IMPOSING LIMITS ON PROSECUTORIAL DISCRETION IN

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 1 16-OCT-07 14:37

IMPOSING LIMITS ON PROSECUTORIAL

DISCRETION IN CORPORATE

PROSECUTION AGREEMENTS

RIK

AULSEN

In late 2006, the Department of Justice (DOJ) overhauled its internal policy

regarding the prosecution of corporate entities. The new policy—expressed in the

“McNulty Memo”—was issued as a direct response to charges that the DOJ had

abused its leverage over the companies it targeted for criminal prosecution, specifi-

cally with regard to compelled cooperation. The McNulty Memo addressed these

charges, in part, by restricting the discretion of individual prosecutors and requiring

approval by the Deputy Attorney General on significant prosecutorial decisions.

While the changes ushered in by the McNulty Memo are a promising first step, they

will not remedy all of the potentially abusive practices that mar the prosecution of

corporate entities—particularly in regard to the use of deferred and non-

prosecution agreements. At first blush, deferred and non-prosecution agreements

appear to present a win-win situation: They offer the benefits of criminal punish-

ment without the negative collateral consequences that flow from criminal charges.

Their increased use, however, has revealed a different picture. By removing the

threat of collateral consequences, deferred and non-prosecution agreements allow

individual prosecutors to take full advantage of the unique weaknesses of corpora-

tions in the criminal justice system. These weaknesses provide prosecutors with a

dangerous amount of leverage over the corporations they target, creating a bar-

gaining imbalance and a new threat of abuse.

The potential for abuse that flows from the use of deferred and non-prosecution

agreements should be addressed by restricting the discretion of individual prosecu-

tors. This Note argues that the DOJ should look to the solution offered in the

McNulty Memo and require that individual prosecutors receive permission from

the Deputy Attorney General before entering into any deferred or non-prosecution

agreement.

NTRODUCTION

Early in the morning of December 12, 2006, Paul McNulty, then

Deputy Attorney General of the Department of Justice (DOJ), strode

to the podium at the Lawyers for Civil Justice Conference in New

* Copyright



2007 by Erik Paulsen. J.D., 2007, New York University School of Law;

B.A., 1998, Brown University. I would like to express my gratitude to Emily Bishop, Mitra

Ebadolahi, David Kamen, Mario Mendalaro, Paul Monteleoni, Andrew Purcell, Matt

Reagan, and the staff of the New York University Law Review, for their collective work on

this Note. Special thanks are owed to Mathew Miller for his tireless efforts throughout the

production process. Lastly, I am greatly indebted to Denis McInerney and Professor

Harry First, both of whom graciously offered their time and insight to this project.

1434

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 2 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1435

York City.

It was perhaps fitting that Mr. McNulty was surrounded

by an audience attired in dark suits, for he came to give a eulogy. He

told the audience that the “Thompson Memo,” scourge of corporate

America, had been laid to rest.

For nearly four years, the Thompson Memo had served as the

touchstone for all DOJ efforts to root out corporate fraud. Written in

the wake of the accounting scandals that engulfed companies like

Enron, Arthur Andersen, and WorldCom, it guided prosecutors in

deciding whether or not to pursue criminal charges against a corpora-

tion.

Use of the Thompson Memo guidelines, however, led to allega-

tions of abuse. These allegations focused principally on two kinds of

cooperation requests permitted under the Thompson Memo: requests

for waiver of corporate attorney-client privilege and requests for com-

pany refusal to pay employee legal fees. A diverse range of interests

argued that the DOJ was abusing its power by repeatedly demanding

these kinds of cooperation.

Condemnation came from all directions,

and the Thompson Memo could not withstand the collective weight of

the attacks.

When McNulty took the podium on December 12, 2006,

he was there to announce a new DOJ policy—expressed in the

Paul J. McNulty, U.S. Deputy Att’y Gen., Prepared Remarks of Deputy Attorney

General Paul J. McNulty at the Lawyers for Civil Justice Membership Conference

Regarding the Department’s Charging Guidelines for Corporate Fraud Prosecutions (Dec.

12, 2006), available at http://www.usdoj.gov/dag/speech/2006/dag_speech_061212.htm

[hereinafter McNulty Speech].

Id. (“Today I’m announcing revisions to the Thompson Memorandum. The previous

guidance is superseded, and a new memorandum, under my signature, will provide revised

guidance to corporate fraud prosecutors across the country.”).

See infra notes 77–85 and accompanying text.

See Jason McLure, Coalition Scores Major Victory in Policy Retreat: How Business,

Civil Liberties Groups Pushed DOJ To Alter Fraud Stance, L

EGAL

IMES

, Dec. 18, 2006, at

A wide variety of organizations and individuals criticized the cooperation provisions

in the Thompson Memo. The American Civil Liberties Union and American Bar

Association (ABA) combined lobbying efforts with strange bedfellows like the National

Association of Manufacturers and the U.S. Chamber of Commerce. McLure, supra note 4.

Ex-attorneys general and solicitors general from previous administrations, both

Republican and Democrat, co-signed a letter asking for a repeal of the practice. Jason

McLure, Ex-DOJ Officials Blast Current Policy: Letter Signed by Bell, Thornburgh, and

Others Criticizes Approach Outlined in Thompson Memo, L

EGAL

IMES

, Sept. 11, 2006, at

18. Senator Arlen Specter introduced legislation to change DOJ policy. Lynnley

Browning, Senator Calls for an Easing of Corporate-Wrongdoing Rules, N.Y. T

IMES

, Dec.

8, 2006, at C3. A federal judge in New York declared the Memo, in part, unconstitutional.

United States v. Stein, 435 F. Supp. 2d 330, 365 (S.D.N.Y. 2006). Even Thompson himself,

the eponymous author of the memorandum, expressed reservations. See Lynnley

Browning, Former Deputy at Justice Dept. Would Limit Legal Disclosure, N.Y. T

IMES

, Dec.

1, 2006, at C6 (quoting Thompson as saying that use of waiver should be “extremely

limited”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 3 16-OCT-07 14:37

1436 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

“McNulty Memo”

—that addressed the issues that had provoked such

vigorous opposition.

While the modifications in the McNulty Memo may indeed prove

significant, it would be foolish to think that they will completely solve

the problem of abuse in the prosecution of corporate entities. The

world of corporate prosecution has undergone a revolution during the

past decade, ushering in changes that will survive the replacement of

the Thompson Memo. Foremost among these changes is the wide-

spread use by federal prosecutors of deferred and non-prosecution

agreements. The use of these agreements has created the potential for

a different kind of abuse, which the McNulty Memo does not address.

Deferred and non-prosecution agreements are essentially proba-

tionary agreements between a prosecutor and a corporation suspected

of criminal conduct.

Prosecutors agree to waive or defer criminal

prosecution of the corporate entity in exchange for a variety of corpo-

rate promises. These concessions often include massive fines, internal

reforms, and assistance in prosecuting corporate employees.

The use

of these agreements exploded after the demise of the corporate

accounting giant Arthur Andersen. When Andersen collapsed after

its indictment, federal prosecutors realized that prosecution alone

could destroy even the most established of companies.

It is in this context that corporate prosecutors turned to deferred

and non-prosecution agreements, which serve as a middle ground

between full prosecution and declination.

These agreements allow

prosecutors to punish corporations without risking total corporate col-

lapse—an arrangement that appears to benefit both parties. This

seemingly beneficial dynamic, however, hides a new potential for

abuse. It has become increasingly clear that the government holds all

the cards in negotiations over these agreements. As long as the threat

of prosecution lingers over a company, the corporation is compelled

to agree to the prosecutor’s terms, vesting nearly absolute power in

the government’s hands. Unable to risk a potential indictment, the

corporation is thus left at the mercy of the prosecutor.

This bargaining imbalance is not easily rectified—in fact, there

may not be any true solution short of changing the law of corporate

Memorandum from Paul J. McNulty, U.S. Deputy Att’y Gen., to Heads of Dep’t

Components and U.S. Att’ys (Dec. 12, 2006), available at http://www.usdoj.gov/dag/speech/

2006/mcnulty_memo.pdf [hereinafter McNulty Memo].

See infra Part I.A.

See infra Part I.B.

Declination means to decline to prosecute criminally. Benjamin M. Greenblum,

What Happens to a Prosecution Deferred? Judicial Oversight of Corporate Deferred Prose-

cution Agreements, 105 C

OLUM

. L. R

. 1863, 1868 (2005).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 4 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1437

criminal liability. The abusive tendencies of this bargaining imbalance

can, however, be limited by constraining prosecutorial discretion.

This Note argues that decisions related to deferred and non-

prosecution agreements—both the offer of an agreement as well as its

terms—should be channeled through a formal approval process at the

DOJ in Washington, D.C. rather than decided by individual federal

prosecutors.

This process would mirror the procedures already

adopted in the McNulty Memo to handle requests for waiver of

attorney-client privilege. Although this approach would not alter the

bargaining disparity between corporations and prosecutors, it would

curb the potential for abuse by subjecting the decisionmaking of indi-

vidual prosecutors to centralized review. Procedures of this sort

would impose some consistency and uniformity on what has hitherto

been a decentralized and haphazard process.

Part I of this Note details the modern use of corporate prosecu-

tion agreements. Part II reviews the expansive nature of the black

letter law of corporate liability and outlines the procedures developed

by the DOJ to guide prosecutorial discretion. Part III highlights the

steadily growing power of individual prosecutors in the negotiation of

corporate prosecution agreements and points out some early signs of

abuse. Part IV criticizes various remedies that have been proposed to

rein in potential abuse and proposes an expansion of the remedy pro-

vided in the McNulty Memo.

ROSECUTION

GREEMENTS

This Part provides a detailed look at the use of deferred and non-

prosecution agreements. Part I.A lays out the basics of prosecution

agreements: what they are and how they work. Part I.B reviews some

of the common terms of prosecution agreements. Part I.C makes two

significant observations about these agreements: First, prosecutors

retain almost complete discretion in crafting the terms; and second,

these agreements are becoming ever more common.

A. What Are Prosecution Agreements?

A federal prosecutor seeking a middle ground between a prose-

cution and declination has two principal options: a deferred prosecu-

tion agreement or a non-prosecution agreement.

A deferred

See infra Part IV.B.

This is, of course, a bit simplistic. A prosecutor can arrange for civil penalties with

the Securities and Exchange Commission (SEC), for example, instead of pursuing any kind

of criminal discipline. Such options, however, essentially amount to declinations.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 5 16-OCT-07 14:37

1438 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

prosecution agreement is essentially a probationary agreement

arranged between a corporation and the government.

In lieu of

criminally prosecuting the entity, the government files some kind of

criminal charge—normally a complaint, information, or indictment—

but then agrees to hold the charge open as long as the corporation

successfully fulfills the terms of the agreement.

If the prosecutor

deems the agreement’s terms satisfied, the charges are dismissed.

While non-prosecution agreements are identical to deferred pros-

ecution agreements in most respects, there are two key differences.

First, in a non-prosecution agreement the prosecutor does not file a

charging instrument. In some circumstances, this can be a significant

difference: Mere indictment can trigger penalties for certain corpora-

tions.

Second, non-prosecution agreements normally send a less stri-

dent message to the market—the absence of the charging instrument

often indicates a lower level of culpability.

As a result, potential

defendants prefer non-prosecution agreements.

In most other respects, however, the two types of agreements—

especially those signed since the introduction of the Thompson

Memo—tend to have identical terms and qualities.

In fact, some

non-prosecution agreements can be as harsh as their deferred counter-

parts.

For the purposes of this Note, I will treat them as one and the

Despite their frequent use today, prosecution agreements were not engineered in the

context of corporate liability. Rather, they began as a means to rehabilitate juvenile

offenders. Greenblum, supra note 9, at 1864. It was not until the early 1990s that agree-

ments of this type were used to settle criminal charges against entities. Id.

Eugene Illovsky, Corporate Deferred Prosecution Agreements: The Brewing Debate,

RIM

. J

UST

., Summer 2006, at 36, 36, available at http://www.abanet.org/crimjust/cjmag/21-

2/corporatedeferred.pdf.

Interview with David Pitofsky, C

ORP

. C

RIME

., Nov. 28, 2005, at 8, 13.

See id. (noting that a non-prosecution agreement “communicates something that is

substantially less harsh than a deferred prosecution agreement . . . [and] provides the pros-

ecutor with the ability to send a graded message about the seriousness of the misconduct”).

Id.

