ARTICLE IN PRESS

Journal of Financial Economics 83 (2007) 297–319

www.elsevier.com/locate/jfec

The effect of state laws on capital structure$

John K. Walda,, Michael S. Longb
a
Department of Finance, College of Business Administration, The University of Texas at San Antonio,
One UTSA Circle, San Antonio, TX 78249, USA
b
Department of Finance, Rutgers Business School, Newark and New Brunswick, Rutgers University,
Newark, NJ 07102, USA
Received 2 August 2005; received in revised form 20 October 2005; accepted 28 October 2005
Available online 15 September 2006

Abstract

US manufacturing firms incorporated in states with stronger payout restrictions use less debt,
while antitakeover statutes do not significantly reduce long-run leverage. Correcting for the
endogenously determined choice of where to incorporate, we find that firms sort themselves
according to state laws and capital structure needs. After accounting for self-selection, state
antitakeover laws are positively associated with debt as a fraction of market value, possibly due to
lower market values for these firms. Payout restrictions appear to reduce leverage for firms that have
not reincorporated outside their home states. These constraints explain part of the negative relation
between profitability and leverage.
r 2006 Elsevier B.V. All rights reserved.

JEL classification: G32; K22

Keywords: Leverage; State laws; Profitability

$
The authors would like to thank Ivan Brick, Matthew Clayton, Dhammika Dharmapala, David Haushalter,
Laura Field, Michelle Lowry, Ileen Malitz, Dalia Marciukaityte, Fernando Zapatero, an anonymous referee, and
seminar participants at Moodys and at the 2004 Southern Finance Association meetings for comments on earlier
drafts. The authors also thank the Whitcomb Center at Rutgers University for data support.
Corresponding author.
E-mail address: john.wald@utsa.edu (J.K. Wald).

0304-405X/$ - see front matter r 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfineco.2005.10.008
 ARTICLE IN PRESS
298 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

1. Introduction

Determining the relation between legal institutions and firms’ financing decisions is often
difficult because, internationally, it requires an examination of the effects of tax systems,
bank structure, bankruptcy laws, general economic development, and other factors (see,
e.g., Rajan and Zingales, 1995; Booth, Aivazian, Demirguc-Kunt, and Maksimovij 2001).
Of course, firms incorporating in different US states will be subject to the same bankruptcy
procedures, as bankruptcy laws are national. Differences in state taxes should also have a
minimal impact on leverage because firms largely pay state taxes where they make their
profits,1 and most corporate taxes are federal. Lastly, US firms should have the same firm-
specific determinants of capital structure, have access to the same financial markets, and be
subject to the same S.E.C. regulations. However, antitakeover and payout restriction laws
vary across states, and companies are only subject to the statutes of the state in which they
are incorporated. Variation in antitakeover and payout restrictions across states thus
allows us to measure how these laws impact capital structure choice.
Entrenched firm managers could be better off reducing the riskiness of their firm beyond
what is optimal for firm value (Fama, 1980). Berger, Ofek, and Yermack (1999) analyze the
relation between managerial entrenchment and capital structure and show that more
entrenched managers choose lower levels of debt. Heron and Lewellen (1998) and Bebchuk
and Cohen (2003) suggest that antitakeover statutes are important in firms’ decisions
about where to incorporate, and that these statutes decrease the effectiveness of oversight
of managers. Garvey and Hanka (1999) explicitly study the impact of state antitakeover
legislation on leverage and find that firms incorporated in states that add antitakeover laws
significantly reduce their leverage, consistent with antitakeover legislation increasing
managerial entrenchment. In contrast, Comment and Schwert (1995) find that antitake-
over laws have little effect on the frequency of takeovers across states, although they do
increase the premiums firms pay in a takeover. The impact of differences in payout
restrictions across states on capital structure has not previously been addressed by the
literature.
We add to Garvey and Hanka’s analysis of antitakeover laws and capital structure by
extending their sample and by endogenizing the firm’s decision about where to
incorporate. We are therefore able to examine whether the decrease in leverage for firms
in states that enact antitakeover legislation persists, and how it relates to firms’
incorporation decisions. Because the minimum amount of capital required for a firm to
be able to make a payout (dividend payment, share repurchase, or other payout) also
varies with state laws, we control for this additional factor. Restrictions on payouts or
debt-to-capital ratios, if they are binding, would suggest that firms incorporated in states
with more stringent payout restrictions would have lower levels of debt. Further, if payout
restrictions are binding, firm profitability could be more negatively related to debt-to-value
ratios for firms incorporated in states with tighter restrictions. We therefore test whether
payout restrictions are binding, and whether they are able to explain part of the negative
relation between profitability and leverage.
In contrast to Garvey and Hanka (1999), we find that manufacturing firms incorporated
in states with more antitakeover laws do not choose lower levels of debt after 1987.
However, firms incorporated in states with more stringent restrictions on payouts choose

1
Some details on state taxes are available in Forsberg (2005).
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 299

significantly lower levels of debt. We then construct an empirical model where firm
managers choose both where to incorporate and the firm’s capital structure. Consistent
with Heron and Lewellen (1998) and Bebchuk and Cohen (2003), managers prefer states
with more antitakeover legislation in place. Additionally, managers prefer states with fewer
payout restrictions. After correcting for this self-selection, we find that antitakeover laws
are positively related to debt as a fraction of market value. This increase in leverage as a
fraction of market value (after endogenous sorting) is consistent with the decrease in
market values found for firms in jurisdictions that enact antitakeover statutes (see Karpoff
and Malatesta, 1989, 1995; Szewczyk and Tsetsekos, 1992). Similarly, payout constraints
appear to reduce leverage only for firms incorporated in their home state. Thus, while state
laws appear to affect firms’ capital structure decisions, much of their impact is on how
different firms sort themselves towards more restrictive antitakeover statutes or away from
binding payout constraints. Correcting for payout restrictions and this endogenous sorting
by firms accounts for part of the negative relation between profitability and leverage,
particularly when leverage is measured by the ratio of debt to market value.
In Section 2 we briefly describe the literature and the legal background for our
hypotheses. Section 3 describes our data and method, and Section 4 presents our results.
We conclude in Section 5. The Appendix presents some excerpts from state laws,
demonstrating the different restrictions on distributions in corporate statutes.

2. Background

The state of incorporation has been an issue since the late 19th century when New Jersey
relaxed its laws on capital structure and corporate combinations. This made New Jersey
the ideal state in the corporate charter market. By the late 1960s, this advantage had
shifted to Delaware. Romano (2005) explains why Delaware maintains its position as the
state of incorporation for over half of public firms. After the restructuring wave of the
1980s, both financial economists and legal scholars took interest in the state of
incorporation decision.2 In 1987 the Supreme Court decision CTS Corp. v. Dynamics
Corp of America (481 US 69) upheld the constitutionality of Indiana’s antitakeover
statutes. State legislatures then responded to corporate demand by providing a variety of
antitakeover laws.
Heron and Lewellen (1998) show that firms often reincorporate in order to erect
takeover defenses, even to the detriment of their shareholders. Karpoff and Malatesta
(1989, 1995) and Szewczyk and Tsetsekos (1992) find that firms incorporated in states that
enact antitakeover legislation suffer statistically significant stock price declines. Bebchuk
and Cohen (2003) also find that antitakeover statutes are significant in the decision about
where to incorporate and that states have an easier time retaining firms originally
2
Daines (2001) presents evidence that Delaware firms are worth more than non-Delaware firms in 12 out of 16
years between 1981 and 1996. In contrast, Gompers, Ishii, and Metrick (2001) find a negative correlation between
Delaware incorporation and firm value after controlling for a firm-level antitakeover index and other factors.
More recently, Subramanian (2004) considers firm valuation from 1991 through 2002 and finds that Delaware
firms with less than $50 million in sales are worth more during the period 1991–1996 but not afterwards. He finds
no valuation difference for large firms for any years in his sample. This result is interesting because larger firms
might have more flexibility on where they incorporate, and thus could be more likely to optimally choose a state of
incorporation. In contrast, smaller firms, which are almost always incorporated where they are originally located,
are worth more in Delaware before 1996.
 ARTICLE IN PRESS
300 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

