Institutional and Legal Context in Natural Experiments: The Case of State Antitakeover Laws

• DOI: http://doi.org/10.1111/jofi.12600
• Published date: 01 April 2018
• Author: JONATHAN M. KARPOFF,MICHAEL D. WITTRY
• Date: 01 April 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 2 •APRIL 2018

Institutional and Legal Context in

Natural Experiments: The Case of State

Antitakeover Laws

JONATHAN M. KARPOFF and MICHAEL D. WITTRY∗

ABSTRACT

We argue and demonstrate empirically that a firm’s institutional and legal context

has first-order effects in tests that use state antitakeover laws for identification. A

priori, the size and direction of a law’s effect on a firm’s takeover protection depends

on (i) other state antitakeover laws, (ii) preexisting firm-level takeover defenses, and

(iii) the legal regime as reflected by important court decisions. In addition, (iv) state

antitakeover laws are not exogenous for many easily identifiable firms. We show that

the inferences from nine prior studies related to nine different outcome variables

change substantially when we include controls for these considerations.

EMPIRICAL CORPORATE FINANCE RESEARCH IS PLAGUED by endogeneity. As Alchian

(1950) points out, the economic environment tends to select the combination

of firm characteristics—including organization, financing, payout policy, own-

ership structure, board structure, compensation, and governance—that mini-

mize total production costs. This implies that each firm’s characteristics and

performance are endogenous outcomes of the competitive process. Researchers

nonetheless seek to explore questions of causality. For example, does better

governance improve firm value, innovation, or payout policy?

To address endogeneity issues, researchers frequently investigate firm char-

acteristics around regulatory changes or court decisions that exogenously affect

firms’ costs or governance. For example, previous studies use banking dereg-

ulations to examine innovation and economic growth (Bertrand, Schoar, and

∗Jonathan M. Karpoff and Michael D. Wittry are with the Foster School of Business, University

of Washington. We thank Jennifer Arlen, Julian Atanassov, Michal Barzuza, Bernie Black, Emil-

iano Catan, Steven Davidoff Solomon, Diane Denis, Laura Field, Mariassunta Giannetti, Iftekhar

Hasan, Wei Jiang, Marcel Kahan, Michael Klausner, Don Margotta, Michael Roberts, Roberta

Romano, Tatyana Sokolyk, Randall Thomas, three anonymous referees, participants at the 2015

Western Finance Association meeting, and seminar participants at the University of Washington,

University of North Carolina, Northeastern University, University of Oklahoma, and the joint

SMU/TCU finance seminar for helpful comments and discussions. We especially thank Wei Jiang

for encouragement to write this paper, Julian Atanassov for providing data from his 2013 Journal

of Finance paper on corporate innovation, and Martijn Cremers and Allen Ferrell for providing

data on firms’ G-indices for the 1977 to 1989 period from their 2014 Journal of Finance paper on

takeover defenses. We have read the Journal of Finance’s disclosure policy and have no conflicts

of interest to disclose.

DOI: 10.1111/jofi.12600

657

658 The Journal of Finance R

Thesmar (2007), Chava et al. (2013)); tax changes to examine firm investment,

financing, and payout decisions (Givoly et al. (1992), Dharmapala, Foley, and

Forbes (2011), Faulkender and Petersen (2012), Tsoutsoura (2015)); securities

and financial reforms to examine firm value, governance, and financing (Coates

(2007), Zhang (2007), Atanasov et al. (2010), Iliev (2010), Vig (2013)); legislated

subsidies to examine financial constraints (Butler and Cornaggia (2011)); and

court decisions to examine the impact of firm governance on firm value, disclo-

sures, and exposure to securities lawsuits (Cohen and Wang (2013), Furchtgott

and Partnoy (2015), Crane and Koch (2016), Amihud and Stoyanov (2017),

Lichtetal.(2017)).1

In this paper, we examine the use of regulatory and legal changes to identify

exogenous variation in economic determinants. We focus on an application that

has become especially popular, namely state laws that affect firms’ takeover

vulnerability.Garvey and Hanka (1999) and Bertrand and Mullainathan (1999)

pioneered the use of state antitakeover laws in empirical tests involving firm

governance, and in an influential paper, Bertrand and Mullainathan (2003)ar-

gue that, among the different types of antitakeover laws, business combination

laws have a particularly pronounced effect on firm governance. Bertrand and

Mullainathan (2003) thus use business combination laws to examine the impact

of firm governance on wages, investment, and productivity. Since then, at least

78 additional published articles and working papers have used business com-

bination and other antitakeover laws to examine the effect of firm governance

on operating performance, payout policy, cash holdings, innovation, bond-

holder returns, stakeholder relations, financial statement informativeness, and

other firm characteristics. Table A1 lists and provides a brief description of

these papers, and Section Iillustrates the empirical construct used in this

literature.

