This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
Author's personal copy
Journal of Financial Economics 110 (2013) 478–502
Contents lists available at ScienceDirect
Journal of Financial Economics
journal homepage: www.elsevier.com/locate/jfec
Human capital, capital structure, and employee pay:
An empirical analysis$
Thomas J. Chemmanur a,n, Yingmei Cheng b, Tianming Zhang c
a
Boston College, Carroll School of Management, 440 Fulton Hall, Chestnut Hill, MA 02467, USA
The Florida State University, Department of Finance, USA
c
The Florida State University, Department of Accounting, USA
b
a r t i c l e in f o
abstract
Article history:
Received 26 June 2010
Received in revised form
28 November 2012
Accepted 4 December 2012
Available online 1 August 2013
We test the predictions of Titman (1984) and Berk, Stanton, and Zechner (2010) by
examining the effect of leverage on labor costs. Leverage has a significantly positive
impact on cash, equity-based, and total compensation of chief executive officers (CEOs).
Compensation of new CEOs hired from outside the firm is positively related to prior-year
firm leverage. In addition, leverage has a positive and significant impact on average
employee pay. The incremental total labor expenses associated with an increase in
leverage are large enough to offset the incremental tax benefits of debt. The empirical
evidence supports the theoretical prediction that labor costs limit the use of debt.
& 2013 Elsevier B.V. All rights reserved.
JEL classification:
G32
Keywords:
Capital structure
Human capital
Labor costs
1. Introduction
The trade-off theory of capital structure points to
bankruptcy costs as the main reason that firms in many
industries do not assume higher levels of leverage to take
advantage of the corporate tax saving benefits of debt.
☆
For helpful comments and discussions, we thank Jonathan Berk, John
Graham, Bing He, Michael Roberts, Zacharias Sautner, and participants in
conference presentations at the American Accounting Association Northeast Region meeting (Best Paper Award), the Center for Research in
Security Prices Forum at the University of Chicago, the Financial Management Association meetings, the Financial Management Association European meetings, the Western Finance Association meetings, and the
American Finance Association meetings. Special thanks to an anonymous
referee and the editor, Bill Schwert, for helpful comments and suggestions that greatly improved the paper. We also appreciate comments and
suggestions from seminar participants at the University of Massachusetts
at Amherst, the University of Texas at Arlington, Boston College, Florida
State University, University of Florida, Qinghua University, and Zhejiang
Business University.
n
Corresponding author. Tel.: +1 617 552 3980; fax: +1 617 552 0431.
E-mail address: chemmanu@bc.edu (T.J. Chemmanur).
0304-405X/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfineco.2013.07.003
However, considerable empirical evidence indicates that
the magnitude of direct bankruptcy costs is too low to be
a sufficient disincentive preventing firms from taking
on higher levels of debt. Some authors have, therefore,
suggested indirect bankruptcy costs as a solution to the
puzzle of the observed underleveraging of firms in many
industries. In an important paper, Titman (1984) develops
a model in which a firm's liquidation decision is causally
linked to its bankruptcy status. He argues that customers,
workers, and suppliers of firms that produce unique or
specialized products are likely to suffer high costs in
the event of liquidation. In particular, in a setting where
employees have firm-specific human capital, the fact that
bankruptcy can impose significant costs on employees (by
reducing the value of their human capital) can significantly
affect firms' capital structures.1 Formalizing the Titman
(1984) arguments, Berk, Stanton, and Zechner (2010; BSZ
(2010) hereafter) develop a model incorporating the idea
1
For an excellent review of empirical research on capital structure,
see Parsons and Titman (2008).
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
that human capital costs associated with financial distress
and bankruptcy could be large enough to be a disincentive
for firms to issue debt.
The objective of this paper is to empirically analyze, for
the first time in the literature, whether human capital
costs are an important determinant of the capital structure
of firms as postulated by the theoretical literature. We do
this by examining the relation between the observed
capital structures of firms and the compensation of their
chief executive officers (CEOs), as well as the relation
between observed capital structures and the average
wages of their work forces. While we use CEO compensation to measure the pay of a critical employee, we use the
average employee wage to measure the compensation of a
collective employee. In the model of BSZ (2010), each firm
faces a risk-averse employee and risk-neutral investors.
In the optimal labor contract between firms and employees, a firm with higher leverage pays a higher wage to its
employee to compensate him for the expected bankruptcy
costs that will be borne by the employee, because the
employee is unable to fully insure his human capital risk.
Firms, therefore, choose not to increase leverage beyond
the point where the marginal tax benefits of debt are
offset by the incremental labor costs associated with
higher levels of debt. The empirical implication here is
that, in the cross section, firms with higher leverage are
associated with higher employee pay.2 We test this prediction (“the Titman-BSZ prediction”) in our empirical
analysis. We also study whether the magnitude of the
additional compensation associated with an increase in
leverage is large enough to at least partially explain the
underleveraging of firms.
In contrast to the theories that focus on the ex ante
relation between leverage and employee pay, Perotti and
Spier (1993) focus on the ex post effect of leverage on
employee pay.3 In particular, they argue that firms are able
to use leverage strategically when current profits are low
and future investment is necessary to guarantee full
payment of the union's claim (wages). By retiring equity
through a junior debt issue, shareholders can credibly
threaten not to undertake valuable new investments
unless the union agrees to wage reductions. The implication of the argument is that, under suitable conditions,
firms with high leverage are associated with lower
employee pay.
The ex post relation between leverage and employee
pay implied by the model of Perotti and Spier (1993),
however, is not inconsistent with the ex ante relation
between the same variables in the Titman-BSZ prediction.
As Perotti and Spier (1993) point out, if workers anticipate
that equity holders could attempt to use higher leverage
to negotiate their wages downward ex post, they will
demand higher expected wages ex ante to compensate
them for bearing this risk. Perotti and Spier (1993) also
2
The models of Jaggia and Thakor (1994) and Berkovitch, Israel, and
Spiegel (2000) also have somewhat similar predictions.
3
Several other papers make similar arguments. See, e.g., Baldwin
(1983), Bronars and Deere (1991), Perotti and Spier (1993), Dasgupta and
Sengupta (1993), Hennessy and Livdan (2009), and Brown, Fee, and
Thomas (2009).
479
point out that a firm will not be able to use leverage as a
bargaining tool to reduce employee wages if their profits
from existing assets are large (i.e., the firm does not face
a significant probability of financial distress). We make
use of these results to empirically disentangle the ex ante
effects suggested by the Titman-BSZ prediction from the
ex post effects suggested by Perotti and Spier (1993).
We accomplish this by splitting our sample between firms
approaching financial distress (distressed firms) and those
that do not face a significant probability of distress (safe
firms).
We find that the debt ratio of a firm positively affects
the magnitude of its CEO compensation. Firms with higher
leverage pay their CEOs more, in terms of total compensation, cash pay, and equity-based pay. In our ordinary least
squares (OLS) regressions, an increase in market leverage
by one standard deviation is associated with an increase of
more than 8% in CEO total compensation, a magnitude that
is economically significant. We recognize that unobserved
CEO characteristics could influence firm leverage as well as
CEO pay, so that the direction of causality can be ambiguous. For example, CEOs who have had more interaction
with the board (and, therefore, have more influence) could
have greater ability to affect their own pay and at the same
time choose the firm's leverage level. To address this issue,
we study the relation between the first-year compensation
of newly appointed CEOs who are hired from outside and
firm leverage in the year prior to their appointment.
Clearly, newly appointed CEOs who are hired from outside
should have no influence on the firm's leverage in the year
prior to their appointment. We show that, even in the case
of new CEOs hired from the outside, compensation is
positively related to leverage.
We also find that leverage has a positive and significant
impact on average employee pay. Further, the incremental
labor expenses associated with an increase in leverage are
large enough to offset all of the incremental tax benefits
arising from such an increase. For a firm with median
values of leverage, average employee pay, total labor
expenses, and total debt, if the market leverage ratio
increases by one standard deviation, total labor expenses
increase by $14.01 million, holding the number of employees constant. Assuming 6% as the average return on debt in
our sample from 1992 to 2006 and assuming a tax rate of
35%, the tax benefits of debt increase by $5.09 million,
smaller than the increase in total labor expenses of $14.01
million. This supports the hypothesis that the incremental
labor costs associated with an increase in leverage are
economically significant and large enough in magnitude to
limit the use of debt.
One potential concern with our baseline analysis is the
endogeneity of leverage. In particular, the assets of a given
firm could be such that they can support a high level of
leverage (for example, the proportion of tangible assets
could be high) and could also require highly paid employees to operate these assets, thus generating a positive
correlation between leverage and employee pay. To deal
with this potential endogeneity problem, we employ an
instrumental variable, namely, the marginal corporate tax
rate, to generate an exogenous variation in leverage.
The theoretical literature in corporate finance suggests
Author's personal copy
480
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
that the tax benefit of debt is positively related to a firm's
marginal tax rate, thus resulting in a positive correlation
between a firm's marginal tax rate and its leverage ratio.
The empirical literature also supports the positive relation
between marginal tax rate and leverage (see, e.g., Leary
and Roberts, 2010). At the same time, no theoretical or
empirical literature indicates that the marginal corporate
tax rate directly affects employee pay. Using the marginal
corporate tax rate as the instrument, we study the relation
between leverage and average employee pay in a twostage least squares (2SLS) regression framework. The
results confirm that, even after accounting for the potential endogeneity of leverage, firms with a higher level of
leverage are associated with a higher level of average
employee pay.4
Using the sample of manufacturing firms in the US over
1974–1982, Titman and Wessels (1988) find that firms
with more specialized labor have lower debt ratios.
Because more specialized workers are paid more, this
suggests a negative relation between leverage and wages.
If labor specialization is related to both leverage and
employee pay, the omission of labor specialization from
the regression of employee pay could cause a bias in the
estimated coefficient of leverage. We address this issue by
examining the quits rate, the percentage of the industry's
total work force that voluntarily left their jobs in the
sample years. Following Titman and Wessels (1988), we
use quits rate as our proxy for labor specialization. A lower
quits rate corresponds to greater labor specialization.
We find that the quits rate is negatively correlated with
average employee pay, consistent with the notion that
more specialized workers are paid more. However, we find
that the correlation between leverage and the quits rate is
not statistically significant. Furthermore, in our multivariate regression of average employee pay in which the quits
rate is included as an explanatory variable, the quits rate is
insignificant, and the coefficient of leverage remains positive and significant.
We also empirically disentangle the ex ante relation
between leverage and employee pay from the ex post
effects suggested by Perotti and Spier (1993). To accomplish this, we split our sample based on each firm's Altman
Z-score and study safe and distressed firms separately.
Consistent with the Titman-BSZ prediction, the relation
between leverage and average employee pay is positive
and significant in the sample of safe firms. Meanwhile, the
coefficient of leverage is negative in the distressed sample,
but not statistically significant. This suggests that, while
the ex ante relation between leverage and employee pay
suggested by Titman-BSZ prediction dominates in our
entire sample and in the subsample of safe firms, in
distressed firms the ex post relation postulated by Perotti
and Spier (1993) could partially or fully offset the effect of
firms compensating employees for their human capital
risk due to higher leverage. This is not surprising, because
it is precisely in distressed firms that we expect the ability
4
However, the instrumental variable that we use for leverage, the
marginal tax rate, has some limitations. We discuss these limitations in
Section 8.
of firms to use leverage as a bargaining tool with employees to be the strongest [as pointed out by Perotti and Spier
(1993)].
Labor expenses, which we use to compute average
employee pay, are missing for a number of firms in the
Compustat database. This creates a potential sample-selection
bias if firms selectively decide whether or not to report this
information. To adjust for this potential selection bias, we
adopt a Heckman (1979) two-step analysis. Our results are
robust to the Heckman procedure. The second stage of our
Heckman two-step analysis indicates that, even after controlling for potential sample selection, leverage has a positive
effect on average employee pay.
Employee entrenchment is an important element in the
model of BSZ (2010). Entrenchment in their model means
that employees are unable to fully insure their human
capital risk. BSZ (2010) argue that employee entrenchment
is the reason that an employee demands a higher wage
from a firm with higher leverage. This allows us to conduct
yet another test of the Titman-BSZ prediction. We expect
to observe a stronger effect of leverage on labor costs
when the employee is more entrenched. To empirically
test the effect of employee entrenchment on the leveragewage relation, we examine technology versus nontechnology firms. Existing evidence (e.g., Anderson, Banker, and
Ravindran, 2000) suggests that employees in nontechnology firms are more entrenched than in technology firms
(in the sense that the potential reduction in employees'
human capital if their firm goes bankrupt is greater). Given
this, the impact of leverage on employee compensation in
nontechnology firms can be expected to be greater than in
technology firms. We, therefore, split our sample between
technology and nontechnology firms and conduct our
analysis separately on these two subsamples.
We find that the influence of leverage on the cash, equitybased, and total compensation of CEOs is positive and
significant in nontechnology firms. In technology firms,
leverage affects the cash pay of CEOs, but it does not have
significant effects on their total or equity-based compensation. The leverage ratio also has a positive and significant
effect on average employee pay in nontechnology firms, but
not in technology firms. Thus, the effect of leverage on CEO
compensation as well as on average employee pay is greater
for nontechnology firms than for technology firms, consistent with the Titman-BSZ prediction.
Our paper is related to the empirical literature examining
the notion that leverage could serve as a bargaining tool for
firms against labor and could thereby have a disciplining
effect on labor. See, e.g., Benmelech, Bergman, and Enriquez
(2009), who show that airlines in financial distress obtain
wage concessions from employees whose pension plans are
underfunded; Matsa (2010), who finds that firms characterized by greater union bargaining power use greater leverage;
and Hanka (1998), who shows that firms using higher levels
of debt reduce employment more often and use more parttime or seasonal employees. Our empirical results do not
necessarily contradict those of the above cited literature.
As pointed out by Perotti and Spier (1993), the disciplining
effect of debt on labor is greater in firms with a significant
chance of financial distress and can coexist with employees
demanding greater wages ex ante (to induce them to join
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
481
firms with greater leverage ratios). These greater wages
could be required not only to compensate employees for
the potential loss of their human capital in the event that the
firm goes bankrupt [as suggested by Titman (1984) and BSZ
(2010)], but also for the potential reduction in wages or other
benefits arising from their lower bargaining power ex post if
the firm enters financial distress subsequent to their joining
it. The fact that the positive relation we find between
leverage ratios and employee pay arises mostly from the
subsample of safe firms (in which the disciplining effects of
debt on the employment relation is likely to be the least),
suggests that both of the effects could be operating in
employee–firm relations in practice.5
Our paper contributes to the literature by showing, for
the first time, that leverage has a positive impact on
employee compensation (as measured by either CEO
compensation or average employee pay) and that, at the
existing median debt level, the incremental labor costs
associated with an increase in leverage are sufficient to
offset the incremental tax benefits of debt. Our study helps
to establish the importance of labor costs in capital
structure decisions and, thus, advance our understanding
of the determinants of corporate leverage. Finally, ours
is the first paper that explicitly analyzes the relation
between executive compensation and capital structure.
While a large prior literature exists on executive compensation as reviewed by, e.g., Frydman and Jenter (2010), to
our best knowledge, no prior research has empirically
analyzed the relation between executive compensation
and capital structure.
