class actions, disabled rights, copyright, attorneys general, online speech, law schools, obesity, New York, mortgages, legal blogs, safety, CPSC, pharmaceuticals, patent trolls, ADA filing mills, international human rights, humor, hate speech, illegal drugs, immigration law, cellphones, international law, real estate, bar associations, Environmental Protection Agency, First Amendment, insurance fraud, slip and fall, smoking bans, emergency medicine, regulation and its reform, dramshop statutes, hotels, web accessibility, United Nations, Alien Tort Claims Act, lobbyists, pools, school discipline, Voting Rights Act, legal services programs

Recently in Corporate Governance Category

Selling short: final comment

March 3, 2006 10:24 AM

Thanks again to the Manhattan Institute for setting up this debate, which I've found interesting.

I don�t want to add much to this already lengthy discussion, and Moin�s last post provides little occasion to do so. He is still flatly wrong that the securities laws prohibit trading on non-misappropriated information. Moin cites the open-ended language of the applicable statute and rule but, incredibly, continues to ignore the Supreme Court�s interpretation of the statute.

(By the way, had Moin paid attention to the Court�s rulings, he might have fashioned an argument that a class action lawyer who is trading on litigation information is misappropriating information from his �client,� the class, which would provide a basis for regulation under current law. I have not analyzed that argument because Moin explicitly says in his initial post that in his view �who does the short selling is immaterial.�)

Moin also continues to assert that the market does not accurately value the litigation on which the plaintiff or his lawyer is trading. (This is relevant to my argument that dumping and suing can provide efficient incentives to plaintiffs and their lawyers.) But the well-known fact that markets reflect noise does not mean that they do not also reflect information, including information about litigation. For an analysis of the relationship between behavioral finance and market efficiency see my article, Fraud on a Noisy Market,

In the end, as I�ve said repeatedly, this debate has not been about whether there is a litigation problem, but about how to solve it. Dumping and suing alone is not the problem. At best it may provide incentives that produce higher-quality lawsuits. At worst, it reflects underlying problems with class actions. To the extent that it is accompanied by fraud and manipulation, we already have the necessary tools. It would be perverse and not a little ironic to try to deal with the problem of excessive litigation by adding a whole new area of securities regulation.

The NYT's Joe Nocera writes about "the anguish of being an analyst." The problem is that it's hard for analysts to get paid in an efficient market, where the value of information is dissipated as soon as it's used. This means, among other things, that analysts are more likely to get paid for buy recommendations than for sells. So what to do?

As discussed here, we should avoid regulation that reduces the few incentives that exist to produce valuable securities market information. Don't restrict use of nonpublic information in an elusive quest for "fairness." Get rid of Regulation FD. And if we're concerned that there's too little incentive to produce negative information, then we shouldn't attack the shorts.

Many thanks to for organizing this debate between me and my friend, Steve Bainbridge. Over the course of this week, we will discuss one of the most important contemporary issues in corporate law: the relative control rights of directors and shareholders. In this post, I will set the stage for the debate, and offer some arguments in favor of increased shareholder participation in corporate governance.

In our prior scholarship, Steve and I have relied on the work of economist Kenneth Arrow in analyzing shareholder participation in corporate governance, and this common starting point will prove useful in highlighting areas of disagreement. Arrow argued that centralized decision making in organizations � which he called "authority" � was valuable because it coordinates the activities of members of the organization, thus economizing information costs. The board of directors is the central decision making body of the corporation.

While authority offers many benefits, it also has costs. Recognizing this, Arrow suggested that a necessary counterweight to authority was "responsibility." Directors may be held responsible via elections or shareholder litigation. In addition, various markets � including capital markets, takeover markets, charter markets, product markets, and labor markets � are said to monitor director performance. One key to organizational success is striking the proper balance between authority and responsibility. Arrow describes the tension between authority and responsibility that animates the present corporate governance debate:

"To serve its functions, responsibility must be capable of correcting errors but should not be such as to destroy the genuine values of authority. Clearly, a sufficiently strict and continuous organ of responsibility can easily amount to a denial of authority. If every decision of A is to be reviewed by B, then all we have really is a shift in the locus of authority from A to B and hence no solution to the original problem. To maintain the value of authority, it would appear that responsibility must be intermittent." KENNETH J. ARROW, THE LIMITS OF ORGANIZATION 77-78 (1974).