incorporated in-state than in attracting new firms. Bebchuk and Cohen further discuss the
political and regulatory advantages firms gain from being incorporated in their home state
and summarize several theories of why firms remain in-state, including reduced costs of
incorporation, the uniformity of laws between states, local favoritism from state
legislatures or courts, and agency costs in choosing a non-local law firm.
Bebchuk and Cohen consider a number of statutes which impact managerial
entrenchment including an antitakeover index based on Gartman (2000) and two extreme
statutes that they term ‘‘recapture’’ and ‘‘stagger.’’ The antitakeover index increases if the
state has a control share statute, a fair-price statute, a no-freeze-out statute, a poison pill
endorsement statute, or a constituencies statute. Briefly, a control share statute requires a
hostile bidder to put its offer to a vote of shareholders early in the process. A fair-price
statute requires a bidder who gains control to pay remaining minority shareholders the
same price it paid for shares acquired in the original bid. No-freeze-out statutes restrict
bidders on merging assets for a specified number of years. Poison pills are rights that entitle
existing holders to significant value in the event of an acquisition without board approval.
Constituencies statutes allow managers to take into account interests of nonshareholders
in defending against a takeover (for further details see Bebchuk and Cohen, 2003;
Gartman, 2000).
States with a recapture statute require that unsuccessful acquirers disgorge all profits on
stock purchases. This law reduces the effectiveness of hostile takeovers as a managerial
discipline device, and was adopted in Ohio and Pennsylvania in 1990 (see, e.g., Szewczyk
and Tsetsekos, 1992). Massachusetts imposed staggered elections on boards of directors in
1990, thus reducing the effectiveness of director elections as a means for managerial change
(see Daines, 2001). Berger, Ofek, and Yermack (1999) find that firms with more entrenched
managers use less debt, and Garvey and Hanka (1999) find that firms subject to
antitakeover statutes have lower levels of leverage.
Restrictions on firm payouts also vary across states, typically by requiring a minimum
ratio between the amount of book capital and debt before a dividend payment or share
repurchase can be made. Restrictions on payouts have previously been studied in the
context of debt covenants. For instance, Kalay (1982) points out that firms would not
finance all the way to the restriction because they would want to retain some degree of
financial flexibility; Gilson and Warner (1997) offer some additional evidence that firms
value financial flexibility. Even without specific covenants from pre-existing debt, almost
all firms are currently subject to some form of payout restriction because such restrictions
have been incorporated into state laws.3 These restrictions on payouts exist in order to
avoid a moral hazard problem: if a firm borrows, it has an incentive to pay the full value of
the firm to shareholders and leave the creditors with nothing. Wald (1999) modifies a
traditional tradeoff model of capital structure by adding restrictions on payouts.4 Firms
subject to payout restrictions are limited in the amount of debt they can issue, and
therefore firms with a higher average product of capital, or higher profitability, will have
less debt as a fraction of value.5 This theory primarily reflects market values, but would
3
See Asquith and Wizman (1990) and Malitz (1994) on the disappearance of debt covenants. Eisenberg (1983)
mentions the reforms in state law corporate statutes after 1979. An introduction to these legal restrictions and
related cases is available in most law texts; see, for instance, Gevurtz (2000, pp. 157–167).
4
Berens and Cuny (1995) discuss how future profits limit debt use.
5
The following example from Wald (1999) illustrates how profitability is impacted by restrictions on payouts.
Consider an owner/entrepreneur who maximizes income and thus maximizes equity. His nascent firm can be
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 301

apply to book values if they are at some point restated to reflect market valuation. We
therefore hypothesize that firms incorporated in states with stricter restrictions will choose
lower levels of debt, and could have a more negative relation between profitability and
leverage. For other explanations for the negative relationship between profitability and
leverage, see Sarkar and Zapatero (2003), who show that mean-reverting earnings can
produce a negative relationship between profitability and leverage, and Hennessy and
Whited (2004), who show that taxes and floatation costs can produce this relation in a
dynamic model.
While it is possible that firms incorporated in less restrictive states would negotiate more
restrictive covenants with debtholders, as long as contracting is costly (see, e.g., Smith and
Warner, 1979; John and Kalay, 1982), firms in states with less restrictive laws would on
average still be subject to fewer restrictions.
Capital structure decisions are explained by a variety of theories, including the tradeoff
theory between tax savings and bankruptcy costs (see Scott, 1976; Kim, 1978; Barnea,
Haugen, and Senbet, 1981) and the pecking order of financing choice due to asymmetric
information between insiders and outsiders (Myers and Majluf, 1984). Our tests can be
seen as a further examination of the tradeoff theory, which has found support in Bradley,
Jarrell, and Kim (1984) and more recently in Hovakimian, Opler, and Titman (2001) and
Frank and Goyal (2003).

3. Data and method

We examine the impact of state laws on capital structure using a sample of

manufacturing firms, SIC codes 2000–3999, available in the COMPUSTAT database.
Because the constitutionality of state antitakeover laws was uncertain until 1987, we limit
our primary analysis to data from 1987 through 2003. We trim the sample so that no
variables are missing for any firm/year and to remove outliers: we delete observations if
debt is more than 95% of market value, debt is more than 800% of total assets, R&D is
more than 200% of market or book value, advertising is more than 100% of market value
or more than 200% of book value, property, plant and equipment is more than 200% of
market or book value, depreciation is more than 100% of market or book value, the
absolute value of profitability is more than 200% of market value or more than 500% of
book value, or the market-to-book ratio is more than 100. (Windsorizing the data at 1%
does not significantly change our results.) This provides us with an unbalanced panel of
5,171 firms and 40,602 observations. Table 1 provides a summary of the variables used in
the analysis. We consider the firm’s debt-to-market-value ratio and the debt-to-total-assets
ratio as our measures of leverage, where debt is the sum of long-term and short-term debt.

(footnote continued)
described by the following investment opportunity: With a $50 capital investment, a sure return of $110 is
produced next year. If the cost of debt and equity are both 10%, the owner could borrow $50 in debt at the market
rate, and collect the remaining $55 as net revenue next year. This net revenue is worth $50 as current equity
interest. In a Modigliani and Miller world, the owner could finance his firm by borrowing $100 of debt and paying
himself $50 in the beginning of the period. The firm would then be all debt financed. However, the $50 payment to
the owner is illegal under the net worth or balance sheet restriction because it reduces the firm’s capital below the
level of debt. The all-debt financing implies a payment to shareholders that reduces the value of assets below the
face amount of outstanding debt, and is therefore illegal. Thus, the minimum equity value of this firm can never be
reduced below $50 as long as this investment is undertaken.
 ARTICLE IN PRESS
302 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

Table 1
Summary statistics
Total asset constraint equals the minimum ratio of assets to liabilities. Home equals one if the firm’s
headquarters is in the same state as the firm’s legal incorporation. The antitakeover index is based on McGurn,
Pamepinto, and Spector (1989) and Gartman (2000). Recapture equals one if the firm is incorporated in a state
(Ohio or Pennsylvania after 1989) that requires firms to disgorge short-term profits in the event of a hostile
takeover. The sample includes 40,602 firm-year observations for 5,171 manufacturing firms from 1987 through
2003.

