Boarding a Sinking Ship? An Investigation of Job Applications to Distressed Firms

• Author: JENNIFER BROWN,DAVID A. MATSA
• Date: 01 April 2016
• Published date: 01 April 2016
• DOI: http://doi.org/10.1111/jofi.12367

THE JOURNAL OF FINANCE •VOL. LXXI, NO. 2 •APRIL 2016

Boarding a Sinking Ship? An Investigation of Job

Applications to Distressed Firms

JENNIFER BROWN and DAVID A. MATSA∗

ABSTRACT

We use novel data from a leading online job search platform to examine the impact

of corporate distress on firms’ ability to attract job applicants. Survey responses

suggest that job seekers accurately perceive firms’ financial condition, as measured

by companies’ credit default swap prices and accounting data. Analyzing responses

to job postings by major financial firms during the Great Recession, we find that an

increase in an employer’s distress results in fewer and lower quality applicants. These

effects are particularly evident when the social safety net provides workers with weak

protection against unemployment and for positions requiring a college education.

ACOMPANY’S FINANCIAL CONDITION has far-reaching effects on the firm, including

on its ability to attract and retain human capital. Corporate distress can lead

current employees to search for more stable positions and new recruits to focus

their searches elsewhere. Concerns about short-run solvency strain a firm’s

reputation for treating employees fairly (Maksimovic and Titman (1991)), and

corporate distress often leads firms to pay lower wages (Graham et al. (2013))

and to downsize their workforce. Averageemployment decreases by 10% around

a bond covenant violation (Falato and Liang (2014)), by 27% around a bond

default (Agrawal and Matsa (2013)), and by 50% or more around a bankruptcy

filing (Hotchkiss (1995)).

∗Jennifer Brown and David A. Matsa are at the Kellogg School of Management, Northwestern

University, and NBER. We are grateful to the research and data management teams at the com-

pany that provided the data for this research. For helpful comments, we also thank Emek Basker,

Bo Becker, Effi Benmelech, Meghan Busse, Jason Faberman, Craig Furfine, Craig Garthwaite,

Todd Gormley, Edith Hotchkiss, Hyunseob Kim, Danielle Li, Brian Melzer, Dylan Minor, Her-

nan Ortiz-Molina, Josh Rauh, Michael Roberts (the Editor), Antoinette Schoar, Aaron Sojourner,

Chris Stanton, Amir Sufi, an Associate Editor, and two anonymous referees, as well as seminar

participants at the Bank of Portugal, Federal Reserve Bank of Chicago, HEC Montreal, MIT,

Michigan State University,National University of Singapore, Shanghai University of Finance and

Economics, Tsinghua University, University of Alberta, University of British Columbia, Univer-

sity of Georgia, University of Maryland, University of North Carolina, University of Notre Dame,

University of Southern California, University of Toronto,University of Utah, American Finance As-

sociation Meeting, Bank of Portugal Conference on the Labor Market, CSEF Conference on Finance

and Labor,European Winter Finance Conference, LMU Munich Workshop on Natural Experiments

and Controlled Field Studies, NBER Corporate Finance Program, University of Michigan Mitsui

Finance Symposium, University of Tokyo and RIETI International Workshop on Personnel Eco-

nomics, Western Finance Association Meeting, and Utah Winter Business Economics Conference.

The authors have no conflicts of interest, as identified in the Journal of Finance’s disclosure policy.

DOI: 10.1111/jofi.12367

507

508 The Journal of Finance R

Job loss can be extremely costly for workers and, as a result, workers are

likely to avoid distressed firms.1Unless distressed firms can offer a sufficient

compensating wage premium, they may have difficulty recruiting new talent,

particularly for positions that require firm-specific investments. Although such

costs of distress feature prominently in theoretical explanations of firms’ capital

structure decisions (Titman (1984), Berk, Stanton, and Zechner (2010)), their

empirical relevance is less clear.

Empirical evidence is sparse in part because it is challenging to separately

identify the effects of corporate distress on demand and supply in the market

for labor. Distress often reduces a firm’s scale and labor demand, while also

making the firm less attractive to workers and reducing labor supply. With

data only on employment or wages, it is impossible to separate these channels

empirically. This paper exploits several novel data sets from a large online job

search platform to overcome this identification challenge.2With microdata on

job applications, we hold demand fixed and examine how the supply of workers

to specific jobs at individual firms is affected by firms’ financial condition.

We first examine survey responses of job seekers on the online platform to

evaluate job seekers’ perception of firms’ financial condition. We match job seek-

ers’ assessments of 145 firms between October 2008 and March 2010 with in-

dicators of financial strength, including daily credit default swap (CDS) prices.