The HealthSouth Non-Prosecution Agreement, for example, features a three-year,

six-month duration; over $500 million in penalties and restitution; substantial internal gov-

ernance reforms; and a compliance monitor. Non-Prosecution Agreement Between U.S.

Dep’t of Justice and HealthSouth Corp., paras. 3, 8–9 (May 18, 2006) (on file with the

New York University Law Review) [hereinafter HealthSouth NPA]. The Boeing Non-

Prosecution Agreement features a two-year duration; over $600 million in penalties; sub-

stantial internal governance reforms; and an agreement to provide a limited privilege

waiver. Non-Prosecution Agreement Between U.S. Att’y’s Office for E. Dist. Va. and

Boeing Co., paras. 3–4, 6–7 (June 30, 2006) (on file with the New York University Law

Review) [hereinafter Boeing NPA]. These conditions are comparable to modern deferred

prosecution agreements. E.g., Deferred Prosecution Agreement Between U.S. Att’y’s

Office for S. Dist. of N.Y. and KPMG, para. 3 (Aug. 26, 2005) (on file with the New York

University Law Review) [hereinafter KPMG DPA] (noting substantial penalties, internal

business reforms, and privilege waiver).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 6 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1439

same—when I use the term “prosecution agreements,” I am generally

referring to both deferred and non-prosecution agreements, noting

differences only where relevant.

B. Elements of Prosecution Agreements

Prosecution agreements feature a wide variety of terms. These

include, inter alia, provisions about cooperation, acknowledgment of

criminal conduct, business reforms, penalties, and independent

monitors.

Cooperation: Corporate cooperation has been the most contro-

versial element of prosecution agreements. In particular, negative

attention has focused on two common prosecutorial practices:

requests that the company refuse to pay attorneys’ fees for employees

suspected of criminal conduct,

and requests that the company waive

The overwhelming majority of these agreements share the same features despite

their different titles. If a distinction is to be drawn regarding the agreements’ terms, one

might separate out the early non-prosecution agreements from the rest. For example,

many of the early non-prosecution agreements did not feature any set duration—that is,

once the agreement was signed, the company’s relationship with the prosecutor’s office

was essentially over. Of the fifty-nine agreements signed between 1992 and 2006, seven fit

this mold: the Aurora Agreement, the Lazard Agreement, the Merrill Lynch First Agree-

ment, the John Hancock Agreement, the AETNA Agreement, the Sequa Agreement, and

the Salomon Brothers Agreement. See Lawrence D. Finder & Ryan D. McConnell, Devo-

lution of Authority: The Department of Justice’s Corporate Charging Policies, 51 S

. L

OUIS

U. L.J. 1, 6–7, app. (2006) (indicating that six of seven agreements signed before 1999 did

not contain explicit expiration dates); infra notes 46–47 (noting fifty-nine total agreements

from 1992 to 2006). These seven agreements are probably best described as official state-

ments detailing the reasons why the government declined to prosecute the organization in

question.

Once the Holder Memo was released, nearly all deferred and non-prosecution agree-

ments took on the same form. Memorandum from Eric H. Holder Jr., U.S. Deputy Att’y

Gen., to All Component Heads and U.S. Att’ys (June 16, 1999) (on file with the New York

University Law Review) [hereinafter Holder Memo]. Of the seven agreements noted

above which differ from the modern form, six were signed before 1999, the date that the

Holder Memo was issued. The one exception to this generalization is the Aurora

Agreement, which was signed in 2001 and is perhaps best treated as an anomaly. See Non-

Prosecution Agreement Between U.S. Att’y’s Office for S. Dist. of N.Y. and Aurora Food

(Jan. 22, 2001) (on file with the New York University Law Review). Other than the Aurora

Agreement, all the deferred and non-prosecution agreements issued since 1999 are suffi-

ciently similar that they can be spoken of as a group.

Many corporations, either by explicit employment contract or by historical practice,

provide their employees with attorneys’ fees to cover criminal allegations involving actions

committed in the course of employment. United States v. Stein, 435 F. Supp. 2d 330,

353–55 (S.D.N.Y. 2006) (summarizing historical practice of employers indemnifying

employees for legal expenses). In certain circumstances, prosecutors have frowned on

these agreements, and have pressured companies to waive or limit these arrangements to

demonstrate cooperation. The Thompson Memo explicitly stated that the advancement of

attorneys’ fees might be evidence of non-cooperation as the company is “protecting . . .

culpable employees and agents.” Memorandum from Larry D. Thompson, U.S. Deputy

Att’y Gen., to Heads of Dep’t Components and U.S. Att’ys § VI(B), at 7–8 (Jan. 20, 2003),

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 7 16-OCT-07 14:37

1440 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

attorney-client privilege.

Controversy over these practices eventu-

ally resulted in the replacement of the Thompson Memo with the

McNulty Memo.

Although these issues are of tremendous significance, this Note

will not focus on the potential for abuse in the cooperation provisions

of prosecution agreements for two important reasons. First, com-

pelled cooperation has already received considerable attention. To

the extent that commentators have examined the use and abuse of

prosecution agreements, criticism has mainly focused on the issue of

cooperation.

Second, and more significantly, the abuses identified in

these practices are moving toward a legal remedy: The McNulty

Memo is a clear indication that the DOJ is expending serious effort to

address potential abuse in prosecutorial practices related to coopera-

tion. The same, however, cannot be said for the other provisions of

prosecution agreements. Accordingly this Note will focus on the

potential for prosecutorial abuse in these other provisions.

Acknowledgement of Conduct: The majority of prosecution

agreements, especially those signed since the introduction of the

Thompson Memo, require the company to acknowledge responsibility

available at http://www.usdoj.gov/dag/cftf/business_organizations.pdf [hereinafter

Thompson Memo].

The Thompson Memo stated that a corporation’s willingness to waive attorney-client

privilege was highly relevant evidence of a corporation’s cooperation. Thompson Memo,

supra note 20, § VI(B), at 6–7. While the Thompson Memo stated that such waiver was

not an “absolute requirement,” id., many practitioners found that prosecutors demanded

disclosure of privileged documents in nearly all circumstances. Companies that resisted

risked appearing uncooperative. Stephen W. Grafman & Jeffrey L. Bornstein, New Memo

Won’t Help, N

’

L.J., Nov. 14, 2005, at 31, 31 (“In reality today, a corporation that does

not waive these valuable privilege rights effectively is considered a non-cooperator.”).

See supra note 6 and accompanying text.

See, e.g., Am. Coll. of Trial Lawyers, The Erosion of the Attorney-Client Privilege and

Work Product Doctrine in Federal Criminal Investigations, 41 D

. L. R

. 307, 321–22

(2003) (criticizing government for asking for privilege waiver); Mary Beth Buchanan,

Effective Cooperation by Business Organizations and the Impact of Privilege Waivers, 39

AKE

OREST

L. R

. 587, 595–607 (2004) (summarizing and responding to criticisms of

privilege waiver); George Ellard, Making the Silent Speak and the Informed Wary, 42 A

RIM

. L. R

. 985, 992–93 (2005) (criticizing pressure exerted on companies to waive

attorney-client privilege); Lisa Kern Griffin, Compelled Cooperation and the New Corpo-

rate Criminal Procedure, 82 N.Y.U. L. R

. 311, 319–21 (2007) (discussing Thompson

Memo factors for assessing corporate cooperation, which include corporation’s identifica-

tion and sanctioning of employees); Joseph F. Savage, Jr. & Melissa M. Longo, ‘Waive’

Goodbye to Attorney-Client Privilege, B

. C

RIMES

ULL

., Oct. 2000, at 1, 8 (discussing

waiver in context of inadvertant disclosure); Mary Jo White, Corporate Criminal Liability:

What Has Gone Wrong?, in 2 37

NNUAL

NSTITUTE ON

ECURITIES

EGULATION

815,

821 (PLI Corporate Law & Practice, Course Handbook Series No. B-1517, 2005) (arguing

that waiver has “become a heavy stick and a company, only at its extreme peril, declines a

prosecutor’s request for waiver”); Christopher A. Wray & Robert K. Hur, Corporate Crim-

inal Prosecution in a Post-Enron World: The Thompson Memo in Theory and Practice, 43

. C

RIM

. L. R

. 1095, 1170–74 (2006) (summarizing criticisms of cooperation policy).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 8 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1441

for the conduct in question. The company and the government agree

to a “statement of facts” that details the government’s allegations.

The company stipulates to the conduct alleged in the statement of

facts and consents to admission of the statement as evidence if the

prosecution agreement is breached. If this occurs, there is essentially

a corporate confession to the charge.

The power of these provisions

is worth noting: Detailed admissions to criminal charges typically

result in lucrative civil suits that piggyback on the signing of these

agreements.

For many corporations, “the ripple effect” of these

admissions is catastrophic.

The fact that corporations typically agree

to these admissions—despite the consequences in civil suits—under-

scores the degree to which corporations desire to avoid criminal

indictment and/or conviction.

Monetary Penalties: Modern prosecution agreements generally

feature significant monetary penalties, which can involve a combina-

tion of criminal penalties, civil penalties, and civil settlements to non-

governmental plaintiffs—often in staggering amounts. For example,

Boeing’s non-prosecution agreement called for a $565 million civil set-

tlement and a $50 million criminal penalty.

KPMG was ordered to

pay $456 million—$128 million in disgorged fees, $288 million in resti-

tution to the Internal Revenue Service (IRS), and a $100 million pen-

alty to the IRS.

These numbers are not unusual—numerous

agreements feature settlements worth hundreds of millions of dol-

lars.

Other agreements call for more unorthodox penalties, such as

the provision requiring $4 million worth of free medical services in the

E.g., Deferred Prosecution Agreement Between U.S. Dep’t of Justice and

BankAtlantic, para. 2 (Apr. 26, 2006) (on file with the New York University Law Review);

Deferred Prosecution Agreement Between U.S. Dep’t of Justice and Monsanto Co., para.

4 (Jan. 6, 2005) (on file with the New York University Law Review).

E.g., Boeing NPA, supra note 18, para. 12(a)(4); Deferred Prosecution Agreement

Between U.S. Att’y’s Office for D. of Conn. and Operations Mgmt. Int’l, para. 9 (Feb. 8,

2006) (on file with the New York University Law Review) [hereinafter OMI DPA].

Admissions within a prosecution agreement can have significant consequences on

subsequent or pending litigation, as “a plaintiff will seek to use [them] against the corpora-

tion in later civil proceedings.” Michael R. Sklaire & Joshua G. Berman, Deferred Prose-

cution Agreements: What Is the Cost of Staying in Business?, L

EGAL

PINION

ETTER

June 3, 2005, at 2, available at http://www.sonnenschein.com/docs/Cost_of_Staying_in_B.

pdf.

See Stephanie Martz, Trends in Deferred Prosecution Agreements, C

HAMPION

, Nov.

2005, at 45, 45 (“[F]or corporations, the ripple effect of these admissions could be

devastating.”).

Boeing NPA, supra note 18, para. 4(b)–(c).

KPMG DPA, supra note 18, para. 3.

See Finder & McConnell, supra note 19, app. (noting various prosecution agree-

ments with nine-figure fines).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 9 16-OCT-07 14:37

1442 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

prosecution agreement with Roger Williams Hospital.

While such

settlements are not the norm, they evince a potential for abuse: Pros-

ecutors have wide flexibility and few barriers in crafting penalties.

Business Reforms: Prosecution agreements often require the

target company to make significant changes to the way it conducts

business. Some of the more common undertakings involve personnel.

The agreement might call for the termination or disciplining of

employees for either contributing to the underlying criminal conduct

or for failure to cooperate with the investigation.

Required changes

might also include the hiring of new senior management,

new board

members,

or new auditors.

The other common type of business reform involves operational

changes. Perhaps the most common are enhancements to compliance

programs, which can include, among other measures, amended finan-

cial controls, hotlines for whistleblowers, training programs to under-

score legal behavior, new personnel hiring policies, and ethics

officers.

These business reforms have also raised concern. Shareholders

traditionally control decisions regarding the staffing of boards of

directors, which in turn control the staffing of management. Prosecu-

tors are not businessmen, and some commentators have questioned

the wisdom of permitting prosecutorial intrusion into the internal

workings of large corporations.

Deferred Prosecution Agreement Between U.S. Att’y’s Office for Dist. of R.I. and

Roger Williams Med. Ctr., para. 12 (Jan. 27, 2006) (on file with the New York University

Law Review) [hereinafter Roger Williams DPA].