The task of the corporate governance system, therefore, is to draw the line between authority and responsibility. Though boards of directors generally are said to favor shareholders over other corporate constituencies, the formal powers of shareholders to demand such favored treatment are meager. Shareholders elect directors, but contested elections are extraordinary, and in most corporations, boards of directors are self-perpetuating. Shareholders also vote on other matters � including fundamental transactions, like mergers � but only after the board of directors has recommended such matters to the shareholders. Shareholder lawsuits are more common than contested director elections, but lawsuits tend to arise in a narrow band of cases involving securities frauds or corporate acquisitions. The bottom line is that the power to initiate corporate change lies almost exclusively with the directors or officers of the corporation.

In a series of articles published over the past four years, Steve has defended the status quo, in the process coining the term "director primacy." His neologism was a play on "shareholder primacy," the widely accepted notion that corporations should be operated for the benefit of shareholders rather than other corporate constituencies. D. Gordon Smith, The Shareholder Primacy Norm, 23 J. CORP. L. 277 (1998). According to Steve, however, the notion of "shareholder primacy" is not only about the ends of corporate law, but about the means. In his words, "shareholder primacy contends not only that shareholders are the principals on whose behalf corporate governance is organized, but also that shareholders do (and should) exercise ultimate control of the corporate enterprise." Stephen M. Bainbridge, Director Primacy: The Means and Ends of Corporate Governance, 97 NW. U. L. REV. 547, 563 (2003).

Steve embraces shareholder wealth maximization as the appropriate "end" of corporate law, but he rejects the notion that shareholders control directors in any meaningful way. Perhaps more importantly for present purposes, Steve rejects the notion that shareholders should control directors in any meaningful way. This is the point at which Steve and I diverge.

In my view, shareholders should be allowed to participate meaningfully in director elections and in certain fundamental business decisions. The motivation for my position is my view that the various market mechanisms listed above, though powerful within their domains, do not adequately monitor the competence of corporate directors, and in other instances, do not protect shareholders from the costs of managerial slack.

Last week, in AFSCME v. AIG, the United States Court of Appeals for the Second Circuit opened the door slightly for increased shareholder participation in director elections, though the Securities and Exchange Commission may be positioning itself to slam the door shut. AFSCME involved a shareholder challenge to the SEC�s interpretation of its own Rule 14a-8, also known as the "shareholder proposal rule." Under Rule 14a-8, shareholders are allowed to place certain proposals on the corporate ballot. The rule also contains a long list of "exclusions" � reasons why a corporation might exclude a shareholder proposal from the ballot. Among the reasons is that the proposal "relates to an election for membership on the company's board of directors." Because of this exclusion, shareholders cannot use the shareholder proposal rule to nominate candidates for the board of directors. But can shareholders force the inclusion of proposals that change the rules of the game?

AFSCME wants the shareholders of AIG to amend the company�s bylaws to provide that 3% shareholders may nominate one candidate for election to the board of directors and have that nominee placed on the corporation�s ballot. According to the Court, this sort of bylaw "would simply establish a process for shareholders to wage a future election contest," and thus, it is distinguishable from a proposal that "would result in an immediate election contest." While this language (which was suggested by the SEC�s brief in this case) may need some refinement, the gist of the opinion is sound, despite Steve�s judgment that the Court�s reasoning is "hogwash." According to Steve, "The language is clearly sweeping, encompassing any proposal related to the process by which directors are elected." Nevertheless, as noted by the Second Circuit, the SEC has long allowed shareholder proposals relating to cumulative voting rights or general qualifications for directors, and this practice undermines the argument that the exclusion is "sweeping." (By the way, even the SEC did not take that position, preferring to argue that the exclusion covered some election-related proposals but not others.)

If the SEC does not amend Rule 14a-8 to prohibit AFSCME-like bylaws, we can expect to see company-by-company reform of the election process. Global reform of the director nomination process died � at least for the foreseeable future � when the SEC proposed, then quietly dropped, a controversial rule that would have allowed director nominations by shareholders under limited circumstances.