The rest of this paper is organized as follows. Section 2
reviews the relevant theory in more detail and develops
testable hypotheses. Section 3 describes our data and
sample selection procedures. Section 4 presents our
empirical analysis of the relation between capital structure
and CEO compensation. Section 5 presents our empirical
analysis of the relation between capital structure and
average employee pay. Section 6 compares our empirical
results for technology versus nontechnology firms. Section
7 presents some additional robustness tests. Section 8
summarizes our results, discusses the limitations of our
instrumental variable analysis, and concludes.
of their human capital) can significantly affect firms'
capital structures. The model of BSZ (2010) formalizes
the arguments of Titman (1984). In their model, each firm
has only one employee, who is risk averse; investors in the
firm are risk neutral. The employee is averse to bearing his
own human capital risk. It is also assumed that the firm
operates in competitive capital and labor markets. If the
firm is in financial distress, the employee has to take a pay
cut to ensure full repayment of debt. Further, if the firm is
forced into bankruptcy, the employee could be terminated.
Therefore, the employee faces substantial costs in the
event of financial distress and bankruptcy. Because a
higher debt level implies a higher probability of bankruptcy and the employee is unable to insure fully his
human capital risk, firms with higher leverage have to pay,
in equilibrium, a higher wage to the employee to compensate him for the expected bankruptcy costs borne
by him.
We make use of two measures of labor costs to test the
above theories: CEO compensation and average employee
pay. CEO compensation measures the pay of the most
important employee. In the model of BSZ (2010), there is
only one employee per firm. A company's CEO plays a
critical role in affecting corporate performance, and his
productivity is more difficult to evaluate than that of lower
level employees. Therefore, the single employee in the
model of BSZ (2010) can be best interpreted as the CEO
in empirical tests. Average employee pay measures the
compensation of a collective employee. Because average
employee pay is calculated as total labor expenses divided
by the number of employees, we are able to use this
measure to directly derive the marginal impact of leverage
on total labor expenses and, therefore, to compare the
marginal effect of debt on labor costs with the incremental
tax benefits of debt.
Based on the implications of the theoretical models
discussed above and using the above test variables, we
have the following testable hypotheses.
2. Development of hypotheses
Hypothesis 3. At the existing debt level, the additional
labor costs associated with an increase in leverage are
large enough to offset the incremental tax benefits of debt.
Titman (1984) develops a model in which a firm's
liquidation decision is causally linked to its bankruptcy
status. He argues that customers, workers, and suppliers of
firms that produce unique or specialized products are
likely to suffer high costs in the event of liquidation. In
particular, in a setting where employees have firm-specific
human capital, the fact that bankruptcy can impose
significant costs on employees (through reducing the value
5
Our paper is also broadly related to the large literature studying the
factors that could contribute to the apparent underleveraging of firms.
See, e.g., Graham and Tucker (2006), who find that tax sheltering
activities help to explain the low debt ratio of the firms in their sample.
The literature on the role of human capital in asset pricing is also
indirectly related. See, e.g., Fama and Schwert (1977).
Hypothesis 1. Firms with higher leverage will incur larger
CEO compensation.
Hypothesis 2. Firms with higher leverage will incur larger
average employee pay.
Perotti and Spier (1993) argue that labor unions will
bargain less aggressively and could be more willing to take
pay cuts if highly levered firms run a greater risk of
bankruptcy. Although their model implies that workers,
ex ante, will demand a higher expected wage in compensation for bearing the risk (Proposition IV of their paper),
another empirical implication of their theory is that, ex
post, a negative correlation will exist between leverage
and wage when a firm faces substantial financial distress.
Thus we have the following testable hypothesis.
Hypothesis 4. Firms with higher leverage will incur lower
average employee pay when they are in financial distress.
Author's personal copy
482
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
One important element of the model of BSZ (2010) is
the degree of job entrenchment. Different from the same
term used in the literature on corporate governance,
entrenchment in this context means the degree to which
employees are able to insure their human capital risk
(lower their ability to insure, greater the extent of
entrenchment). Job entrenchment in this sense is the
reason that the employee demands a higher pay from a
firm with higher leverage in BSZ (2010). To empirically
study the impact of employee job entrenchment on the
leverage-wage relation, we examine technology versus
nontechnology firms. Evidence suggests that employees
in nontechnology firms are more entrenched compared
with those in technology firms.6 Given this, we expect
leverage to have a stronger impact on labor costs in
nontechnology firms than in technology firms. This yields
our fifth testable hypothesis.
Hypothesis 5. The effect of leverage on CEO compensation
as well as on average employee pay will be greater in
nontechnology firms than in technology firms.
3. Data and summary statistics
In this section, we provide details of the sample
selection and the preliminary summary of the variables
in the sample.
outside or promoted from inside.7 We identify 373 outside
hires using this methodology.
3.2. Sample of average employee pay
We use information from the Compustat Industrial
Annual database between 1992 and 2006 to study the
impact of leverage on average employee pay.8 We exclude
financial and utilities companies, and we exclude firms
with fewer than one hundred employees. We also drop
firms with nonpositive book values of equity. We calculate
average employee pay as total labor expenses divided
by the number of employees. Compustat provides “labor
and related expenses” (data item 42) and the number of
employees (data item 29). According to the Compustat
data manual, data item 42 includes salaries and wages,
pension costs, payroll taxes, incentive compensation, profit
sharing, and other benefit plans. Data item 42 thus
represents a firm's total labor expenses. This suits our
purpose, as we need to estimate the impact of leverage on
total labor costs. About 10% of firms recorded in the
Compustat have valid information on data item 42. This
could introduce a sample-selection bias (see Section 5).
There are 5,269 firm-year observations that have the
necessary information to be included in our OLS regression
of average employee pay.
3.3. Other data sources
3.1. Sample of CEO compensation
We gather information on CEO pay from the ExecuComp database. It provides detailed information on the
compensation of the top five executives of Standard &
Poor's (S&P) 1,500 firms since 1992. We focus on the CEOs.
We merge ExecuComp with the Compustat Industrial
Annual database from 1992 to 2006. We delete firms with
nonpositive book value of equity and exclude financial and
utilities companies. A total of 17,173 firm-year observations
satisfy these criteria, and 14,891 observations have all the
necessary information to be included in our OLS regressions of CEO compensation.
During our sample period (1992–2006), there are 1,952
new CEOs. To determine whether a new CEO is an outside
hire, we use the following two-step procedure. First, we
search for his previous employer in the ExecuComp
database. If his prior employer is not the same as the
current firm, then he is an outside hire. Second, if we
cannot identify his previous employer in the ExecuComp
database (ExecuComp reports information only on the top
five executives in S&P 1,500 firms), we search the Lexis
Nexis Academic Universe by the name of the executive and
of the company to determine whether he is hired from
6
Anderson, Banker, and Ravindran (2000) show that the demand for
executives and other critical employees in technology firms is intense,
leading to higher employee turnovers than in nontechnology firms.
Ittner, Lambert, and Larcker (2003), using proprietary compensation
survey data, find that technology firms rank “employee retention
objectives” as the most important goal of their equity grant program.
Overall, this evidence indicates that employees in technology firms suffer
a lower loss of human capital if their firms enter financial distress
compared with those in nontechnology firms.
We obtain quits rates from the database of Job Openings and Labor Turnover Survey (JOLTS) provided by the
U.S. Bureau of Labor Statistics. The quits rate is the number
of quits (voluntary separations) during the entire year as a
percent of annual average employment. The data are
available at the industry level from 2001. The industry
classification is based on the North American Industry
Classification System (NAICS). Appendix Table A1 reports
annual quits rates by industry and year.
Corporate governance could play a role in CEO compensation, and it could also matter in determining average
employee pay.9 Therefore, we examine whether corporate
governance is a factor in determining average employee pay
and CEO compensation. We use the G-Index constructed by
Gompers, Ishii, and Metrick (2003) as a measure of corporate governance. They compute the G-Index using a total of
24 possible antitakeover provisions. The data source is the
Investor Responsibility Research Center (IRRC) database,
which provides annual information on corporate antitakeover provisions for the years 1990, 1993, 1995, 1998, 2000,
2002, 2004, 2006, and 2008. We fill in observations in the
missing years using information from the most recent year.
7
Lexis-Nexis Academic Universe provides comprehensive information contained in major US and world publications (including Wall Street
Journal, New York Times, Washington Post, USA Today, among many
others), Securities and Exchange Commission filings, news wire services,
web publications, TV and radio broadcast transcripts, major company
profiles and reports, court cases, law reviews, and even blogs.
8
This ensures that our samples of CEO compensation and employee
pay cover the same time period.
9
Cronqvist, Heyman, Nilsson, Svaleryd, and Vlachos (2009) find that
CEOs with more control pay their workers more.
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
For example, we use information from 2004 for year 2005.
A greater value of the G-Index corresponds to weaker
shareholder rights and stronger managerial power.
Throughout our empirical analysis of both CEO compensation and average employee pay, all dollar amounts
are adjusted to 1992 dollars using the consumer price
index (CPI).10 We use the Fama and French 48-industry
classification to categorize firms into their respective
industries (the classification is obtained from Kenneth
French's website).11
4. Empirical tests and results on capital structure and
CEO compensation
In this section, we describe our empirical tests of the
impact of leverage on the magnitude of CEO compensation. We start with OLS regressions of CEO compensation
in the whole sample. We then perform additional tests to
identify causality. To accomplish this, we examine the
impact of leverage in the prior year on the compensation
of newly appointed CEOs who are hired from outside.
4.1. Summary statistics
In Table 1, we present summary statistics for the
variables used in our analysis of CEO compensation.
ExecuComp provides two measures of total compensation:
one includes the value of the options granted, and the
other includes the value of options exercised. We use the
total compensation including the value of options exercised in our analysis. The results remain qualitatively the
same when the value of options granted is considered.
Cash compensation is the sum of salary and bonus, as
provided by ExecuComp. We compute equity-based compensation as the total compensation minus salary, bonus,
other annual pay, and LTIP (long-term incentive plan). The
most common forms of equity-based compensation are
stock options and restricted stocks. Market capitalization is
computed as the stock price multiplied by the number of
shares outstanding at the end of a fiscal year. Market-tobook ratio is the market capitalization divided by the book
value of equity. All continuous variables except leverage
are winsorized at the 1st and 99th percentiles.12
Leverage is the variable of interest. We measure leverage in four ways. The market leverage, as used widely in
the literature (e.g., Leary and Roberts, 2010), is computed
as the total debt divided by the sum of total debt and
market value of equity. The book leverage, also used
commonly in the literature, is computed as the total debt
divided by the sum of total debt and book value of equity.
Total debt is the sum of long-term debt and debt in current
liabilities (data item 9 plus data item 34). Debt in current
liabilities (data item 34) includes notes payable (data item
10
CPI data are taken from the website of the Bureau of Labor
Statistics: http://www.bls.gov/cpi/.
11
French's website is http://mba.tuck.dartmouth.edu/pages/faculty/
ken.french/data_library.html.
12
Another way to identify outliers is by employing the Hadi (1992,
1994) procedure. The exclusion of outliers does not affect the results of
our multivariate analysis.
483
206) and debt due in 1 year (data item 44). Welch (2011)
argues that the liabilities that are nonfinancial debt should
not be included in the computation of leverage ratio. We
follow Welch (2011) and introduce two additional measures of leverage, which we refer to as “alternative market
leverage” and “alternative book leverage”. We calculate
alternative market leverage as (total long-term debt+debt
due in one year)/(total long-term debt+debt due in one
year+market value of equity) and calculate alternative
book leverage as (total long-term debt+debt due in one
year)/(total long-term debt+debt due in one year+book
value of equity).13 Due to space limitations, we report
results only from our analysis using market leverage,
alternative market leverage, and alternative book leverage.
Results from our analysis using book leverage are available
upon request.
CEOs' cash compensation (salary plus bonus) has a
mean of $972,330 and a median of $736,490, with a 1%
cutoff of $109,090 and 99% cutoff of $4.531 million. The
equity-based compensation has a larger mean but a
smaller median than the cash compensation. The reason
is that equity-based pay has a wider range across firms
than cash pay, and some CEOs have extremely large
equity-based pay. For example, the 99% cutoff of equitybased compensation is about $27 million, while the 1%
cutoff is only $12,500. We use the natural log of the
compensation variables in our multivariate regression of
CEO compensation to reduce the potential impact of outliers. The one-year return to shareholders (including
dividends), a measure of firm performance, has a median
of 10.33%.
Turning to CEO characteristics, the median CEO age is
65, and the median length of CEO tenure is four years. Only
2% of the CEOs in our sample are female, and 64% of the
CEOs also serve as chairman of the board. The G-Index has
a mean of 9.26 and a median of 9.
4.2. OLS regressions
In our reduced form analysis, we model CEO compensation as:
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t þ γ 2 Leveragei;t þ γ 3 MTBi;t
þγ 4 RET i;t þ γ 5 Agei;t þ γ 6 Tenurei;t þ γ 7 Chair i;t
þγ 8 MALEi;t þ εi;t ;
ð1Þ
CEOPayi,t is the CEO compensation of firm i in year t, and it
is measured in three ways: cash, equity-based, and total
compensation. Sizei,t is the natural log of market capitalization of firm i as of year t. We expect Sizei,t to be a positive
and significant determinant of CEO compensation. As
Murphy (1999) points out, the best-documented stylized
fact regarding CEO pay is that CEO pay is higher in larger
firms. Leveragei,t is the leverage ratio of firm i as of year t.
If firms with higher leverage pay a higher wage to their
CEOs, γ2 is positive. MTBi,t is the market-to-book ratio of
firm i as of year t, which is used as a proxy for firms'
13
We thank an anonymous referee for bringing Welch (2011) to our
attention and for suggesting that we also report our analysis using these
two alternative measures of market and book leverage.
Author's personal copy
484
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Table 1
Summary statistics of variables used in the analysis of CEO compensation.
This table summarizes the variables used in the analysis of CEO compensation. Cash (salary plus bonus) and total compensations are provided by
Execucomp. We compute equity-based compensation as the total compensation minus salary, bonus, other annual pay, and long-term incentive plan.
Market leverage is computed as (debts in current liabilities+long-term debt)/(debts in current liabilities+long-term debt+market value of equity).
Alternative market leverage is computed as (debts due in one year+long-term debt)/(debts due in one year+long-term debt+market value of equity).
Alternative book leverage is computed as (debts due in one year+long-term debt)/(debts due in one year+long-term debt+book value of equity). Market
capitalization is computed as the stock price multiplied by the number of shares outstanding as of the end of a fiscal year. Market-to-book ratio is the
market capitalization divided by the book value of equity. All continuous variables except leverage are winsorized at the 1st and 99th percentiles. All dollar
amounts are adjusted to 1992 dollars using the consumer price index. The G-Index was constructed by Gompers, Ishii, and Metrick (2003) as a measure of
corporate governance.