E.g., Non-Prosecution Agreement Among U.S. Att’y’s Office for E. Dist. of N.Y.,

U.S. Att’y’s Office for S. Dist. of N.Y. and Bank of New York, para. 10(a) (Nov. 8, 2005)

(on file with the New York University Law Review); HealthSouth NPA, supra note 18,

para. 8(a).

Non-Prosecution Agreement Between U.S. Att’y’s Office for E. Dist. of N.Y. and

Symbol Techs., para. 9(b) (June 3, 2004) (on file with the New York University Law

Review) [hereinafter Symbol NPA].

E.g., HealthSouth NPA, supra note 18, para. 8(b); Symbol NPA, supra note 33, para.

9(c).

E.g., HealthSouth NPA, supra note 18, para. 8(d); Deferred Prosecution Agreement

Between U.S. Att’y’s Office for S. Dist. of N.Y. and Prudential Secs., Inc. 3 (Oct. 27, 1994)

(on file with the New York University Law Review).

E.g., HealthSouth NPA, supra note 18, para. 8(c); Symbol NPA, supra note 33, para.

9(d).

E.g., OMI DPA, supra note 25, paras. 5–6.

Professor John C. Coffee argues, for example, that there is little to stop a prosecutor

from demanding as a condition of deferred prosecution that a board without much diver-

sity reconstitute itself as half female and at least one-third minority. John C. Coffee Jr.,

Deferred Prosecution: Has It Gone Too Far? N

’

L.J., July 25, 2005, at 13, 13. Coffee

believes that shareholders should have the right to choose their own directors, and such

interference by prosecutors goes beyond their “competence or entitlement.” Id.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 10 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1443

Independent Monitors: A large percentage of modern prosecu-

tion agreements require compliance monitors—independent experts

retained to supervise compliance with the agreement who report peri-

odically to the prosecutor.

These monitors are normally either com-

pliance specialists or august members of the legal community.

The

corporation pays the costs of the monitor and his/her staff and is

expected to give the monitor relatively unfettered access to business

operations.

For some, the monitor’s very presence assuages con-

cerns about prosecutorial expertise in business regulation.

Others,

however, generally doubt the wisdom of allowing prosecutors to inter-

pose their power into sensitive boardroom decisions, arguing that such

business decisions are beyond prosecutors’ expertise.

Miscellaneous Penalties: Some of these agreements contain unu-

sual provisions which defy categorization. Prosecutors have signifi-

cant leeway as to what conditions they demand of the target

corporations, and there is little to stop a prosecutor from crafting

unique penalties. For example, the agreement between WorldCom

and the Attorney General of Oklahoma required the company to

create sixteen hundred new jobs over a ten-year period.

Similarly,

the Bristol-Meyers Squibb Agreement mandated the creation of an

endowed chair in business ethics at Seton Hall Law School, the local

U.S. Attorney’s alma mater.

See Finder & McConnell, supra note 19, app. (tabulating terms of deferred and non-

prosecution agreements, including monitors).

In the Bristol-Meyers Squibb Agreement, for example, the prosecutor and company

agreed to use Frederick B. Lacey, a former U.S. Attorney and federal judge. Deferred

Prosecution Agreement Between U.S. Att’y’s Office for D.N.J. and Bristol-Meyers Squibb

Co., para. 5 (June 15, 2005) (on file with the New York University Law Review) [herein-

after Bristol-Meyers DPA].

E.g., Roger Williams DPA, supra note 31, paras. 21–31; HealthSouth NPA, supra

note 18, para. 11.

See Christopher J. Christie & Robert M. Hanna, A Push Down the Road of Good

Corporate Citizenship: The Deferred Prosecution Agreement Between the U.S. Attorney for

the District of New Jersey and Bristol-Meyers Squibb Co., 43 A

. C

RIM

. L. R

. 1043, 1055

(2006) (“A strong, independent monitor is in a far better position to ride herd over a

mammoth corporation than any U.S. Attorney’s Office or Probation Office. . . . Monitors

are able to observe and understand the business they oversee, . . . in ways that federal

prosecutors never could or should.”); Interview with David Pitofsky, supra note 14, at 9

(“One of the reasons why the deferred prosecution agreements require a monitor to be put

in place is that the prosecutor’s office has no experience or skills to analyze whether a

company is reforming its internal governance practices.”).

See supra notes 32–38 and accompanying text.

Deferred Prosecution Agreement Between State of Oklahoma and WorldCom, Inc.

2 (Mar. 12, 2004) (on file with the New York University Law Review).

Bristol-Meyers DPA, supra note 40, para. 20. The OMI Agreement contains a sim-

ilar provision with the Coast Guard Academy. OMI DPA, supra note 25, app. A.

Christopher Christie, the U.S. Attorney who negotiated the Bristol-Meyers Squibb Agree-

ment, states that the idea of endowing a chair in ethics came from defense counsel. Fur-

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 11 16-OCT-07 14:37

1444 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

C. The Use and Frequency of Prosecution Agreements

There are two initial observations that should be made about the

use of prosecution agreements in the corporate criminal context.

First, prosecutors are not significantly limited in what they might

include as conditions in these agreements. As noted above, prosecu-

tion agreements feature a wide variety of terms, often including enor-

mous fines, serious overhauls of business practices, and full admissions

of criminal responsibility. Despite the severity of these conditions,

however, corporations nearly always accept the prosecutor’s terms.

The content of the agreement depends almost entirely on the discre-

tion of the prosecutor; what the prosecutor asks for, the prosecutor

gets.

A second observation relates to the frequency of these agree-

ments. From 1992 to the end of 2002, the ten years following the first

use of prosecution agreements in the corporate criminal context, there

were sixteen deferred and non-prosecution agreements—an average

of fewer than two a year.

From the beginning of 2003 to the end of

2006, however, there were a staggering forty-three prosecution agree-

ments—an average of more than ten a year.

And during this latter

ther, Seton Hall was chosen because Rutgers University School of Law already had a chair

in business ethics endowed by Prudential. Christie & Hanna, supra note 42, at 1058 n.29.

Finder & McConnell, supra note 19, app. Lawrence Finder and Ryan McConnell

have assembled the most comprehensive index of known agreements. In the appendix to

their article they count ten deferred and non-prosecution agreements during the period

from 1992 to 2002. They intentionally excluded the agreements with Coopers & Lybrand,

Salomon Brothers, and Sequa because those agreements were never made public. Id. at 36

n.193. For the purpose of this counting exercise, however, I am including those three

agreements as well as three agreements not considered by Finder and McConnell—the

agreements with JB Oxford & Co., Hilton Hotels, and Doyon Drilling. Deferred Prosecu-

tion Agreement Among U.S. Att’y’s Office for Cent. Dist. of Cal., JB Oxford Holdings,

Inc., and JB Oxford & Co. (Feb. 12, 2000) (on file with the New York University Law

Review); Press Release, Office of U.S. Att’y for W. Dist. of Mo., Hilton Kansas City

Corporation and Hildeon Gaming Enter Pretrial Diversion Agreement Over Investigation

of $1.25 Million Grant (Aug. 12, 1998) (on file with the New York University Law Review);

Deferred Prosecution Agreement Between U.S. Att’y’s Office for Dist. of Ala. and Doyon

Drilling Inc. (Apr. 1998) (on file with the New York University Law Review).

Finder & McConnell, supra note 19, app. Finder and McConnell counted thirty-

eight deferred and non-prosecution agreements between 2003 and 2006. They intention-

ally excluded two agreements from their list: the Hilfiger Agreement, because the agree-

ment was never released to the public, and the WorldCom Agreement, because it was

negotiated with the Attorney General of Oklahoma. For the purposes of this counting

exercise, I have included both agreements. Further, three agreements were signed in the

final months of 2006, too late to be included in Finder and McConnell’s article. These

three agreements—the Statoil Agreement, the SSI Korea Agreement, and the Intermune

Agreement—are also included in my final count of forty-three agreements. Press Release,

U.S. Dep’t of Justice, U.S. Resolves Probe Against Oil Company that Bribed Iranian Offi-

cial (Oct. 13, 2006) (on file with the New York University Law Review); Press Release,

U.S. Dep’t of Justice, Schnitzer Steel Industries Inc.’s Subsidiary Pleads Guilty to Foreign

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 12 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1445

period, the rate of use was still increasing—2006 alone saw eighteen

such agreements.

The first of these two observations—that prosecutors possess

wide discretion in crafting agreements—points to a developing

problem of abuse, a concern that will be addressed in Parts III and IV

of this Note. Before addressing the issue of abuse, however, we need

to understand why prosecution agreements have become so common

in the area of corporate criminal liability. The peculiar nature of cor-

porate prosecution—in particular the broad scope of entity liability

and the concern for “collateral consequences”—has provided prose-

cutors with clear incentives to use these agreements with increasing

frequency.

ROSECUTION

GREEMENTS AND

ORPORATE

RIMINAL

ONDUCT

This Part explores reasons why prosecutorial agreements have

become so common in the area of corporate criminal liability. Part

II.A lays out the basic law of corporate criminal liability, which is

astonishingly broad: Any criminal act committed by an employee of a

corporation, if committed within the scope of employment and in part

to benefit the corporation, can be imputed to the corporation itself.

Part II.B details some common concerns with entity prosecution, and

goes on to discuss federal prosecutors’ responses. Part I.C focuses on

the most significant of these concerns: the fear of collateral conse-

quences. Typified by the Arthur Andersen collapse, corporate prose-

cution can visit massive harms upon innocent parties who depend

upon the corporation. In part to avoid these effects, prosecutors have

turned to prosecution agreements as a means to punish companies

while sidestepping collateral consequences.

A. The Remarkably Broad Scope of Corporate Criminal Liability

Modern corporate criminal liability was born in 1909 in New York

Central & Hudson River Railroad Co. v. United States.

New York

Bribes and Agrees to Pay a $7.5 Million Criminal Fine (Oct. 16, 2006) (on file with the New

York University Law Review); Deferred Prosecution Agreement Between U.S. Att’y’s

Office for N. Dist. of Cal. and Intermune, Inc. (Oct. 28, 2006) (on file with the New York

University Law Review).

This count includes the fifteen 2006 agreements listed in Finder and McConnell’s

matrix, as well as the Statoil Agreement, the SSI Korea Agreement, and the Intermune

Agreement. See supra note 47 and accompanying text.

See infra notes 54–66 and accompanying text.

212 U.S. 481 (1909); Developments in the Law—Corporate Crime: Regulating Corpo-

rate Behavior Through Criminal Sanctions, 92 H

ARV

. L. R

. 1226, 1246 & n.12 (1979).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 13 16-OCT-07 14:37

1446 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

Central was a rate-fixing case: An employee of the defendant railroad

violated federal tariff laws on sugar shipments.

Instead of filing

criminal charges solely against the employee, however, the govern-

ment took the radical step of charging both the employee and the cor-

porate employer. Prior to New York Central, no federal standard

existed for charging a corporation with a crime that required intent.

Rather, the legal community generally accepted that corporations

could not be charged with criminal behavior because corporations are

fictional persons.

New York Central drastically altered the legal landscape. Bor-

rowing the tort law principle of respondeat superior,

the Supreme

Court held that a corporate entity could be criminally liable for its

employees’ actions.

The imposition of entity liability through the

respondeat superior model addressed two fundamental concerns asso-

ciated with corporate crime. First, entity liability augmented the

deterrent effect of the law: Because corporations often derived bene-

fits from the unlawful practices of their employees, criminal conduct

could only be deterred effectively if the corporation, in addition to its

employees, was subject to prosecution.

Second, vicarious corporate

N.Y. Cent., 212 U.S. at 489–91.

The decision in New York Central was not completely without precedent; there had

been some development of corporate liability for public nuisances and strict liability crimes

before 1909 in non-federal forums. See V. S. Khanna, Corporate Criminal Liability: What

Purpose Does It Serve?, 109 H

ARV

. L. R

. 1477, 1479–82 (1996) (describing evolution of

corporate criminal liability). In addition, Justice Day, the author of the New York Central

opinion, cited a variety of non-federal sources for the proposition that corporations could

be held liable for crimes of intent, including a contemporaneous criminal treatise and court

decisions from Massachusetts and England. N.Y. Cent., 212 U.S. at 492–93. However, as

New York Central was the first case where the Supreme Court explicitly used these prece-

dents in formulating federal law, commentators generally consider this case to be the pro-

genitor of entity liability in the United States. See, e.g., Shaun P. Martin, Intracorporate

Conspiracies, 50 S

TAN

. L. R

. 399, 406–07 (1998) (noting that Supreme Court ratified

previous developments in corporate liability in New York Central, signaling “beginning of

the end of corporate immunity from criminal liability”).