Director elections are an important part of the debate over increased shareholder participation in corporate governance, but they do not exhaust the debate. Perhaps the most visible academic proponent of increasing shareholder power has been Harvard law professor Lucian Bebchuk, who has argued forcefully that shareholders should be allowed to adopt charter provisions that would give them power to pass binding resolutions regarding business issues, such as "decisions to merge, sell all assets, or dissolve the company, [or] decisions [to reduce] the company's size by ordering a cash or in-kind distribution." Lucian Arye Bebchuk, The Case for Increasing Shareholder Power, 118 HARV. L. REV. 833, 837 (2005). This proposal expands a proposal that Bob Thompson and I made in 2001 to allow shareholders to "initiate action that would put the company up for sale, leaving it to the board to conduct the auction." Robert B. Thompson & D. Gordon Smith, Toward a New Theory of the Shareholder Role: "Sacred Space" in Corporate Transactions, 80 TEXAS L. REV. 261, 308 (2001).

The brevity required by this forum does not allow for a complete explication of the case for increased shareholder participation in corporate governance, but I hope that this post will serve as a useful starting point for our debate. As noted above, the central issue of contemporary corporate governance debates is whether the current regime of "director primacy" should be modified. Arrow identified the tension between authority and responsibility, but this is only a starting point, not a conclusion. Acknowledging that tension does not tell us where to draw the lines.

My thanks to for organizing this debate and to my friend Gordon Smith for getting us off to such a great start. Gordon did an excellent job of laying out the basic divide between us; i.e., we both start with Kenneth Arrow's authority versus consensus model, but diverge in its application to the question of shareholder primacy.

I agree with Gordon's observation that:

"Arrow identified the tension between authority and responsibility, but this is only a starting point, not a conclusion. Acknowledging that tension does not tell us where to draw the lines."

A complete theory of the firm requires one to balance the virtues of discretionary fiat on the part of the board of directors against the need to ensure that the power of fiat is used responsibly. Neither fiat nor accountability can be ignored, because both promote values essential to the survival of business organizations. Unfortunately, however, because the power to hold to account differs only in degree and not in kind from the power to decide, fiat and accountability also are antithetical. As Kenneth Arrow explained:

"[Accountability mechanisms] must be capable of correcting errors but should not be such as to destroy the genuine values of authority. Clearly, a sufficiently strict and continuous organ of [accountability] can easily amount to a denial of authority. If every decision of A is to be reviewed by B, then all we have really is a shift in the locus of authority from A to B and hence no solution to the original problem."

So, on second thought, perhaps the tension between authority and accountability in fact does tell us something about where to draw the line.

One of the striking things about American corporate law (by which, of course, I mean Delaware corporate law) is the extent to which it consistently draws that line in favor of authority.

Indeed, judicial recognition of the proposition that directors cannot be held accountable without undermining their authority pervades many aspects of corporation law. The business judgment rule, for example, substantially insulates director decisions from being challenged by shareholders in litigation. According to the Delaware Supreme Court, the rule in fact exists precisely to prevent shareholders from holding boards of directors to account:

"... the business judgment rule is the offspring of the fundamental principle, codified in [Delaware General Corporation Law] � 141(a), that the business and affairs of a Delaware corporation are managed by or under its board of directors. ... The business judgment rule exists to protect and promote the full and free exercise of the managerial power granted to Delaware directors." Smith v. Van Gorkom, 488 A.2d 858, 872 (Del. 1985).

The New York Supreme Court likewise opined that:

"To encourage freedom of action on the part of directors, or to put it another way, to discourage interference with the exercise of their free and independent judgment, there has grown up what is known as the 'business judgment rule.'" Bayer v. Beran, 49 N.Y.S.2d 2, 6 (Sup. Ct. 1944).

(I advanced this understanding of the business judgment rule in Stephen M. Bainbridge, The Business Judgment Rule as Abstention Doctrine, 57 Vanderbilt L. Rev. 83 (2004). Former Delaware Supreme Court Chief Justice Veasey recently cited that article�s analysis, deeming it to be "approach is consistent with the Delaware doctrine that the rule is a presumption that courts will not interfere with, or secondguess, decision making by directors." E. Norman Veasey & Christine T. di Guglielmo, What Happened in Delaware Corporate Law and Governance From 1992�2004? A Retrospective on Some Key Developments, 153 U. Penn. L. Rev. 1399, 1422 (2005).)