Cash (salary+bonus) (thousands)
Equity-based compensation (thousands)
Total compensation including options exercised (thousands)
Market leverage
Alternative market leverage
Alternative book leverage
Market capitalization (millions)
Market-to-book ratio
One-year return to shareholders (%)
CEO age
CEO tenure (years as CEO in the firm)
CEO is male
CEO is also the chairman
G-Index
N
Mean
Median
Standard deviation
1% Cutoff
99% Cutoff
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
14,891
11,527
972.33
1,715.02
2,809.71
0.19
0.18
0.28
4,762.54
3.42
17.84
65.01
6.38
0.98
0.64
9.26
736.49
189.99
1,201.47
0.14
0.12
0.28
916.74
2.45
10.33
65
4
1
1
9.0
789.83
4,131.79
4,739.57
0.19
0.19
0.23
17,368
3.25
54.41
8.75
7.30
0.12
0.48
2.70
109.09
12.5
158
0
0
0
28.59
0.45
77.55
45
0
0
0
4
4,531
26,656
32,099
0.80
0.79
0.88
78,204
21
249
86
35
1
1
15
growth opportunities. RETi,t is the return to shareholders of
firm i in year t, a popular measure of the performance of
firm i in year t. The existing literature shows a positive
relation between CEO pay and firm performance.14 Hence,
we expect γ4 to be positive. In addition, we control for
individual CEO characteristics that could affect CEO compensation. Agei,t is the age of the CEO of firm i as of year t;
Tenurei,t is the number of years the executive has acted as
the CEO in firm i prior to year t; Chairi,t is one if the CEO is
also the chairman and zero otherwise; and MALEi,t is one if
the CEO is male and zero otherwise. We include year
dummies to control for time-specific variation in CEO pay.
As shown by the literature, CEO compensation has
increased tremendously during the past few decades. We
include industry dummies due to the significant variation
in CEO pay across industries.
In Table 2, we report the estimated coefficients and
standard errors obtained from the OLS regression of
Eq. (1). The standard errors are clustered by firm. Estimation results from using market leverage, alternative market
leverage, and alternative book leverage are reported in
Panel A, Panel B, and Panel C, respectively. Columns 1–3 in
each panel exclude the G-Index, and columns 4–6 in each
panel include the G-Index. Including the G-Index in the
regression reduces the sample size. Firm size has a positive
impact on all three measures of CEO compensation.
A larger firm pays its CEO, on average, more than a smaller
firm does, which is consistent with the literature. A higher
one-year return to shareholders is associated with greater
CEO pay. This is consistent with the positive relation
between CEO pay and firm performance as shown by the
literature.
14
Murphy (1999) provides a comprehensive review of the relation
between firm performance and CEO compensation.
On average, an older CEO earns a larger pay. Being the
chairman has a positive and significant effect on CEO
compensation. Gender does not have a significant effect on
CEO pay. The coefficient on CEO tenure is not significant in
the regression of total and cash compensation, but it is
negative and significant (at the 5% level) in that of equitybased compensation. Market-to-book ratio is not significant
in the regressions of total compensation, but it is negative in
the regression of cash pay and is positive in that of equitybased compensation. This suggests that growth firms pay less
cash but more stock-based compensation to their CEOs than
value firms.
The leverage ratio has a positive and significant effect
on cash, equity-based, and total compensations. According
to Column 1 of Panel A, if market leverage goes up by
o1 standard deviation (0.19, as reported in Table 1),
the natural log of CEO total compensation increases by
0.19 0.42¼0.080, which translates to more than 8.3%
increase in total pay. Therefore, starting at the median
total CEO compensation of $1.20 million, the total CEO pay
increases by about $100,000, an economically significant
amount. If market leverage increases by 1 standard deviation (0.19), the CEO's cash pay goes up by more than 12%
and the CEO's equity-based pay goes up by more than 8%.
In summary, the effect of leverage on CEO compensation is
economically as well as statistically significant.
The G-Index is a positive and significant factor in
determining CEOs' cash, equity-based, and total pay, suggesting that stronger managerial power is associated with
greater CEO compensation. Leverage continues to have a
positive and significant effect on CEO compensation in the
presence of the G-Index.
We also estimate Eq. (1) by year, in the spirit of Fama
and Macbeth (1973). Table 3 reports the coefficient of
leverage in the regression of CEO compensation for every
year between 1992 and 2006. The coefficients of all three
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
485
Table 2
Ordinary least square regressions of CEO compensation.
This table reports the coefficients and standard errors obtained from OLS estimation of the following model:
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t þ γ 2 Leveragei;t þ γ 3 MTBi;t þ γ 4 RET i;t þ γ 5 Agei;t
þγ 6 Tenurei;t þ γ 7 Chairi;t þ γ 8 MALEi;t þ εi;t ;
CEOPayi,t is measured in three ways: the log of CEO total compensation, the log of CEO cash compensation (salary plus bonus), and the log of equity-based
compensation. Sizei,t is the log of market capitalization of firm i as of year t. Leveragei;t is the leverage ratio of firm i as of year t. Market leverage, alternative
market leverage, and alternative book leverage are defined the same as in Table 1. MTBi;t is the market-to-book ratio of firm i as of year t. RETi,t is the return
to the shareholders of firm i in year t. Agei,t is the age of the CEO of firm i as of year t; Tenurei,t is the number of years the executive has been acting as CEO in
firm i prior to year t; Chairi,t is one if the CEO is also the chairman and zero otherwise; and MALEi,t is one if the CEO of firm i as of year t is male and zero
otherwise. Numbers in the parentheses are the standard errors. The standard errors are clustered by firm and are also robust to heteroskedasticity.
Regressions in Panel A use market leverage, regressions in Panel B use alternative market leverage, and regressions in Panel C use alternative book leverage.
nnn nn
, , and n indicate significance at the 1%, 5%, and 10% level, respectively. The G-Index was constructed by Gompers, Ishii, and Metrick (2003) as a
measure of corporate governance.
Total
compensation
(4)
Cash
compensation
(5)
Equity-based
compensation (6)
Yes
Yes
1.57nn
(0.62)
0.43nnn
(0.07)
0.41nnn
(0.01)
0.011nn
(0.005)
0.002nnn
(0.0002)
0.006nnn
(0.002)
0.001
(0.003)
0.18nnn
(0.03)
0.07
(0.13)
0.02nnn
(0.005)
Yes
Yes
2.91nnn
(0.31)
0.55nnn
(0.06)
0.28nnn
(0.01)
0.01nn
(0.004)
0.001nnn
(0.0001)
0.004nn
(0.002)
0.001
(0.002)
0.14nnn
(0.02)
0.07
(0.09)
0.02nnn
(0.004)
Yes
Yes
3.68nnn
(0.30)
0.45nn
(0.20)
0.66nnn
(0.03)
0.02nn
(0.01)
0.003nnn
(0.001)
0.019nnn
(0.005)
0.01
(0.01)
0.24nnn
(0.07)
0.49
(0.31)
0.05nnn
(0.01)
Yes
Yes
1.86nn
(0.74)
14,891
14,891
11,527
11,527
11,527
0.49
0.23
0.43
0.47
0.24
Total
compensation
(1)
Cash
compensation
(2)
Equity-based
compensation (3)
0.42nnn
(0.07)
0.41nnn
(0.01)
0.006
(0.005)
0.001nnn
(0.0002)
0.006nnn
(0.002)
0.002
(0.003)
0.18nnn
(0.03)
0.05
(0.13)
0.59nnn
(0.05)
0.29nnn
(0.01)
0.02nnn
(0.003)
0.001nnn
(0.0001)
0.006nnn
(0.001)
0.002
(0.002)
0.14nnn
(0.02)
0.09
(0.08)
0.42nn
(0.18)
0.66nnn
(0.03)
0.023nn
(0.010)
0.002nnn
(0.0004)
0.018nnn
(0.005)
0.015nn
(0.006)
0.24nnn
(0.07)
0.32
(0.31)
–
–
–
Yes
Yes
3.15nnn
(0.23)
Yes
Yes
3.73nnn
(0.21)
14,891
0.42
Panel A: Market leverage
Market leverage
Firm size
Market-to-book ratio
One-year return to
shareholders
CEO age
CEO tenure
CEO is also the
chairman
CEO is male
G-Index
Year effects
Industry effects
Intercept
Number of
observations
R-squared
Panel B: Alternative market leverage
Alternative market leverage
Firm size
Market-to-book ratio
One-year return to shareholders (%)
CEO age
CEO tenure
CEO is also the chairman
CEO is male
G-Index
Year effects
Industry effects
Intercept
0.41nnn
(0.07)
0.41nnn
(0.01)
0.006
(0.005)
0.001nnn
(0.0002)
0.006nnn
(0.002)
0.002
(0.003)
0.18nnn
(0.03)
0.04
(0.13)
0.58nnn
(0.05)
0.29nnn
(0.01)
0.02nnn
(0.003)
0.001nnn
(0.0001)
0.006nnn
(0.001)
0.003
(0.002)
0.15nnn
(0.02)
0.09
(0.08)
0.40nn
(0.18)
0.66nnn
(0.03)
0.023nn
(0.010)
0.002nnn
(0.0004)
0.018nnn
(0.005)
0.015nn
(0.006)
0.24nnn
(0.07)
0.32
(0.31)
–
–
–
Yes
Yes
3.16nnn
Yes
Yes
3.74nnn
Yes
Yes
1.56nn
0.43nnn
(0.08)
0.41nnn
(0.01)
0.011nn
(0.005)
0.002nnn
(0.0002)
0.006nnn
(0.002)
0.001
(0.003)
0.18nnn
(0.03)
0.07
(0.13)
0.02nnn
(0.005)
Yes
Yes
2.92nnn
0.54nnn
(0.06)
0.28nnn
(0.01)
0.01nnn
(0.003)
0.001nnn
(0.0001)
0.004nn
(0.002)
0.001
(0.002)
0.14nnn
(0.02)
0.07
(0.09)
0.02nnn
(0.004)
Yes
Yes
3.69nnn
0.45nn
(0.20)
0.66nnn
(0.03)
0.02nn
(0.01)
0.003nnn
(0.001)
0.018nnn
(0.005)
0.01
(0.01)
0.25nnn
(0.07)
0.49
(0.31)
0.05nnn
(0.01)
Yes
Yes
1.85nn
Author's personal copy
486
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Table 2 (continued )
Panel B: Alternative market leverage
Number of observations
R-squared
(0.23)
14,891
0.42
(0.21)
14,891
0.49
(0.62)
14,891
0.23
(0.31)
11,527
0.43
(0.30)
11,527
0.47
(0.74)
11,527
0.24
0.25nnn
(0.06)
0.40nnn
(0.01)
0.0002
(0.005)
0.001nnn
(0.0002)
0.006nnn
(0.002)
0.002
(0.002)
0.18nnn
(0.03)
0.04
(0.13)
0.49nnn
(0.04)
0.28nnn
(0.01)
0.03nnn
(0.003)
0.001nnn
(0.0001)
0.005nnn
(0.001)
0.002
(0.002)
0.14nnn
(0.02)
0.09
(0.08)
0.29nn
(0.15)
0.65nnn
(0.02)
0.01
(0.01)
0.002nnn
(0.0004)
0.018nnn
(0.005)
0.014nn
(0.006)
0.24nnn
(0.07)
0.32
(0.31)
0.26nnn
(0.07)
0.40nnn
(0.01)
0.005
(0.006)
0.002nnn
(0.0002)
0.006nnn
(0.002)
0.001
(0.003)
0.18nnn
(0.03)
0.07
(0.13)
0.02nnn
(0.005)
Yes
Yes
3.02nnn
(0.29)
11,527
0.43
0.44nnn
(0.05)
0.27nnn
(0.01)
0.02nnn
(0.004)
0.001nnn
(0.0001)
0.004nn
(0.002)
0.001
(0.002)
0.14nnn
(0.02)
0.07
(0.09)
0.02nnn
(0.004)
Yes
Yes
3.80nnn
(0.29)
11,527
0.47
0.34nn
(0.17)
0.65nnn
(0.03)
0.01
(0.01)
0.002nnn
(0.0005)
0.02nnn
(0.005)
0.01
(0.01)
0.24nnn
(0.07)
0.48
(0.31)
0.05nnn
(0.01)
Yes
Yes
1.76nn
(0.74)
11,527
0.24
Panel C: Alternative book leverage
Alternative book leverage
Firm size
Market-to-book ratio
One-year return to shareholders (%)
CEO age
CEO tenure
CEO is also the chairman
CEO is male
G-Index
Year effects
Industry effects
Intercept
Number of observations
R-squared
–
–
–
Yes
Yes
3.23nnn
(0.22)
14,891
0.42
Yes
Yes
3.83nnn
(0.20)
14,891
0.52
Yes
Yes
1.49nn
(0.62)
14,891
0.23
measures of leverage in the regression of CEOs' total pay
and cash pay are positive in all of the 15 years. The
coefficient of alternative book leverage is positive in the
regression of CEOs' equity-based pay in 13 out of 15 years,
and the other two measures of leverage have a positive
coefficient in the regression of CEOs' equity-based pay in
14 out of 15 years.
4.3. New CEOs hired from outside
Some unobservable and thus uncontrolled CEO characteristics could affect both leverage and compensation
in the same direction, thus resulting in the positive
coefficient of leverage in the OLS regression of CEO
compensation. For example, CEOs who have had more
interaction with the board (and, therefore, have more
influence) could have greater ability to affect their own
pay and at the same time choose the firm's leverage level.
To address potential concerns regarding causality, we
study the subset of newly appointed CEOs who are hired
from outside. We examine how the first-year compensation of these new CEOs is affected by firm leverage in the
year prior to their appointment. CEOs hired from outside
should have no influence on their firms' capital structure
in the year prior to their appointment, so that this is a
clean test of the relation between leverage and CEO
compensation, allowing us to deal with the potential
causality problem.
We model the relation between the first-year compensation of newly appointed CEOs hired from outside and the
leverage ratio in the year prior to their appointment as:
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t1 þ γ 2 Leveragei;t1 þ γ 3 MTBi;t1
þγ 4 RET i;t þ γ 5 Agei;t þ γ 6 Chair i;t þ γ 7 MALEi;t þ εi;t :
ð2Þ
In Eq. (2), firm size, leverage, and market-to-book ratio
are computed as of the fiscal year prior to the appointment
of the new CEO. CEO tenure is omitted from Eq. (2),
because we estimate Eq. (2) on the sample of newly
appointed CEOs hired from outside (all of them have zero
tenure, by definition). Titman (1984) and BSZ (2010)
predict that a firm with higher leverage will pay its
employees more. In the case of a newly hired CEO, he will
demand and obtain a higher pay from a firm with higher
leverage. Therefore, we expect γ2 to be positive.
In Table 4, we present the coefficients and standard
errors obtained from estimating Eq. (2) on the subset of
newly appointed CEOs who are hired from outside. Firm
size is a strong factor in determining the pay of newly
appointed CEOs (cash, equity-based, and total compensation). The coefficient on the market-to-book ratio is
negative and significant in all three types of compensation,
suggesting that growth firms pay their new CEOs less than
value firms do. The coefficient of stock return during the
first year of a new CEO is positive and significant in the
regression of equity-based compensation. CEO age has a
negative effect on equity-based compensation, different
from what is in Table 2. This is due to the differences
between the underlying samples. In Table 2, the same CEO
in the same firm appears in multiple years, and the pay
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
487
Table 3
Fama-MacBeth analysis of CEO compensation.
