Expert Commentary

Why do workers’ earnings stay lower after job displacement? A review of the major theories

2015 paper from researchers at the Congressional Budget Office examining various theories of earnings loss from job displacement.

The Great Recession of 2007-2009 left significant scars on the United States economy, from a slowdown in productivity to challenges in getting back to work and recovering wage growth. Among the myriad problems, job loss was among the most significant. The Bureau of Labor Statistics estimated that “at its lowest point, February 2010, U.S. employment had declined by 8.8 million from its prerecession peak.”

There continue to be many signs of general, long-term improvement: In a 2014 survey, the BLS found that 4.3 million people lost their jobs of three years or more between 2011 and 2013, decreasing from 6.1 million in 2009-2011. However, as the economy starts to recover, the job market is not the same as it was before the recession. Moreover, job displacement comes with significant wage loss and numerous other issues. Yet the precise mechanism for these losses is still unknown.

In a 2015 paper for the Congressional Budget Office (CBO), “Do We Know Why Earnings Fall with Job Displacement?” William J. Carrington of the CBO and Bruce C. Fallick of the Federal Reserve Bank of Cleveland examine various theories of earnings loss from job displacement. The authors consider several models and analyze whether the theoretical arguments can explain the empirical evidence observed, and whether the costs incurred are merely for individual workers (a “private” loss) or have wider “social” dimensions. They note that “after being displaced from their jobs, workers experience reduced earnings for many years” but the precise explanations for this remain “scattered.”

Their main findings include:

  • No one model seems to explain the complex data associated with this phenomenon: “A range of theories have something to say about why earnings fall for displaced workers, but little of the evidence is strongly differentiating and no theory provides a full explanation on its own.”
  • Often the most-cited theories of earnings loss, including work from Gary Becker of the University of Chicago, Walter Oi of the University of Rochester, and Jacob Mincer of Columbia University, focus on “job-specific human capital.” These theorize that job loss leads to lower earnings because of lower demand for job-specific attributes. This can be generalized more widely to include specific human capital by location, industry or occupation. These skills are refined during tenure at a firm and thus individuals experience reduced earnings when they lose the specific job and are left with only their general skills. These theories can explain empirical evidence that employees with high tenure see higher earnings loss, but not why some wages actually increase after job displacement.
  • In the “specific human capital” model, it is unclear whether losses are completely private in nature or if they have a social cost as well. This would depend on the cause of the loss: A technological advancement would lead to higher overall social benefit, but is a cost to the individual worker who is displaced. The development of an industry cartel, on the other hand, would carry both a social and private cost.
  • New York University’s Boyan Jovanovic’s theory of “job matching” contends that productivity can be derived from a skill or personal characteristic that makes for a better combination of worker and firm. Although the implications of a job loss in this scenario are similar to that of the specific human capital model, this model allows for the possibility of a worker finding a better organizational match once she leaves her job, and thus attaining higher earnings. Matching theories can explain the higher losses associated with tenure and why some workers gain higher wages after displacement, but not the persistent nature of wage loss.
  • Various others have presented models of “back-loaded compensation,” such as those from Edward Lazear of the Stanford Graduate School of Business. He speculates that employers, concerned about providing incentives for employee shirking, can delay compensation and cement an employer-employee bond. A permutation of the model from Steven Salop of Georgetown University and Joanne Salop for the Board of Governors of the Federal Reserve System is that back loaded compensation can even out high employee training costs in the beginning of their employment. There is less of a social cost for this model, assuming that workers could be just as productive at another job. These theories can explain the effect of tenure seen in the empirical evidence, but not the role of local market conditions.
  • Carrington and Fallick look at two models of “rent” and how employers share profits with workers. In one, employers pass these rents on to employees and in another, only some employers pay rents. Rent models can help explain the persistence of earnings loss, but “cannot easily explain wage gains upon displacement, the variation of earnings losses with ‘degrees’ of displacement, or, obviously, the presence of losses among workers who were likely not receiving pre-displacement rents.”
  • “Information revelation” theories contend that current employers have knowledge about their employees. Once a worker is displaced, this potentially negative information is signaled to future employers. Paul Oyer of the Stanford Graduate School of Business and Scott Schaefer of the University of Utah assert that employers may delay firing overpaid workers for some time and instead use collective layoffs, sending the message that these employees performed lower than their wages would indicate. Thus, the social cost of displacement is unclear, as there may be productivity gains to the employer once this employee is no longer there. This would depend on their productivity at their next job. These theories demonstrate the persistence of earnings loss seen in the evidence, but not the effect of tenure.
  • Some theories also consider the effect on a household, with many concluding that when one spouse is displaced, often the other works more to compensate. The extent of the resulting loss depends on the dynamics of the household as well as the availability of social insurance.
  • The experience of job displacement may also have implications on health. As there is not much evidence of the loss, Carrington and Fallick speculate that this is more likely to be a loss of intangible skills, such as confidence, than of hard skills.

Carrington and Fallick determine that although these theories can explain some aspects of why earnings fall after job displacement, no single theory explains it fully. “Understanding why earnings fall with displacement can be helpful in formulating policy responses,” the authors conclude, “much of the policy discussion focuses on the role of short-run policy interventions such as unemployment insurance. But there is clearly a longer-run issue.”

Related research: A 2014 research brief looks at the distress felt by displaced American workers and examines criticisms of unemployment benefits and social insurance.


Keywords: jobs, workers, unemployment, economy, Great Recession

About The Author