In many cases, of course, the defendants never need invoke the business judgment rule because plaintiff's case founders on the procedural limitations on derivative litigation, which thus further insulate boards of directors from shareholder litigation. Again, however, as the New York Court of Appeals seemingly recognized, these limitations follow precisely because corporate law must strive to balance authority and accountability:

"By their very nature, shareholder derivative actions infringe upon the managerial discretion of corporate boards. . . . Consequently, we have historically been reluctant to permit shareholder derivative suits, noting that the power of courts to direct the management of a corporation�s affairs should be 'exercised with restraint.'" Marx v. Axers, 644 N.Y.S.2d 121, 124 (1996).

See also Pogostin v. Rice, 480 A.2d 619, 624 (noting that "the derivative action impinges on the managerial freedom of directors").

We observe similar restrictions designed to protect the board of director's authority with respect to statutory provisions such as those governing interested director transactions. Similar restrictions also may be observed with respect to management buyouts, which involve a significant conflict of interest and therefore tend to get close judicial scrutiny, but which will receive judicial deference in appropriate cases. Stephen M. Bainbridge, Independent Directors and the ALI Corporate Governance Project, 61 Geo. Wash. L. Rev. 1034, 1075-79 (1993). Likewise, corporation statutes grant the board sweeping authority with respect to negotiated acquisitions, which the business judgment rule then largely insulates from judicial intervention. Stephen M. Bainbridge, Mergers and Acquisitions 162-65 (2003). As a final example, the Delaware courts allow the target�s board of directors a substantial gatekeeping role in unsolicited tender offers, which again is attributable to the courts' recognition of the importance of preserving the board's authority. Id. at 352-53.

Given the pervasiveness in corporate law of rules that (a) strike a balance between authority and accountability and (b) are biased towards deference to the board�s authority rather than accountability, it would be surprising if the rules on shareholder voting rights were any more empowering than they are. Put simply, corporate law is designed from the ground up to make holding directors to account very difficult. Proponents of expanded shareholder power thus need to explain why the voting rules need to be changed to create an exception to this general rule.

In my introductory post in this debate, I observed that acknowledging the tension between �authority� and �responsibility� does not tell us where to draw the lines in the corporate governance system. Steve initially seemed to agree, but �on second thought,� he suggested that �the tension between authority and accountability in fact does tell us something about where to draw the line.� Reasoning from existing corporate law � which �consistently draws that line in favor of authority� � he concluded with a challenge:

Put simply, corporate law is designed from the ground up to make holding directors to account very difficult. Proponents of expanded shareholder power thus need to explain why the voting rules need to be changed to create an exception to this general rule.

In this post, I will challenge Steve�s challenge. That is, rather than accepting the premises of the challenge, I will offer two responses that undermine those premises.

First, the lesson that Steve draws from Arrow�s analysis of authority and responsibility might be summarized as follows: introducing any (additional) quantum of responsibility via shareholder oversight into the corporate governance system risks obliterating authority. This may be a bit of a caricature, but not much. In his most recent article on this subject, Steve wrote:

Boards of directors are subject to a pervasive network of accountability mechanisms that are more or less independent of shareholder oversight. The capital and product markets, the internal and external employment markets, and the market for corporate control all constrain shirking by firm agents. [T]hese incentive structures induce directors to behave generally in ways consistent with shareholder wealth maximization.

Stephen M. Bainbridge, Director Primacy and Shareholder Disempowerment, 119 Harv. L. Rev. 1735, 1746-47 (2006).

The problem with reasoning from the status quo, as Steve does in his last post and throughout his work on director primacy, is that the status quo contains two mechanisms of direct shareholder oversight: shareholder litigation and shareholder voting. To sustain his argument in favor of director primacy, Steve must contend that these mechanisms, like the human vermiform appendix, either have no function or have a function that is hidden from our understanding. (Perhaps shareholder litigation and shareholder voting are vestigial organs of an ancient corporate governance system?)