This table reports the coefficient of leverage obtained from OLS regression of CEO pay in each fiscal year during 1992–2006:
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t þ γ 2 Leveragei;t þ γ 3 MTBi;t þ γ 4 RET i;t
þγ 5 Agei;t þ γ 6 Tenurei;t þ γ 7 Chairi;t þ γ 8 MALEi;t þ εi;t ;
CEOPayi,t is measured in three ways: the log of CEO total compensation, the log of CEO cash compensation (salary plus bonus), and the log of equity-based
compensation. Sizei,t is the log of market capitalization of firm i as of year t. Leveragei;t is the leverage ratio of firm i as of year t. Market leverage, alternative
market leverage, and alternative book leverage are defined the same as in Table 1. MTBi;t is the market-to-book ratio of firm i as of year t. RETi,t is the return
to the shareholders of firm i in year t. Agei,t is the age of the CEO of firm i as of year t; Tenurei,t is the number of years the executive has been acting as CEO in
firm i prior to year t; Chairi,t is one if the CEO is also the chairman and zero otherwise; MALEi,t is one if the CEO of firm i as of year t is male and zero
otherwise.
Compensation measure
Total
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Mean (t-stat)
0.43
0.19
0.42
0.55
0.55
0.59
0.54
0.55
0.35
0.45
0.41
0.39
0.45
0.65
0.41
0.46
(15.73)
Cash
Equity-based
Using market leverage
0.63
0.47
0.50
0.50
0.55
0.68
0.58
0.52
0.58
0.68
0.54
0.60
0.61
0.85
0.34
0.58
(19.39)
0.004
0.03
0.69
0.98
0.69
0.37
0.64
0.77
0.24
0.01
0.47
0.30
0.92
1.22
0.08
0.49
(4.84)
Compensation measure
Total
Cash
Equity-based
Using alternative market leverage
0.35
0.23
0.45
0.58
0.54
0.55
0.48
0.52
0.36
0.46
0.40
0.38
0.44
0.62
0.44
0.45
(17.33)
often increases with the CEO's age. In Table 4, all the CEOs are
newly hired from outside and the compensation information
is based on their first-year pay only. The negative coefficient
on CEO age in the regression of equity-based compensation
suggests that a younger newly hired CEO, in his first year,
earns more equity-based pay than an older newly hired CEO,
after controlling for other factors.
The variable of interest is the leverage ratio. The coefficient on leverage is positive and significant in the regressions
of all three forms of compensation for the newly hired CEOs,
for all three measures of the leverage ratio. The impact of
leverage on CEO compensation is also economically significant. An increase of 1 standard deviation in market leverage
corresponds to a 19% increase in cash pay, a 27% increase in
equity-based pay, and an 18% increase in the total compensation of a new CEO.
The results in Tables 2–4 demonstrate that leverage has
a strong and positive effect on the level of CEO compensation, supporting Hypothesis 1. Firms with higher leverage
incur a greater amount of CEO compensation, which is
consistent with the Titman-BSZ prediction.
5. Empirical tests and results on capital structure and
average employee pay
In this section, we present results on the effect of leverage
on average employee pay. In the multivariate analysis, we
start with OLS regressions. We then utilize instrumental
variable regressions of average employee pay to address
0.62
0.46
0.53
0.51
0.45
0.65
0.52
0.48
0.57
0.67
0.52
0.63
0.60
0.84
0.34
0.56
(18.48)
0.04
0.06
0.72
1.03
0.65
0.29
0.67
0.77
0.25
0.01
0.48
0.19
0.89
1.18
0.02
0.48
(4.65)
Compensation measure
Total
Cash
Equity-based
Using alternative book leverage
0.41
0.17
0.35
0.40
0.44
0.30
0.23
0.30
0.04
0.33
0.36
0.19
0.19
0.47
0.24
0.29
(9.69)
0.52
0.37
0.42
0.41
0.55
0.49
0.47
0.50
0.48
0.57
0.51
0.44
0.41
0.67
0.27
0.47
(19.55)
0.35
0.04
0.66
0.69
0.57
0.10
0.16
0.38
0.15
0.25
0.67
0.28
0.40
0.99
0.17
0.35
(4.11)
potential concerns about the endogeneity of leverage. Finally,
we deal with a potential sample selection problem using a
Heckman (1979) two-step analysis.
5.1. Summary statistics
Table 5 provides the summary statistics of the variables
used in our analysis of average employee pay. Average
employee pay is computed as labor expenses (data item
42) divided by the number of employees (data item 29).
Three measures of leverage are defined as before. Market
capitalization is the stock price multiplied by the number of
shares outstanding as of the fiscal year end. We compute
average sales per employee by dividing the amount of total
sales (data item 12) by the number of employees (data item
29). Market-to-book ratio is the market capitalization
divided by the book value of equity. Physical capital
intensity is computed as gross property, plant, and equipment scaled by total assets (data item 6). All continuous
variables except leverage are winsorized at the 1st and 99th
percentiles.
The mean (median) of average employee pay is $32,760
($32,000). The 1% cutoff is $1,490, and the 99% cutoff is
$95,580. Market capitalization has a wide range, from
$2.08 million (the 1% cutoff) to $82,827 million (the 99%
cutoff). To reduce the potential influence of outliers, we
use the log of average employee pay and the log of market
capitalization in our analysis. The mean of sales per
employee is about $166,260. The market-to-book ratio
Author's personal copy
488
Table 4
Ordinary least squares regressions of the compensation of newly appointed CEOs who are hired from outside.
This table presents the coefficients and standard errors obtained from estimation of the following model of CEO compensation on the sample of newly appointed CEOs who are hired from outside:
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t1 þ γ 2 Leveragei;t1 þ γ 3 MTBi;t1 þ γ 4 RET i;t þ γ 5 Agei;t þ γ 6 Chairi;t þ γ 7 MALEi;t þ εi;t ;
Market leverage
Alternative market
leverage
Alternative book
leverage
Firm size
Market-to-book
ratio
One-year return to
shareholders
CEO age
CEO is also the
chairman
CEO is male
Year effects
Industry effects
Intercept
Number of
observations
R-squared
Log of total
compensation
Log of cash
compensation
0.75nn
(0.30)
0.81nnn
(0.25)
Log of equity-based
compensation
1.11nnn
(0.41)
Log of total
compensation
Log of cash
compensation
–
–
0.71nnn
(0.26)
Log of equity-based
compensation
Log of total
compensation
Log of cash
compensation
Log of equity-based
compensation
–
–
–
–
–
–
–
0.64nn
(0.31)
0.97nn
(0.42)
–
–
–
–
–
–
0.50nnn
(0.04)
0.07nnn
(0.02)
0.002nn
(0.001)
0.01
(0.01)
0.01
(0.13)
0.15
(0.63)
Yes
Yes
4.67nnn
(0.96)
0.34nnn
(0.03)
0.04nn
(0.02)
0.0015n
(0.0009)
0.01
(0.01)
0.01
(0.10)
0.09
(0.35)
Yes
Yes
4.79nnn
(0.59)
0.62nnn
(0.06)
0.12nnn
(0.03)
0.004nnn
(0.001)
0.021n
(0.011)
0.09
(0.18)
0.33
(0.90)
Yes
Yes
3.38nn
(1.44)
0.50nnn
(0.04)
0.07nnn
(0.02)
0.002nn
(0.001)
0.01
(0.01)
0.02
(0.13)
0.15
(0.64)
Yes
Yes
4.78nnn
(0.97)
0.34nnn
(0.03)
0.04nn
(0.02)
0.0015n
(0.0009)
0.01
(0.01)
0.01
(0.10)
0.10
(0.37)
Yes
Yes
4.90nnn
(0.60)
0.62nnn
(0.06)
0.13nnn
(0.03)
0.004nnn
(0.001)
0.021n
(0.011)
0.10
(0.18)
0.33
(0.92)
Yes
Yes
3.54nn
(1.45)
–
–
–
0.62nn
(0.30)
0.49nnn
(0.03)
0.08nnn
(0.02)
0.0021
(0.0012)
0.01
(0.01)
0.02
(0.13)
0.17
(0.64)
Yes
Yes
4.93nnn
(0.98)
0.62nnn
(0.23)
0.33nnn
(0.03)
0.05nnn
(0.01)
0.0015n
(0.0009)
0.01
(0.01)
0.001
(0.10)
0.13
(0.38)
Yes
Yes
5.10nnn
(0.61)
0.92nn
(0.41)
0.60nnn
(0.06)
0.14nnn
(0.03)
0.005nnn
(0.001)
0.021n
(0.011)
0.10
(0.19)
0.36
(0.91)
Yes
Yes
3.77nn
(1.46)
373
373
373
373
373
373
373
373
373
0.49
0.47
0.50
0.49
0.47
0.50
0.49
0.47
0.50
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
CEOPayi,t is measured in three ways: the natural log of total compensation, the natural log of cash compensation (salary plus bonus), and the natural log of equity-based compensation of the newly appointed CEO
of firm i in year t who is hired from outside. Sizei,t 1 is the log of market capitalization of firm i in year t 1; Leveragei,t 1 is the leverage ratio of firm i in year t 1; Market leverage, alternative market leverage, and
alternative book leverage are defined the same as in Table 1; MTBi,t 1 is the market-to-book ratio of firm i in year t 1; RETi,t is the return to the shareholders of firm i in year t; Agei,t is the age of the CEO of firm i as
of year t; Chairi,t is one if the CEO is also the chairman of firm i in year t and zero otherwise; MALEi,t is one if the CEO of firm i as of year t is male and zero otherwise. The numbers in parentheses are the standard
errors. The standard errors are robust to heteroskedasticity and are clustered by firm. nnn and nn indicate significance at the 1% and 5% level, respectively.
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
489
Table 5
Summary statistics of variables used in the analysis of average employee pay.
This table summarizes the variables used in the analysis of average employee pay. Average employee pay is computed as labor expenses (Compustat data
item 42) divided by the number of employees (data item 29). Market leverage, alternative market leverage, and alternative book leverage are defined the
same as in Table 1. Market capitalization is the stock price multiplied by the number of shares outstanding as of the fiscal year end. Average sales per
employee is the amount of total sales (data item 12) divided by the number of employees. Market-to-book ratio is the market capitalization divided by the
book value of equity. Physical capital intensity is computed as gross property, plant, and equipment scaled by total assets. Marginal tax rate (MTRB) is the
marginal tax rate based on income before the deduction of interest expenses. All continuous variables except leverage are winsorized at the 1st and 99th
percentiles. All dollar amounts are adjusted to 1992 dollars using the consumer price index.
Average employee pay (thousands)
Market leverage
Alternative market leverage
Alternative book leverage
Market capitalization (millions)
Average sales per employee (thousands)
Market-to-book ratio
Physical capital intensity
G-Index
Marginal tax rate (MTRB)
Number of observations
Mean
Median
Standard deviation
1% Cutoff
99% Cutoff
5,269
5,269
5,269
5,269
5,269
5,269
5,269
5,269
1,326
2,902
32.76
0.25
0.25
0.33
5,226
166.26
2.94
0.69
9.22
0.31
32.00
0.20
0.17
0.31
590
111.76
2.06
0.67
9.0
0.35
19.76
0.23
0.22
0.24
13,035
198.41
3.30
0.40
2.61
0.09
1.49
0
0
0
2.08
12.74
0.28
0.04
4
0
95.58
0.91
0.90
0.93
82,827
1,253
20.54
1.76
15
0.38
has a mean of 2.94 and a median of 2.06. On average, the
gross amount of property, plant, and equipment is about
69% of total assets.
5.2. OLS regressions
Our objective here is to estimate the effect of leverage on
average employee pay. In our reduced form analysis (the base
case), we use following specification:
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t
þβ4 MTBi;t þ β5 PCI i;t þ εi;t ;
ð3Þ
AEPi,t is the natural log of average employee pay of firm i in
fiscal year t. Sizei,t is the log of market capitalization of firm
i at the end of year t. Prior empirical studies have shown
that larger firms tend to pay higher wages to their employees than smaller firms, so we expect β1 to be positive.
AvgSalei,t is the sales per employee. We use AvgSalei,t to
directly measure the productivity of the average employee
of firm i in year t, and we expect β3 to be positive. MTBi,t is
the market-to-book ratio of firm i as of year t. We control
for the market-to-book ratio, as it is a common proxy for a
firm's growth opportunity. PCIi,t is the physical capital
intensity of firm i as of year t. We include the measure of
physical capital intensity for two reasons. First, capital
intensive firms tend to be more productive (Cronqvist,
Heyman, Nilsson, Svaleryd, and Vlachos, 2009). Second,
BSZ (2010) predict a positive correlation between physical
capital intensity and employee wage. We include the year
dummies to control for the aggregate variation in employee
pay. We also include the industry dummies because a great
deal of heterogeneity in pay practices is evident across
industries. The effect of leverage on average employee pay
is of particular interest. If firms of higher leverage pay their
employees more, β2 is positive.
Panel A of Table 6 presents the estimated coefficients and
standard errors obtained from the OLS regression of Eq. (3) for
all firms. The standard errors are clustered by firm and are
also robust to heteroskedasticity. Larger firms pay their
employees more, consistent with the literature (e.g., Brown
and Medoff, 1989). Average sales per employee affects average
employee pay positively, consistent with our expectation, as
sales per employee is a measure of employee productivity.
Neither physical capital intensity nor the market-to-book ratio
has a significant impact on average employee pay. Most
important, after controlling for other factors, the leverage
ratio has a positive effect on average employee pay. The
coefficients on all three leverage ratios are positive and
significant at the 1% or 5% level. This supports Hypothesis 2.
We find that the G-Index is not a statistically significant factor
in determining average employee pay. Further, even in the
presence of the G-Index, leverage is a positive and significant
determinant of average employee pay.
We now examine the subset of financially distressed
versus safe firms. Since its introduction by Altman (1968),
the Z-score has been used for the prediction of bankruptcy.
Following the original formula, we compute the Z-score as:
Z ¼ 1:2T 1 þ 1:4T 2 þ 3:3T 3 þ :6T 4 þ T 5 ;
ð4Þ
here T1 ¼ working capital/total assets, where working capital is
computed as current assets minus current liabilities;
T2 ¼retained earnings/total assets; T3 ¼earnings before interest and taxes/total assets; T4 ¼market value of equity/book
value of total liabilities; and T5 ¼sales/total assets. A lower
Z-score corresponds to a greater probability of bankruptcy.
Firms with a Z-score above 2.99 are considered to be safe,
those with a Z-score of 1.8 or lower are considered distressed,
and those with Z scores in between the two threshold values
are considered in the gray zone.
Panel B of Table 6 reports the results from estimating
Eq. (3) on the two subsets: distressed firms and safe firms.
When firms are financially distressed, average employee pay
is not significantly related to leverage; when firms are safe,
average employee pay increases with leverage. In summary,
the evidence supporting Hypothesis 4 is weak or nonexistent.
This indicates that while the ex ante relation between
leverage and employee pay suggested by Titman-BSZ prediction dominates in our entire sample and in the subsample of
safe firms, in distressed firms the ex post relation postulated
by Perotti and Spier (1993) could partially or fully offset the
effect of firms compensating employees for the reduction in
the value of their human capital due to higher leverage.
Author's personal copy
490
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Table 6
Ordinary least square regressions of average employee pay.