See N.Y. Cent., 212 U.S. at 492 (“Some of the earlier writers on common law held the

law to be that a corporation could not commit a crime.”). Specifically, Justice Day noted

that Blackstone’s Commentaries stated, “‘A corporation cannot commit treason, or felony,

or other crime in its corporate capacity, though its members may, in their distinct indi-

vidual capacities.’” Id. (quoting W

ILLIAM

LACKSTONE

, 1 C

OMMENTARIES

*476).

See id. at 493–94 (describing liability of corporation for acts of employee). Respon-

deat superior is defined as “[t]he doctrine holding an employer or principal liable for the

employee’s or agent’s wrongful acts committed within the scope of employment or

agency.” B

LACK

’

ICTIONARY

1338 (8th ed. 2004).

N.Y. Cent., 212 U.S. at 495–96.

See id. (“[T]o give [corporations] immunity from all punishment because of the old

and exploded doctrine that a corporation cannot commit a crime would virtually take away

the only means of effectually controlling the subject-matter and correcting the abuses

aimed at.”); Khanna, supra note 52, at 1483 (describing deterrence as motive for change in

jurisprudence).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 14 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1447

liability sidestepped the thorny issue of intent: By using the indi-

vidual’s criminal conduct as the basis for the criminal charge against

the corporation, the law could satisfactorily address the issue of how

to establish the mens rea of an inanimate entity.

The New York Central Court suggested a two-part test to effec-

tuate this criminal version of respondeat superior: If an employee

committed a crime (1) within the scope of his or her employment and

(2) for the benefit of the corporation, the corporation could be liable

for the employee’s criminal act.

This standard remains the law to

this day.

The two elements of the New York Central test—“scope of

employment” and “for the benefit”—are not difficult to satisfy. Con-

sequently, virtually every case of employee misconduct could result in

corporate criminal liability. Courts have construed the “scope of

employment” very broadly, covering actions authorized by both

actual

and apparent

authority. Federal courts have considered

employee actions to be within the scope of employment even where

the corporation expressly forbade the behavior in question, either

with a general policy or with individual admonitions to the culpable

employees.

“For the benefit” has also been construed broadly.

Courts do not require that the corporation actually receive a benefit;

it is sufficient that the employee merely intends to benefit the corpo-

Cf. Developments in the Law—Corporate Crime: Regulating Corporate Behavior

Through Criminal Sanctions, supra note 50, at 1241 (noting that “mental state has no

meaning when applied to a corporate defendant”).

See N.Y. Cent., 212 U.S. at 493–95 (allowing criminal prosecution of corporation

where agents act “within the authority conferred upon them” and where “the corporation

. . . profits by the transaction”).

See, e.g., United States v. Philip Morris USA, Inc., 449 F. Supp. 2d 1, 892–93 (D.D.C.

2006) (applying New York Central test of corporate criminal liability).

An employee acts with actual authority if “a corporation knowingly and intentionally

authorizes [the] employee to act on its behalf.” Michael Viano & Jenny R. Arnold, Corpo-

rate Criminal Liability, 43 A

. C

RIM

. L. R

. 311, 314 (2006); see also Joel M. Androphy

et al., General Corporate Criminal Liability, 60 T

. B. J. 121, 122 (1997) (discussing actual

authority).

An employee acts with apparent authority “if a third party reasonably believes that

the [employee] has the authority to perform the act in question.” Viano & Arnold, supra

note 60, at 314; see also Androphy, supra note 60, at 121–22 (discussing apparent

authority).

See United States v. Automated Med. Labs., Inc., 770 F.2d 399, 407 (4th Cir. 1985)

(“The fact that many of [the employees’] actions were unlawful and contrary to corporate

policy does not absolve [the defendant-corporation] of legal responsibility for their acts.”);

United States v. Hilton Hotels Corp., 467 F.2d 1000, 1007 (9th Cir. 1972) (holding that

company can be found liable “for the acts of its agents in the scope of their employment,

even though contrary to general corporate policy and express instructions to the agent”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 15 16-OCT-07 14:37

1448 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

ration.

Nor do courts require “that the employee be primarily con-

cerned with benefiting the corporation;”

employees may act

principally for their own gain and intend to benefit the corporation

only in a secondary manner.

Generally speaking, an employee’s

behavior must be completely inimical to the corporation’s interests—

for example, violating a fiduciary duty owed to the company itself—

for courts to determine that the corporation received no benefit.

The New York Central standard thus offers a simple method for

finding a corporation criminally liable. It is a powerful weapon in the

hands of prosecutors since nearly any predicate criminal act by an

employee can lead to a corporate criminal charge. Given the low

threshold of liability established by New York Central, it is natural to

ask why so few corporations are convicted of federal crimes. The

answer is prosecutorial discretion.

B. DOJ Efforts To Guide Prosecutorial Discretion in Corporate

Prosecution: The Holder, Thompson, and McNulty Memos

There are a variety of reasons why a prosecutor might be disin-

clined toward using New York Central’s vicarious criminal liability

standard. Some concerns with vicarious criminal liability border on

the philosophical: The prosecutor may struggle with conceptualizing

corporations as entities capable of intent.

For most, however, con-

cerns with vicarious criminal liability are moral or instrumental.

See Automated Med. Labs., 770 F.2d at 407 (“[W]hether the agent’s actions ulti-

mately redounded to the benefit of the corporation is less significant than whether the

agent acted with the intent to benefit the corporation.”); Viano & Arnold, supra note 60, at

316 (“The corporation [to be held liable] need not have actually received a benefit; the

employee’s mere intention to bestow such benefit suffices.”).

Viano & Arnold, supra note 60, at 316.

See Automated Med. Labs., 770 F.2d at 407 (upholding conviction of corporation

based on employee’s acting “at least in part” to benefit corporation); Viano & Arnold,

supra note 60, at 316 (noting that “courts recognize that many employees act primarily for

their own personal gain”).

See Standard Oil Co. of Tex. v. United States, 307 F.2d 120, 129 (5th Cir. 1962)

(reversing conviction of corporate defendants because criminal acts of corporate

employees were intended to defraud company itself); Viano & Arnold, supra note 60, at

317 & n.41 (noting that corporation cannot be subject to criminal liability on basis of

employee behavior that is “expressly contrary to the interests of the corporation and for

which the corporation derives no benefit”).

Edward, First Baron Thurlow, is famously said to have quipped that a corporation

has “no soul to be damned, and no body to be kicked.” M

ERVYN

ING

, P

UBLIC

OLICY

AND THE

ORPORATION

1 (1977) (attributing quote to Thurlow); John C. Coffee, Jr., “No

Soul to Damn: No Body to Kick”: An Unscandalized Inquiry into the Problem of Corpo-

rate Punishment, 79 M

ICH

. L. R

. 386, 386 & n.1 (1981) (same). On this subject, see

Samuel W. Buell, The Blaming Function of Entity Criminal Liability, 81 I

. L.J. 473, 475

& n.8, 517 & n.191 (2006). Buell observes that scholars have long doubted the logic of

respondeat superior criminal liability, “pointing to illogic in retribution toward objects and

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 16 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1449

Overuse of the vicarious liability standard runs the risk of criminally

condemning an entity that may have had little to do with the criminal

behavior in question.

Vicarious criminal liability may overdeter,

compelling risk-averse corporations to steer employees away from

borderline but permissible conduct out of uncertainty over punish-

ment.

Efficiency may be best served by reserving criminal liability

for individuals and subjecting corporations to civil penalties.

Finally,

and perhaps most significantly, criminal charges against a corporation

can negatively affect shareholders, employees, and communities

dependent on the corporation—parties which may have had no

involvement in the alleged misconduct.

These negative effects are

typically described as “collateral consequences”—damages to those

innocent of the wrongdoing in question.

In response to these various concerns surrounding the use of cor-

porate liability, the DOJ issued a series of memoranda outlining fac-

tors that federal prosecutors should consider in determining whether

to indict a corporation. These memos, named after their Deputy

Attorney General authors, have served as touchstones for discre-

tionary decisions on corporate prosecution.

The first of these memos, named after Deputy Attorney General

Eric Holder,

was released in 1999 without much fanfare.

At that

time, the United States was experiencing unprecedented business

prosperity, and corporate prosecution was hardly en vogue.

Within

a short period of time, however, this dramatically changed. In the

years following the issuance of the Holder Memo, a series of corpo-

the impossibility of fitting liberal concepts about responsibility with nonhuman actors.” Id.

at 475.

Samuel W. Buell notes that although the rule of entity liability allows virtually all

crimes committed by corporate agents to be imputed to the corporation, the law is not

actually applied in such an aggressive manner: Prosecutors have developed a social prac-

tice of evaluating an institution’s “real responsibility” in determining whether or not to

charge the institution. Buell, supra note 67, at 526; see also Thompson Memo, supra note

20, § IV, at 5 (noting that it may not be appropriate to apply respondeat superior liability

for an employee’s “single isolated act” and that prosecutors “should exercise sound discre-

tion in evaluating the pervasiveness of wrongdoing within a corporation”).

See infra note 123 and accompanying text.

See generally Khanna, supra note 52 (comparing efficacy of corporate civil and crim-

inal liability and concluding that civil liability can provide more effective deterrence at

lower cost).

See Thompson Memo, supra note 20, § IX, at 12–13 (expressing this concern).

See id. (describing collateral effects of corporate prosecution).

Holder Memo, supra note 19.

See Finder & McConnell, supra note 19, at 14 (observing that Holder Memo did not

receive as much attention as later Thompson Memo).

See id. (noting that Holder Memo was “drafted in a period of unprecedented

prosperity”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 17 16-OCT-07 14:37

1450 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

rate accounting scandals occurred, including the notable failures of

Enron and WorldCom.

As the facts of these scandals came to light,

it became clear that the corporations’ own aggressive business cul-

tures, not merely the actions of a few bad employees, were in large

part driving the criminality.

In response to this wave of business crime, the DOJ reprioritized

the prosecution of corporate entities. The Thompson Memo, the most

significant of the three Deputy Attorney General memos, was

released in early 2003 as part of this renewed DOJ effort.

The

Memo noted that prosecutions against individuals could be supple-

mented by the unique benefits that flow from an additional charge

against the corporate entity.

It also posited that an indictment

against a corporation could change the culture and behavior of the

entity, as well as that of other corporations in the industry.

Industry

peers “are likely to take immediate remedial steps” when another cor-

poration in its respective industry is indicted for criminal conduct, so

that one prosecution can create opportunities for “deterrence on a

massive scale.”

To effectuate these purposes, the Thompson Memo laid out a

multifactor test to guide the decision to charge a corporation.

Although the DOJ specified that all federal prosecutors were required

Id.

See John C. Coffee, Jr., What Caused Enron? A Capsule Social and Economic His-

tory of the 1990s, 89 C

ORNELL

L. R

. 269, 270, 281–82 (2004) (arguing “that Enron and

related scandals were neither unique nor idiosyncratic; rather, pervasive problems arose

that undercut existing systems of corporate governance”). See generally B

ETHANY

EAN

& P

ETER

LKIND

, T

MARTEST

UYS IN THE

OOM

: T

MAZING

ISE

AND

CANDALOUS

ALL OF

NRON

(2003) (providing detailed history of Enron scandal

and description of business culture that generated it). Coffee notes, among other explana-

tions for the spate of corporate scandals, that emphasis on earnings and stock price, with

executive compensation tied to success in those metrics, induced managers to engage in

risky behavior. Coffee, supra, at 275.

See Carmen Couden, The Thompson Memorandum: A Revised Solution or Just a

Problem?, 30 J. C

ORP

. L. 405, 413 (2005) (“[I]n response to the corporate fraud scandals

. . . Deputy Attorney General Larry D. Thompson issued a revision of the DOJ’s guide-

lines on corporate prosecution.”); Wray & Hur, supra note 23, at 1097 (“The spate of

corporate scandals that began with Enron’s meltdown in 2001 prompted the Bush Admin-

istration to dramatically increase the federal government’s focus on rooting out corporate

fraud . . . .”).

Thompson Memo, supra note 20, § I(B), at 1.

Id.

The factors are:

1. the nature and seriousness of the offense, including the risk of harm to the

public, and applicable policies and priorities, if any, governing the prosecu-

tion of corporations for particular categories of crime;

2. the pervasiveness of wrongdoing within the corporation, including the com-

plicity in, or condonation of, the wrongdoing by corporate management;

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 18 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1451

to use these factors,

the Memo was ambiguous about their practical

application. It stated that while the nine factors listed were meant to

be “illustrative of those that should be considered,” they were not

meant to be an exhaustive list.