Why should we prefer this account of shareholder litigation and shareholder voting to an alternate account? We might view shareholder litigation and shareholder voting as potentially useful mechanisms that have been crippled by misguided legislatures, agencies, and courts. Or we might view each mechanism as highly specialized and appropriate for use only in extreme circumstances. For example, shareholder litigation may be useful in policing conflict-of-interest transactions, but may be relatively unhelpful in policing negligence, while voting may be important for effecting changes in control, but not for charting new corporate strategies.

My point is simply that shareholder oversight may be valuable, even if it operates only at the margins of the corporate governance system. And if we perceive instances when shareholder oversight improves corporate governance, we should remain open to the possibility of value-enhancing reforms. Despite the phrasing of Steve�s challenge quoted above, the theory of director primacy is more than just a call for rigorous justification of reforms. Steve uses director primacy as a basis for arguing against reform, which leads to my second response to Steve�s challenge: the theory of director primacy does not support the notion that increased shareholder participation in corporate governance is presumptively bad.

In my initial post, I focused on director elections, and Steve�s challenge mentions �voting rules.� But every example of director primacy in his post was drawn from the context of shareholder litigation. Of course, arguments in favor of increased shareholder participation in corporate governance sometimes extend to shareholder litigation � this has been particularly noticeable in connection with the recent Disney litigation in the Delaware courts � but my focus and the recent proposals made by Lucian Bebchuk relate to shareholder voting.

Obviously, Steve recognized this, and I don�t intend to imply that he missed the point. Instead, I think his decision to focus on litigation rules reveals something important about his approach to the corporate governance system. Steve could have approached the issue of shareholder voting directly by pointing to numerous legal rules that limit the efficacy of shareholder voting as a monitoring device and suggesting that such rules bolster his case for director primacy. Instead, he chose to observe that directors are rarely held accountable in shareholder litigation, then to assert, �[g]iven the pervasiveness in corporate law of rules that (a) strike a balance between authority and accountability and (b) are biased towards deference to the board�s authority rather than accountability, it would be surprising if the rules on shareholder voting rights were any more empowering than� the rules relating to litigation.


Steve wants us to believe that �deference to the board�s authority� is the defining characteristic of corporate law. He perceives such deference in the litigation rules, and suggests that similar deference should be reflected in voting rules. I see two problems with this line of reasoning: (1) the notion of deference is too diffuse to be a useful analytical tool, and (2) one cannot justifiably infer a policy of "deference to authority" merely from the fact of shareholder disablement. I elaborate on each point in turn.

What constitutes �deference�? Delaware courts are quite reluctant to impose personal liability on directors in any circumstance, but they are much more willing to grant injunctive or other equitable relief, especially in cases involving managerial conflicts of interest. Thus, the deference to directors displayed by the Delaware courts varies depending on the remedy sought and the duty that is alleged to have been breached. Are the courts being "deferential" in all of these circumstances? Or are they being intrusive to different degrees?

Another fundamental objection relates to Steve�s attempt to compare �deference� in litigation to �deference� in elections. Limits on shareholder litigation have different origins and rationales from limits on shareholder voting. As a result, assuming one could compare the respective levels of shareholder power in litigation and elections, one might reasonably hold the view that judicial interventions on behalf of shareholder litigants should be less expansive than direct shareholder interventions through elections. (Or vice versa?) The fact � if it is a fact � that shareholders are substantially hindered in both litigation and voting is not necessarily evidence of an overriding policy of �deference to authority� in the corporate governance system.

In this post, I have argued that Steve�s account of the corporate governance system seems unrealistic to the extent that it depends on the existence of elaborate and costly mechanisms of shareholder oversight that have no meaningful function. In addition, I have argued that �deference to authority� is not necessarily the defining characteristic of corporate law, even if we observe that shareholders are relatively impotent to influence most business decisions. Obviously, these are arguments against the theory of director primacy, rather than an affirmative case for shareholder participation. More in that latter vein in another post.

In U.S. corporation law, it is very, very difficult for shareholders to challenge the decisions made by the board of directors of their portfolio companies. Perhaps this difficulty is a vestige of some evolutionary deadend, like the appendix, but when one sees the same sort of difficulties cropping up in rule after rule and context after context, it becomes fair to at least ponder the possibility that it has some sort of survival benefit.