This table presents the coefficients and standard errors obtained from the OLS regression of the following model of average employee pay:
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t þ β4 MTBi;t þ β5 PCI i;t þ εi;t ;
where AEPi,t is the log of average employee pay of firm i in fiscal year t, and it is calculated as the log of the total labor expenses (data item 42) divided by
the number of employees (data item 29); Sizei,t is the log of market capitalization of firm i in year t; and Leveragei,t is the leverage ratio of firm i in year t.
Market leverage, alternative market leverage, and alternative book leverage are defined the same as in Table 1. AvgSalei,t is the average sales (in thousand
dollars) per employee, i.e., the amount of total sales divided by the number of employees; MTBi,t is the market-to-book ratio of firm i in year t; and PCIi,t is
the physical capital intensity of firm i in year t, computed as gross property, plant, and equipment scaled by total assets. Numbers in the parentheses are the
standard errors. The standard errors are robust to heteroskedasticity and are clustered by firm. In Panel A, we examine all firms. In Panel B, we study
financially safe and distressed firms separately. In Panel C, we add quits rates to the regressions. nnn, nn, and n indicate significance at the 1%, 5%, and 10%
level, respectively.
Panel A: All firms
Market leverage
All firms
1
All firms
2
All firms
3
All firms
4
All firms
5
All firms
6
0.23nnn
(0.08)
–
–
0.28nnn
(0.09)
–
–
nn
0.29
(0.08)
–
Alternative market leverage
Alternative book leverage
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
G-Index
Year effects
Industry effects
Intercept
Number of observations
R-squared
–
0.28
(0.10)
–
–
0.04nn
(0.02)
0.001nnn
(0.0002)
0.002
(0.005)
0.001
(0.09)
0.01
(0.01)
Yes
Yes
2.88nnn
(0.18)
1,326
0.77
0.04nn
(0.02)
0.001nnn
(0.0002)
0.002
(0.005)
0.002
(0.09)
0.01
(0.01)
Yes
Yes
2.89nnn
(0.18)
1,326
0.77
–
nnn
–
–
0.08nnn
(0.01)
0.001nnn
(0.0002)
0.003
(0.003)
0.05
(0.06)
0.08nnn
(0.01)
0.001nnn
(0.0001)
0.003
(0.004)
0.04
(0.06)
0.22
(0.07)
0.07nnn
(0.01)
0.001nnn
(0.0001)
0.008nn
(0.004)
0.04
(0.06)
–
–
–
Yes
Yes
1.53nnn
(0.07)
5,269
0.52
Yes
Yes
1.53nnn
(0.06)
5,269
0.52
Yes
Yes
1.57nnn
(0.06)
5,269
0.52
nnn
–
0.21nn
(0.09)
0.030n
(0.017)
0.001nnn
(0.0002)
0.008
(0.006)
0.02
(0.09)
0.01
(0.01)
Yes
Yes
2.94nnn
(0.18)
1,326
0.77
Panel B: Safe versus distressed firms
Market leverage
Alternative market leverage
Alternative book leverage
Distressed firms
1
Distressed firms
2
Distressed firms
3
0.01
(0.12)
–
–
–
0.15
(0.11)
–
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
Year effects
Industry effects
Intercept
Number of observations
R-squared
–
nnn
0.08
(0.01)
0.002nnn
(0.0003)
0.002
(0.005)
0.07
(0.07)
Yes
Yes
2.16nnn
(0.17)
977
0.56
0.08
(0.01)
0.002nnn
(0.0003)
0.003
(0.005)
0.07
(0.07)
Yes
Yes
2.09nnn
(0.16)
977
0.56
Safe firms
5
Safe firms
6
–
–
0.22nn
(0.10)
0.32nnn
(0.11)
–
0.004
(0.10)
0.08nnn
(0.01)
0.002nnn
(0.0003)
0.002
(0.006)
0.07
(0.07)
Yes
Yes
2.17nnn
(0.16)
977
0.56
–
nnn
Safe firms
4
–
–
nnn
0.07nnn
(0.01)
0.001nnn
(0.0001)
0.006n
(0.004)
0.02
(0.05)
Yes
Yes
2.49nnn
(0.11)
2,197
0.55
0.07
(0.01)
0.001nnn
(0.0001)
0.006n
(0.004)
0.01
(0.05)
Yes
Yes
2.48nnn
(0.11)
2,197
0.55
–
0.17nn
(0.07)
0.07nnn
(0.01)
0.001nnn
(0.0001)
0.009nn
(0.004)
0.02
(0.05)
Yes
Yes
2.50nnn
(0.11)
2,197
0.55
Panel C: Quits rates
Market leverage
All firms
1
All firms
2
All firms
3
All firms
4
All firms
5
–
–
0.43nnn
(0.13)
–
–
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
491
Table 6 (continued )
Panel C: Quits rates
All firms
1
All firms
2
All firms
3
All firms
4
–
–
–
0.55nnn
(0.13)
Alternative market leverage
–
Alternative book leverage
–
Firm size
–
–
Average sales per employee
–
–
Market-to-book ratio
–
–
–
–
0.026nnn
(0.002)
–
–
3.88nnn
(0.07)
1,993
0.16
0.002
(0.008)
Yes
Yes
1.73nnn
(0.23)
1,993
0.37
Physical capital intensity
Quits rate
Year effects
Industry effects
Intercept
Number of observations
R-squared
–
–
nnn
0.09
(0.02)
0.001nnn
(0.0002)
0.002
(0.007)
0.02
(0.09)
0.01
(0.01)
Yes
Yes
1.07nnn
(0.25)
1,993
0.49
0.09nnn
(0.02)
0.001nnn
(0.0002)
0.002
(0.007)
0.03
(0.09)
0.01
(0.01)
Yes
Yes
1.06nnn
(0.24)
1,993
0.49
All firms
5
–
0.44nnn
(0.11)
0.08nnn
(0.02)
0.001nnn
(0.0002)
0.001
(0.008)
0.04
(0.09)
0.01
(0.01)
Yes
Yes
1.17nnn
(0.24)
1,993
0.49
In Panel C, we include the quits rate in the OLS
estimation. Column 1 includes the quits rate only. The
coefficient is significant and negative, suggesting that
more specialized labor gets paid more. Column 2 adds
industry and year fixed effects to the regression, and the
coefficient of quits rate becomes insignificant. This is not
surprising, given that the annual quits rate is measured at
the industry level. In Columns 3–5, the coefficient of quits
rate remains insignificant, but that of all three leverage
ratios is still positive and statistically significant.
As a robustness test, we also estimate Eq. (3) by year,
in the spirit of Fama and MacBeth (1973). Table 7 reports
the estimated coefficient on leverage in every year during
1992–2006. The coefficient on market leverage ranges from
0.02 to 0.66, and it is positive in 13 out of 15 years. Its mean
is 0.22, statistically larger than zero. The impact of leverage is
somewhat weaker prior to year 2000 than after year 2000. To
understand why, we examine the percentage of nontechnology firms by year. We find that the percentage of nontechnology firms is below sample mean in five out of nine years
during 1992–2000 while the percentage of nontechnology
firms is below sample mean in only two out of six years
during 2001–2006. The effect of leverage on average
employee pay is stronger in nontechnology firms, so that
the smaller coefficients on leverage prior to the year 2000
could be due to the lower fraction of nontechnology firms
prior to year 2000.
to generate an exogenous variation in leverage. A valid
instrumental variable for leverage needs to satisfy two conditions. It is correlated with the leverage ratio (the validity
requirement) but is uncorrelated with the residual in the
regression of employee pay (the exclusion restriction). The
marginal corporate tax rate satisfies both requirements. The
theoretical literature in corporate finance indicates that the
tax benefit of debt is positively related to a firm's marginal tax
rate, thus resulting in a positive correlation between a firm's
marginal tax rate and its leverage ratio. The empirical
literature supports this view (for example, Leary and
Roberts, 2010). At the same time, no theoretical or empirical
literature indicates that the marginal corporate tax rate
directly affects average employee pay.
Following Graham, Lemmon, and Schallheim (1998), we
use the marginal tax rates based on income before interest is
deducted (MTRB) from the database of marginal tax rates
provided by John Graham (for more details, see Graham
(1996a, 1996b)). When examining the effect of firms' leverage
on bond ratings, Molina (2005) also uses marginal tax rate as
an instrument for leverage. We implement the instrumental
variable regressions by using the 2SLS procedure in STATA
(Wooldridge, 2002). In the first stage, leverage is regressed
onto the instrumental variable and control variables; in the
second stage, average employee pay is regressed onto the
instrumented leverage and control variables. The first stage
regression specification is given by
5.3. Instrumental variable regressions
Leveragei;t ¼ α0 þ α1 MTRBi;t þ α2 Sizei;t þ α3 AvgSalei;t
The assets of a given firm could be such that they can
support a high level of leverage (for example, the proportion
of tangible assets could be high) and could also require highly
paid employees to operate these assets, thus generating a
positive correlation between leverage and employee pay. To
deal with this potential endogeneity problem, we employ an
instrument variable, namely, the marginal corporate tax rate,
þα4 MTBi;t þ α5 PCI i;t þ α6 ðEBIT=TAÞi;t
þα7 STDðEBIT=TAÞi;t þ δi;t :
ð5Þ
The second stage regression specification is given by
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t
þβ4 MTBi;t þ β5 PCI i;t þ β6 ðEBIT=TAÞi;t
þβ7 STDðEBIT=TAÞi;t þ εi;t :
ð6Þ
Author's personal copy
492
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
MTRB is the marginal tax rate based on income before
interest is deducted. EBIT/TA is earnings before depreciation, interest, and taxes divided by total assets, and STD
(EBIT/TA) is the standard deviation of EBIT/TA in the past
five years. The results from the instrumental variable
regression are presented in Table 8. In the first stage
analysis (leverage is the dependent variable), marginal tax
rate is an important determinant of debt ratio, significant at
the 1% level. In their survey of the weak-instrument
literature, Stock, Wright, and Yogo (2002) develop benchmarks for the necessary magnitude of the F-statistic. When
the number of instruments is 1, 2, 3, 5, and 10, the
suggested critical F-values are 8.96, 11.59, 12.83, 15.09, and
20.88, respectively. If the first-stage partial F-statistic falls
below these critical values, the instruments are considered
to be weak and inference problems are potentially serious.
The partial F-statistics of our instrument in the regressions
of all three leverage ratios are above the critical value of
8.96. The results in the first stage and the partial
F-test confirm that the marginal tax rate is a strong
instrument (i.e., it satisfies the validity requirement).
In the second stage analysis, firm size and average sales
per employee are positive and significant determinants of the
average employee pay, consistent with the results from our
OLS regressions presented in Table 6. More important, we find
from our second stage regression that, even after accounting
for the potential endogeneity of leverage, leverage continues
to be a positive and significant determinant of average
employee pay.
In the first-stage regression of alternative book leverage,
the market-to-book ratio has a positive coefficient, which
seems to contradict the negative relation between leverage
and the market-to-book presented in the existing literature.15
However, Chen and Zhao (2006) show that the negative
relation that has been found between book leverage and
market-to-book ratio is driven by a few small firms with very
large market-to-book ratios. In particular, they note that a
positive relation between market-to-book and leverage holds
for 88% of all firms, accounting for more than 95% of the total
market capitalization.16
5.4. Missing data on labor expenses: a Heckman (1979)
two-step analysis
Labor expenses are missing for a number of firms in
Compustat, creating a potential sample-selection bias, if
15
Even papers in the existing literature show that the negative
coefficient on market-to-book has a significantly smaller magnitude in
the regression of book leverage than in the regression of market leverage.
For example, in Table 5 of Hovakimian, Kayhan, and Titman (2012),
the coefficient of market-to-book is 0.024 (z-statistic is 3.8) in the
regression of book leverage while it is 0.097 (z-statistic is 15.2) in the
regression of market leverage. The difference in the sign of the marketto-book coefficient when using book leverage between our result and the
previous work could be due to the fact that our sample size is smaller.
The limited availability of CEO compensation and average employee pay
data significantly reduces our sample size, compared with other studies.
16
The review article by Parsons and Titman (2008) has a detailed
discussion of papers studying the relation between leverage and the
market-to-book ratio, including the paper by Chen and Zhao (2006). They
suggest caution when using and interpreting market-to-book ratios in
leverage specifications.
Table 7
Fama-MacBeth analysis of average employee pay.
This table reports the coefficient on leverage obtained from the OLS
regression of average employee pay for each fiscal year in 1992–2006:
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t þ β4 MTBi;t
þβ5 PCI i;t þ εi;t ;
where AEPi,t is the log of average employee pay of firm i in fiscal year t,
which is calculated as the log of the total labor expenses (data item 42)
divided by the number of employees (data item 29); Sizei,t is the log of
market capitalization of firm i in year t; Market leverage, alternative
market leverage, and alternative book leverage are defined the same as in
Table 1; AvgSalei,t is the average sales per employee, i.e., the amount of
total sales divided by the number of employees; MTBi,t is the market-tobook ratio of firm i in year t; and PCIi,t is the physical capital intensity of
firm i in year t, computed as gross property, plant, and equipment scaled
by total assets.
Year
Market
leverage
Alternative market
leverage
Alternative
book leverage
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
0.16
0.10
0.06
0.28
0.01
0.12
0.01
0.02
0.18
0.08
0.29
0.66
0.43
0.36
0.62
0.09
0.06
0.11
0.26
0.02
0.07
0.04
0.12
0.24
0.22
0.40
0.76
0.67
0.49
0.60
0.07
0.04
0.004
0.17
0.02
0.09
0.05
0.10
0.24
0.26
0.35
0.52
0.49
0.32
0.39
Mean (t-stat)
0.22
(3.96)
0.27
(4.14)
0.20
(4.12)
firms selectively decide whether or not to report labor
expense information. To control for this potential sampleselection bias, we adopt a Heckman (1979) two-step
analysis in this section.
In the first step, we estimate a probit model of whether
or not a firm reports labor expenses. The dependent
variable is one if the data on labor expenses are nonmissing and zero otherwise. The independent variables
include the dummies of the firm's listing exchange, in
addition to the original control variables in the regression
of average employee pay. The listing exchange is the
identifying variable. We assume that firms on different
exchanges have different reporting behavior (the results in
the first-step probit analysis confirm this assumption), and
exchange listing does not affect the reported average
employee pay (to verify this condition, we add the dummies of exchange listing to the OLS regression of average
employee pay and find that they are jointly insignificant,
with an F-statistic of 1.29 and p-value of 0.28). In the
second step, we examine the effect of leverage on average
employee pay. The inverse Mills ratio (Lambda) derived
from the selection model is included in the second step as
a regressor, and all other independent variables are as
specified in Eq. (3).
The estimated coefficients and standard errors are
reported in Table 9. From the estimation of the selection
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
493
Table 8
Instrumental variable regressions of average employee pay.
