonemployment durations without adjusting for censoring. Results were qualitatively similar to the results adjusting for censoring that we report here.
The ratio (d/d+H) is analogous to a steady state rate of unemployment. As shown by Shimer [2005], steady state unemployment rates track actual unemployment rates quite closely over the business cycle. Although exit rates from nonemployment for displaced workers are, on average, lower than exit rates from unemployment for all workers—leading to a larger discrepancy between actual and steady state rates—the steady state approximation is still informative.
This equation follows from the steady state condition that e t =H × (1−e t )−d × e t =0, where e represents employment and the labor force is normalized to equal 1. We implicitly assume that nonemployed displaced workers were available for work and thus contributed to reduced resource utilization.
Following the literature (e.g., Katz and Murphy, 1992), we multiply topcoded earnings by 1.5, based on the assumption of a Pareto distribution for earnings above the topcode limit with Pareto parameter k=2. About 1.4 percent of the observations relevant to the calculations in Table 6 were adjusted in this way.
Given the evidence presented in Section 2, we interpret changes in value added price inflation in downsizing industries as due to changes in demand. It is also possible that low-cost imports and falling commodity prices put more downward pressure on gross output (as opposed to value added) prices in downsizing industries over 2000–03 than in the 1990s.
As shown in Section 2, correlations between real value added and the price deflator suggest that the difference in average deflator changes from 1990–2000 to 2000–03 primarily reflects a relative change in demand for the output of downsizing industries.
For an analysis of the large increases in productivity during this period, see Gordon [2003] and Oliner, Sichel, and Stiroh [2007].
Such a decomposition can be intuited from a simple two-sector model with sector-specific capital. Tapp [2007] takes a somewhat different approach to estimating the costs of restructuring. He calibrates a modified Mortensen and Pissarides [1994] style model to include industry-specific human capital and uses simulations from the model to calculate the welfare cost of a relative demand shock favoring resource-producing industries in Canada.
Note that, in our calculations we are not ignoring the potential benefits of reallocation to consumers who are made better off by the reallocation of resources toward the production of goods that they value more highly. We subtract these benefits (p n × mp k × k n T) when computing the net cost of reallocation.
This figure is the average difference in the price levels in the two sectors resulting from the estimated cumulative relative decline in prices in the restructuring sector (1.73 percent per year for three years).
As new specific human capital is acquired in the worker's new industry, earnings should move back up toward previous levels. However, this reacquisition of capital may take some time and considerable resources, and for some workers (for example older workers) the costs of acquiring this new capital may exceed the benefits. As a result, the earnings loss from destroyed specific capital may extend beyond 2003.
The total number of workers displaced in downsizing industries from 2001–03 was 5.1 million. Total employment (including self-employment and government workers) over these years averaged 136.2 million. Dividing 5.1 by 136.2 yields 3.74 percent. Assuming that that displacement rates occurred approximately uniformly across 2001–03, the average displacement rate per year was 1.25 percent and the average cumulative displacement rate was 2.5 percent. Multiplying each number by 0.67 (the share of displacements switching industries) yields 0.84 and 1.68, respectively.
To derive this estimate of labor's share, we add two-thirds of proprietor's income to aggregate compensation and divide by gross domestic income minus taxes on production and imports less subsidies, using average annual data for the year 2000.
In an environment where demand is increasing and worker displacement is limited, it is likely that only workers with relatively small amounts of specific capital will choose to find new employment outside of their original industries. This selection effect reduces differences in earnings changes between workers remaining in and workers switching out of downsizing industries relative to environments where demand is declining and displacement rates are high.
See, for example, Ramey and Shapiro [2001] and Phelan and Trejos [2000].
This calculation is derived using estimates of potential output and the output gap developed by the Congressional Budget Office.
References
Aaronson, Daniel, Rissman, Ellen R., and Sullivan, Daniel G. 2004. “Can Sectoral Reallocation Explain the Jobless Recovery?” Federal Reserve Bank of Chicago, Economic Perspectives, 28 (2): 36–49.
Abraham, Katharine G., and Katz, Lawrence F. 1986. “Cyclical Unemployment: Sectoral Shifts or Aggregate Disturbances?” Journal of Political Economy, 94 (3-1): 507–522.
Bahk, Byong-Hyong, and Gort, Michael 1993. “Decomposing Learning by Doing at New Plants.” Journal of Political Economy, 101 (4): 561–583.
Caballero, Ricardo, Hoshi, Takeo, and Kashyap, Anil 2006. “Zombie Lending and Depressed Restructuring in Japan,” NBER Working Paper no. 12129.
Corrado, Carol, Hulten, Charles, and Sichel, Daniel 2005. “Measuring Capital and Technology: An Expanded Framework,” in Measuring Capital in the New Economy, NBER Studies in Income and Wealth, Eds. Carrol Corrado, John Haltiwanger and Daniel Sichel, Vol. 65, Chicago and London: University of Chicago Press, 11–41.
Eisfeldt, Andrea L., and Rampini, Adriano A. 2003. “Capital Reallocation and Liquidity.” Journal of Monetary Economics, 53 (3): 369–399.