Furthermore, the Memo was silent

on how much weight a prosecutor should put on any individual factor,

stating, without further explanation, that “in some cases one factor

may override all others.”

In short, the Thompson Memo balancing

test entrusted a massive amount of discretion to the individual prose-

cutors who would determine whether to employ the New York Central

vicarious liability standard.

In December 2006, the DOJ replaced the Thompson Memo with

the McNulty Memo,

which contains two significant modifications

relating to corporate cooperation with criminal investigations. As

noted in the Introduction, the cooperation provisions of the

Thompson Memo produced an enormous outcry: The Memo was crit-

icized by businesses, advocacy organizations, senators, ex-attorneys

general, and even Thompson himself.

The McNulty Memo

addresses these criticisms by making two key changes. First, the

McNulty Memo states that prosecutors generally may not consider

whether a corporation is advancing an employee attorneys’ fees in

3. the corporation’s history of similar conduct, including prior criminal, civil,

and regulatory enforcement actions against it;

4. the corporation’s timely and voluntary disclosure of wrongdoing and its

willingness to cooperate in the investigation of its agents, including, if neces-

sary, the waiver of corporate attorney-client and work product protection;

5. the existence and adequacy of the corporation’s compliance program;

6. the corporation’s remedial actions, including any efforts to implement an

effective corporate compliance program or to improve an existing one, to

replace responsible management, to discipline or terminate wrongdoers, to

pay restitution, and to cooperate with the relevant government agencies;

7. collateral consequences, including disproportionate harm to shareholders,

pension holders and employees not proven personally culpable and impact

on the public arising from the prosecution;

8. the adequacy of the prosecution of individuals responsible for the corpora-

tion’s malfeasance;

9. the adequacy of remedies such as civil or regulatory enforcement actions.

Id. § II(A), at 3 (internal cross-references omitted). For an analysis of each factor, see

Couden, supra note 78, at 407–14.

U.S. D

’

TOF

USTICE

, C

RIMINAL

ESOURCE

ANUAL

§ 163 (2005) (“The

Thompson Memorandum sets forth nine factors that federal prosecutors must consider in

determining whether to charge a corporation or other business organization.”); see also

United States v. Stein, 435 F. Supp. 2d 330, 338 (S.D.N.Y. 2006) (noting that, unlike

Holder Memo, Thompson Memo was binding on all federal prosecutors).

Thompson Memo, supra note 20, § II(B), at 4.

Id.

McNulty Memo, supra note 6.

See supra notes 4–5 and accompanying text.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 19 16-OCT-07 14:37

1452 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

assessing cooperation.

This change addresses a recent constitutional

challenge to the attorneys’ fees aspect of the Thompson Memo.

The

second modification relates to prosecutorial requests for waiver of

attorney-client privilege—a highly contentious aspect of the

Thompson Memo.

The McNulty Memo does not ban these waiver requests. Rather,

it restricts their use in three significant ways. First, the Memo clarifies

that such waivers should only be requested when a “legitimate need”

exists.

Second, the McNulty Memo states that if the government

makes a request for waiver, the waiver should be “the least intrusive

waiver necessary.”

Furthermore, the Memo precludes prosecutors

from considering a corporation’s refusal to release “attorney-client

communications or non-factual attorney work product” in their

charging decision.

Third—and most importantly for the present dis-

cussion—the Memo imposes a certain amount of uniformity by

requiring individual prosecutors to vet their waiver decisions with

upper-level DOJ officials.

As discussed below, this latter proposal—

vetting by senior DOJ officials—may provide guidance for remedying

threats of abuse in the use of prosecution agreements with

corporations.

C. Accounting Fraud and Extreme Collateral Consequences

Although each of the three DOJ memos offered some guidance

for prosecutors faced with the decision of whether or not to prosecute

a corporation, the memos did not—and perhaps could not—tell a

prosecutor how to make a final discretionary determination. In

making these determinations, one concern in particular became para-

mount: the massive collateral consequences to innocent shareholders

and employees that often result from corporate indictment and/or

prosecution. Even in those situations when entity prosecution could

Specifically, the McNulty Memo states that prosecutors should not consider a corpo-

ration’s advancement of legal fees as indicative of non-cooperation except in rare cases

where the totality of the circumstances indicates that the corporation’s actions were

intended to impede a government investigation. McNulty Memo, supra note 6,

§ VII(B)(3), at 11–12; see also McNulty Speech, supra note 1 (emphasizing that McNulty

Memo “generally prohibit[s] prosecutors from considering whether a corporation is

advancing attorneys’ fees”).

See United States v. Stein, 435 F. Supp. 2d 330, 362–69 (S.D.N.Y. 2006) (holding that

government violated Fifth and Sixth Amendments by causing company to cut off payment

of legal fees to employees).

McNulty Memo, supra note 6, § VII(B)(2), at 8–9.

Id. § VII(B)(2), at 9.

Id. § VII(B)(2), at 10.

Id. § VII(B)(2), at 9.

See infra Part IV.B.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 20 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1453

be justified because of a corrupt corporate culture (as was arguably

the case with Enron), prosecutors were rightly concerned with the

negative effects on innocent third parties such as shareholders and

employees. Although each of the DOJ memos explicitly counsels that

prosecutors should consider the issue of collateral consequences, they

provide little guidance on how that interest should be balanced

against the others.

The concern with collateral consequences is especially acute

when dealing with charges related to financial fraud.

It has often

been said that no major American financial services institution has

survived criminal indictment.

The prosecution of Arthur Andersen

is perhaps the most dramatic recent example of the collateral conse-

quences of prosecuting a financial services company: After indict-

ment for its role in the Enron accounting scandal, more than twenty-

eight thousand Arthur Andersen employees were laid off in the

United States alone.

A company like Arthur Andersen could not

survive such a public blow to its reputation and integrity.

The collateral consequences of financial fraud are not limited,

however, to financial services companies. Disastrous collateral conse-

quences can result when any major company that depends on public

reputation—publicly traded companies, in particular—engages in

fraud related to its finances. Criminal charges involving financial

fraud call into question the basic information relayed by the company

about its financial health and prospects—information that various

David Pitofsky, the lead prosecutor in the prosecution agreement negotiations with

Computer Associates, believes that a distinction needs to be drawn between prosecutions

of corporations for financial fraud and prosecutions for other kinds of crimes:

You have to draw a distinction between the types of violations.

If it is a Foreign Corrupt Practices Act violation, or an antitrust or an

environmental violation, that’s different than if it is an accounting fraud viola-

tion. The market receives an accounting fraud violation as a black cloud over

the entire company. Investors don’t know what they are buying when they buy

a share of stock of the company.

With an FCPA violation, it doesn’t undermine your confidence in the

entire company, top to bottom. Its [sic] something that can be carved out.

You can see that the company can go forward, that it has good products and

strong finances and that it is worth investing in. But when you are talking

about a securities fraud case, you really have to ask yourself—is this company

a total sham?

Interview with David Pitofsky, supra note 14, at 12.

E.g., Ken Brown et al., Called to Account: Indictment of Andersen in Shredding Case

Puts Its Future in Question: Obstruction of Justice Count May Speed the Departure of Cli-

ents and Partners: Firm Calls It ‘Death Penalty,’ W

ALL

. J., Mar. 15, 2002, at A1; Monica

Langley & Ian McDonald, Marsh Averts Criminal Case with New CEO, W

ALL

. J., Oct.

26, 2004, at A1.

Finder & McConnell, supra note 19, at 14–15.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 21 16-OCT-07 14:37

1454 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

markets rely upon in evaluating the company.

Charges involving

inaccurate financial data send shockwaves through the investing and

creditor communities and often cause the market to lose faith in the

targeted corporation.

The heightened stakes of these kinds of offenses thrust a tremen-

dous responsibility upon the government. When confronted with a

corporation that measures poorly against the DOJ memo factors, the

government has to decide not only whether the company’s behavior

merits sanction, but whether it wants to risk putting the company out

of business and inflicting heavy losses on innocent third parties.

100

It is in this context that the use of prosecution agreements has

become so significant. Prosecutors—even when convinced that a

company’s actions merit criminal sanction—have become hesitant to

risk destroying the company by prosecuting it, thereby harming share-

holders and employees. Prosecution agreements provide a solution to

this problem, permitting the imposition of criminal penalties without

the risk of dire collateral consequences.

101

III

ORPORATE

ULNERABILITY AND THE

PECTER OF

ROSECUTORIAL

BUSE

This Part explores the ramifications of unrestrained prosecutorial

power in the use of prosecution agreements. Part III.A notes the

respective reasons why both a prosecutor and a corporation might

Id.

See infra notes 104–07 and accompanying text.

100

Deputy Attorney General James Comey has remarked that corporate criminal pros-

ecutions should only be used in cases where the prosecutor is willing to “put down” the

corporation. James Comey, U.S. Deputy Att’y Gen., Statement at Justice Dep’t News

Conference (Sept. 22, 2004) (LexisNexis Media Transcripts). In this vein, then-New York

Attorney General Eliot Spitzer backed off prosecuting Marsh & McLennan when told that

the company “was in an ‘Arthur Andersen situation.’” B

ROOKE

A. M

ASTERS

, S

POILING

FOR A

IGHT

: T

ISE OF

LIOT

PITZER

218 (2006).

101

Of course, this particular advantage of prosecution agreements may not be the only

explanation for their increased use. Part of this increase can perhaps be attributed to the

general reprioritization of corporate crime prevention following the accounting scandals of

the early part of the decade. On July 9, 2002, President George W. Bush formed a

Corporate Fraud Task Force, which focused the DOJ and U.S. Attorneys’ Offices on the

issue of corporate crime. See Exec. Order No. 13,271, 3 C.F.R. 245 (2002), reprinted as

amended in 28 U.S.C. § 509 (Supp. IV 2004) (establishing Task Force). At the state level,

attorneys general like Eliot Spitzer reinvigorated efforts to root out fraud. See generally

ASTERS

, supra note 100 (highlighting Eliot Spitzer’s campaigns against white collar

fraud). However, while the reallocation of resources aimed at detection and prosecution

of corporate crime certainly explains the attention paid to corporate crime, it does not

explain the increased use of one particular tool: prosecution agreements. See infra notes

118–19 and accompanying text (arguing that powers prosecution agreements offer to pros-

ecutors incentivize their increased use).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 22 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1455

desire to enter into a prosecution agreement. There is, however, a

bargaining imbalance in these negotiations: Corporations are

uniquely vulnerable to criminal sanction, which shifts power and lev-

erage to the prosecutor. Part III.B addresses the implications of this

bargaining imbalance, including the threat of abuse that results from

one-sided negotiations. I focus in particular on two signs of abuse—

lack of predictability in punishment and prosecutorial overreaching—

as examples of problems that result when prosecutorial discretion is

unchecked.

A. Inequality of Bargaining Position in the Negotiation of

Prosecution Agreements

Prosecution agreements, like other settlements, offer benefits for

all parties. Companies avoid criminal liability; prosecutors reform and

punish while conserving resources. Beneath this apparent mutuality

of benefits, however, lies a distinct power imbalance. Though both

the corporation and the prosecutor have reason to prefer settlement

to prosecution, their interests are not equivalent.

102

A company does

not just want to avoid prosecution—it may need to avoid prosecution

in order to save itself.

Corporations are uniquely vulnerable to criminal prosecution for

a variety of reasons. First, as explained previously, the law of corpo-

rate liability is remarkably broad in scope—once the government has

established criminal conduct by an employee, the corporation is

nearly strictly liable.

103

Corporations thus have great difficulty win-

ning at trial, no matter what precautions they have taken.

Even in the small percentage of cases where a company might

have a promising defense, publicly traded corporations are still tre-

mendously vulnerable to criminal charges.

104

This is primarily due to

corporate reliance on public perception—the necessity of preserving

102

Prosecutors may also want to avoid the collateral consequences of corporate prose-

cution—but not to the same extent as the shareholders and employees of a targeted com-

pany. Prosecutors must accept that their job may sometimes require the use of the

ultimate penalty against a corporate offender, despite the collateral consequences. See

Thompson Memo, supra note 20, § IX(B), at 12 (“[T]he mere existence of [collateral con-

sequences] is not sufficient to preclude prosecution of the corporation.”).

103

See White, supra note 23, at 817 (“[T]he sweep of corporate criminal liability could

hardly be broader. . . . It is essentially absolute liability.”); supra Part II.A.