This table presents the coefficients and standard errors obtained from the two-stage instrumental variable regression:
Leveragei;t ¼ α0 þ α1 MTRBi;t þ α2 Sizei;t þ α3 Avgsalei;t
þα4 MTBi;t þ α5 PCI i;t þ α6 ðEBIT=TAÞi;t þ α7 STDðEBIT=TAÞi;t þ δi;t
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t
þβ4 MTBi;t þ β5 PCIi;t þ β6 ðEBIT=TAÞi;t þ β7 STDðEBIT=TAÞi;t þ εi;t
Leveragei,t in the second stage is instrumented by marginal tax rate based on income before interest expense has been deducted (MTRB). EBIT/TA is earnings
before depreciation, interest, and taxes divided by total assets, and STD (EBIT/TA) is the standard deviation of EBIT/TA in the past five years. Market leverage,
alternative market leverage, and alternative book leverage are defined the same as in Table 1. Numbers in the parentheses are the standard errors. The
standard errors are robust to heteroskedasticity and are clustered by firm. nnn, nn, and n indicate significance at the 1%, 5%, and 10% level, respectively.
First stage: leverage is the dependent variable
Variable
Marginal tax rate (MTRB)
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
EBIT/TA
STD(EBIT/TA)
Year effects
Industry effects
Intercept
Number of observations
R-squared
Market leverage
Alternative market leverage
Alternative book leverage
0.11nnn
(0.04)
0.019nnn
(0.002)
0.00002
(0.00002)
0.005nnn
(0.001)
0.101nnn
(0.015)
0.80nnn
(0.05)
0.01
(0.01)
Yes
Yes
0.38nnn
(0.04)
2,902
0.34
0.14nnn
(0.04)
0.016nnn
(0.002)
0.00002
(0.00002)
0.005nnn
(0.001)
0.122nnn
(0.015)
0.76nnn
(0.05)
0.01
(0.01)
Yes
Yes
0.29nnn
(0.04)
2,902
0.32
0.20nnn
(0.05)
0.0002
(0.002)
0.00001
(0.00003)
0.018nnn
(0.001)
0.152nnn
(0.018)
0.96nnn
(0.06)
0.01
(0.02)
Yes
Yes
0.19nnn
(0.04)
2,902
0.24
10.62
(0.0011)
14.98
(0.0001)
12.26
Partial F-test of MTRB F-statistic (p-value)
(0.0005)
Second stage: average employee pay is the dependent variable
Variable
Market leverage (instrumented)
Alternative market leverage (instrumented)
Alternative book leverage (instrumented)
Average employee pay
2.77n
(1.49)
–
–
nnn
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
EBIT/TA
STD(EBIT/TA)
Year effects
Industry effects
Intercept
Number of observations
R-squared
0.13
(0.03)
0.001nnn
(0.0001)
0.01
(0.01)
0.01
(0.16)
1.48
(1.13)
0.02
(0.05)
Yes
Yes
1.39nn
(0.62)
2,902
0.22
model in the first step, we observe that larger firms with
higher leverage, lower sales per employee, lower market-tobook ratio, and higher physical capital intensity are more
Average employee pay
Average employee pay
–
–
2.15nn
(1.02)
–
0.11nnn
(0.02)
0.001nnn
(0.0001)
0.01
(0.01)
0.02
(0.13)
0.90
(0.72)
0.02
(0.05)
Yes
Yes
1.82nnn
(0.35)
2,902
0.40
–
1.56nn
(0.70)
0.07nnn
(0.01)
0.001nnn
(0.0001)
0.01nn
(0.01)
0.05
(0.12)
0.75
(0.63)
0.02
(0.04)
Yes
Yes
2.14nnn
(0.35)
2,902
0.45
likely to report labor expenses. The exchange dummies are
jointly significant. In the second step, the coefficients on
firm size and average sales per employee are positive and
Author's personal copy
494
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
significant. More important, the impact of leverage on average
employee pay remains positive and significant after we
control for the potential sample-selection bias.
The Heckman two-step procedure produces consistent
estimation of parameters. The coefficient on the inverse
Mills ratio is statistically distinguishable from zero and
negatively signed, suggesting that the unobserved factors
that make reporting of labor expenses more likely tend to
be associated with lower average employee pay.
5.5. A comparison of the incremental costs and benefits of
leverage
Both our OLS and instrumental variable regressions
provide evidence supporting Hypothesis 2. Based on the
results presented in Panel A of Table 6, we now compute
the incremental tax benefits and labor costs associated
with an increase in leverage. For a firm with the median
values of the leverage ratio, average employee pay, total
labor expenses, and total debt, if the market leverage ratio
increases by 0.23 (1 standard deviation of leverage in the
sample in Panel A of Table 6), the natural log of average
employee pay increases by 0.23 0.23¼0.0529. Starting
at the median level of average employee pay of $32.00
(in thousands), average employee pay then becomes $33.79
(in thousands), an increase of 5.60%. The median total labor
expenses is about $250 million, so the increase in total labor
expenses is about 250n5.60%¼$14.01 million, assuming that
the number of employees does not change.
The return on corporate bonds depends on various factors
such as interest rate, credit rating, and time to maturity,
so that we can use only an average interest rate for our
calculation of the tax benefits of debt. We use 6% as the
average rate of return on corporate bonds in our sample from
1992 to 2006.17 The median level of debt is about $120 million
in our sample. Starting from the median leverage ratio of 0.20,
the level of debt goes up by 202% when we increase market
leverage ratio by 0.23 (1 standard deviation), holding everything else constant. If we assume a marginal tax rate of 35%,
which is the median corporate marginal tax rate as computed
by John Graham (Graham, 1996a, 1996b), interest expenses
increase by $120 202% 0.06¼$14.54 million, and the tax
benefits of debt increase by 14.54 0.35¼$5.09 million,
which is smaller than the increase in total labor expenses
($14.01 million). The calculation shows that the additional
labor costs associated with an increase of 1 standard deviation
in the leverage ratio offsets all of the incremental tax benefits
associated with the leverage increase.
We repeat the above calculation by increasing the market
leverage ratio from the median level of 0.20 all the way to
0.68, an increase of more than 2 standard deviations. Fig. 1
plots the changes in total labor expenses and the tax benefits
of debt as the leverage ratio goes up. The graph shows that the
additional labor expenses offset all of the incremental tax
benefits of debt even when the leverage ratio is increased by
as much as 2 standard deviations.
17
We believe that 6% is a reasonable estimate. The compounded
annual return for long-term US government bonds has averaged less than
3% during the past four decades, and we assume that corporate bonds, on
average, have a 3% premium over long-term US government bonds.
The analysis demonstrates that the incremental labor costs
associated with an increase in leverage are economically
significant. Further, these incremental labor costs are greater
than the additional tax benefits of debt associated with a wide
range of changes in the leverage ratio. Therefore, the results
support our Hypotheses 2 and 3. Overall, this evidence is
consistent with the Titman-BSZ prediction, i.e., risk-averse
employees demand greater compensation from firms
with higher leverage, and such indirect costs of bankruptcy
are economically large enough to limit the use of debt by
these firms.
6. Technology firms versus nontechnology firms
We now study CEO compensation and average employee
pay in two subsets of our sample: technology firms versus
nontechnology firms. The definition of technology firms and
nontechnology firms follows that in Anderson, Banker, and
Ravindran (2000), Ittner, Lambert, and Larcker (2003), and
Murphy (2003). Technology firms are defined as companies in
the computer, software, internet, telecommunications, or
networking fields. Nontechnology firms are firms with Standard Industry Classification (SIC) codes less than 4000 not
otherwise categorized as technology firms.18 We examine
whether the effect of leverage on employee pay is different
between technology and nontechnology firms, as employees
in nontechnology firms are more entrenched than in technology firms (in the sense discussed in Section 2), so that we
expect the effect of leverage on employee pay in nontechnology firms to be greater than that in technology firms
(consistent with Hypothesis 5).
6.1. CEO compensation in technology versus nontechnology
firms
We first examine CEO compensation in technology
versus nontechnology firms. In Table 10, we compare CEO
compensation and various explanatory variables across
technology and non technology firms. The mean of total
compensation for CEOs in technology firms is greater than
that in nontechnology firms, but the median is smaller for
CEOs in technology firms. Although CEOs in technology
firms receive less cash compensation than CEOs in nontechnology firms, the former have a greater mean of equity
compensation than the latter, consistent with Anderson,
Banker, and Ravindran (2000), Ittner, Lambert, and Larcker
(2003), and Murphy (2003). Technology firms have lower
leverage than nontechnology firms. Further, CEOs in technology firms are younger than CEOs in nontechnology
18
Technology firms are defined as companies with primary SIC
designations of 3570 (Computer and Office Equipment), 3571 (Electronic
Computers), 3572 (Computer Storage Devices), 3576 (Computer Communication Equipment), 3577 (Computer Peripheral Equipment), 3661
(Telephone & Telegraph Apparatus), 3674 (Semiconductor and Related
Devices), 4812 (Wireless Telecommunication), 4813 (Telecommunication), 5045 (Computers and Software Wholesalers), 5961 (Electronic
Mail-Order Houses), 7370 (Computer Programming, Data Processing),
7371 (Computer Programming Service), 7372 (Prepackaged Software),
and 7373 (Computer Integrated Systems Design). Nontechnology firms
are firms with SIC codes less than 4000 not otherwise categorized as
technology firms.
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
495
Table 9
Heckman two-step analysis of average employee pay.
This table reports the coefficients and standard errors obtained from a Heckman two-step analysis of average employee pay. In the first step, we estimate
a probit model of whether a firm reports labor expenses. The dependent variable is one if the data on labor expenses are non-missing and zero otherwise.
The independent variables include the dummies of the firm's listing exchange, in addition to other firm characteristics. In the second step, we examine the
impact of leverage on average employee pay. The inverse mills ratio (Lambda) derived from the selection model is included in the second step as a
regressor. Numbers in the parentheses are the standard errors. The standard errors are robust to heteroskedasticity and are clustered by firm. Market
leverage, alternative market leverage, and alternative book leverage are defined the same as in Table 1. nnn and nn indicate significance at the 1% and 5% level,
respectively.
First stage: probit model of firms reporting information on labor expenses
Variable
Firm size
Market-to-book ratio
Market leverage
Alternative market leverage
Alternative book leverage
Coefficient (standard error)
Coefficient (standard error)
Coefficient (standard error)
0.26nnn
(0.01)
0.03nnn
(0.003)
0.30nnn
(0.05)
0.26nnn
(0.01)
0.03nnn
(0.003)
0.26nnn
(0.01)
0.03nnn
(0.003)
–
–
0.15nnn
(0.05)
–
–
Physical capital intensity
Exchange dummies
Year effects
Industry effects
0.0004nnn
(0.0001)
0.49nnn
(0.03)
Jointly significant
Yes
Yes
–
–
(0.04)
0.27nnn
–
–
nnn
Average sales per employee
–
0.12nnn
(0.04)
0.0004nnn
(0.0001)
0.49nnn
(0.03)
Jointly significant
Yes
Yes
0.0004
(0.0001)
0.48nnn
(0.03)
Jointly significant
Yes
Yes
Second stage: average employee pay is the dependent variable
Market leverage
Alternative market leverage
0.20nnn
(0.04)
–
–
–
0.08nnn
(0.004)
0.001nnn
(0.0001)
0.003
(0.003)
0.047n
(0.027)
Yes
Yes
1.56nnn
(0.33)
0.08nnn
(0.004)
0.001nnn
(0.0001)
0.002
(0.003)
0.037
(0.027)
Yes
Yes
1.55nnn
(0.33)
0.20nnn
(0.04)
0.07nnn
(0.004)
0.001nnn
(0.0001)
0.008nn
(0.003)
0.037
(0.027)
Yes
Yes
1.59nnn
(0.33)
0.044nnn
(0.012)
0.043nnn
(0.012)
0.045nnn
(0.012)
Number of observations
Censored observations
Uncensored observations
49,357
44,088
5,269
49,357
44,088
5,269
49,357
44,088
5,269
Wald chi-square (p-value)
5,653.61
(0.0000)
5,688.90
(0.0000)
5,644.72
(0.0000)
Alternative book leverage
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
Year effects
Industry effects
Intercept
Inverse mills ratio (Lambda)
firms. Finally, CEOs in technology firms are less likely to
serve as the Chairmen of the board and are more likely to
be female than those in nontechnology firms.
Panel A of Table 11 reports the results from OLS regressions
of CEO compensation for nontechnology firms. Firm size is
positively related to all three types of compensation: cash,
equity-based, and total compensation. Market-to-book ratio is
negatively related to cash compensation, but positively related
to equity-based pay. One-year return is positive and
significant in the regressions of all three measures of CEO
compensation. CEO age is positive and significant in the
regression of total and equity-based compensation. Serving
as the chairman of the board increases the CEO's cash, equitybased, and total compensation. The leverage ratio has a
positive and significant effect on CEOs' cash, equity-based,
and total compensation, for all three measures of leverage.
Panel B of Table 11 presents the regression results for
technology firms. Size and CEO age are positive and significant
Author's personal copy
496
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Fig. 1. Changes in total labor expenses and tax benefit of debt as leverage
increases. This graph plots the changes in total labor expenses and tax benefit
of debt (in million $) for a firm with median values of leverage ratio, average
employee pay, total labor expenses, and total debt in the sample of Panel A,
Table 6. We start with the median value of market leverage, 0.20, and increase
it by 0.48 (more than two standard deviations), by increments of 0.04.
determinants of all three types of compensation. One-year
return has positive and significant influences on total and cash
compensation. A higher market-to-book ratio is associated
with a lower cash pay. Leverage has a positive and significant
effect on cash compensation, but not on equity-based or total
compensation.
We use a Wald test to examine whether the coefficient of
leverage is statistically different across the two groups. The
value of chi-square is 22.40 (10.89) with a p-value of 0.0000
(0.0010) for the regression of total compensation (equitybased compensation) on market leverage, and the value of
chi-square is 1.24 with a p-value of 0.26 for the regression of
cash compensation on market leverage. The Wald tests
suggest that the effect of leverage on total and equity-based
CEO compensation is different between technology and nontechnology firms, although the effect of leverage on CEO cash
compensation is not statistically different between the two
groups. Overall, the effect of leverage on CEO compensation is
greater for nontechnology firms than for technology firms,
consistent with our Hypothesis 5, thus providing further
support for the Titman-BSZ prediction.
6.2. Average employee pay in technology versus
nontechnology firms
In Table 12, we analyze the effect of leverage on average
employee pay in technology versus nontechnology firms.
Consistent with the existing literature, technology firms
have lower physical capital intensity and lower leverage
ratio than nontechnology firms. Technology firms are also
smaller than nontechnology firms, and they have smaller
sales per employee than nontechnology firms. The mean
of average employee pay is not significantly different
between the two groups, but the median of average
employee pay is greater for nontechnology firms. Although
the mean leverage ratio in technology firms is low, the
cross-sectional variation of leverage ratio is still large, e.g.,
alternative book leverage ranges from 0 to 0.90 with a
standard deviation of 0.21 (not tabulated). Similar to the
sample in Table 10, the technology firms in Table 12 also
have lower leverage than nontechnology firms. Different
from Table 10, an average technology firm in Table 12 is
significantly smaller than an average nontechnology firm.
The reason is that Tables 10 and 12 contain different
samples. Table 12 has 2,101 nontechnology and 298
technology observations due to the missing information
on “labor and related expenses” (data item 42) in Compustat, while Table 10 has 8,527 nontechnology and 2,345
technology observations from S&P 1,500 firms.