Variable Mean Standard deviation

Market value variables

Debt/total market value 0.189 0.223
R&D/ market value 0.052 0.098
Advertising/ market value 0.013 0.046
Property, plant, & equipment/market value 0.462 0.426
Depreciation/ market value 0.043 0.043
Profitability/ market value 0.039 0.211
Size—log(market value) 4.707 2.184
Log(employees) (in thousands) 0.637 2.151
Market value/book value 2.162 3.784
Book value variables
Debt/book value 0.197 0.270
R&D/book value 0.085 0.162
Advertising/book value 0.015 0.049
Property, plant, & equipment/book value 0.467 0.306
Depreciation/book value 0.046 0.035
Profitability/book value 0.003 0.394
Size—log(book value) 4.386 2.175
Other variables
Home 0.361 0.480
Antitakeover index 2.046 1.587
Recapture 0.038 0.191
Total asset constraint 0.426 0.512
Total asset constraint ¼ 1 0.388 0.487
Total asset constraint ¼ 1.25 0.055 0.228

The average debt-to-market-value ratio is 0.189 while the average debt-to-total-assets ratio
is 0.197, consistent with similar studies.
As in Bebchuk and Cohen (2003), we include the index of antitakeover statutes from
Gartman (2000). We extend this variable historically by collecting additional antitakeover
statutes from McGurn, Pamepinto, and Spector (1989) for the late 1980s. Takeover laws
change slowly over this time period; the correlation between the antitakeover index for
1989 and 2000 is 0.75. Firms in some states (such as Pennsylvania) are able to optout of
some of the antitakeover laws. We gather additional data on individual firm decisions from
the Investor Responsibility Research Center (IRRC) data base. Because the IRRC data set
is more limited than COMPUSTAT, we probably miss some firms’ decisions to optout of
antitakeover statutes; however, given the small fraction of total firms that can and do
optout, we believe any selection bias from these missing values is small. We gather data on
state statutes restricting payouts from a variety of sources, primarily Lexis/Nexis.
Approximately 54.5% of our observations are for firms incorporated in Delaware, with
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 303

California and New York the next most frequently chosen states for incorporation at 5.5%
and 4.6% of the sample, respectively. We follow Bebchuk and Cohen (2003) in defining
home as equal to one if the firm’s main headquarters is in the same state as the state of
incorporation and zero otherwise. While COMPUSTAT has variables for the firm’s main
headquarters and for the current state of incorporation, we gather details on historical
reincorporation decisions by searching Mergent online.
Because the debt-to-value ratios are censored at zero for approximately 17% of the
sample, we employ tobit regressions (the results are similar using ordinary least squares).
Since the observations make up a panel, we expect that the errors will be autocorrelated;
that is, the firm’s capital structure decisions will be correlated across time because of
transaction costs and unobserved variables. We therefore bootstrap the standard errors by
estimating the coefficients 1,000 times from the resampled data set with clustering of
observations by firm. Thus, each sample in a random draw is a set of observations for a
given firm, and this procedure captures the correlation structure within firms. For
additional robustness checks, we also examine only data after 1990 (when further state
laws had been upheld by higher courts) and only cross-sectional regressions for a few
years, with results that are consistent with our full sample.
Besides these tobit regressions, we also consider a two-stage self-selection model in
which the coefficients are allowed to vary between firms that incorporate in their home
states and those that incorporate in other states. Self-selection models, originally suggested
by Heckman (1976), are described in detail in Maddala (1983). Maddala specifies a choice
between two regimes:
Regime 1:
yi ¼ b001 X 1i þ u1i iff g0 Z i Xvi , (1)
Regime 2:
yi ¼ b002 X 2i þ u2i iff g0 Z i ovi , (2)
where vi is correlated with u1i and u2i. In our case, we consider the two regimes to be
whether the firm is incorporated in its home state or has incorporated away from its
headquarters. Thus, home equals one if gZiXvi, and zero otherwise. Consistent estimation
of the parameters in Eqs. (1) and (2) can be accomplished through a two-stage procedure,
where we first estimate a probit on whether home equals one, and then calculate inverse
Mills ratios:
W 1i ¼ fðg0 Z i Þ=Fðg0 Z i Þ, (3)

W 2i ¼ fðg0 Z i Þ=½1  Fðg0 Z i Þ. (4)

We can then estimate
yi ¼ b01 X 1i  s1v W 1i þ 1i for home ¼ 1, (5)

yi ¼ b02 X 2i  s2v W 2i þ 2i for home ¼ 0. (6)

The covariance between vi and u1i equals s1v, and the covariance between vi and u2i equals
s2v. In our case, Eqs. (5) and (6) are estimated with a tobit because of lower censoring.
Standard errors in these regressions need to be adjusted for self-selection, and thus we
again calculate standard errors using a bootstrap with clustering by firms and re-estimating
both the self-selection probit and the tobit with each draw of observations.
 ARTICLE IN PRESS
304 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

Our independent variables are scaled by either the market or book value of the firm to
correspond with the variable used to scale the dependent variable. The independent
variables include the firm’s research and development expenses (R&D) scaled by firm
value. Because many firms do not undertake enough R&D to require reporting it, we set
R&D equal to zero if this variable is missing in COMPUSTAT. This variable captures
both the intangible nature of the firm’s assets and the specificity of the firm’s products (see
Long and Malitz, 1985; Titman and Wessels, 1988). Similarly, we include the advertising-
to-value ratio. We include the firm’s property, plant, and equipment (PP&E) scaled by the
firm’s value to capture the firm’s tangible property. A firm’s depreciation tax shields can
also impact the optimal debt ratio (DeAngelo and Masulis, 1980), and we therefore include
scaled depreciation as an additional independent variable. We include the log of the firm’s
value (as measured by either total market value or total asset value) to capture any size
effects, the market-to-book ratio (or average Tobin’s Q) to capture growth opportunities,
and year and industry dummies for two-digit SIC codes. (As a robustness check, in
unreported regressions, we use the average industry debt-to-value ratio instead of industry
dummies, with similar results.)
As in the existing literature, we examine the relation between profitability and capital
structure, where we define profitability as the firm’s EBITDA divided by either the book
value of total assets or market value, depending on whether we define leverage as debt to
book value or as debt to market value. In order to include a measure of cash flow risk, we
also consider the standard deviation of first-differences in the EBITDA-to-value ratio for
the prior ten years. Because this variable is missing for a significant fraction of the
observations, and because it does not significantly change our other results, our reported
regressions do not include this variable.
We include a dummy variable for whether a firm is subject to the recapture statute (i.e.,
firms incorporated in Ohio and Pennsylvania after 1990), and we use the IRRC data to
determine whether a firm has opted out of these antitakeover provisions. Bebchuk and
Cohen (2003) consider the Massachusetts staggered board provision separately; however,
approximately 60% of the firms covered by IRRC have staggered board elections
regardless of the state of incorporation in 2004. Because so many firms have adopted
staggered board elections, we do not believe this statute serves as an important additional
antitakeover provision.6 Including a separate dummy for incorporation in Massachusetts
does not change our other results.
We consider regressions that include the minimum asset-to-liability ratio required by
state law for the firm to pay out funds, as well as dummy variables that classify states
according to the type of restrictions. Specifically, New York and many other states require
a minimum one-to-one ratio between book capital and debt before a payout can be made,
so we set the total asset constraint equal to one for firms incorporated in these states.
Alaska and California have laws specifying a minimum 1.25 ratio of the book value of
capital to debt for a payout. The most important state without either of these restrictions is
Delaware, and thus the total asset constraint is set to zero for firms incorporated in

6
Mayers and Smith (2005) find that mutual insurance company boards are staggered more frequently than their
non-mutual counterparts. As mutual insurance companies are not subject to takeovers, Mayers and Smith suggest
that a major purpose of staggered boards is to foster board continuity, increase long-term business focus, and
increase firm-specific human capital investment by directors.
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 305

Delaware. Delaware, New York, and California laws on payouts are included in the
Appendix.
We also consider the interactions between the minimum capital-to-debt ratio and firm
profitability or the state dummies and firm profitability. These interaction terms let us see if
firms in more restrictive jurisdictions have a more negative relation between current
profitability and capital structure.

4. Empirical analysis

The most striking aspect of the relation between capital structure and the state of
incorporation is how clearly some differences are reflected in the raw data. In Table 2A, we
provide a comparison of mean capital structures for firms incorporated in states with
forced recapitalization (i.e., Ohio and Pennsylvania) and for firms incorporated in states
with different restrictions on payouts. While the former show only slight differences from

Table 2A
Capital structure by state law: two-sided t-tests, unequal variances assumed
Recapture is equal to one if the firm is incorporated in a state (i.e., Pennsylvania and Ohio after 1989) that
requires recapture of profits in the event of a hostile takeover. Total asset constraint equals the minimum ratio of
assets to liabilities.

