University of California, Hastings College of the Law
UC Hastings Scholarship Repository
1)04 %,)./%&-
Coerced Confessions Self Policing in the Shadow
of the Regulator
Jodi L. Short
UC Hastings College of the Law/%,.0'1 %/0&+$/"!1
Michael W. Toel
,)),30%&/+!!!&0&,+)3,.(/0 %8-."-,/&0,.41 %/0&+$/"!1# 1)04/ %,)./%&-
7&/.0& )"&/.,1$%00,4,1#,.#.""+!,-"+ "//4/0&+$/ %,)./%&-"-,/&0,.40%/""+ "-0"!#,.&+ )1/&,+&+ 1)04 %,)./%&-
4+10%,.&5"!!*&+&/0.0,.,#/0&+$/ %,)./%&-"-,/&0,.4,.*,."&+#,.*0&,+-)"/" ,+0 0 *. 1/ 1 %/0&+$/"!1
" ,**"+!"!&00&,+
,!&%,.0+!& %"),6") Coerced Confessions Self Policing in the Shadow of the Regulator Journal of Law, Economics &
Organization
2&))"0%8-."-,/&0,.41 %/0&+$/"!1# 1)04/ %,)./%&-
Coerced Confessions: Self-Policing in the
Shadow of the Regulator
Jodi L. Short
*
University of California, Berkeley
Michael W. Toffel
**
Harvard Business School
As part of a recent trend toward more cooperative relations between regulators
and industry, novel government programs are encouraging firms to monitor their
own regulatory compliance and voluntarily report their own violations. In this
study, we examine how regulatory enforcement activities influence organiza-
tions’ decisions to self-police. We created a comprehensive data set for the
‘‘Audit Policy,’’ a United States Environmental Protection Agency program that
encourages companies to self-disclose violations of environmental laws and
regulations in exchange for reduced sanctions. We find that facilities are more
likely to self-disclose if they were recently subjected to one of several different
enforcement measures and if they were provided with immunity from prosecu-
tion for self-disclosed violations.
The pitched political battles over regulation in the 1970s and 1980s, from de-
regulation to Reagan’ s vow to get government ‘‘off the back’’ of industry, have
given way in recent years to a new wave of voluntary self-regulation programs
based on a more cooperative approach between government and industry. Reg-
ulatory agencies are embracing programs that see firms as active partic ipants in
their own governance. And from the industry side, talk is increasingly about
companies regulating themselves rather than trying to avoid regulation alto-
gether. These new cooperative arrangements enjoy wide support, both from
industry proponents who see self-regulation as the most flexible and efficient
*University of California, Berkeley. Email: jlshort@berkeley.edu.
**Harvard Business School. Email: mtoffel@hbs.edu.
Insightful comments by Neil Fligstein, Robert A. Kagan, David I. Levine, Howard Shelanski,
and Jason Snyder are gratefully acknowledged, as are comments from participants in the Fifth
Annual Strategy and the Business Environment Conference at Stanford University and the First
Annual Conference on Institutional Mechanisms for Industry Self-Regulation at Dartmouth Uni-
versity. We appreciate the assistance of Phil Milton, Audit Policy Coordinator at the US EPA, for
providing and interpreting some EPA data. Research funding was provided by the Center for Re-
sponsible Business and the Institute of Business and Economic Research at the Haas School of
Business. Ara Abrahamian provided excellent research assistance.
The Journal of Law, Economics, & Organization, Vol. 24, No. 1,
doi:10.1093/jleo/ewm039
Advance Access publication August 30, 2007
The Author 2007. Published by Oxford University Press on behalf of Yale University.
All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org
JLEO, V24 N1 45
at UC Hastings College of the Law on June 24, 2015http://jleo.oxfordjournals.org/Downloaded from
way to achieve regulatory goals (Murray 1999) and in a substantial body
of academic literature touting the virtues of a more cooperative regulatory sys-
tem (Bardach and Kagan 1982; Scholz 1984; Ayres and Braithwaite 1992;
Gunningham and Grabosky 1998).