Director Primacy is Pervasive

In his latest post, Gordon complains that "every example of director primacy in [Steve's] post was drawn from the context of shareholder litigation." But so what? I'm not sure how we would measure degrees of difficulty in this context, but it's fair to say that shareholders have a hard time holding directors to account via the vote, just like they do via the lawsuit.

The vast majority of corporate decisions are made by the board of directors acting alone, or by persons to whom the board has properly delegated authority. Shareholders have virtually no right to initiate corporate action and, moreover, are entitled to approve or disapprove only a very few board actions.

Under the Delaware code, for example, shareholder voting rights are essentially limited to the election of directors and approval of charter or bylaw amendments, mergers, sales of substantially all of the corporation�s assets, and voluntary dissolu�tion. As a formal matter, only the election of directors and amending the bylaws do not require board approval before shareholder action is possible. In practice, of course, even the election of directors (absent a proxy contest) is predetermined by the existing board nominating the next year�s board.

The statutory decision-making model thus is one in which the board acts and shareholders, at most, react.

As I detail in my article, The Case for Limited Shareholder Voting Rights, 53 UCLA L. Rev. 601 (2006), corporation law�s direct restrictions on shareholder power are supplemented by a host of other rules that indirectly prevent shareholders from exercising significant influence over corporate decision making. Three sets of statutes are especially important: (1) disclosure requirements pertaining to large holders; (2) shareholder voting and communication rules; (3) insider trading and short swing profits rules. These laws affect shareholders in two respects. First, they discourage the formation of large stock blocks. Second, they discourage communication and coordination among shareholders.

Indeed, Harvard law professor Lucian Bebchuk, perhaps the leading proponent of shareholder empowerment, says that under the current regime the shareholder franchise is just a "myth." (See also my friend and UCLA law school colleague Lynn Stout's new paper, The Mythical Benefits of Shareholder Control, which compares shareholder control to "vampires" and "alligators in sewers.")

Gordon suggests that I "must contend that" voting and litigation "like the human vermiform appendix, either have no function or have a function that is hidden from our understanding." I decline to do either. Instead, as I explained in The Case for Limited Shareholder Voting Rights:

... like all accountability mechanisms, shareholder voting must be constrained in order to preserve the value of authority. As Arrow observes:

To maintain the value of authority, it would appear that [accountability] must be intermittent. This could be periodic; it could take the form of what is termed �management by exception,� in which authority and its decisions are reviewed only when performance is sufficiently degraded from expectations. . . .

The function of shareholder voting thus should be apparent. Like shareholder litigation, it "is properly understood not as an integral aspect of the corporate decision-making structure, but rather as an accountability device of last resort to be used sparingly, at best."

In sum, it strikes me as eminently fair to conclude that �Corporate governance is best characterized as based on �director primacy.�� Larry Ribstein, Why Corporations?, 1 Berkeley Bus. L.J. 183, 196 (2004). Cf. Harry G. Hutchison, Director Primacy And Corporate Governance: Shareholder Voting Rights Captured By The Accountability/Authority Paradigm, 36 Loy. U. Chi. L.J. 1111, 1194 (2005) (�Although 'Delaware has not explicitly embraced director primacy,' the relevant statutory provisions and the Unocal/Revlon/Unitrin paradigm have largely intimated that directors retain authority and need not passively allow either exogenous events or shareholder action to determine corporate decision-making.�); Kevin L. Turner, Settling The Debate: A Response To Professor Bebchuk's Proposed Reform Of Hostile Takeover Defenses, 57 Ala. L. Rev. 907, 927-28 (2006) ("Delaware jurisprudence favors director primacy in terms of the definitive decisionmaking power, while simultaneously requiring directors to be ultimately concerned with the shareholders' interest. ... The Delaware jurisprudence, while not explicitly affirming "director primacy," does implicitly leave the directors to make decisions with shareholders expressing their views only in specific and limited situations.").)

Establishing that director primacy is the status quo, of course, does not dispose of the question of whether it ought to be the status quo.