Number of Legal characteristic Difference in means

observations (t-statistic)
(equal ¼ 0/
equal ¼ 1)

For no recapture For recapture Difference between

no recapture and
recapture

Debt/market value 39,063/1,539 0.188 0.203 0.014

(0.224) (0.196) (2.780)
Debt/book value 39,063/1,539 0.197 0.197 0.000
(0.273) (0.181) (0.048)
For no total asset For total asset Difference between
constraint constraint ¼ 1.0 unconstrained and
(Standard total asset
deviation) constraint ¼ 1.0

Debt/market value 22,628/15,743 0.198 0.189 0.009

(0.230) (0.216) (3.766)
Debt/book value 22,628/15,743 0.207 0.193 0.014
(0.263) (0.277) (4.852)
For No total asset For total asset Difference between
constraint constraint ¼ 1.25 unconstrained and
(Standard total asset
deviation) constraint ¼ 1.25

Debt/market value 22,628/2,231 0.198 0.094 0.104

(0.230) (0.166) (27.114)
Debt/book value 22,628/2,231 0.207 0.121 0.086
(0.263) (0.288) (13.523)
 ARTICLE IN PRESS
306 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

Table 2B
Capital structure deviation from industry/year means by state law: two sided t-tests, unequal variances assumed
Recapture is equal to one if the firm is incorporated in a state (i.e., Pennsylvania and Ohio after 1989) that
requires recapture of profits in the event of a hostile takeover. Total asset constraint equals the minimum ratio of
assets to liabilities. Deviations from debt/total market value or debt/book value are deviations from the average
for other firms in the same two-digit SIC code and year.

Number of Legal characteristic Difference in means

observations (t-statistic)
(equal ¼ 0/
equal ¼ 1)

For no recapture For recapture Difference between

no recapture and
recapture

Deviation in debt/ 39,063/1,539 0.000 0.009 0.009

market value (0.209) (0.192) (1.844)
Deviation in debt/ 39,063/1,539 0.000 0.002 0.002
book value (0.268) (0.179) (0.327)
For no total asset For total asset Difference between
constraint constraint ¼ 1.0 unconstrained and
(Standard total asset
deviation) constraint ¼ 1.0

Deviation in debt/ 22,628/15,743 0.009 0.003 0.012

market value (0.212) (0.206) (5.724)
Deviation in debt/ 22,628/15,743 0.008 0.004 0.012
book value (0.256) (0.273) (4.392)
For no total asset For total asset Difference between
constraint constraint ¼ 1.25 unconstrained and
(Standard total asset
deviation) constraint ¼ 1.25

Deviation in debt/ 22,628/2,231 0.009 0.072 0.081

market value (0.212) (0.165) (21.543)
Deviation in debt/ 22,628/2,231 0.008 0.057 0.066
book value (0.256) (0.284) (10.507)

firms incorporated in other jurisdictions, firms incorporated in states with a minimum 1.25
capital-to-debt ratio (i.e., Alaska and California) have 9.4% debt as a fraction of market
value on average versus 19.8% for firms incorporated in states with no payout restrictions.
While many of the differences in mean capital structures between states are statistically
significant, these are clearly economically significant as well, and the question then is
whether part of these differences is due to state laws or to firm-specific characteristics such
as the prevalence of high-tech, low debt-to-value firms in California. Additionally, as we
show below, some of these differences can also be explained by the endogenous
incorporation choice.
As a first approximation, we control for industry and time variation by considering
deviations from mean leverage levels for the given industry (as given by the two-digit
SIC code) in a given year. We compare these deviations across different jurisdictions in
Table 2B. The results are consistent with those in Table 2A, although they are slightly
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 307

more muted. For instance, the average debt as a fraction of market value for states with a
minimum 1.25 capital-to-debt ratio is 7.2 percentage points below the mean for that
industry/year. The difference between this sample and the sample of firms incorporated in
states with no payout restrictions is 8.1 percentage points with a t-statistic of 21.5. Thus,
industry and year averages alone are insufficient to explain the differences in average
leverage across jurisdictions.

4.1. Reincorporations

We next examine capital structure changes for firms that reincorporate from a state with
one type of statute to another. A total of 213 firms change their state of incorporation
during the years in our sample, and out of those, 81 firms change from states with a total
asset constraint equal to 1.0 to a total asset constraint equal to zero; 112 firms move from
states with a total asset constraint equal to 1.25 to those with a constraint equal to zero.
(No firms move from states with a 0 or 1.0 constraint to one with a 1.25 constraint, and
seven firms move from states with a 0 constraint to a state with a 1.0 constraint.) To see
whether the relaxation in the capital constraint associated with these reincorporations
corresponds with an increase in the firms’ leverage, we graph the average leverage for the
years around the reincorporation in Fig. 1. Both the debt-to-market-value and debt-to-
book-value ratios increase for the years around reincorporation, suggesting that firms that
switch are constrained on average and rebalance their capital structure to reflect their new
legal environment.
In order to control for average year and industry capital structures, we look at the
deviations from average debt-to-value ratios for an industry in a given year in Fig. 2.
Again, considering average deviations, firms that reincorporate in states with less
restrictive payout laws increase their leverage. The typical reincorporation captured in

Change in Debt/Book for Total Asset

0.07 Constraint Change from1.0 to 0
Change in Debt/Market for Total Asset
Constraint Change from1.0 to 0
Increase in Leverage in Percentage Points

0.06 Change in Debt/Book for Total Asset

Constraint Change from1.25 to 0
0.05 Changein Debt/Market for Total Asset
Constraint Change from1.25 to 0

0.04

0.03

0.02

0.01

0
1 2 3 4 5 6
-0.01 Number of Years around a Reincorporation

-0.02

Fig. 1. Mean changes in capital structure around a reincorporation from a more restrictive to a less restrictive
jurisdiction.
 ARTICLE IN PRESS
308 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

Change in Deviations of Debt/Book for Total

Asset Constraint Change from 1.0 to 0
0.06 Change in Deviations of Debt/Market for Total
Change in Leverage in Percentage Points Asset Constraint Change from 1.0 to 0
Change in Deviations of Debt/Book for Total
0.05 Asset Constraint Change from 1.25 to 0
Change in Deviations of Debt/Market for Total
0.04 Asset Constraint Change from 1.25 to 0

0.03

0.02

0.01

0
1 2 3 4 5 6
-0.01 Number of Years around a Reincorporation

-0.02

Fig. 2. Mean changes in capital structure around a reincorporation from a more restrictive to a less restrictive
jurisdiction in deviations from industry/year averages.

these graphs is a move by a firm from California (minimum 1.25 ratio of capital to debt) or
Texas (minimum 1.0 ratio of capital to debt) to Delaware (no minimum restriction).
Because California and Texas have an antitakeover index lower than or equal to that of
Delaware (which has an antitakeover index of only 1.0 after 1988) over this time period,
the increase in leverage cannot be due to a decrease in antitakeover provisions.

4.2. Capital structure regressions

Table 3 presents tobit regressions on debt to market value and debt to book value for
our unbalanced panel. The first column presents debt-to-market-value regressions for our
basic variables, the antitakeover statute index, a dummy if the state of incorporation has
forced recapitalization, the minimum capital-to-debt ratio (total asset constraint), and the
interaction between this constraint and profitability. The second column replaces the total
asset constraint with dummies for whether the minimum capital-to-debt constraint equals
1.0 or 1.25, and interactions between these dummies and profitability. The third and fourth
columns present the same regression but replacing market value with book value as the
denominator of the dependent and independent variables (and size is defined as the
logarithm of market value or book value, respectively, for these regressions). As Barclay,
Morellec, and Smith (2005) point out, debt-to-book-value ratios should be interpreted as
the ratio of debt to the value of assets in place, rather than as an alternative measure of
leverage.
We consider several standard variables such as research and development (R&D),
property, plant and equipment (PP&E), advertising, depreciation, profitability, size, and
market-to-book. Consistent with theory and prior results, we find negative coefficients on
R&D and advertising, and positive coefficients on PP&E and the number of employees.
The coefficient on profitability is negative and significant, particularly in the debt-to-book-
value regressions. The coefficient on depreciation is positive, possibly because of the high
correlation between physical assets and depreciation, and the coefficient on size is negative
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 309

Table 3
Capital structure regressions

Debt/market Debt/book value

value

With total asset With dummies With total asset With dummies for
constraint for total asset constraint total asset
constraints constraints