104

See Andrew Weissmann with David Newman, Rethinking Criminal Corporate Lia-

bility, 82 I

. L.J. 411, 426 (2007) (describing “devastating consequences” of criminal

indictment of corporation); Richard A. Epstein, Op-Ed, The Deferred Prosecution Racket,

ALL

. J., Nov. 28, 2006, at A14 (stating that even companies that believe they have

reasonable defenses at trial are “helpless to protect [themselves],” since indictment can be

very damaging).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 23 16-OCT-07 14:37

1456 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

“the reputation of a corporation in the eyes of third parties.”

105

As a

result, mere indictment can put a company in a tremendously vulner-

able position: Stock prices plummet, lines of credit dry up, and clients

are scared off.

106

Without the promise of a prosecution agreement, a

company may not be able to “avoid[ ] the paralysis that can grip an

indicted corporation unsure of the impending resolution of its legal

problems.”

107

These harms—which can result from a mere charge,

regardless of the result at trial—can destroy a business entity, as

occurred with Arthur Andersen.

108

The effective result of this vulnerability to mere indictment, not

just conviction, is that corporations cannot typically risk going to

trial.

109

This is a tremendous weakness. Prosecutors enjoy wide dis-

cretion regarding whom they charge and what crimes they allege—

decisions which are generally unreviewable.

110

The trial by jury is the

105

Greenblum, supra note 9, at 1887 & n.167. This concern is more pressing for corpo-

rations than for individuals. Id. at 1887, 1895 (“The adverse publicity and collateral conse-

quences of a conviction are tantamount to a death penalty for corporations, but not for

individuals.”).

106

See id. at 1886 (“The adverse publicity that accompanies a prosecution can devastate

a corporation, particularly one that relies heavily on its reputation in the marketplace,

because of the effect on relationships with customers, creditors, and the public at large.”

(citation omitted)); Interview with David Pitofsky, supra note 14, at 11 (“[U]pon the

announcement of a criminal investigation, companies regularly lose half of their market

value. If the price remains depressed long enough, the capital markets dry up, the ability

to hire quality people dries up. The company’s oxygen supply is cut off.”).

107

Greenblum, supra note 9, at 1887.

108

The Supreme Court overturned the Arthur Andersen conviction on May 31, 2005.

Arthur Andersen LLP v. United States, 544 U.S. 696, 698 (2005) (“We hold that the jury

instructions failed to convey properly the elements of a ‘corrup[t] persua[sion]’ conviction

under § 1512(b), and therefore reverse.” (alteration in original) (quoting 18 U.S.C.

§ 1512(b) (2000 & Supp. IV 2004))). The decision proved to be too little, too late, as the

indictment itself had effectively driven Arthur Andersen out of business. Finder &

McConnell, supra note 19, at 14–15 & n.66.

It is commonly assumed that the fate of Arthur Andersen—collapse following indict-

ment—would be shared by many other companies were they to face similar indictment.

See Greenblum, supra note 9, at 1888–89 (“[Arthur Andersen] suggests the potentially

catastrophic result that may await at trial.”); Weissmann, supra note 104, at 426 (noting

that indictment “risks the market imposing what is in effect a corporate death penalty”);

Wray & Hur, supra note 23, at 1097 (noting that “indictment often amounts to a virtual

death sentence for business entities”). This is believed to be true even though Andersen

was a partnership and not a publicly traded corporation. Id. at 1097 (noting that Andersen

case “sent a clear and unmistakable message to Corporate America in general”). As the

number of non-deferred prosecutions of publicly traded companies following the collapse

of Andersen has been minimal, there is very little empirical support for this claim. Regard-

less, an evaluation of this widespread assumption is beyond the scope of this Note.

109

Weissmann, supra note 104, at 426 (“[C]orporations as a practical matter can rarely

afford to take criminal cases to trial . . . .”).

110

See, e.g., Wayte v. United States, 470 U.S. 598, 607 (1985) (“In our criminal justice

system, the Government retains ‘broad discretion’ as to whom to prosecute.”);

Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978) (“[S]o long as the prosecutor has prob-

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 24 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1457

principal way we test prosecutorial discretion in our criminal justice

system. It is the primary means for checking overly aggressive

prosecutorial behavior.

111

Without the threat of trial, however, there

is no assurance that the prosecutor is acting in a judicious manner.

B. The Specter of Abuse

The legal and reputational vulnerabilities detailed above make

corporations uniquely weak negotiators in the criminal context. Since

corporations cannot run the risk of going to trial, their choice to

accept a deferred prosecution agreement is not really a choice at all.

112

Given this kind of leverage, prosecutors can negotiate overwhelmingly

favorable prosecution agreements. The benefits that flow to a prose-

cutor as a result of this leverage are evident from the terms of the

agreements: enormous financial penalties, significant internal busi-

able cause to believe that the accused committed an offense defined by statute, the deci-

sion whether or not to prosecute, and what charge to file or bring before a grand jury,

generally rests entirely in his discretion.”); United States v. Dotterweich, 320 U.S. 277, 285

(1943) (“[T]he good sense of prosecutors . . . must be trusted.”); Stolt-Nielsen, S.A. v.

United States, 442 F.3d 177, 183 (3d Cir. 2006) (noting that executive branch, through its

prosecutors, “has exclusive authority and absolute discretion to decide whether to prose-

cute a case” (citing United States v. Nixon, 418 U.S. 683, 693 (1974))). See generally James

Vorenberg, Decent Restraint of Prosecutorial Power, 94 H

ARV

. L. R

. 1521 (1981) (exam-

ining pervasive power of prosecutors in U.S. criminal justice system). The Supreme Court

has identified two separate bases for why courts give prosecutors broad deference in mat-

ters of prosecution. First, because prosecution is an executive act aimed at fulfilling the

President’s constitutional obligation to “take Care that the Laws be faithfully executed,”

U.S. C

ONST

. art. II, § 3, cl. 4, judicial deference helps fulfill separation of powers principles.

United States v. Armstrong, 517 U.S. 456, 464 (1996) (“[Prosecutors help the President]

discharge his constitutional responsibility to take Care that the Laws be faithfully exe-

cuted. As a result, [t]he presumption of regularity supports their prosecutorial decisions

and, in the absence of clear evidence to the contrary, courts presume that they have prop-

erly discharged their official duties.” (alteration in original) (internal quotation marks and

citations omitted)). Second, courts’ lack of competence in matters of prosecutorial discre-

tion militate against intervention. Id. at 465 (“Judicial deference to the decisions of [prose-

cutors] rests in part on an assessment of the relative competence of prosecutors and

courts.”).

111

The Supreme Court has noted the role that a jury trial plays in checking aggressive

prosecutorial behavior: “The purpose of a jury is to guard against the exercise of arbitrary

power—to make available the commonsense judgment of the community as a hedge

against the overzealous or mistaken prosecutor and in preference to the professional or

perhaps overconditioned or biased response of a judge.” Taylor v. Louisiana, 419 U.S. 522,

530 (1975) (citing Duncan v. Louisiana, 391 U.S. 145, 155–56 (1968)). Professor Lisa Kern

Griffin notes that juries have provided a check against prosecutorial overreaching in a

number of recent, high-profile white collar cases. Griffin, supra note 23, at 380 & n.357.

Each of her examples, however, concerns a charge against an individual, not a corporate

entity. As detailed in this Note, corporations cannot typically risk indictment or trial, and

thus cannot test the merits of the prosecution’s case by bringing their cases before a jury.

112

See Greenblum, supra note 9, at 1885, 1895 (“The corporate offender’s unique vul-

nerability to adverse publicity and collateral consequences . . . calls into question whether

the choice to enter into deferral is really a choice at all.”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 25 16-OCT-07 14:37

1458 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

ness reforms, and cooperation in pursuing the corporation’s individu-

ally culpable employees and directors.

113

This last point—assistance

in the pursuit of individuals—is of particular importance to resource-

deficient prosecutors. White collar prosecutions are tremendously dif-

ficult to conduct. The trials often depend on complex financial

records and complicated regulatory schemes

114

—material which may

test the comprehension of a typical jury.

115

The offenses typically take

place in large organizations where decisionmaking is shared among

boards of directors, executives, mid-level managers, and ground-level

employees, making it difficult to identify who should be held respon-

sible for the alleged criminal activity.

116

Furthermore, government

prosecutors often have to tangle with well-financed defendants

capable of hiring sophisticated law firms that can match government

resources.

117

For all of these reasons, corporate cooperation through

prosecution agreements—admitting liability, identifying perpetrators

of the criminal conduct, and occasionally waiving privilege to internal

documents and investigations—can be invaluable in prosecuting a

case against the individuals.

In light of these benefits, one might ask why a prosecutor would

not demand a prosecution agreement. After all, a prosecution agree-

ment removes the threat of corporate collapse, saves resources, and

offers a relatively one-sided negotiation where the prosecutor can ask

for almost anything, including cooperation in prosecuting individual

offenders.

It is precisely this accumulation of unchecked power that is dis-

concerting. Prosecutors have tremendous incentives to use prosecu-

113

See supra Part I.B.

114

Andrew Weissmann & Joshua A. Block, White-Collar Defendants and White-Collar

Crimes, 116 Y

ALE

L.J. P

OCKET

ART

286, 290 n.18 (2007), http://thepocketpart.org/2007/

02/21/weissmann_block.html (“White-collar prosecutions are notoriously difficult to

pursue successfully because they depend on complex financial records and often arcane

regulatory schemes.”).

115

Professor Stuart Green notes that understanding the processes behind white collar

crime often requires sophisticated knowledge of a variety of complex subjects, including,

inter alia, finance, management, and accounting. Stuart P. Green, Moral Ambiguity in

White Collar Criminal Law, 18 N

OTRE

AME

J.L. E

THICS

& P

. P

’

501, 508 (2004).

116

Id. at 510.

117

See Elizabeth Szockyj, Imprisoning White-Collar Criminals?, 23 S. I

. U. L.J. 485,

487–88 (1999) (“Criminologists have long lamented the resources white-collar criminals

have at their disposal: adroit legal counsel, political connections, social legitimacy, alterna-

tive enforcement paths (civil and regulatory routes), and finances that allow them to out-

spend prosecutors.” (citations omitted)); Weissman & Block, supra note 114, at 290 n.18

(“[W]hite collar defendants are often represented by skilled and well-financed

attorneys.”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 26 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1459

tion agreements: The benefits are often too great to refuse.

118

They

offer a means for prosecutors to exploit the vulnerabilities of public

corporations—in particular their dependence on reputation and

public perception for healthy relationships with investors, creditors,

and clients

119

—as a means to attack corporate crime. Corporations—

wary of Arthur Andersen’s fate, and cognizant of New York Central’s

unfavorable liability rule—have little choice but to agree to the prose-

cutor’s demands.

One might argue that there is nothing wrong with this situation:

The reputational concerns of corporations are being leveraged against

them in the legitimate pursuit of ferreting out crime. To a certain

extent, it is satisfying that the same motives that led to the recent

wave of accounting fraud—the desire to craft a strong public image of

financial health—are being leveraged against corporations in an effort

to clean up criminal behavior. The concern, however, is not with the

efforts to stop corporate crime, but rather with the bargaining and

settlement dynamic itself. When the corporation is unwilling to risk

trial and incapable of defending itself in any case, there is no assur-

ance that the prosecutor is using her discretion wisely. It is in this

exploitation of the corporation’s unique vulnerabilities that one

begins to see the first signs of abuse. The two most obvious indica-

tions of the abuse are the lack of predictability and prosecutorial over-

reaching in various agreements.

Lack of Predictability: The first sign of abuse is that there is little

predictability in the terms of prosecution agreements. In the corpo-

rate criminal context, like conduct often does not receive like punish-

ment. Conduct that would merit a declination from one prosecutor

might lead to a prosecution agreement with another prosecutor. To

take a recent example, a billion dollar misreporting scheme received a

prosecutorial declination from one U.S. Attorney’s Office, while a

five-figure bribery attempt produced a tremendously punitive

deferred prosecution agreement from a different U.S. Attorney’s

Office.

120

The difference between deferred prosecution agreements

118

See Richard S. Gruner, Three Painful Lessons: Corporate Experience with Deferred

Prosecution Agreements, in C

ORPORATE

OMPLIANCE

NSTITUTE

2006, at 61, 65–67 (PLI

Corp. Law & Practice, Course Handbook Series No. 891761, 2006) (noting that prosecu-

tion agreements make “the threat of corporate criminal liability more palatable than if a

full fledged prosecution were the federal authorities’ only choice”).