Our main argument is that the institutional and legal environment matters

in empirical research of this type. Focusing on state antitakeover laws, par-

ticularly business combination laws, we show that the institutional, political-

economy, and historical context in which a law is enacted has a large effect on

the appropriate specification and interpretation of tests that use legal changes

for identification.

Our argument is both theoretical and empirical. Sections II to IV begin by

showing that, a priori, the size and direction of an antitakeover law’s effect on

a firm’s takeover protection depends on (i) coverage by a first-generation state

antitakeover law or by any of the other four major types of second-generation

state antitakeover laws, (ii) preexisting firm-level takeover defenses, and (iii)

the legal regime as determined by important court decisions. Coverage by first-

generation laws is a particularly important concern for empirical tests that

include pre-1982 data because it can reverse the inferences from these tests. An

additional concern is that business combination and other state antitakeover

1This is only a partial listing of papers that use external shocks, typically to legal rules, to

identify natural experiments. See Atanasov and Black (2016) for an analysis of shock-based studies

in corporate governance.

Institutional and Legal Context in Natural Experiments 659

Tabl e I

First-Generation State Antitakeover Laws, 1968 to 1982

This table lists in chronological order the first-generation state antitakeover laws adopted by 38

states from 1968 through 1981, which were effective until a U.S. Supreme Court ruling in Edgar

vs. MITE Corp. on June 23, 1982. First-generation laws regulated cash tender offers and gener-

ally imposed extremely strong takeover protections for the covered corporations. The protections

included requirements that the bidding firm files extensive disclosure statements with the state

securities commissioner, provisions that required long and variable open periods for tender of-

fers, state administrative overview of the tender offer, and potential civil and criminal liability for

bidding firms and their managers for violations of the antitakeover provisions.

State Effective Date Repeal Date State Effective Date Repeal Date

Virginia 03/05/1968 07/01/1989 Kentucky 07/01/1976 07/15/1986

Nevada 03/04/1969 10/01/1991 Maryland 07/01/1976 07/01/1986

Ohio 10/09/1969 Michigan 07/01/1976 04/01/1988

Wisconsin 07/01/1972 New York 11/01/1976

Minnesota 08/01/1973 Georgia 03/23/1977 03/28/1986

Hawaii 05/24/1974 04/23/1985 Arkansas 03/24/1977 03/24/2000

Kansas 07/01/1974 04/21/1988 New Hampshire 03/25/1977

Indiana 05/01/1975 Nebraska 04/27/1977 04/08/1988

Colorado 07/01/1975 07/01/1984 New Jersey 04/27/1977

Idaho 07/01/1975 07/01/1986 Texas 05/06/1977

South Dakota 07/01/1975 07/01/1990 North Carolina 06/28/1977 10/01/2001

Utah 02/05/1976 07/01/1983 Mississippi 07/01/1977

Pennsylvania 03/03/1976 Illinois 09/08/1977 07/01/1984

Tennessee 03/17/1976 Florida 10/01/1977 09/01/1979

Delaware 05/01/1976 07/01/1987 Maine 03/24/1978 07/16/1986

Massachusetts 05/22/1976 South Carolina 06/12/1978 01/01/1989

Connecticut 06/02/1976 Missouri 08/13/1978

Alaska 06/12/1976 Iowa 01/01/1979 01/01/2005

Louisiana 06/28/1976 08/15/1987 Oklahoma 07/21/1981 07/22/1985

laws are not exogenous for, or do not apply to, a substantial number of easily

identifiable firms that are incorporated in the states that adopt these laws,

either because these firms lobbied for the adoption of these laws or because

they opted out of coverage by the laws.

In Section V, we argue that these institutional and political-economy con-

siderations can be viewed as omitted variable problems for which there ex-

ist data-oriented solutions. Table Ipresents data on the adoption of first-

generation antitakeover laws in 38 states. Table II presents state-by-state

data on the adoption of all five major types of second-generation state anti-

takeover laws, including business combination laws. Table III presents data

on firms for which the enactment of state antitakeover laws is clearly endoge-

nous. These data supplement those on firms that opted out of coverage that are

available in the Institutional Shareholder Services (ISS) Governance (formerly,

RiskMetrics) database. Cremers and Ferrell (2014) have compiled data on firm-

level defenses for a large number of firms for the period 1977 to 1989, which

complement the post-1989 data on firm-level defenses available in the ISS

database. Together, these data can be used to increase the accuracy and power of