The Argument from the Status Quo

Proponents of Intelligent Design content that "there are natural systems that cannot be adequately explained in terms of undirected natural forces and that exhibit features which in any other circumstance we would attribute to intelligence." I express no opinion on the biological science of Intelligent Design, but when I look at corporate law I see a clockmaker. The system is so pervasively tilted in one direction that it one plausibly infers it was designed to do so. (Cf. Hollinger Inc. v. Hollinger Int'l, Inc., 858 A.2d 342, 374 (Del. Ch. 2004), in which Leo Strine referred to "the director-centered nature of our law, which leaves directors with wide managerial freedom subject to the strictures of equity, including entire fairness review of interested transactions. It is through this centralized management that stockholder wealth is largely created, or so much thinking goes.")

Even if the difficulties shareholders face in holding directors to account are mere evolutionary accident, moreover, it was a very happy accident. Gordon complains that I reason "from the status quo." I admit it freely.

Organizational structures that survive over time deserve the benefit of the doubt. Presumably, they offer features that contribue to their survival.

In fact, of course, the American corporation has not simply survived, it has thrived. John Micklethwait and Adrian Wooldridge opined that the corporation is �the basis of the prosperity of the West and the best hope for the future of the rest of the world.� John Micklethwait & Adrian Wooldridge, The Company: A Short History of a Revolutionary Idea xv (2003). A comprehensive review of the evidence by Holmstrom and Kaplan is temperate only by comparison:

"Despite the alleged flaws in its governance system, the U.S. economy has performed very well, both on an absolute basis and particularly relative to other countries. U.S. productivity gains in the past decade have been exceptional, and the U.S. stock market has consistently outperformed other world indices over the last two decades, including the period since the scandals broke. In other words, the broad evidence is not consistent with a failed U.S. system. If anything, it suggests a system that is well above average." Bengt R. Holmstrom & Steven N. Kaplan, The State of U.S. Corporate Governance: What�s Right and What�s Wrong? (Sept. 2003).

My colleague Lynn Stout's new paper argues that the evidence "suggests shareholders in public firms reap net benefits from board control." In the various articles I've written on director primacy, I've reached the same conclusion and made similar arguments.

Yet, even if the evidence doesn't prove director primacy is a contributing factor in the success of the US corporation, at the very least I can insist on a much stronger showing that we have seen to date from the "reformers."

Political thinkers from Plato to Edmund Burke have taught that prudence is the chief virtue of true statesmen. If nothing else, the law of unintended consequences must be given its due. The prudent legislator is hesitant to promulgate reforms that may give rise to new and unforeseen abuses worse than the evil to be cured.

Consider Sarbanes-Oxley, the great triumph of corporate governance reformers. When it was faced with voter unrest arising from the bursting of the tech stock bubble and high-profile misconduct by corporate directors and officers (Enron et al.), Congress decided that Something Must be Done. So Congress put together a package of reforms of varying provenance that had been kicking around Capitol Hill for ages and sent them up to President Bush for signing. And, ever since, it has been one unintended consequence after another.

Roberta Romano calls SOX "quack" corporate governance. Henry Butler and Larry Ribstein call it a "debacle."

The prospect of more such "reform" is precisely why I think we ought to assume that legal changes designed to promote "increased shareholder participation" are presumptively bad." Guilty until proven innocent by clear and convincing evidence.

What's the Case for Change?

Where is the proof? The observant reader will note that Gordon's posts have been mainly a critique of my "approach to the corporate governance system." What is the affirmative case for reforms to promote "increased shareholder participation"? Or, for that matter, the negative case against the status quo?

Shareholder Voting: The Case for Change

September 15, 2006 4:00 AM

In his most recent post, Steve refers to voting as an �accountability device of last resort.� Compare this appellation with Steve�s description of the shareholders� voting rights:

Under the Delaware code, � shareholder voting rights are essentially limited to the election of directors and approval of charter or bylaw amendments, mergers, sales of substantially all of the corporation�s assets, and voluntary dissolution. As a formal matter, only the election of directors and amending the bylaws do not require board approval before shareholder action is possible. In practice, of course, even the election of directors (absent a proxy contest) is predetermined by the existing board nominating the next year�s board.