119

Interview with David Pitofsky, supra note 14, at 11 (“[U]pon the announcement of a

criminal investigation, companies regularly lose half of their market value. If the price

remains depressed long enough, the capital markets dry up, the ability to hire quality

people dries up. The company’s oxygen supply is cut off.”).

120

Comparing the facts surrounding criminal investigations is something like comparing

apples and oranges. However, there is still observable inconsistency. F. Joseph Warin and

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 27 16-OCT-07 14:37

1460 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

and non-prosecution agreements has largely disappeared: The terms

of non-prosecution agreements, which are thought to convey a lower

degree of culpability than deferred prosecution agreements, actually

can be as harsh as their deferred counterparts.

121

Much of this, of

course, is the result of the language in the Holder, Thompson, and

McNulty Memos: They give broad discretion to prosecutors to deter-

mine whether or not to file criminal charges against a corporation,

resulting in almost total leeway in crafting such agreements.

122

A cor-

poration under investigation might compare its behavior with that of

past corporations, in a hope that analogous criminality will receive

analogous treatment. The prosecutor, however, is not required to

follow any kind of precedent and has virtually unlimited discretion in

deciding what kind of action to take.

This lack of predictability has two implications. First, unpredict-

ability prevents corporate planning. If a corporation has no clear

sense of what employee crimes will be imputed to the entity or what

punishment those crimes might produce, the corporation will have dif-

ficulty properly allocating its resources. Corporations will be unable

to predict, ex ante, what compliance measures prosecutors will con-

sider adequate ex post, which in turn “breed[s] over deterrence and

risk aversion, both of which increase the company’s (and society’s)

costs.”

123

Deferred and non-prosecution agreements further exacer-

bate this tendency toward overdeterrence: Onlooker corporations

Peter E. Jaffe offer the following comparison of two 2005 prosecutorial actions, the Mon-

santo Deferred Prosecution Agreement and the declination to prosecute Shell Oil:

So in Shell and Monsanto we have two blue-chip, highly regarded public com-

panies: each discovered a violation of federal law; each immediately initiated

an internal investigation and promptly self-reported to federal authorities who,

in each instance, had theretofore been unaware of the conduct; each cooper-

ated fully with the investigations of both the DOJ and the SEC; and each sub-

stantially remedied its respective compliance program. Yet one corporation

walked away with the disconcerting prospect of conducting 36 months of busi-

ness under the shadow of a deferred criminal information and a corporate

monitor, while the other was let off with a good talking to. Oh, and did we

mention that Shell, the one admonished to “go forth and sin no more,”

admitted to a misreporting scheme that allegedly cost investors billions of dol-

lars, while Monsanto, the one with the hammer-shaped cloud hanging over its

head, admitted to a failed five-figure bribery attempt that, in the end, cost no

one but itself?

F. Joseph Warin & Peter E. Jaffe, The Deferred-Prosecution Jigsaw Puzzle: A Modest Pro-

posal for Reform, A

NDREWS

ITIG

. R

.: W

HITE

OLLAR

RIME

, Sept. 2005, at 1, 1 (foot-

note omitted).

121

See supra note 18 and accompanying text.

122

See supra notes 82–85 and accompanying text.

123

Miriam H. Baer, Insuring Corporate Crime, 83 I

. L.J. (forthcoming 2008) (manu-

script at 22), available at http://lsr.nellco.org/cgi/viewcontent.cgi?article=1050&context=

nyu/plltwp.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 28 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1461

notice the drastic penalties which accompany these agreements, giving

the agreements a deterrent power that belies their relatively small

number.

124

Corporations may respond to these vague deterrents by

erring on the side of caution, which could result in the overdeterrence

of aggressive but legitimate business practices.

125

Second, unpredictability violates basic notions of fairness. When

similar offenses give rise to widely disparate punishments, a defendant

has good reason to feel that the system is unjust. Our legal system has

faced this issue before: In 1984, Congress reformed the Federal Sen-

tencing Guidelines to address the disparities in punishment that

resulted from the wide sentencing discretion enjoyed by federal

judges.

126

With the development of prosecution agreements, however,

the power to punish corporations has shifted from judges to prosecu-

tors. These prosecutors, like federal judges prior to 1984, have no

guidelines to cabin their discretion.

Prosecutorial Overreaching: The second indication of abuse

comes from the quixotic nature of some terms in the agreements.

Some of the more bizarre conditions have already been addressed—

the requirement to create sixteen hundred jobs in the WorldCom

Agreement,

127

or the provision to endow a chair in legal ethics in the

Bristol-Meyers Squibb Agreement.

128

These examples are, admit-

tedly, still quite unusual. The infrequency of these unusual terms is

likely due to the restraint of the prosecutors that author the agree-

ments: The majority of these agreements have been negotiated by the

DOJ and U.S. Attorneys’ Offices, organizations generally known for

their integrity and restraint. Of the two aforementioned bizarre con-

ditions, only one of the two—the endowed chair of ethics—was

authored by a federal prosecutor, and that term was allegedly pro-

posed by the defense counsel, not the government.

129

Integrity, however, can only go so far: The massive, unfettered

power that prosecution agreements offer to prosecutors ushers in the

possibility of these kinds of bizarre conditions. While such conditions

124

Id. at 1.

125

See id. (arguing that corporations will overpay to avoid collateral consequences of

criminal indictment).

126

See Adam Lamparello, Implementing the “Heartland Departure” in a Post-Booker

World, 32 A

. J. C

RIM

. L. 133, 152–54 (2005) (noting that concerns with sentencing dispar-

ities motivated creation of guidelines).

127

See supra note 44 and accompanying text.

128

See supra note 45 and accompanying text.

129

The WorldCom Agreement, which featured the job-creation provision, was negoti-

ated by the Attorney General of Oklahoma, not the DOJ. Finder & McConnell, supra

note 19, at 36. The Bristol-Meyers Squibb Agreement was negotiated by the U.S.

Attorney’s Office in New Jersey. See supra note 45.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 29 16-OCT-07 14:37

1462 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

are fairly innocuous at present, they could become more offensive in

the future. Further, it is disturbing to see these conditions coming

from government prosecutors, individuals who are trained in prosecu-

tion, not corporate regulation.

130

Mary Jo White, the former U.S.

Attorney for the Southern District of New York, has claimed that

prosecutors are transforming into “super regulators”—capable of

being proactive in any direction they please.

131

If a prosecutor can

require the creation of jobs in the local community, or the endowment

of a chair at a favored law school, what limits exist—especially when a

corporation under threat of indictment will do almost anything to sat-

isfy the government?

132

As one commentator has said, prosecutors

are “starting to possess something close to absolute power.”

133

solution that limits this discretion is sorely needed.

HAT

ONE

This Part reviews potential solutions to the threat of prosecutorial

abuse. Part IV.A lays out two solutions suggested in the scholarship

that address the imbalance in bargaining leverage between corpora-

tions and prosecutors and examines their weaknesses. Part IV.B pro-

poses an alternate solution. Rather than seeking to alter the

underlying law or the enforceability of the agreements, this solution

restrains prosecutorial discretion by centralizing the review of prose-

cution agreements in the office of the Deputy Attorney General.

Such a framework, similar to that employed by the McNulty Memo

with regard to waivers of attorney-client privilege, would impose uni-

formity and consistency on the choices of individual prosecutors,

addressing the problem of abuse noted in Part III. Centralized over-

sight would address the concerns identified in Part III by ensuring

consistency and restraining overreaching by individual prosecutors.

A. Proposed Fixes

Armed with an easy-to-satisfy legal standard and near limitless

discretion, the prosecutor who desires to impose corporate criminal

130

Professor Coffee argues that acts of regulation are beyond the “competence or enti-

tlement” of prosecutors. See supra note 38.

131

White, supra note 23, at 818.

132

A company’s objective, as described in the KPMG case, was “to be able to say at the

right time with the right audience, we’re in full compliance with the Thompson Guide-

lines.” United States v. Stein, 435 F. Supp. 2d 330, 348 (S.D.N.Y. 2006) (internal quotation

marks omitted).

133

Coffee, supra note 38, at 13 (“[T]he deeper problem lies in the danger that power

corrupts and that prosecutors are starting to possess something close to absolute power.”).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 30 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1463

liability is not significantly restrained by the law. Prosecutors do not

need to push the envelope in order to threaten liability against corpo-

rate entities: The stigma of criminal liability is so powerful, yet

remarkably easy to apply.

Commentators have proposed two solutions to address the bar-

gaining imbalance between corporations and prosecutors. The first

proposed solution—perhaps both the cleanest and the least likely to

occur—is simply to change the law of vicarious corporate liability.

The foundation of prosecution agreements is the near absolute lia-

bility standard of New York Central coupled with the corporation’s

inability to go to trial. Scholars and practitioners have long criticized

the scope of corporate criminal liability.

134

Some have argued that

New York Central was a wrong turn from which the law never recov-

ered.

135

Andrew Weissman, ex-director of the Enron Task Force, has

argued that before charging a corporation with a crime, the govern-

ment should have to bear the burden of establishing that the corpora-

tion failed to have reasonably effective policies and procedures to

prevent the conduct, a standard which would mirror Supreme Court

developments in corporate civil liability.

136

Weissman argues that

adding this new element to the black letter law of liability would

effectively separate the blameworthy corporations from the non-

blameworthy corporations by preventing prosecutors from threat-

ening corporate charges against companies with satisfactory

compliance programs.

137

There is merit to this suggestion. At the base of all prosecution

agreements is the remarkable power granted by the New York Central

standard—a long term solution, therefore, may need to address the

standard itself. Such a proposal, however, would represent a radical

shift in nearly one hundred years of jurisprudence. This is not to say

that such a change would be impossible—the McNulty Memo, after

all, was the result of a massive outcry over what was perceived as

prosecutorial abuse stemming from the Thompson Memo.

138

If the

use of prosecution agreements continues at the current breathtaking

pace, the courts and Congress might feel the need to step in. Congress

has already shown itself willing to intervene in issues related to

134

See, e.g., White, supra note 23, at 817 (stating “the sweep of corporate criminal lia-

bility could hardly be broader”).

135

See Buell, supra note 67, at 474, 478 (noting that enterprise liability had “a weak

start nearly a century ago when common law courts . . . imported respondeat superior

liability from tort law into the criminal law” and that this decision was “a wrong turn for

the criminal law”).

136

Weissman, supra note 104, at 412–13.

137

Id. at 414–15.

138

See supra notes 4–6 and accompanying text.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 31 16-OCT-07 14:37

1464 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

prosecutorial discretion in corporate charging decisions.

139

Perhaps

that desire can evolve into a willingness to reassess the basic black

letter law.

A second proposed solution is to require judicial review of prose-

cution agreements. Most corporations agree to terms whereby the

prosecutor retains the right to determine whether breach has

occurred.

140

If the corporation breaches, the agreement is revoked

and the corporation will be subject to prosecution. However, corpora-

tions are probably unwilling to demand such judicial review them-

selves. The corporation has signed the agreement because it wants to

avoid prosecution and trial, and the agreement, even if harsh, satisfies

this core concern. With the power to determine whether the agree-

ment has been breached, the prosecutor can effectively play judge and

jury over the corporation’s criminal conduct.

141

While there are some

instances where prosecution agreements permit judicial review of

determinations of breach, these agreements are the exception rather

than the rule.

142

In light of this prosecutorial power, some commentators have

pondered methods for imposing judicial review over the issue of

breach, regardless of the terms of the contract.

143

Judicial review of

breach would no doubt offer some measure of security to the corpora-

tions that sign prosecution agreements. The threat of unilateral revo-

cation has exacerbated the leverage exercised by prosecutors—

companies such as KPMG have been pressured to cooperate

throughout the deferral period, knowing that the government would

be the sole arbiter of whether the terms of the agreement had been

met.

144

139

Browning, supra note 5 (“[T]he departing chairman of the Senate Judiciary

Committee proposed legislation yesterday calling for a rollback of the tactics adopted by

federal prosecutors to combat corporate wrongdoing after the Enron collapse.”).

140

See, e.g., OMI DPA, supra note 25, para. 15. Since prosecution agreements are pri-

vate contractual agreements, it is not clear that judicial review could start any earlier than

the occurrence of breach. See Greenblum, supra note 9, at 1898–99 (noting difficulties of

judicial involvement in negotiation phase).

141

See Greenblum, supra note 9, at 1864 (“Deferral is a powerful prosecutorial tool

because it is negotiated and implemented exclusively by the prosecutor.”).