Table 13 presents the coefficients and standard errors
obtained from our OLS regressions of average employee
pay for nontechnology and technology firms. We find that
the leverage ratio has a positive and significant effect on
average employee pay for nontechnology firms. For technology firms, the coefficient on leverage is not statistically
significant. We use a Wald test to examine whether the
coefficient on leverage is statistically different across the
two groups. Wald tests suggest that alternative market and
book leverage ratios have differential effects on average
employee pay in technology versus nontechnology firms.
The results in this section demonstrate that leverage has a
greater effect on both CEO compensation and average
employee pay for nontechnology firms than for technology
firms, consistent with our Hypothesis 5. This is because
employees in nontechnology firms are more entrenched than
those in technology firms. Faced with a greater degree of
entrenchment, employees or CEOs in nontechnology firms are
more fearful of a potential bankruptcy. Therefore, in equilibrium, their compensation is more sensitive to their firm's
leverage ratio.
7. Additional robustness tests
In this section, we report additional empirical analysis
that examine the robustness of the previous results.
7.1. Issues relating to the use of panel data and the use of
alternative measures
In our empirical analysis, we use panel data sets, with a
significant number of firms appearing in multiple years. When
we estimate a linear model on panel data, the standard OLS
assumption of independence among the observations is very
likely violated. Therefore, we need to consider both a firm
effect and a time effect in our regressions. As defined by
Petersen (2009), the firm effect refers to the correlation within
the same firm across different years, and the time effect is the
cross-sectional correlation among different firms in the same
year. Petersen (2009) compares different approaches in estimating standard errors using financial panel data. He finds
that, in the presence of a firm effect only, clustering by firms
generates unbiased estimates of standard errors. In the
presence of both firm and time effects, clustering by firms
after including time dummies yields unbiased estimates of
standard errors.
Consistent with Petersen (2009), we find that the standard
errors of the estimated coefficients are significantly smaller if
we do not cluster them by firm. The difference in standard
errors strongly indicates the existence of a firm effect. We
control for both firm and time effects in our empirical analysis.
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
497
Table 10
CEO compensation in technology and nontechnology firms: univariate tests.
This table summarizes and compares CEO compensation and firm characteristics in technology and nontechnology firms. Technology firms are defined as
companies with primary SIC designations of 3570, 3571, 3572, 3576, 3577, 3661, 3674, 4812, 4813, 5045, 5961, 7370, 7371, 7372, and 7373. Nontechnology
firms are firms with SIC codes less than 4000 not otherwise categorized as technology firms. Total and Cash (salary plus bonus) compensations are
provided directly by Execucomp. We compute equity-based compensation as the total compensation minus salary, bonus, other annual pay, and long-term
incentive plan. Market leverage, alternative market leverage, and alternative book leverage are defined the same as in Table 1. Market capitalization is the
stock price multiplied by the number of shares outstanding as of the fiscal year end. Market-to-book ratio is the market capitalization divided by the book
value of equity. All continuous variables except leverage are winsorized at the 1st and 99th percentiles. All dollar amounts are adjusted to 1992 dollars
using the consumer price index.
Number of observations
Total compensation (thousands)
Cash compensation (thousands)
Equity-based compensation (thousands)
Market leverage
Alternative market leverage
Alternative book leverage
Market capitalization (millions)
Market-to-book ratio
One-year return to shareholders (%)
Years as CEO in the firm
CEO age
CEO is also the chairman
CEO is male
Nontechnology firms
mean (median)
Technology firms
mean (median)
8,527
2,690
(1,259)
1,025
(791)
1,512
(202)
0.21
(0.17)
0.20
(0.15)
0.32
(0.32)
4,585
(972)
3.30
(2.40)
16.16
(10.22)
5.99
(4.00)
66.62
(67.00)
0.67
(1.00)
0.99
(1.00)
2,345
3,310
(988)
780
(539)
2,439
(171)
0.08
(0.01)
0.07
(0.01)
0.13
(0.02)
6,413
(905)
4.36
(3.06)
25.72
(10.02)
6.10
(4.00)
60.40
(60.00)
0.51
(1.00)
0.97
(1.00)
In our multivariate regressions, we include year dummy
variables and cluster the standard errors by the firm. The
standard errors are also robust to heteroskedasticity. Another
way to control for both firm and time effects is clustering
by both year and firm. As a robustness test, we repeat our
analysis by adopting such an approach and find that the
results are very similar to those we report in earlier sections.
In our multivariate analysis of average employee pay, we
do not include operating income volatility as one of the
independent variables. Its computation requires five years of
data, which reduces our sample size. As a robustness test, we
now include this variable in our analysis. We find that our
results remain qualitatively the same and that the coefficient
of operating income volatility is insignificant. Further, in our
analysis of CEO pay, we have been using the total compensation including the value of options exercised. Our results
remain qualitatively the same if, as another robustness test,
the total compensation including the value of options granted
(instead of options exercised) is used in the analysis.
7.2. Leverage and the Z-score
An important assumption underlying the Titman (1984)
and the BSZ (2010) models is that higher leverage is associated
with a greater probability of bankruptcy, resulting in firms
t-Test
t-Stat
(p-value)
Kruskal-Wallis test
Chi-square
(p-value)
4.66
( o 0.0001)
13.40
( o 0.0001)
7.89
( o 0.0001)
36.11
( o 0.0001)
39.32
( o 0.0001)
39.16
( o 0.0001)
3.42
(0.0006)
13.70
( o 0.0001)
5.68
( o 0.0001)
0.67
(0.50)
31.59
( o 0.0001)
15.02
( o 0.0001)
4.84
( o 0.0001)
43.01
( o 0.0001)
441.65
( o 0.0001)
6.58
(0.01)
1,458.51
( o 0.0001)
1,401.87
( o 0.0001)
1,356.92
( o 0.0001)
5.48
(0.02)
186.70
( o 0.0001)
0.05
(0.82)
1.97
(0.16)
852.48
( o 0.0001)
221.10
( o 0.0001)
23.35
( o 0.0001)
with higher leverage having to compensate employees for the
effect of this increased probability of bankruptcy on their
human capital. Consequently, to further understand the role of
leverage on labor costs, we examine the correlation of
leverage with the Altman Z-score, which is a measure of a
firm's bankruptcy probability. In the sample of average
employee pay, the Pearson correlation coefficient between
the Z-score and alternative book leverage is 0.54; in the
sample of CEO compensation, the Pearson correlation coefficient between the Z-score and alternative book leverage is
0.47. As an additional robustness test, we replace leverage
with the Altman Z-score in our regressions and find that the
Z-score has a negative and significant impact on average
employee pay (at the 1% level). The Z-score also affects the
cash and total compensation of CEOs negatively and significantly, although the effect of the Z-score on CEOs' equitybased pay is not significant. These results are available upon
request.
The results of the robustness tests confirm that
leverage and bankruptcy probability are positively correlated and that the probability of bankruptcy affects both
average employee pay and CEO compensation. They add
support to the idea that the prospect of bankruptcy has an
important influence on the human capital costs incurred
by firms.
Author's personal copy
498
Table 11
OLS regressions of CEO compensation in technology and nontechnology firms.
Panel A and Panel B report the estimated coefficients and standard errors obtained from OLS regressions of CEO compensation in technology and nontechnology firms, respectively.
CEOPayi;t ¼ γ 0 þ γ 1 Sizei;t þ γ 2 Leveragei;t þ γ 3 MTBi;t þ γ 4 RET i;t þ γ 5 Agei;t þ γ 6 Tenurei;t þ γ 7 Chair i;t þ γ 8 MALEi;t þ εi;t :
Market leverage, alternative market leverage, and alternative book leverage are defined the same as in Table 1. The numbers in the parentheses are the standard errors. The standard errors are robust to
heteroskedasticity and are clustered by firm. We use Wald test to examine whether the coefficient on leverage is statistically different in the regressions of the two groups. nnn, nn, and n indicate significance at the
1%, 5%, and 10% level, respectively.
Variable
Log of cash
compensation
(5)
Log of equity-based
compensation (6)
Log of total
compensation
(7)
Log of cash
compensation
(8)
Log of equity-based
compensation (9)
–
–
–
–
–
–
Log of equity-based
compensation (3)
0.51nnn
(0.09)
–
0.61nnn
(0.07)
–
0.51nnn
(0.23)
–
–
–
–
0.51
(0.09)
–
0.61
(0.07)
–
0.53
(0.23)
–
0.41nnn
(0.01)
0.013nn
(0.005)
0.001nnn
(0.0002)
0.004
(0.003)
0.005nn
(0.002)
0.20nnn
(0.03)
0.02
(0.18)
Yes
Yes
2.97nnn
(0.27)
8,527
0.50
0.31nnn
(0.01)
0.01nnn
(0.003)
0.001nnn
(0.0001)
0.003
(0.002)
0.002
(0.002)
0.16nnn
(0.02)
0.003
(0.14)
Yes
Yes
3.62nnn
(0.23)
8,527
0.56
0.64nnn
(0.03)
0.04nnn
(0.01)
0.002nnn
(0.0006)
0.01
(0.01)
0.017nnn
(0.006)
0.26nnn
(0.08)
0.06
(0.42)
Yes
Yes
1.71nn
(0.71)
8,527
0.27
0.41nnn
(0.01)
0.010nn
(0.005)
0.001nnn
(0.0002)
0.004
(0.003)
0.005nn
(0.002)
0.21nnn
(0.03)
0.02
(0.18)
Yes
Yes
2.97nnn
(0.27)
8,527
0.50
0.31nnn
(0.01)
0.01nnn
(0.003)
0.001nnn
(0.0001)
0.003
(0.002)
0.002
(0.002)
0.16nnn
(0.02)
0.004
(0.14)
Yes
Yes
3.62nnn
(0.23)
8,527
0.56
0.65nnn
(0.03)
0.04nnn
(0.01)
0.002nnn
(0.0006)
0.01
(0.01)
0.017nnn
(0.006)
0.26nnn
(0.08)
0.06
(0.42)
Yes
Yes
1.72nn
(0.72)
8,527
0.27
Panel A: CEO compensation in nontechnology firms
Market leverage
Alternative market leverage
Alternative book leverage
Firm size
Market-to-book ratio
One-year return to
shareholders (%)
Years as CEO in the firm
CEO age
CEO is also the chairman
CEO is male
Year effects
Industry effects
Intercept
Number of observations
R-squared
nnn
nnn
nn
–
–
–
0.30nnn
(0.08)
0.40nnn
(0.01)
0.005
(0.005)
0.001nnn
(0.0002)
0.003
(0.003)
0.005nn
(0.002)
0.21nnn
(0.03)
0.03
(0.19)
Yes
Yes
3.07nnn
(0.27)
8,527
0.49
0.46nnn
(0.05)
0.30nnn
(0.01)
0.02nnn
(0.003)
0.001nnn
(0.0001)
0.003
(0.002)
0.002
(0.002)
0.16nnn
(0.02)
0.001
(0.14)
Yes
Yes
3.72nnn
(0.23)
8,527
0.56
0.37nn
(0.19)
0.64nnn
(0.03)
0.028nn
(0.013)
0.002nnn
(0.0006)
0.01
(0.01)
0.017nnn
(0.006)
0.26nnn
(0.08)
0.06
(0.42)
Yes
Yes
1.62nn
(0.71)
8,527
0.27
Panel B: CEO compensation in technology firms
Market leverage
Alternative market leverage
Alternative book leverage
Firm size
Market-to-book ratio
0.26
(0.24)
–
0.56nnn
(0.16)
–
0.96
(0.66)
–
–
0.31
(0.25)
–
nnn
0.47
(0.17)
–
–
–
–
0.91
(0.67)
–
–
–
0.18
(0.18)
0.40nnn
(0.03)
0.01
0.38nnn
(0.13)
0.26nnn
(0.02)
0.04nnn
0.47
(0.46)
0.65nnn
(0.07)
0.01
–
–
–
–
–
–
0.40nnn
(0.03)
0.01
0.26nnn
(0.02)
0.04nnn
0.65nnn
(0.07)
0.02
0.40nnn
(0.03)
0.01
0.26nnn
(0.02)
0.04nnn
0.65nnn
(0.07)
0.02
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Log of total
compensation
(4)
Log of cash
compensation
(2)
Log of total
compensation
(1)
Author's personal copy
6.99
(0.0082)
(0.28)
1.15
15.15
(0.0001)
(0.0015)
10.14
1.39
(0.24)
(0.0000)
23.65
10.89
(0.0000)
(0.0010)
1.24
(0.26)
22.40
Wald test whether the coefficient of leverage is different
between technology and nontechnology
firms: Chi-square (p-value)
Number of observations
R-squared
Intercept
CEO is male
Year effects
Industry effects
CEO is also the chairman
CEO age
Years as CEO in the firm
(0.01)
0.001nnn
(0.0003)
0.006
(0.005)
0.012nnn
(0.004)
0.12nn
(0.05)
0.15
(0.10)
Yes
Yes
4.36nnn
(0.28)
2,345
0.36
One-year return to shareholders (%)
(0.01)
0.0008n
(0.0005)
0.002
(0.007)
0.010n
(0.006)
0.13n
(0.07)
0.01
(0.30)
Yes
Yes
3.61nnn
(0.47)
2,345
0.32
(0.03)
0.001
(0.001)
0.012
(0.017)
0.03nnn
(0.01)
0.14
(0.17)
0.57
(0.66)
Yes
Yes
1.93n
(1.12)
2,345
0.17
(0.01)
0.0007n
(0.0004)
0.002
(0.007)
0.010n
(0.006)
0.13n
(0.07)
0.01
(0.30)
Yes
Yes
3.60nnn
(0.47)
2,345
0.32
(0.01)
0.001nnn
(0.0003)
0.006
(0.005)
0.011nnn
(0.004)
0.12nn
(0.05)
0.16
(0.10)
Yes
Yes
4.37nnn
(0.28)
2,345
0.36
(0.03)
0.001
(0.001)
0.012
(0.017)
0.03nnn
(0.01)
0.14
(0.17)
0.57
(0.67)
Yes
Yes
1.95n
(1.12)
2,345
0.17
(0.01)
0.0008n
(0.0005)
0.002
(0.007)
0.011n
(0.006)
0.13n
(0.07)
0.01
(0.30)
Yes
Yes
3.58nnn
(0.47)
2,345
0.33
(0.01)
0.001nnn
(0.0003)
0.006
(0.005)
0.011nnn
(0.004)
0.12nn
(0.05)
0.16
(0.10)
Yes
Yes
4.41nnn
(0.28)
2,345
0.36
(0.03)
0.001
(0.001)
0.012
(0.017)
0.032nn
(0.013)
0.14
(0.17)
0.56
(0.67)
Yes
Yes
1.99
(1.13)
2,345
0.17
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
499
8. Conclusion
Given the potentially large tax benefits of debt, why do
firms adopt a low level of leverage? The existing literature
shows that direct bankruptcy costs are not large enough to
justify the empirically observed low leverage ratios of firms.
Titman (1984) and Berk, Stanton, and Zechner (2010) argue
theoretically that a particular form of indirect bankruptcy
cost, namely, the incremental employee pay associated with
an increase in debt, is large enough to prevent firms from
increasing their leverage ratios. In this paper, we empirically
test this prediction. Specifically, we answer the following
questions: First, does higher leverage result in greater
employee compensation? Second, are the additional labor
costs associated with higher leverage large enough to offset
the incremental tax benefits of debt?