Groshen, Erica, and Potter, Simon 2003. “Has Structural Change Contributed to a Jobless Recovery?” Federal Reserve Bank of New York Current Issues in Economics and Finance, 9 (8): 1–7.
Gordon, Robert 2003. “Exploding Productivity Growth: Context, Causes, and Implications.” Brookings Papers on Economic Activity, 2: 207–279.
Hopenhayn, Hugo, and Rogerson, Richard 1993. “Job Turnover and Policy Evaluation: A General Equilibrium Analysis.” Journal of Political Economy, 101 (5): 915–938.
Katz, Lawrence F., and Murphy, Kevin M. 1992. “Changes in Relative Wages, 1967–87: Supply and Demand Factors.” Quarterly Journal of Economics, 107 (1): 35–78.
Kirkegaard, Jacob F. 2009. “Structural and Cyclical Trends in Net Employment over U.S. Business Cycles, 1949–2009: Implications for the Next Recovery and Beyond,” Working paper no. 09-5, Peterson Institute for International Economics.
Koenders, Kathryn, and Rogerson, Richard 2005. “Organizational Dynamics Over the Business Cycle: A View on Jobless Recoveries.” Federal Reserve Bank of St. Louis Review, 87 (4): 555–579.
Lilien, David M. 1982. “Sectoral Shifts and Cyclical Unemployment.” Journal of Political Economy, 90 (4): 777–793.
Loungani, Prakesh, and Rogerson, Richard 1989. “Cyclical Fluctuations and the Sectoral Reallocation of Labor: Evidence from the PSID.” Journal of Monetary Economics, 23 (2): 259–273.
Mann, Catherine L. 2003. “Globalization of IT Services and White Collar Jobs: The Next Wave of Productivity Growth,” Institute for International Economics International Economics Policy Briefs, Number PB03-11, December.
Mortensen, Dale, and Pissarides, Christopher 1994. “Job Creation and Job Destruction in the Theory of Unemployment.” Review of Economic Studies, 61: 397–415.
Neal, Derek 1995. “Industry-Specific Human Capital: Evidence from Displaced Workers.” Journal of Labor Economics, 13 (4): 653–677.
Oliner, Stephen, Sichel, Daniel, and Stiroh, Keven 2007. “Explaining a Productive Decade.” Brookings Papers on Economic Activity, 1: 81–137.
Phelan, Christopher, and Trejos, A. 2000. “The Aggregate Effects of Sectoral Allocations.” Journal of Monetary Economics, 45 (2): 249–268.
Phelps, Edmund 2008. “In Search of a More Dynamic Economy,” The Financial Times (July 21).
Rabin, Matthew 1998. “Psychology and Economics.” Journal of Economic Literature, 3 (1): 11–46.
Ramey, Valerie A., and Shapiro, Matthew D. 2001. “Displaced Capital: A Study of Aerospace Plant Closings.” Journal of Political Economy, 109 (5): 958–92.
Rapping, Leonard 1965. “Learning and World War II Production Functions.” Review of Economics and Statistics, 47 (1): 81–86.
Ruhm, Christopher 1991. “Are Workers Permanently Scarred by Job Displacements?” American Economic Review, 81 (1): 319–323.
Shimer, Robert 2005. “The Cyclicality of Hires, Separations, and Job-to-Job Transitions.” Federal Reserve Bank of St. Louis Review, 87 (4): 493–507.
Tapp, Stephen 2007. “Lost in Transition: The Costs and Consequences of Sectoral Labour Adjustment”, QED Working Paper 1192, November.
Topel, Robert 1991. “Specific Capital, Mobility and Wages: Wages Rise with Job Seniority.” Journal of Political Economy, 99 (1): 145–176.
Williams, John 2009. “How Big is the Output Gap?” Economic Letter, Federal Reserve Bank of San Francisco, Number 2009-19, June 12.
Acknowledgements
We express our appreciation to Leslie Carroll for excellent research assistance.
Additional information
The views expressed in this paper are the authors’ and do not necessarily represent those of the Federal Reserve Board.
*Andrew Figura is a group manager at the Federal Reserve Board, where he focuses on issues related to the household sector and the labor market. He started working at the Federal Reserve in 2000, after receiving a Ph.D. in economics from the University of Maryland. Prior to that, he worked as a Peace Corps volunteer, an economist at the Bureau of Labor Statistics, and a Foreign Service Officer.William Wascher is Senior Associate Director in the Division of Research and Statistics at the Federal Reserve Board, where he has oversight responsibilities for the forecast and analysis of the U.S. economy. He came to the Federal Reserve in 1983 and has also served as a senior economist with the President's Council of Economic Advisers and as a visiting economist with the Bank for International Settlements. His research covers a range of topics, including minimum wages and other labor market issues.
Rights and permissions
About this article
Cite this article
Figura, A., Wascher, W. The Causes and Consequences of Sectoral Reallocation: Evidence from the Early 21st Century. Bus Econ 45, 49–68 (2010). https://doi.org/10.1057/be.2009.42
Published:
Issue Date:
DOI: https://doi.org/10.1057/be.2009.42