Firm characteristics
R&D 0.484c 0.480c 0.319c 0.317c
(15.645) (15.655) (6.668) (6.437)
Advertising 0.117b 0.118b 0.182b 0.183b
(2.066) (2.107) (2.136) (2.162)
PP&E 0.139c 0.138c 0.140c 0.138c
(13.588) (13.334) (6.959) (6.674)
Depreciation 0.503c 0.507c 0.642c 0.650c
(5.758) (5.791) (4.480) (4.402)
Profitability 0.040c 0.039b 0.173c 0.173c
(2.644) (2.551) (6.997) (6.690)
Size 0.042c 0.042c 0.027c 0.026c
(13.205) (12.338) (5.575) (5.403)
Log(employees) 0.056c 0.056c 0.048c 0.048c
(16.690) (16.061) (9.769) (9.289)
Market-to-book 0.005c 0.005c 0.003b 0.003b
(6.787) (6.936) (2.300) (2.368)

State law variables

Antitakeover index 0.003 0.003 0.004 0.005
(1.463) (0.902) (1.228) (1.229)
Recapture 0.017 0.012 0.004 0.003
(1.278) (0.841) (0.289) (0.177)
Total asset constraint 0.045c 0.040c
(6.588) (4.686)
Total asset constraint*profitability 0.054c 0.051
(2.820) (1.633)
TA constraint ¼ 1 0.024b 0.011
(2.387) (0.823)
TA constraint ¼ 1.25 0.079c 0.083c
(6.094) (5.106)
TA constraint ¼ 1*profitability 0.056b 0.045
(2.437) (1.231)
TA constraint ¼ 1.25*profitability 0.079b 0.096a
(2.201) (1.797)

Tobit regressions where the dependent variable is either total debt/total market value or total debt/book value.
The independent variables R&D, advertising, PP&E, depreciation, and profitability are all normalized by the
same value as the dependent variable, i.e., either market or book value. Size equals log(Book Value) in the book
value regression, and log(Market Value) in the market value regression. The Antitakeover Index applies to the
state of incorporation and is based on McGurn, Pamepinto, and Spector (1989) and Gartman (2000). Recapture
equals one if the firm is incorporated in a state (Ohio or Pennsylvania after 1989) that requires bidding firms to
disgorge short-term profits in the event of an unsuccessful hostile takeover. The total asset constraint equals the
minimum book value of assets required for a payout to occur—either 1.0 or 1.25 depending on the state of
incorporation. All regressions include year and two-digit SIC code industry dummies. t-statistics are reported in
parentheses, and are calculated using bootstrap with clustering by firm and 1,000 repetitions. The sample includes
40,602 firm-year observations for 5,171 manufacturing firms from 1987 through 2003 for all regressions. 7,090
firm-years in this sample have zero debt. Superscripts a, b, and c denote significance at the 10%, 5%, and 1%
levels, respectively.
 ARTICLE IN PRESS
310 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

and significant, but this variable is also highly correlated with log(Employees) which has a
positive and significant coefficient. All regressions also include dummies for year and
industry.
Bebchuk and Cohen (2003) suggest that an antitakeover index and the recapture statutes
are particularly important in decreasing the effectiveness of oversight of managers. In
Table 3 we examine whether firms incorporated in states with these laws choose different
capital structure levels. Garvey and Hanka (1999) use a dummy variable equal to one if the
state has any antitakeover legislation, whereas we use an antitakeover index as in Bebchuk
and Cohen (2003). Using a dummy variable in our regressions does not change our results.
In contrast to Garvey and Hanka, we find that neither the antitakeover index nor the
recapture dummy is significant in any of the regressions. However, the coefficient on the
antitakeover index becomes negative and significant if, like Garvey and Hanka, we
consider the 1983–1993 time period, and if the total asset constraint variables are not
included (these regressions are available upon request). This implies that Garvey and
Hanka’s findings are, as they also suggest, at least partially due to unusually high levels of
leverage for firms in some states prior to the imposition of antitakeover legislation.
Overall, we find no evidence that antitakeover legislation is associated with a decrease in
leverage in the period 1987–2003.
Table 3 also provides coefficients on the total asset constraint, parameterized either as
the minimum capital-to-debt ratio (0, 1.0, or 1.25) or with two dummy variables that
capture the different types of laws. Except for the coefficient on the dummy for the
total asset constraint equal to 1.0 in the debt-to-total-assets regression, the coefficients
on the constraints are negative and significant at the 5% or 1% level. Examining
the coefficients on whether the constraint equals 1.0 or 1.25 also reveals significantly
more negative coefficients for 1.25 than 1.0. Lastly, the coefficients on the interaction
between profitability and the total asset constraint are all significant for the debt-
to-market-value regression, but only significant at the 10% level in one case (in the
regression with the dummy for the total asset constraint equal to 1.25) when considering
debt-to-book-value. Overall, state laws appear to impact capital structure choice, with
firms in states with stricter capital restrictions choosing lower levels of debt. Further, more
profitable firms choose lower levels of debt-to-market value in states with stricter capital
restrictions.
If we do not include the total asset constraint variables or their interactions, the
coefficients on profitability are more negative at 0.068 in the debt-to-market regression
and 0.202 in the debt-to-book regression, with t-statistics of 5.628 and 10.161,
respectively (these regressions are available upon request). Thus, including state laws
decreases the magnitude and significance of the profitability coefficients, and state laws
appear to explain part of the relation between profitability and leverage.

4.3. A self-selection model

Instead of state restrictions being binding on firms, firms could instead be sorting away
from states when statutes are costly. In this case, the firm makes a simultaneous decision
about its capital structure and its state of incorporation, and any simple regression, such as
those given in Table 3, is potentially biased. As we report in Table 1, 36.1% of our firm-
year observations are for firms incorporated in their home states. Of those not
incorporated in their home states, 84.8% are in Delaware, but this still provides us with
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 311

Table 4
Probit regression on whether the firm chooses to incorporate in-state

Independent variable Probit regression

Firm characteristics
Log(market value) 0.016
(0.392)
Market-to-book 0.001
(0.110)
PP&E/market value 0.149a
(1.761)
Depreciation/market value 1.802c
(2.935)
EBITDA/market value 0.060
(0.721)
R&D/market value 0.563c
(2.780)
Advertising/market value 0.205
(0.465)
Tech 0.871a
(1.681)
Log(employees) 0.139c
(4.988)
Log(book value) 0.231c
(4.789)
PP&E/book value 0.312c
(3.025)
Depreciation/book value 0.495
(0.770)
EBITDA/book value 0.411c
(6.563)
R&D/book value 0.207
(1.417)
Advertising/book value 0.020
(0.043)
State law variables
State antitakeover index 0.113c
(5.821)
State has recapture 0.205b
(2.331)
State’s total asset constraint ¼ 1 0.382c
(2.775)
State’s total asset constraint ¼ 1.25 0.225
(1.392)

The dependent variable is one if the home state equals the state of incorporation, and zero otherwise. The state
antitakeover index applies to the state where the firm is headquartered, and is based on McGurn, Pamepeuto, and
Spector (1989) and Gartman (2000). State recapture equals one if the state where the firm has headquarters (Ohio
or Pennsylvania after 1989) requires bidding firms to disgorge short-term profits in the event of a hostile takeover.
The total asset constraint dummies equal the minimum book value of assets required for a payout to occur in the
state where the firm is headquartered. Tech is a dummy equal to one if the SIC code is between 3500 and 3699, or
3800 and 3899. Other two-digit SIC code industry dummies and year dummies are also included. t-statistics are
reported in parentheses, and are adjusted for clustering by firm. The sample includes 40,602 firm-year
observations for 5,171 manufacturing firms from 1987 through 2003. 14,659 firm-years are for firms incorporated
within their home states. Superscripts a, b, and c denote significance at the 10%, 5%, and 1% levels, respectively.
 ARTICLE IN PRESS
312 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