We conduct our empirical analysis using two measures
of employee compensation: the magnitude of CEO compensation and average employee pay. We find that the
effect of leverage on the magnitude of CEO compensation
is economically and statistically significant. For the whole
sample, leverage has a positive effect on cash, equitybased, and total compensation of CEOs in our multivariate
regressions. To establish causality, we also study the
relation between the compensation of newly appointed
CEOs who are hired from outside and firm leverage in the
year before their appointment. We find that leverage has a
significant effect on the magnitude of the compensation of
a new CEO. An increase of one standard deviation in
leverage corresponds to a 19% increase in the total compensation of a new CEO.
In both OLS and instrumental variable regressions, we find
that leverage also influences average employee pay positively
and significantly. Further, we show that, for a firm with the
median level of leverage, the incremental tax benefits arising
from increased leverage are offset by the additional labor costs
associated with such an increase. The effect of leverage on
average employee pay is positive and significant for financially safe firms, but the impact is insignificant for financially
distressed firms. We also find that, while leverage has a
positive and significant influence on CEOs' cash, equity-based,
and total compensation in nontechnology firms, it does not
have a significant influence on CEOs' total or equity-based
compensation in technology firms. The effect of the leverage
ratio on average employee pay is also greater in nontechnology firms than in technology firms. Because employees in
nontechnology firms can be viewed as more entrenched
(in the sense of BSZ (2010)), this provides additional support
for the Titman-BSZ prediction.
Our instrumental variable analysis to control for the
endogeneity of leverage has some limitations. In particular,
one potential criticism of the instrument we use, namely,
the marginal tax rate, is that the independent variation in
this variable can arise only from past losses and relatively
recent investments with investment tax credits, both of
which could independently influence wages. However,
even though the marginal tax rate could be an imperfect
instrument, it allows us to address the endogeneity of
leverage to a significant degree, in the absence of a natural
experiment that could have allowed us to account for the
potential endogeneity of leverage unambiguously.
Author's personal copy
500
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Table 12
Comparison of average employee pay in technology and nontechnology firms: univariate tests.
This table summarizes and compares average employee pay and financial characteristics in technology and nontechnology firms. Technology firms are
defined as companies with primary SIC designations of 3570, 3571, 3572, 3576, 3577, 3661, 3674, 4812, 4813, 5045, 5961, 7370, 7371, 7372, and 7373.
Nontechnology firms are firms with SIC codes less than 4000 not otherwise categorized as technology firms. Average employee pay is computed as labor
expenses (data item 42) divided by the number of employees (data item 29). Market leverage, alternative market leverage, and alternative book leverage
are defined the same as in Table 1. Market capitalization is the stock price multiplied by the number of shares outstanding as of the fiscal year end. Average
sales per employee is the amount of total sales (data item 12) divided by the number of employees (data item 29). Market-to-book ratio is the market
capitalization divided by the book value of equity. Physical capital intensity is computed as gross property, plant, and equipment scaled by total assets. All
continuous variables except leverage are winsorized at the 1st and 99th percentiles. All dollar amounts are adjusted to 1992 dollars using the consumer
price index.
Variable
Nontechnology firms
mean (median)
Technology firms
mean (median)
2,101
39.20
(40.01)
0.23
(0.19)
0.20
(0.16)
0.33
(0.32)
10,298
(2,631)
235.11
(174.55)
3.17
(2.24)
0.72
(0.67)
298
36.87
(34.56)
0.11
(0.17)
0.09
(0.01)
0.14
(0.03)
2,303
(187)
152.51
(130.40)
3.41
(2.16)
0.34
(0.25)
Number of observations
Average employee pay (thousands)
Market leverage
Alternative market leverage
Alternative book leverage
Market capitalization (millions)
Average sales per employee (thousands)
Market-to-book ratio
Physical capital intensity
t-Test
t-Stat
(p-value)
Kruskal-Wallis test
chi-square
(p-value)
1.55
(0.12)
10.35
( o 0.0001)
12.96
( o 0.0001)
14.62
( o 0.0001)
15.17
( o 0.0001)
9.72
( o 0.0001)
0.99
(0.32)
20.16
( o 0.0001)
8.62
(0.003)
186.79
( o 0.0001)
210.78
( o 0.0001)
231.97
( o 0.0001)
202.47
( o 0.0001)
60.53
( o 0.0001)
1.02
(0.31)
306.41
( o 0.0001)
Table 13
Leverage and average employee pay: OLS regressions for technology and nontechnology firms.
This table reports estimated coefficients and standard errors obtained from OLS regressions of average employee pay for technology and nontechnology
firms.
AEP i;t ¼ β0 þ β1 Sizei;t þ β2 Leveragei;t þ β3 AvgSalei;t þ β4 MTBi;t þ β5 PCI i;t þ εi;t :
Market leverage, alternative market leverage, and alternative book leverage are defined the same as in Table 1. The numbers in parentheses are the
standard errors. The standard errors are robust to heteroskedasticity and are clustered by firm. We use Wald test to examine whether the coefficient on
leverage is statistically different across the two groups. nnn And nn indicate significance at the 1% and 5% level, respectively.
Variable
Market leverage
Alternative market leverage
Alternative book leverage
Log of average
employee pay in
nontechnology
firms (1)
Log of average
employee pay in
technology firms
(2)
0.29n
(0.18)
0.06
(0.53)
–
Firm size
Average sales per employee
Market-to-book ratio
Physical capital intensity
Year effects
Industry effects
Intercept
Number of observations
R-squared
Wald test whether the coefficient of
leverage is different between
technology and nontechnology
firms: chi-square (p-value)
–
0.23
(0.06)
0.002nn
(0.001)
0.06nnn
(0.02)
0.03
(0.31)
Yes
Yes
2.40nn
(0.38)
298
0.32
nnn
0.10
(0.02)
0.001nnn
(0.0003)
0.005
(0.005)
0.13
(0.13)
Yes
Yes
1.55nnn
(0.12)
2,101
0.45
–
–
–
nnn
–
nnn
Log of average Log of average
employee pay in employee pay
in technology
nontechnology
firms (6)
firms (5)
–
0.08
(0.64)
–
nnn
0.10
(0.02)
0.001nnn
(0.0003)
0.003
(0.005)
0.13
(0.13)
Yes
Yes
1.59nnn
(0.13)
2,101
0.44
–
0.56
(0.20)
–
nnn
Log of average
employee pay in
technology firms
(4)
nnn
–
–
Log of average
employee pay in
nontechnology
firms (3)
0.23
(0.06)
0.002nn
(0.001)
0.06nnn
(0.02)
0.03
(0.30)
Yes
Yes
1.81nn
(0.75)
298
0.32
0.13
(0.48)
0.23nnn
(0.06)
0.002nn
(0.001)
0.06nnn
(0.02)
0.03
(0.31)
Yes
Yes
1.86nn
(0.76)
298
0.32
0.53
(0.17)
0.09nnn
(0.02)
0.001nnn
(0.0002)
0.007
(0.006)
0.15
(0.13)
Yes
Yes
1.59nnn
(0.11)
2,101
0.45
2.00
3.87
4.07
(0.16)
(0.05)
(0.04)
Author's personal copy
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
501
Table A1
Annual quits rates by industry and year.
The quits rate is obtained from the database of Job Openings and Labor Turnover Survey (JOLTS) provided by US Bureau of Labor Statistics. The data are
available from 2001. The quits rate is the number of quits (voluntary separations) during the entire year as a percent of annual average employment. The
industry classification is based on North American Industry Classification System (NAICS).
Quits rate
Industry
2001
2002
2003
2004
2005
2006
Mining and logging
Construction
21.8
28.7
18.9
27.5
17.1
26.4
19.1
28.1
19.3
32.2
20.6
29.2
Manufacturing
Durable goods
Nondurable goods
15.3
13.3
18.8
15.0
14.0
16.8
14.5
14.2
14.9
16.3
15.7
17.2
16.6
15.9
17.8
17.7
16.2
20.4
Trade, transportation, and utilities
Wholesale trade
Retail trade
Transportation, warehousing, and utilities
32.9
19.8
40.9
23.8
27.6
17.9
34.4
17.8
25.3
16.6
31.5
16.0
27.8
17.0
35.3
17.1
30.3
17.9
38.3
20.0
31.1
17.9
39.1
22.1
Information
27.9
21.2
18.6
18.3
22.9
26.1
Financial activities
Finance and insurance
Real estate and rental and leasing
20.7
18.6
26.7
18.1
15.6
25.4
17.6
14.7
26.2
20.0
17.0
28.9
19.2
17.3
24.4
21.2
19.1
27.3
Professional and business services
38.4
37.3
29.2
30.8
33.0
34.1
Education and health services
Educational services
Health care and social assistance
23
12.9
25
20.3
12.3
21.8
19.6
13.5
20.7
19.8
12.9
21.2
21.3
14.1
22.7
21.3
15.3
22.5
Leisure and hospitality
Arts, entertainment, and recreation
Accommodation and food services
59.7
39.1
63.4
52.8
38.2
55.3
48.6
31.3
51.7
50.3
29.4
54.0
55.5
34.7
59.1
57.1
32.1
61.4
Other services
27.3
27.2
26.0
30.0
32.2
25.0
Total
31.0
28.0
25.5
27.3
29.6
30.0
Appendix A
See Appendix Table A1.
References
Altman, E., 1968. Financial ratios, discriminant analysis, and the prediction of corporate bankruptcy. Journal of Finance 23, 189–209.
Anderson, M., Banker, R., Ravindran, S., 2000. Executive compensation in the
information technology industry. Management Science 46, 530–547.
Baldwin, C., 1983. Productivity and labor unions: an application of the
theory of self-enforcing contracts. Journal of Business 56, 155–185.
Benmelech, E., Bergman, N., Enriquez, R., 2009. Negotiating with labor
under financial distress. Unpublished working paper. Harvard
University, Cambridge, MA.
Berk, J., Stanton, R., Zechner, J., 2010. Human capital, bankruptcy, and
capital structure. Journal of Finance 65, 891–925.
Berkovitch, E., Israel, R., Spiegel, Y., 2000. Managerial compensation and
capital structure. Journal of Economics and Management Strategy 9,
549–584.
Bronars, S., Deere, D., 1991. The threat of unionization, the use of debt,
and the preservation of shareholder wealth. Quarterly Journal of
Economics 106, 231–254.
Brown, D., Fee, E., Thomas, S., 2009. Financial leverage and bargaining
power with suppliers: evidence from leveraged buyouts. Journal of
Corporate Finance 15, 196–211.
Brown, C., Medoff, J., 1989. The employer size-wage effect. Journal of
Political Economy 97, 1027–1059.
Chen, L., Zhao, X., 2006. On the relation between the market-to-book
ratio, growth opportunities, and leverage ratio. Finance Research
Letters 3, 253–266.
Cronqvist, H., Heyman, F., Nilsson, M., Svaleryd, H., Vlachos, J., 2009.
Do entrenched managers pay their workers more? Journal of Finance
64, 309–339.
Dasgupta, S., Sengupta, K., 1993. Sunk investment, bargaining, and
choice of capital structure. International Economic Review 34,
203–220.
Fama, E., MacBeth, J., 1973. Risk, return, and equilibrium: empirical tests.
Journal of Political Economy 81, 607–636.
Fama, E., Schwert, W., 1977. Human capital and capital market equilibrium. Journal of Financial Economics 4, 95–125.
Frydman, C., Jenter, D., 2010. CEO compensation. Unpublished working
paper. Massachusetts Institute of Technology, Cambridge, MA.
Gompers, P., Ishii, J., Metrick, A., 2003. Corporate governance and equity
prices. Quarterly Journal of Economics 118, 107–155.
Graham, J., 1996a. Debt and the marginal tax rate. Journal of Financial
Economics 41, 41–73.
Graham, J., 1996b. Proxies for the corporate marginal tax rate. Journal of
Financial Economics 42, 187–221.
Graham, J., Lemmon, M., Schallheim, J., 1998. Debt, leases, taxes, and the
endogeneity of corporate tax status. Journal of Finance 53, 131–162.
Graham, J., Tucker, A., 2006. Tax shelters and corporate debt policy.
Journal of Financial Economics 81, 563–594.
Hadi, A., 1992. Identifying multiple outliers in multivariate data. Journal
of the Royal Statistical Society Series B54, 761–771.
Hadi, A., 1994. A modification of a method for the detection of outliers in
multivariate samples. Journal of the Royal Statistical Society Series
B56, 393–396.
Hanka, G., 1998. Debt and the terms of employment. Journal of Financial
Economics 48, 245–282.
Heckman, J., 1979. Sample selection bias as a specification error. Econometrica 47, 153–161.
Hennessy, C., Livdan, D., 2009. Debt, bargaining, and credibility in firm–
supplier relationships. Journal of Financial Economics 93, 382–399.
Hovakimian, A., Kayhan, A., Titman, S., 2012. Are corporate default
probabilities consistent with the static trade-off theory? Review of
Financial Studies 25, 315–340.
Ittner, C., Lambert, R., Larcker, D., 2003. The structure and performance
consequences of equity grants to employees of new economy firms.
Journal of Accounting and Economics 34, 89–127.
Author's personal copy
502
T.J. Chemmanur et al. / Journal of Financial Economics 110 (2013) 478–502
Jaggia, P., Thakor, A., 1994. Firm-specific human capital and optimal
capital structure. International Economic Review 35, 283–308.
Leary, M., Roberts, M., 2010. The pecking order, debt capacity, and information asymmetry. Journal of Financial Economics 95, 332–355.
Matsa, D., 2010. Capital structure as a strategic variable: evidence from
collective bargaining. Journal of Finance 65, 1197–1232.
Molina, C., 2005. Are firms underleveraged? An examination of the
effect of leverage on default probabilities. Journal of Finance 60,
1427–1459.
Murphy, K., 1999. Executive compensation. In: Ashenfelter, O., Card, D.
(Eds.), Handbook of Labor Economics, vol. 3. , North Holland,
Amsterdam, pp. 2485–2563.
Murphy, K., 2003. Stock-based pay in new economy firms. Journal of
Accounting and Economics 34, 129–147.
Parsons, C., Titman, S., 2008. Empirical capital structure: a review.
Foundations and Trends in Finance 3, 1–93.
Perotti, E., Spier, K., 1993. Capital structure as a bargaining tool: the role
of leverage in contract renegotiation. American Economic Review 83,
1131–1141.
Petersen, M., 2009. Estimating standard errors in finance panel data sets:
comparing approaches. Review of Financial Studies 22, 435–480.
Stock, J., Wright, J., Yogo, M., 2002. A survey of weak instruments and
weak identification in generalized method of moments. Journal of
Business and Economics Statistics 20, 518–529.
Titman, S., 1984. The effect of capital structure on a firm's liquidation
decision. Journal of Financial Economics 13, 137–151.
Titman, S., Wessels, R., 1988. The determinants of capital structure choice.
Journal of Finance 43, 1–19.
Welch, I., 2011. Two common problems in capital structure research: the
financial-debt-to-assets ratio and issuing activity versus leverage
changes. International Review of Finance 11, 1–17.
Wooldridge, J., 2002. Econometric Analysis of Cross Section and Panel
Data. MIT Press, Cambridge, MA.