Why Do Earnings Fall with Job Displacement?

• Author: Bruce Fallick,William J. Carrington
• Date: 01 October 2017
• Published date: 01 October 2017
• DOI: http://doi.org/10.1111/irel.12192

Why Do Earnings Fall with Job Displacement?*

WILLIAM J. CARRINGTON and BRUCE FALLICK

Displaced workers experience reduced earnings for many years. While this empiri-

cal phenomenon is well established, the theory of displacement-induced earnings

loss is scattered. Policy discussion often interprets displacement-induced losses

through the lens of specific human capital theory but there are other credible theo-

ries with different causal mechanisms and different interpretations. This paper

reviews theories of costly job displacement and discusses their consistency with the

available empirical evidence for the United States. We find that specific human cap-

ital theory and matching theory have considerable but far from conclusive empirical

support. We suggest avenues for better discriminating among theories.

Introduction

The recession of 2007–2009 displaced millions of workers in the United

States and left them with the largest average earnings reductions since the

Bureau of Labor Statistics (BLS) began the Displaced Worker Survey in 1984

(BLS 2014). The BLS defines displaced workers as “persons 20 years of age

and older who lost or left jobs because their plant or company closed or

moved, there was insufficient work for them to do, or their position or shift

was abolished,”and we will stick closely to that definition in this review.

The theory most commonly associated with the earnings losses of displaced

workers is the specific human capital model of Gary Becker (1962) and Walter

Oi (1962). In that model, job loss reduces earnings because job-specific human

capital is lost with displacement or, more precisely, because demand for that

particular human capital declines. Similar reasoning has been applied to human

capital that may be specific to workers’locations, industries, or occupations—

broader dimensions along which labor demand may fall. Workers in these

models invest in specific human capital and the loss in the value of those

*The authors’affiliations are, respectively, Congressional Budget Office, Washington, DC. E-mail: william.

carrington@cbo.gov; and Federal Reserve Bank of Cleveland, Cleveland, Ohio. E-mail: Bruce.Fallick@

clev.frb.org. The views expressed here should no t be interpreted as those of the C ongressional Budget

Office, the Federal Reserv e Bank of Cleveland, or th e Board of Governors of the F ederal Reserve

System. The authors thank Pawe l Krolikowski of the Federal Res erve Bank of Cleveland, Xiaoto ng Niu

and Jeffrey Kling of CBO, an d two anonymous referees fo r helpful comments and sug gestions.

INDUSTRIAL RELATIONS, Vol. 56, No. 4 (October 2017). ©2017 Regents of the University of California

Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

688

investments that accompanies displacement is largely due (as far as the worker

is concerned) to bad luck. Because the loss is mostly due to bad luck, ex post

transfers from people whose investments paid off to people whose investments

did not pay off provide insurance to all and may encourage more investment

in socially productive but privately risky specific human capital (Acemoglu

and Shimer 2000). While the connection is rarely explicitly drawn, specific

human capital theory appears to supply much of the logic behind efforts to

assist displaced workers (e.g., von Wachter 2010), such as unemployment

insurance, the Trade Adjustment Assistance program, and proposals such as

wage insurance (Corak 2014; Kletzer and Rosen 2006; LaLonde 2007; Litan

2015).

However, other theories may explain why workers’earnings tend to fall

with displacement. In theories of matching and search, workers climb from job

to higher-earning job over time, and displacement can set them back in this

progression. Theories of backloaded compensation posit that wages rise with

job tenure because firms require that workers build up de facto bonds with the

firm through deferred compensation, and that tenure premium in current com-

pensation is lost upon displacement. Earnings fall with displacement in these

alternative models, but the mechanisms differ significantly from that of specific

human capital theory. Moreover, in models like those of backloaded compen-

sation the economic notion of what is lost may be only weakly tied to the

reduction in earnings as conventionally measured.

This paper reviews these and other theories of displaced workers’earnings

losses and considers each theory’s ability to explain a range of outcomes iden-

tified in the now-rich empirical literature on displaced workers in the United

States. The next section reviews broad classes of theories and examines their

causal mechanisms and distinctive empirical predictions. We then confront the

various theories with the empirical evidence on the experience of displaced

workers. We do not attempt to comprehensively review empirical studies,

either overall or since the reviews of Fallick (1996) and Kletzer (1998); rather,

our empirical review is narrowly focused on findings with implications for the

relative validity and applicability of the theories reviewed earlier. The final

sections summarize our findings and suggest some avenues for further research

that may help to further differentiate among alternative theories.

Theories of Costly Job Displacement

The idea that job loss is costly is almost as old as empirical labor eco-

nomics, and has been well established in the modern literature on displaced

workers (Couch and Placzek 2010; Jacobson, LaLonde, and Sullivan 1993;

Why Do Earnings Fall? / 689

Ruhm 1991; von Wachter, Song, and Manchester 2009).

1

The earliest work on

human capital earnings functions posited that a worker’s earnings rose with

education, with general labor market experience, and with the tenure on a

specific job (Becker 1962; Mincer 1962; Oi 1962). While it was not typically

emphasized at the time, an implication of those studies is that being forcibly

deprived of job tenure would entail a loss in earnings. That notion was initially

related to the theory of firm-specific human capital but over time a variety of

other interpretations arose, the most prominent of which are summarized in

Table 1. This section lays out the basic architecture of those theories and

points out some of their empirical implications.

TABLE 1

MODELS OF COSTLY JOB DISPLACEMENT

Model Source of Costly Job Loss

Representative Theoretical

Papers

Specific human capital Specific training or learning-by-doing creates

human capital that is only valuable at the current

employer or at a subset of firms or jobs.

Reduction in the rental price of that human

capital, to zero, entails an income loss

Becker (1962), Fallick

(1993), Mincer (1962),

Neal (1995), Shaw (1984)

Job matching Workers match to some jobs better than others Jovanovic (1979a)

Backloaded

compensation

High-tenure workers receive pay that exceeds

productivity as part of a tilted wage/tenure profile

that firms use either to constrain moral hazard or

to attract more stable workers

Lazear (1981), Salop and

Salop (1976)

Rents Workers receiving rents—because of unionism,

threat of unionism, or rent-sharing by product

market monopolists—lose those rents upon

displacement

Hildreth and Oswald (1997)

Revelation of

information

Displacement reveals to potential new employers

that the displaced worker is low quality on some

dimension, thereby reducing their prospective

earnings

Gibbons and Katz (1991),

Oyer and Schaefer (2000)

Intrahousehold

reallocation

Families substitute toward work performed by

nondisplaced family members

Lundberg (1985)

Health Job displacement reduces mental and physical

health, which in turn reduces earnings

Kessler, House, and Turner

(1987)

Peer effects Workers are often displaced in bunches, and

workers that lose their job and their job contacts

(who are also displaced) may see their long-term

opportunities fall

Calvo-Armengol and Jackson

(2007), Granovetter (1995)

1

More recently, however, a few studies have called into question the notion that job loss itself leads to

large and persistent earnings losses. See Fallick, Haltiwanger, and McEntarfer (2012); Flaaen, Shapiro, and

Sorkin (2013); and Cooper (2013), whose findings suggest that presence or absence of displacement, at least

as conventionally identified in the data, is less important for earnings losses than the presence of an unem-

ployment or nonemployment spell following separation.

690 / WILLIAM J. CARRINGTON AND BRUCE FALLICK