adequate variability in the state law variables, with 3,936 firm-year observations for firms
incorporated out-of-state and not in Delaware.
Table 4 shows the estimated coefficients from a probit model of whether a firm chooses
to incorporate out of its home state. We follow Heron and Lewellen (1998) and Bebchuk
and Cohen (2003) in our choice of variables, and where possible include variables
normalized by both market and book values. As in Heron and Lewellen, we report the
coefficient on the technology industry dummy (equal to one if the SIC code is between
3500–3599, 3600–3699 or 3800–3899, and zero otherwise), but we also include dummies for
other SIC codes and years. As in Bebchuk and Cohen, we include the antitakeover index
for the state where the firm is headquartered, and a dummy equal to one if the state where
the firm is headquartered has recapture laws (note that for this regression, the antitakeover
and recapture variables are defined by the location of the firm’s headquarters, not by where
the firm is incorporated as in the capital structure regressions). We also test to see whether
including the home state’s total asset constraint (again, not the state of incorporation but
where the state has its headquarters) impacts the decision to reincorporate out-of-state.
We find that firms with many employees and more physical capital (PP&E) are more
likely to remain in their home state, whereas high book value firms are more likely to
incorporate out-of-state. Possibly these relations imply that firms with many employees
gain from increased political influence in their home states, while firms with greater capital
are more likely to benefit from Delaware statutes. As in Bebchuk and Cohen (2003), firms
headquartered in states with a higher antitakeover index are also more likely to remain in-
state. Consistent with Heron and Lewellen (1998), managers choose the state of
incorporation at least partly to defend against hostile takeovers. Firms in states with a
recapture statute are also more likely to remain in their home state. Additionally, firms are
more likely to incorporate out-of-state if their home state has a tighter payout restriction,
suggesting that these constraints are at times costly to firms. Surprisingly, this result holds
more strongly when the minimum capital-to-debt ratio equals 1.0 than for the 1.25
constraint. If a single total asset constraint variable is used rather than the two dummies, it
is also significantly different from zero. Lastly, more profitable firms, i.e., firms with a
higher ratio of EBITDA to book value, are much more likely to remain in-state. As we
show below, this decision by more profitable firms to remain in-state impacts the relation
between profitability and leverage.
Table 5 provides the estimated coefficients corresponding to the self-selection model of
Eqs. (5) and (6) for the dependent variable equal to the ratio of debt to market value, and
Table 6 provides the corresponding coefficients when the dependent variable is debt to
book value. In Table 5, the coefficients on the self-selection controls, W 1 and W 2, are all
significant at the 1% level, and these coefficients are significant at either the 5% or 10%
levels in the debt-to-book-value regressions.
The coefficient on the antitakeover index is positive and significant when the firm is
incorporated in its home state (home equals one) for both debt-to-market-value
specifications and in one of the debt-to-total-assets specifications. The antitakeover index
is not significant for firms incorporated out-of-state. The coefficients on the recapitaliza-
tion dummy are also not significant in Table 5 or 6 for home equals zero, and are positive
(and significant at the 10% level in three our of four cases) for home equals one. Thus,
these statutes appear to have no impact on the capital structure of firms incorporating out-
of-state. For firms that choose to stay in their home states, being in a state with a higher
antitakeover index, or being in the recapture states of Ohio or Pennsylvania, is associated
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 313

Table 5
Regressions including endogenous incorporation choice for debt/market value

Debt/total market value With total asset constraint With dummies for total asset constraints

Home equal 0 Home equal 1 Home equal 0 Home equal 1

Firm characteristics
R&D 0.461c 0.455c 0.461c 0.454c
(11.594) (6.454) (11.574) (6.474)
Advertising 0.159b 0.073 0.159b 0.074
(2.144) (0.514) (2.142) (0.522)
PP&E 0.162c 0.189c 0.162c 0.187c
(11.637) (9.005) (11.645) (9.027)
Depreciation 0.238b 0.020 0.237a 0.007
(1.966) (0.107) (1.955) (0.039)
Profitability 0.030 0.153b 0.030 0.161a
(1.495) (2.024) (1.491) (1.953)
Size—log(market value) 0.056c 0.067c 0.056c 0.066c
(10.082) (7.596) (10.079) (7.345)
Log(employees) 0.068c 0.063c 0.068c 0.063c
(13.520) (7.341) (13.516) (7.217)
Market-to-book 0.005c 0.003a 0.005c 0.003b
(5.318) (1.934) (5.313) (1.971)

State law variables

Antitakeover index 0.000 0.026c 0.001 0.024c
(0.083) (4.891) (0.134) (3.340)
Recapture 0.005 0.047a 0.005 0.048a
(0.115) (1.896) (0.128) (1.925)
Total asset constraint 0.009 0.075b
(0.649) (1.997)
Total asset constraint * profitability 0.030 0.112
(0.980) (1.602)
TA constraint ¼ 1 0.013 0.068
(0.730) (1.625)
TA constraint ¼ 1.25 0.002 0.097b
(0.058) (2.020)
TA constraint ¼ 1 * profitability 0.033 0.122
(1.058) (1.451)
TA constraint ¼ 1.25 * profitability 0.093 0.151a
(0.742) (1.728)

Self-selection controls
W1 0.363c 0.358c
(7.920) (7.714)
W2 0.139c 0.139c
(3.658) (3.660)

Tobit regressions where the dependent variable is debt/total market value. t-statistics are reported in parentheses,
and are calculated using bootstrap with clustering by firm and 1,000 repetitions. The independent variables R&D,
advertising, PP&E, depreciation, and profitability are all normalized by market value. The antitakeover index
applies to the state of incorporation and is from McGurn, Pamepinto, and Spector (1989) and Gartman (2000).
Recapture equals one if the firm is incorporated in a state (Ohio or Pennsylvania after 1989) that requires bidding
firms to disgorge short-term profits in the event of an unsuccessful hostile takeover. Home equals one if the firm’s
main headquarters is in the same state as the state of incorporation and zero otherwise. All regressions include
year and two-digit SIC code industry dummies. The sample includes 40,602 firm-year observations for 5,171 firms
for all regressions. 14,659 firm-years are for firms incorporated within their home states. Of firms incorporated out
of their home state, 13.8%, or 3,571 firm-years are for firms incorporated in a state with a total asset constraint
equal to 1.0, and 121 or 0.5% are for firms incorporated in a state with a total asset constraint equal to 1.25. W 1
and W 2 are inverse Mills ratios calculated from the probit regression in Table 4. Superscripts a, b, and c denote
significance at the 10%, 5%, and 1% levels, respectively.
 ARTICLE IN PRESS
314 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

Table 6
Regressions including endogenous incorporation choice for debt/book value

Debt/book value With total asset constraint With dummies for total asset constraints

Home equal 0 Home equal 1 Home equal 0 Home equal 1

Firm characteristics
R&D 0.345c 0.159 0.344c 0.164
(6.400) (1.367) (6.370) (1.391)
Advertising 0.239b 0.054 0.239b 0.058
(2.316) (0.293) (2.313) (0.303)
PP&E 0.119c 0.272c 0.119c 0.275c
(3.794) (3.850) (3.792) (2.900)
Depreciation 0.719c 0.073 0.716c 0.061
(3.874) (0.170) (3.858) (0.115)
Profitability 0.213c 0.077 0.212c 0.205
(7.589) (0.379) (7.569) (0.914)
Size– log(book value) 0.017b 0.085c 0.017b 0.086b
(2.242) (2.868) (2.242) (2.053)
Log(employees) 0.046c 0.077c 0.046c 0.077c
(6.774) (3.487) (6.771) (2.628)
Market-to-book 0.003a 0.003 0.003a 0.003
(1.905) (1.255) (1.915) (1.163)

State law variables

Antitakeover index 0.005 0.032c 0.006 0.034
(0.719) (2.903) (0.765) (1.555)
Recapture 0.033 0.068a 0.035 0.068
(0.728) (1.711) (0.756) (1.479)
Total asset constraint 0.007 0.118b
(0.351) (2.053)
Total asset constraint * Profitability 0.013 0.023
(0.276) (0.127)
TA constraint ¼ 1 0.013 0.136
(0.476) (1.641)
TA constraint ¼ 1.25 0.001 0.161b
(0.025) (2.181)
TA constraint ¼ 1 * profitability 0.016 0.167
(0.327) (0.812)
TA constraint ¼ 1.25 * profitability 0.174 0.132
(1.635) (0.625)