This cooperative approach has inspired regulatory agencies to experiment
with an array of voluntary self-regulation programs that engage firms as part-
ners in regulatory activities, from achieving ‘‘beyond compliance’’ results to
policing their own noncompliance. Beyond compliance programs, such as the
United States Environmental Protection Agency’ s (US EPA) Greenlights, 33/
50 and Project XL, as well as industry initiatives, such as the chemical man-
ufacturers’ Responsible Care, recognize and reward firms for performance and
management practices that go above and beyond what the law requires. Al-
though there is little evidence that such programs actually improve perfor-
mance (King and Lenox 2000; Welch et al. 2000; Rivera and de Leon
2004), they have been popular for their feel-good, ‘‘win-win’’ approach to reg-
ulation. However, administrative agencies can justify outsourcing more and
more of their responsibilities to regulated entities only if corporations are will-
ing to admit and correct their failures as well as tout their successes.
‘‘Self-policing’’ programs push the envelope of self-regulation by shifting
the burden of monitoring regulatory compliance and reporting noncompliance
from the government to the private sector. Several regulatory agencies have
developed self-policing programs that provide incentives to encourage com-
panies to self-disclose their legal violations. For exam ple, through its Hazard
Analysis and Critical Control Point program, the US Department of Agricul-
ture ‘‘shifted much of the responsibility for safety to the plants, requiring them
to identify vulnerable points in their production lines and build in steps to kill
germs’’ (Peterson and Drew 2003:A1). In addition, the US Department of Jus-
tice, the US Department of Defense, and the Securities and Exchange Com-
mission offer incentives including amnesty, limited liability, prosecutorial
leniency, and confidentiality to encourage companies to disclose fraudulent
or other illegal behavior (Fleder 1999; Duggin 2003; Medinger 2003).
These types of initiatives carry promise as well as pitfalls. On the one hand,
the incentives of self-policing programs have encouraged many companies to
report and correct problems that regulators never would have disc overed, sug-
gesting the potential for real improvements in compliance. If compliance does
improve at self-policing firms, regulators could shift their scarce enforcement
resources to more recalcitrant companies. On the other hand, there is some
evidence that self-repor ted violations are often minor, perhaps masking more
serious unreported violations (Pfaff and Sanchirico 2004). Without
any evidence that self-policing improves compliance, such programs may give
industry an unprecedented and unwarranted level of control over its own reg-
ulation, raising ‘‘fears of the Ôfox guarding the henhouseÕ’’ (Cox 2004:28). How
can regulators establish self-policing programs that balance thes e concerns?
Although the academic literature has long explored regulatory strategies that
rely on some mix of voluntary reporting and governmental enforcement
(e.g., Malik 1993; Kap low and Shavell 1994; Innes 1999a, 1999 b; Pfaff
46 The Journal of Law, Economics, & Organization, V24 N1
at UC Hastings College of the Law on June 24, 2015http://jleo.oxfordjournals.org/Downloaded from
and Sanchirico 2000; Innes 2001), this literature is largely theoretical; little
empirical evidence exists to suggest what mix most effectively encourages
self-policing.
Among the first empirical studies to address self-policing behavior , this ar-
ticle seeks to understand how a mix of regulatory enforcement activities and
statutory protections that shield companies from enforcement activities can be
used to encourage organizations to police their own operations and ‘‘turn them-
selves in’’ by self-disclosing violations. We develop a panel data set of vol-
untary disclosures under the US EPA’s Audit Policy, which provides rich data
on how violators behave when offered the option of voluntarily self-reporting.
We find that, despite the rhetoric of cooperation surrounding sel f-policing pro-
grams, self-disclosures are motivated by coercive regulatory enforcement ac-
tivities. Specifically, facil ities were more lik ely to self-disclose violations if
they were recently inspected, subjected to an enforcement action, or narrowly
targeted for heightened scrutiny by a US EPA Compliance Incentive Program.
We also find some evidence that facil ities are more likely to turn themselves in
when statutory immunity shields their self-disclosed violations from prosecu-
tion, but no evidence that facilities protected by audit privilege are more likely
to self-disclose.
The article proceeds as follows. In the next section, we review the literature
on self-policing and the related literature that describes how deterrence mea-
sures affect compliance. In Section 2, we describe the US EPA Audit Policy,
the empirical setting of our research. In Section 3, we hypothesize how various
enforcement activities and legal protections that shield self-reporters from
enforcement activities may influence facilities’ decisions whether to self-
police. Section 4 describes our sample and measures, and Section 5 details
our empirical methods and presents our results. We discuss our results in
Section 6 and offer suggestions for future researc h. Finally, we conclude in
Section 7.