Across the various measures, we consistently find that job seekers’ perception

is highly positively correlated with firms’ true financial condition.

Our second analysis examines whether applicants act on these perceptions.

We exploit a second proprietary data set that describes all of the jobs posted by

40 high-profile financial services firms to the job search platform between April

2008 and December 2009.3Similar to Holzer, Katz, and Krueger (1991), we

examine the volume of applications to open positions as a measure of workers’

interest in these jobs. We match the application data with firms’ CDS prices

to assess the relationship between a firm’s financial condition and workers’

willingness to apply.

During the Great Recession, corporate financial conditions varied substan-

tially both between and within financial firms over time. Our identification

strategy exploits these changes. The richness of the data allows us to develop

1Workers face reductions in consumption during unemployment and less stability in earnings

and employment even 10 years later (Gibbons and Katz (1991), Jacobson, LaLonde, and Sullivan

(1993), Gruber (1997), Sullivan and von Wachter (2009)). With job displacement also leading

to worse health outcomes, higher mortality, lower achievement by children, and other adverse

consequences (see Davis and von Wachter (2011) for a survey), even the prospect of job insecurity

imposes sizeable psychological costs on workers (Sverke and Hellgren (2002)).

2We were provided the data under a nondisclosure agreement, which restricts us from identi-

fying the online platform. This agreement places no constraints on the conclusions of the analysis.

3Our data provider requested that we restrict the application sample to one industry and a

limited period of time. We chose the financial sector and this 21-month period because the appli-

cation data have strong coverage, many firms have traded CDS contracts, and firms experienced

various degrees of distress in this period—some sample firms experienced extreme distress, some

less extreme, and some none at all. The survey results show that job seekers in other industries

are aware of firms’ financial conditions as well.

Boarding a Sinking Ship? 509

a compelling counterfactual in our analysis. We compare job seekers’ interest

in a given posting to their interest in other postings for the same type of job

in the same geographic area and the same month, and we test whether the

firm’s credit default risk at the time of the posting is related to the number

of applications.4The detailed fixed effects rule out many possible alternative

interpretations, such as a decline in the social status of finance jobs or changes

in unobserved local industry- and occupation-specific investment opportunities

or labor market conditions. All of our analyses also control for firm fixed effects

so that we are measuring only time-varying relative differences in the supply

of labor to each employer.

We find that firms attract significantly fewer applications per job opening

during periods of corporate distress. On average, a 1,000 basis point increase

in a firm’s CDS price is associated with about 20% fewer applications to a given

position.5The results are not driven by outliers and are robust to alternative

measures of firms’ financial position. The reduction in applications occurs de-

spite the fact that job seekers typically search more broadly and submit more

applications in tight labor markets (Kuhnen (2011)). Given the low cost of sub-

mitting an online application to these jobs, we would expect to find even larger

effects on potential applicants’ propensity to accept a job offer at a distressed

firm.6

Analyses of potential mechanisms confirm that the decline in applications

cannot be fully explained by shifts in labor demand. We find no evidence that

positions that firms post during periods of distress draw from smaller pools

of potential applicants or pay lower salaries. If anything, advertised salaries

increase when a firm is in distress.

Heterogeneity in the effect along two dimensions further supports the inter-

pretation that firms’ distress affects labor supply—that is, holding fixed other

job-specific and labor market-specific characteristics, a worker prefers to work

for a firm that is in a stronger financial position. First, we exploit state-level

variation in unemployment insurance (UI) and find that workers are less sensi-

tive to potential employers’ distress in locations in which unemployment costs

are lessened by a stronger social safety net. Second, we find that large upfront

costs make job seekers reluctant to apply to distressed firms. Specifically, we

find that distressed firms receive a lower proportion of job applications from

out-of-state applicants—individuals for whom new employment is typically

particularly costly and disruptive.

4We are not suggesting that job seekers track individual firms’ actual CDS price over time.

After establishing that job seekers’ perception of a firm’s financial stability is correlated with the

firm’s CDS price, we use this price to proxy for the firm’s (actual and perceived) overall financial

condition.

5A 1,000 basis point increase in a firm’s CDS price is large but not unusual during our sample

period. The increase is associated with a five-year cumulative default rate increase of about 17 per-

centage points (see footnote 21). Among all firms in the financial sector with total assets greater

than $25 billion for which CDS prices are available, 20% experienced CDS price swings exceeding

1,000 basis points during the Great Recession.

6Our data provider does not observe who is interviewed or hired for the open positions.