Self-selection controls
W1 0.455b 0.464a
(2.526) (1.763)
W2 0.103b 0.103b
(2.171) (2.168)

Tobit regressions where the dependent variable is debt/book value. All standard errors are calculated using
bootstrap with clustering by firm and 1,000 repetitions. The independent variables R&D, advertising, PP&E,
depreciation, and profitability are all normalized by book value. The antitakeover index applies to the state of
incorporation and is from McGurn, Pamepinto, and Spector (1989) and Gartman (2000). Recapture equals one if
the firm is incorporated in a state (Ohio or Pennsylvania after 1989) that requires bidding firms to disgorge short-
term profits in the event of an unsuccessful hostile takeover. Home equals one if the firm’s main headquarters is in
the same state as the state of incorporation and zero otherwise. All regressions include year and two-digit SIC
code industry dummies. The sample includes 40,602 firm-year observations for 5,171 firms for all regressions.
14,659 firm-years are for firms incorporated within their home states. Of firms incorporated out of their home
state, 13.8% or 3,571 firm-years are for firms incorporated in a state with a total asset constraint equal to 1.0, and
121 or 0.5% are for firms incorporated in a state with a total asset constraint equal to 1.25. W 1 and W 2 are
inverse Mills ratios calculated from probit regressions. Superscripts a, b, and c denote significance at the 10%,
5%, and 1% levels, respectively.
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 315

with a larger rather than a smaller debt-to-market-value ratio. Overall, after self-selection
is taken into account, state laws that facilitate managerial entrenchment do not seem to
have a negative impact on debt ratios. Instead, there seems to be a slight positive impact
for firms in states with stronger antitakeover indexes. Karpoff and Malatesta (1989) find
that firms in states that pass new antitakeover legislation suffer a decline in shareholder
value, and Heron and Lewellen (1998) find that firms incorporating in states with stronger
antitakeover defenses lose value. Thus, the higher debt-to-market-value ratios we find for
firms incorporating in states with more antitakeover statutes could be driven by the lower
market values of these firms. As book values only somewhat reflect these lower market
values, the antitakeover statutes have a weaker relation with firms’ debt-to-book value.
Intuitively, firms with abnormally low levels of debt are more likely to be takeover
targets (see, e.g., Jensen, 1986; or Billett, 1996) and thus are more likely to sort themselves
into jurisdictions with greater antitakeover protection. Once this sorting is accounted for,
the impact of antitakeover statutes is to decrease firm value, and thus to increase the debt-
to-value ratio. Thus, while Garvey and Hanka (1999) find that state antitakeover laws
decrease firm leverage, we find that antitakeover laws are associated with higher levels of
leverage after controlling for how firms sort themselves into different jurisdictions.
Similarly, the coefficients on total asset constraints in Tables 5 and 6 are now only
significant for firms incorporated in their home states. That is, firms that move out-of-state
do not appear to be constrained by state statutes on minimum capital requirements. Only
firms stuck in states with a more stringent capital requirement appear to have their capital
structure choice impacted. Moreover, the interaction between capital constraint and
profitability is only negative and significant at the 10% level in one case, for firms in-state
with the more stringent 1.25 constraint. Thus, payout restrictions in the debt-to-market-
value analyses appear to be binding for only those firms that have not moved out-of-state,
and in these cases, other advantages of the particular state of incorporation must outweigh
the implicit cost of these constraints on the firm.
Further, the coefficients on profitability itself are not negative and significant in any of
the debt-to-market-value regressions in Table 5. The coefficients on profitability remain
negative and significant (with an almost identical magnitude as in a model with no state
laws or endogenous sorting) in the debt-to-total-assets regression for firms incorporated
out-of-state. For firms incorporated in-state, the coefficients on profitability in the debt-to-
book-value regression are also not significant. This change in the coefficients on
profitability in the capital structure regression can be explained by the determinants of
the state of incorporation. From Table 4, more profitable firms are much more likely to
remain in-state. Because firms incorporated in-state are also likely to use less debt, this
endogenous sorting creates part of the negative relation between profitability and leverage.
Correcting for this endogenous sorting removes the standard negative and significant
relation between profitability and leverage in debt-to-market-value regressions and in debt-
to-total-assets regressions for firms incorporated in-state.

5. Conclusion

We examine the impact of antitakeover and payout restriction laws on firms’ capital
structure decisions. While prior research has found that the addition of antitakeover
legislation decreases the leverage of affected firms, we find no direct impact on leverage
over a longer time period. On the other hand, laws that constrain payout appear to
 ARTICLE IN PRESS
316 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

significantly impact capital structure, with firms incorporated in states subject to stricter
payout laws choosing significantly lower levels of debt. After correcting for self-selection,
we find that a primary impact of these laws is in causing firms to sort themselves into less
restrictive jurisdictions. After self-selection is taken into account, firms incorporated in
states with more antitakeover provisions have higher debt-to-market-value ratios, possibly
because the antitakeover statutes decrease firm value. Payout restrictions appear to be
binding only for firms that have not moved out-of-state. Correcting for self-sorting and the
effect of state laws reduces the explanatory power of profitability on leverage (when
leverage is measured by the ratio of debt to market value) or for firms incorporated in-
state. Further, these findings suggest that self-sorting by firms into different jurisdictions
helps explain the negative relation between profitability and leverage.

Appendix A. Some examples of state constraints on distributions to shareholders

California (the laws in Alaska are similar):

In California, a distribution is defined as follows:

‘‘Distribution to its shareholders’’ means the transfer of cash or property by a corporation

to its shareholders without consideration, whether by way of dividend or otherwise, except
a dividend in shares of the corporation, or the purchase or redemption of its shares for
cash or property, including the transfer, purchase, or redemption by a subsidiary of the
corporation (Cal Corp Code y 166, 2004).

The restrictions on distributions in California include the following language:

‘‘Neither a corporation nor any of its subsidiaries shall make any distribution to the
corporation’s shareholders (Section 166) except as follows:

(a) The distribution may be made if the amount of the retained earnings of the corporation
immediately prior thereto equals or exceeds the amount of the proposed distribution.
(b) The distribution may be made if immediately after giving effect thereto:
(1) The sum of the assets of the corporation (exclusive of goodwill, capitalized research
and development expenses and deferred charges) would be at least equal to 1 1/4
times its liabilities (not including deferred taxes, deferred income and other
deferred credits); and
(2) The current assets of the corporation would be at least equal to its current liabilities
or, if the average of the earnings of the corporation before taxes on income and
before interest expense for the two preceding fiscal years was less than the average
of the interest expense of the corporation for those fiscal years, at least equal to
1 1/4 times its current liabilities; provided, however, that in determining the amount
of the assets of the corporation profits derived from an exchange of assets shall not
be included unless the assets received are currently realizable in cash; and provided,
further, that for the purpose of this subdivision ‘‘current assets’’ may include net
amounts which the board has determined in good faith may reasonably be expected
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 317

to be received from customers during the 12-month period used in calculating

current liabilities pursuant to existing contractual relationships obligating those
customers to make fixed or periodic payments during the term of the contract or, in
the case of public utilities, pursuant to service connections with customers, after in
each case giving effect to future costs not then included in current liabilities but
reasonably expected to be incurred by the corporation in performing those
contracts or providing service to utility customers (Cal Corp Code y 500, 2004).’’

Delaware law is relatively less restrictive. A few states, such as Maine, Oklahoma, and
South Dakota, have similar laws.
(a) The directors of every corporation, subject to any restrictions contained in its
certificate of incorporation, may declare and pay dividends upon the shares of its capital
stock, or to its members if the corporation is a nonstock corporation, either (1) out of its
surplus, as defined in and computed in accordance with yy 154 and 244 of this title, or (2) in
case there shall be no such surplus, out of its net profits for the fiscal year in which the
dividend is declared and/or the preceding fiscal year. If the capital of the corporation,
computed in accordance with yy 154 and 244 of this title, shall have been diminished by
depreciation in the value of its property, or by losses, or otherwise, to an amount less than
the aggregate amount of the capital represented by the issued and outstanding stock of all
classes having a preference upon the distribution of assets, the directors of such
corporation shall not declare and pay out of such net profits any dividends upon any
shares of any classes of its capital stock until the deficiency in the amount of capital
represented by the issued and outstanding stock of all classes having a preference upon the
distribution of assets shall have been repaired. Nothing in this subsection shall invalidate
or otherwise affect a note, debenture or other obligation of the corporation paid by it as a
dividend on shares of its stock, or any payment made thereon, if at the time such note,
debenture or obligation was delivered by the corporation, the corporation had either
surplus or net profits as provided in clause (1) or (2) of this subsection from which the
dividend could lawfully have been paid (8 Del. C. y 170 2003).