1. Literature Review
The most robust discussion of self-policing occurs in the economic literature,
which models self-repor ting of legal violations as a way to optimize enforce-
ment regimes by reducing monitoring and compliance costs. Kaplow and
Shavell (1994) develop a model that integrates a self-reporting component into
Becker’s (1968) classic theory of probabilistic law enforcement. They dem-
onstrate that self-reporting reduces government monitoring and enforcement
costs because, to the extent that violators admit their wrongdoing, ‘‘enforce-
ment effort need not be spent identifying them’’ (Kaplow and Shavell
1994:584). Innes (2001) extends this model, demonstrating that self-reporting
optimizes the allocation of enforcement resources by lowering avoidance costs
for the regulators and the regulated and by increasing the likelihood and low-
ering the costs of remediation (Innes 1999a).
A few prior studies have identified factors that lead firms to self-police.
Innes (2000) , for example, develops an economic model that suggests that
Coerced Confessions 47
at UC Hastings College of the Law on June 24, 2015http://jleo.oxfordjournals.org/Downloaded from
violators with a greater risk of appre hension are more likely to self-disclose
their wrongdoing, but his model does not address what factors constitute and
influence perception of this risk. Others have proposed how regulators can set
fines to optimize compliance using self-reporting. Malik (1993), for instance,
argues that an enforcement regime that relies on mandatory self-reporting of
emissions data will be optimized only if regulators impose sufficiently large
fines for violations that were not self-reported. Pfaff and Sanchirico (2000)
demonstrate that more firms would voluntarily conduct self-audits if fines
for violations were made contingent on the firm’s investigative effort.
Several studies in the economic, legal, and policy literatures suggest that
enforcement activities tend to discourage self-policing and self-reporting. Ar-
guing that exposu re to enforcement actions is a powerful disince ntive to self-
policing, some commentators call for greater statutory protections to shield
would-be self-policers from legal liability for their violations (Hunt and
Wilkins 1992; Goldsmith and King 1997; Grayson and Landgraf 1997; Murray
1999; Innes 2001). Innes (2001:253), for example, argues that the internal
compliance audit information regulators may obtain in connection with vol-
untary disclosures ‘‘can place a self-reporter at risk for other civil enforcement
actions.’’ Most commentators agree that this risk ‘‘is a major stumbling block to
the widespread participation of American companies in the system’’ (Murray
1999:56). This logic suggests that the government must provide legal protec-
tions to mitigate exposure to enforcement activities in order to encourage firms
to self-police.
The related literature that describes how enforcement activities affect reg-
ulatory compliance, however, suggests a different result. There is broad con-
sensus in this literature that enforcement activities improv e complianc e, even
when they are confrontational and coercive. For example, numerous studies
have shown that regulatory inspections improve compliance at targeted firms
(Magat and Viscusi 1990; Braithwaite and Makkai 1991; Kuperan and Sutinen
1998; Gray and Shadbegian 2005; Gunningham et al. 2005; Shimshack and
Ward 2005), including compliance with legal requirements to monitor emis-
sions data and self-report permit violations (Laplante and Rilstone 1996;
Helland 1998). And growing empirical evidence suggests that more severe
deterrence measures such as penalties and enforcement actions also improv e
facilities’ regulatory compliance (Gray and Scholz 1991; Aoki and Coiffi
2000; Gray and Shadbegian 2005; Gunningham et al. 2005; Mendelhoff
and Gray 2005; Shimshack and Ward 2005).
Our research expands on the existing literature in three important ways.
First, we add an empirical dimension to a literature on self-policing that
has been studied almost exclusively through the use of economic models.
Our data on violations voluntarily self-reported under the Audit Policy pro-
vide a unique window on self-policing behavior that is notoriously difficult to
observe. For example, the data enable us to test the competing claims in the
literature about whether stronger enforcement encourages or deters self-
policing. They also allow us to measure the incentive effects of a broad array
of enforcement tools, including inspections, violation citations, enforcement
48 The Journal of Law, Economics, & Organization, V24 N1
at UC Hastings College of the Law on June 24, 2015http://jleo.oxfordjournals.org/Downloaded from