142

See Warin & Jaffe, supra note 120, at 4 & n.25 (“[I]n those cases where the District

Court does not agree with the DOJ’s breach assessment, a policy of pre-indictment judicial

intervention would save everyone’s time and resources.”).

143

Benjamin Greenblum, for example, argues that judges should be permitted to inter-

vene during the implementation of the deferral terms, which would bring judicial involve-

ment into the definition of breach. This intervention, argues Greenblum, would place an

outer boundary on the “sole discretion” exercised by prosecutors and thus ensure that the

deferral mechanism is not abused. Greenblum, supra note 9, at 1899–1903.

144

See United States v. Stein, 435 F. Supp. 2d 330, 350 (S.D.N.Y. 2006) (noting that if

KPMG did not comply with demands for cooperation, “it will be open to the risk that the

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 32 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1465

It is worth noting, however, that neither the DOJ nor any U.S.

Attorney’s Office has ever revoked a prosecution agreement. While

judicial review over breach of the agreement would be a positive step

for corporations, it would not change the underlying issue of bar-

gaining leverage or prosecutorial discretion. Prosecutors would still

be free to demand almost anything they desire. Moreover, if breach

were to be determined by a judge rather than a prosecutor, one might

expect prosecutors to ask for more detailed agreements. Certain

terms, such as “cooperation,” are currently kept vague—in no small

part because the prosecutor knows he or she reserves the right to

interpret the meaning of “cooperation” and thus determine breach. If

that were not the case, prosecutors would likely respond by crafting

agreements which would more explicitly detail their demands.

The two solutions presented offer limited assistance to a corpora-

tion facing an aggressive prosecutor. The reformation of black letter

law would indeed redistribute the negotiating leverage, but this shift

in jurisprudence is unlikely to occur. Imposing judicial review of

breach has the opposite problem—it may indeed become a feasible

option, but it still does not substantially alter the relationship between

prosecutor and defendant. A more effective solution will have to mit-

igate the bargaining imbalance by limiting the scope of prosecutorial

discretion.

B. Expanding the McNulty Solution

The concerns over the Thompson Memo’s treatment of coopera-

tion, particularly with regard to attorneys’ fees and privilege waiver,

found a remedy in the political sphere. The DOJ responded to pres-

sure—some political, some legal—and issued the McNulty Memo.

145

The legal community decided that these bargaining imbalances were

unacceptable, and addressed it as a matter of policy.

government will declare that KPMG breached [the deferred prosecution agreement] and

prosecute the criminal information to verdict”). In the KPMG prosecution agreement, the

government retained the “sole discretion” to determine whether the company had violated

any of the provisions, including the requirement of cooperation. KPMG DPA, supra note

18, para. 12.

145

This sort of solution was appropriate. The determination of prosecutorial priorities

and the allocation of prosecutorial resources are best determined by individuals account-

able to the public because issues of prosecution are typically questions of politics. Ross E.

Wiener, Inter-branch Appointments After the Independent Counsel: Court Appointment of

United States Attorneys, 86 M

INN

. L. R

. 363, 427 (2001) (noting that prosecutor’s

appointment is “intrinsically a political decision” and that “[s]electing a U.S. Attorney

requires making political judgments about what a district’s prosecutorial priorities should

be and who is best suited to realize those goals”). Courts lack the tools to guide prosecu-

tors’ discretion when they are operating within the boundaries of the law. Prosecutorial

discretion is virtually unfettered. See supra note 110 and accompanying text.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 33 16-OCT-07 14:37

1466 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

The McNulty Memo solution may offer a blueprint for addressing

concerns relating to the accumulation of prosecutorial power high-

lighted in this Note. The McNulty solution requires consultation with

senior attorneys in the DOJ before waiver requests can be made. If a

prosecutor wants access to communications between attorney and

client, the local United States Attorney must get approval from the

Deputy Attorney General.

146

Without explicit permission there can

be no request for waiver.

Some critics of the McNulty Memo have argued that these modi-

fications represent only a “modest improvement.”

147

On the contrary,

I believe that these changes will prove to be significant. On several

prior occasions, the DOJ has sought to constrain prosecutorial discre-

tion by requiring the approval of the Deputy Attorney General, and

this procedural hurdle has been largely effective. Procedures under

the Foreign Intelligence Surveillance Act (FISA)

148

are an example:

FISA warrants, like waiver requests under the McNulty Memo, can

only be issued after approval from either the Attorney General, the

Deputy Attorney General, or the Assistant Attorney General for

National Security.

149

While FISA court proceedings have been criti-

cized as rubber-stamp affairs,

150

various post-9/11 inquiries revealed

DOJ’s internal vetting policies as “much more demanding than out-

siders had imagined.”

151

Centralized review has had a similar effect in

regard to the Economic Espionage Act, a statute that criminalizes

theft of trade secrets and other forms of proprietary information.

152

The Act restrains prosecutorial discretion by mandating centralized

146

McNulty Memo, supra note 6, § VII, at 2; McNulty Speech, supra note 1.

147

See Martha Neil, Thompson Memo Changes Not Enough, ABA Says, A.B.A. J.

EPORT

, Dec. 15, 2006, http://www.abanet.org/journal/ereport/d15specter.html (quoting

ABA President Karen J. Mathis calling McNulty Memo “modest improvement” and

stating that changes do not go far enough).

148

Originally enacted as the Foreign Intelligence Surveillance Act of 1978, Pub. L. No.

95-511, 92 Stat. 1783 (codified as amended at 50 U.S.C. §§ 1801–1863).

149

Matthew R. Hall, Constitutional Regulation of National Security Investigation: Mini-

mizing the Use of Unrelated Evidence, 41 W

AKE

OREST

L. R

. 61, 84 (2006) (citing 50

U.S.C. § 1801(g) (2000) (amended 2006)); see also 50 U.S.C.A. § 1801(g) (West Supp.

2007) (adding Assistant Attorney General for National Security to list of those who may

approve FISA applications).

150

Laura K. Donohue, Anglo-American Privacy and Surveillance, 96 J. C

RIM

. L. &

RIMINOLOGY

1059, 1097 (2006) (“[A] legitimate question could be raised as to whether

the court merely serves as a rubber stamp function.”); Stephen J. Schulhofer, The New

World of Foreign Intelligence Surveillance, 17 S

TAN

. L. & P

’

. 531, 535 (2006)

(“[C]ivil liberties groups understandably viewed the FISA Court as little more than a

rubber stamp.”).

151

Schulhofer, supra note 150, at 535.

152

See generally James M. Fischer, Note, An Analysis of the Economic Espionage Act of

1996, 25 S

ETON

ALL

EGIS

. J. 239 (2001).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 34 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1467

DOJ approval, which, according to commentators, has resulted in a

lower incidence of prosecution.

153

These results should not be surprising, for two key reasons. First,

the Deputy Attorney General is a single individual with limited

resources and time. Naturally, any approval process that requires the

imprimatur of a single individual will be more restrictive than one that

allows decentralized discretion. Second, and perhaps more impor-

tantly, the Deputy Attorney General is charged with “providing

overall supervision and direction to all the organizational units of the

[DOJ].”

154

Funneling decisionmaking through a figure with such

responsibility will be of particular import where, as with prosecution

agreements, the unfettered discretion of individual prosecutors is part

of the problem.

Constraining prosecutorial discretion will begin to rectify the bar-

gaining imbalance that exists in crafting the terms of prosecution

agreements. While the principal reason prosecution agreements can

be abused is the expansive nature of vicarious liability law and corpo-

rate vulnerability to bad publicity—two factors that likely will not

change—the potential for abuse is exacerbated by prosecutors’ near

limitless discretion. An approval process akin to that required in the

McNulty Memo could address both of the concerns identified above.

Lack of Predictability: As discussed in Part III.B, there is little

standardization in the penalties negotiated in prosecution agree-

ments,

155

which can result in unfairness and overdeterrence of legiti-

mate employee behavior. The Deputy Attorney General’s approval

of penalties could help remedy these problems. That office could pro-

vide centralized review, use past agreements as rough precedent, and

ensure that penalties are proportionate to the misconduct.

153

See Peter E. Calamari, Protection of Confidential Business Information, in

ORPORATE

AIDING

2001, at 35, 76 (PLI Corp. Law & Practice, Course Handbook Series

No. B0-00VS, 2001) (“[P]rosecutions are likely to be few and far between in the short term.

The Attorney General has stated that, until October 2001, any prosecution under the

Espionage Act will require the personal approval of the Attorney General, the Deputy

Attorney General, or the Assistant Attorney General for the Criminal Division.”); D.

Peter Harvey, IP Maintenance: Protecting Intellectual Property Assets Through Registra-

tion, Proper Use and Contractual Provisions, in P

ROTECTING

OUR

NTELLECTUAL

ROPERTY

SSETS

2002, at 33, 77 (PLI Patents, Copyrights, Trademarks, & Literary Prop-

erty, Course Handbook Series No. G0-0120, 2002) (“Because the Justice Department has

required the personal approval of the Attorney General, the Deputy Attorney General or

the Assistant Attorney General for the Criminal Division before filing a charge under the

[Economic Espionage Act of 1996], only a relatively modest number of prosecutions under

the Act have taken place.”).

154

Office of the Deputy Attorney General, United States Department of Justice, http://

www.usdoj.gov/dag/index.html (last visited June 16, 2007).

155

See supra notes 18, 120.

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 35 16-OCT-07 14:37

1468 NEW YORK UNIVERSITY LAW REVIEW [Vol. 82:1434

Prosecutorial Overreaching: One also sees the potential for abuse

in the seemingly inappropriate terms that have appeared in some of

the more recent prosecution agreements.

156

These sorts of terms push

the boundaries of what should be included in criminal punishment,

but remain available due to individual prosecutors’ virtually unfet-

tered discretion. Deputy Attorney General approval could also help

remedy this problem. Centralized review of agreement terms could

ensure that bizarre and inappropriate conditions,

157

especially those

pandering to local interests, would be detected and removed.

To some extent, consultation of this sort with central DOJ prob-

ably does happen.

158

It is not, however, official policy, and the hap-

hazard nature of these agreements testifies to the lack of any

centralizing standard. Central DOJ can and should provide such a

standard. This approval process will no doubt cause some administra-

tive headaches, but the costs will be a small price to pay to ensure

consistency and fairness in corporate prosecutions.

Of course, even with such an approval system, corporations will

still be vulnerable due to their legal and social relationship to the

criminal justice system. Such reform would, however, at least allow

corporations to expect a more uniform approach to punishment. Fur-

thermore, if such a vetting system failed to direct prosecutorial discre-

tion in a principled manner, Congress or the courts would have a

clearer mandate to step in and contemplate changes to the black letter

law.

ONCLUSION

The black letter law of entity liability is remarkably broad,

though it has traditionally been limited by prosecutors’ discretion.

The use of prosecution agreements, however, has enabled prosecutors

to overcome many of the traditional concerns with corporate liability,

thus permitting a more aggressive application of criminal law against

corporations. While the prosecution of corporate criminality is a posi-

tive development, the unique weaknesses of corporations in the crim-

inal justice system has vested a dangerous amount of power in the

hands of individual prosecutors. The DOJ has recognized this and has

156

See supra notes 127–31 and accompanying text.

157

See supra notes 127–33 and accompanying text.

158

The former U.S. Attorney for the Southern District of New York, David Kelley,

would have prosecuted KPMG criminally if not for the intervention of Deputy Attorney

General James Comey, who insisted on a deferred prosecution agreement. Corporate

Crime Reporter, Crime Without Conviction: The Rise of Deferred and Non-Prosecution

Agreements, http://www.corporatecrimereporter.com/deferredreport.htm (last visited

August 16, 2007).

\\server05\productn\N\NYU\82-5\NYU505.txt unknown Seq: 36 16-OCT-07 14:37

November 2007] CORPORATE PROSECUTION AGREEMENTS 1469

attempted to constrain some aspects of prosecutorial discretion with

the McNulty Memo. This is a worthwhile development, but more

needs to be done. The consultation framework established in the

McNulty Memo should be expanded to include decisions relating to

prosecution agreements. Such a change will not erase the bargaining

inequities between corporations and prosecutors, but it will ensure

that the use of prosecutorial discretion is exercised in a more consis-

tent and fair manner. The current relationship between corporate

prosecutors and their target companies is ripe for abuse—the DOJ

needs to expand its supervision efforts to ensure that corporate crim-

inal sanctions are used in a principled and consistent manner.