New York law is quite typical of most states including Alabama, Arizona, Arkansas, and
others. The specific language varies from state to state, but the restrictions remain similar.
y 510. Dividends or other distributions in cash or property
(a) A corporation may declare and pay dividends or make other distributions in cash or
its bonds or its property, including the shares or bonds of other corporations, on its
outstanding shares, except when currently the corporation is insolvent or would thereby be
made insolvent, or when the declaration, payment or distribution would be contrary to any
restrictions contained in the certificate of incorporation.
(b) Dividends may be declared or paid and other distributions may be made out of
surplus only, so that the net assets of the corporation remaining after such declaration,
payment or distribution shall at least equal the amount of its stated capital; except that a
corporation engaged in the exploitation of natural resources or other wasting assets,
including patents, or formed primarily for the liquidation of specific assets, may declare
and pay dividends or make other distributions in excess of its surplus, computed after
taking due account of depletion and amortization, to the extent that the cost of the wasting
or specific assets has been recovered by depletion reserves, amortization or sale, if the net
 ARTICLE IN PRESS
318 J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319

assets remaining after such dividends or distributions are sufficient to cover the liquidation
preferences of shares having such preferences in involuntary liquidation (NY CLS Bus
Corp y 510, 2003).

References

Asquith, P., Wizman, T., 1990. Event risk, covenants, and bondholder returns in leveraged buyouts. Journal of
Financial Economics 27, 195–213.
Barclay, M., Morellec, E., Smith, C., 2005. On the debt capacity of growth options. Journal of Business 79, 37–59.
Barnea, A., Haugen, R., Senbet, L., 1981. An equilibrium analysis of debt financing under costly tax arbitrage and
agency problems. Journal of Finance 36, 569–581.
Bebchuk, L., Cohen, A., 2003. Firms’ decisions where to incorporate. Journal of Law and Economics 46, 383–425.
Berens, J., Cuny, C., 1995. The capital structure puzzle revisited. Review of Financial Studies 8, 1185–1208.
Berger, P., Ofek, E., Yermack, D., 1999. Managerial entrenchment and capital structure decisions. Journal of
Finance 52, 1411–1438.
Billett, M., 1996. Targeting capital structure: The relationship between risky debt and the firm’s likelihood of
being acquired. Journal of Business 69, 173–192.
Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovij, V., 2001. Capital structure in developing countries.
Journal of Finance 56, 87–130.
Bradley, M., Jarrell, G., Kim, E., 1984. On the existence of an optimal capital structure: Theory and evidence.
Journal of Finance 39, 857–880.
Comment, R., Schwert, G., 1995. Poison or placebo? Evidence on the deterrence and wealth effects of modern
antitakeover measures. Journal of Financial Economics 39, 3–43.
Daines, R., 2001. Does Delaware law improve firm value? Journal of Financial Economics 62, 525–558.
DeAngelo, H., Masulis, R., 1980. Optimal capital structure under corporate and personal taxation. Journal of
Financial Economics 8, 3–29.
Eisenberg, M., 1983. The modernization of corporate law: An essay for Bill Cary. University of Miami Law
Review 37, 187–212.
Fama, E., 1980. Agency problems and the theory of the firm. Journal of Political Economy 88, 288–307.
Forsberg, M., 2005. A question of balance: Taxing business in the 21st century. New Jersey Policy Perspective,
June 20, 2005.
Frank, M., Goyal, V., 2003. Capital structure decisions. Unpublished working paper. University of British
Columbia.
Gartman, G., 2000. State Takeover Laws. Investor Responsibility Research Center, Washington, DC.
Garvey, G., Hanka, G., 1999. Capital structure and corporate control: The effect of antitakeover statutes on firm
leverage. Journal of Finance 54, 519–546.
Gevurtz, F., 2000. Corporation Law. Westgroup, St. Paul, MN.
Gilson, S., Warner, J., 1997. Junk bonds, bank debt, and financial flexibility. Unpublished working paper.
Harvard University.
Gompers, P., Ishii, J., Metrick, A., 2001. Corporate governance and equity prices, NBER Working Paper No.
8449.
Heckman, J., 1976. The common structure of statistical models of truncation, sample selection and limited
dependent variables and a simple estimator for such models. Annals of Economic and Social Measurement 5,
475–492.
Hennessy, C., Whited, T., 2004. Debt dynamics. Journal of Finance 60, 1129–1165.
Heron, R., Lewellen, W., 1998. An empirical analysis of the reincorporation decision. Journal of Financial and
Quantitative Analysis 33, 549–568.
Hovakimian, A., Opler, T., Titman, S., 2001. The debt-equity choice, Journal of Financial and Quantitative
Analysis 36, 1–24.
Jensen, M., 1986. Agency costs of free cash flow, corporate finance and takeovers, American Economic Review
76, 323–329.
John, K., Kalay, A., 1982. Costly contracting and optimal payout constraints. Journal of Finance 37, 457–470.
Kalay, A., 1982. Shareholder–bondholder conflict and dividend constraints. Journal of Financial Economics 10,
211–233.
 ARTICLE IN PRESS
J.K. Wald, M.S. Long / Journal of Financial Economics 83 (2007) 297–319 319

Karpoff, J., Malatesta, P., 1989. The wealth effects of second-generation state takeover legislation. Journal of
Financial Economics 25, 291–322.
Karpoff, J., Malatesta, P., 1995. State takeover legislation and share values: The wealth effects of Pennsylvania’s
Act 36. Journal of Corporate Finance 1, 367–382.
Kim, E., 1978. A mean-variance theory of optimal capital structure and corporate debt capacity. Journal of
Finance 33, 45–63.
Long, M., Malitz, I., 1985. Investment patterns and financial leverage. In: Friedman, B. (Ed.), Corporate Capital
Structure in the United States. University of Chicago Press, Chicago, IL, pp. 325–348.
Maddala, G.S., 1983. Limited-Dependent and Qualitative Variables in Econometrics. Cambridge University
Press, Cambridge.
Malitz, I., 1994. The Modern Role of Bond Covenants. The Research Foundation of the Institute of Chartered
Financial Analysts. Charlottesville, VA.
Mayers, D., Smith, C., 2005. Agency problems and the corporate charter. Journal of Law, Economics, and
Organization 21, 417–440.
McGurn, P., Pamepinto, S., Spector, A., 1989. State Takeover Laws. Investor Responsibility Research Center,
Washington, DC.
Myers, S., Majluf, N., 1984. Corporate financing decisions when firms have investment information that investors
do not. Journal of Financial Economics 13, 187–221.
Rajan, R., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data.
Journal of Finance 50, 1421–1460.
Romano, R., 2005. The states as a laboratory: Legal innovations and state competition for corporate charters.
Yale Law School Research Paper #310.
Sarkar, S., Zapatero, F., 2003. The trade-off model with mean reverting earnings: Theory and empirical tests.
Economic Journal 113, 834–860.
Scott, J., 1976. A theory of optimal capital structure. The Bell Journal of Economics 7, 33–54.
Subramanian, G., 2004. The disappearing Delaware effect. Journal of Law, Economics and Organization 20,
32–59.
Smith, C., Warner, J., 1979. On financial contracting: An analysis of bond covenants. Journal of Financial
Economics 7, 117–161.
Szewczyk, S., Tsetsekos, G., 1992. State intervention in the market for corporate control: The case of
Pennsylvania Senate Bill 1310. Journal of Financial Economics 31, 3–23.
Titman, S., Wessels, R., 1988. The determinants of capital structure choice. Journal of Finance 63, 1–19.
Wald, J., 1999. Capital structure with dividend restrictions. Journal of Corporate Finance 5, 193–208.