When Steve says that voting is a last resort, he really means last resort! Let�s make sure that we all understand this. Under state corporate law, shareholders can initiate only two votes: an election of directors and an amendment of the bylaws. (Actually, shareholders also can initiate shareholder proposals unrelated to director elections or bylaw amendments, but such proposals are not binding on the board of directors.) Director elections require a proxy context, which is still extremely expensive. Bylaw amendments would be less expensive if placed on the corporate ballot through Rule 14a-8, but many bylaw proposals fail to pass the SEC�s review process and are excluded by companies. Most importantly, as discussed in my first post above, the SEC has allowed companies to exclude any proposals that relate to the election of directors, including most bylaw amendments that change the rules governing director elections. In short, shareholders can initiate only certain bylaw amendments without incurring substantial expense, and casual observation suggests that shareholders rarely pursue the more aggressive course of action.

According to Steve, impotent shareholders make for good corporate governance: �shareholder voting must be constrained in order to preserve the value of authority.� And thus we return to the place where we started, drawing the line between authority and responsibility.

Why draw the shareholder voting line in a way that offers shareholders no meaningful voting rights? Could it have something to do with the fact that corporate executives are a powerful and well-organized political force, while shareholders traditionally have been diffuse? I suspect Steve would find this explanation unconvincing, given his views on the market for governance terms:

If those governance terms are unfavorable, investors will discount the price they are willing to pay for that firm's securities. As a result, the firm's cost of capital rises, leaving it, inter alia, more vulnerable to bankruptcy or hostile takeover. Corporate managers therefore have strong incentives to offer investors attractive governance arrangements, either via the corporation's own organic documents or by incorporating in a state offering such arrangements off the rack. Likewise, competition for corporate charters deters states from adopting excessively pro-management statutes.

Stephen M. Bainbridge, Director Primacy and Shareholder Disempowerment, 119 HARV. L. REV. 1735, 1736 (2006).

I am less sanguine about the ability of markets to produce optimal governance terms. As noted recently by Lucian Bebchuk, �I do not view the U.S. corporate governance system, nor that of many other countries with developed stock markets, as largely dysfunctional. But between the dysfunctional and the optimal lies a rather large gap, and a developed stock market that grows over time is consistent with a governance system that lies in that gap and could be significantly improved.� Lucian Bebchuk, Letting Shareholders Set the Rules, 119 HARV. L. REV. 1784, 1791 (2006).

Steve reasonably requests a statement of the affirmative case for increased shareholder power, so here it is in heavily abridged form. Though I share Steve�s upbeat assessment of capital markets, takeover markets, charter markets, product markets, and labor markets as corporate governance mechanisms, each of these markets is useful in that context primarily for creating positive incentives for shareholder wealth maximization. As I noted in my earliest work, these markets are not well-equipped to deal with the problem of managerial incompetence:

External forces solve managerial incompetence in only two instances: (1) when the market for corporate control directly attacks managerial incompetence by replacing the incompetent manager, and (2) when product markets directly attack managerial incompetence by forcing companies out of business. Both of these solutions to managerial incompetence are effective at replacing managers but are slow and costly responses when compared with action by the board of directors.

D. Gordon Smith, Corporate Governance and Managerial Incompetence: Lessons From Kmart, 74 N.C. L. REV. 1037 (1996).

In my view, increased shareholder initiation rights in the election of directors would go some distance toward addressing managerial incompetence and would also provide an additional incentive to avoid any residual managerial slack. Of course, the devil is in the details, and the design of such a reform is crucial. Vice-Chancellor Leo Strine recently offered some hopeful thoughts on director election reform. Leo E. Strine, Jr., Toward a True Corporate Republic: A Traditionalist Response to Bebchuk's Solution for Improving Corporate America, 119 HARV. L. REV. 1759, 1778-82 (2006).

Whether voting reforms should extend beyond director elections is more controversial. Lucian Bebchuk has made the most ambitious case for such reforms. Lucian Arye Bebchuk, The Case for Increasing Shareholder Power, 118 HARV. L. REV. 833, 837 (2005). Though Lucian offers many ideas that I find appealing, my sense is that such reforms are very far off the radar for lawmakers.

Rep. Baker (R-La.) on the election

November 2, 2006 4:14 PM

"I just never thought I�d live to see the day when liberals would, as they have with relation to the markets, so openly advocate more taxation, more regulation, and more litigation, and, with a straight face, argue that most Americans would find this to be an attractive prescription for their financial well-being." Also:

Frivolous lawsuits are undeniably and unnecessarily raising the costs of doing business in America, and frightening off investors. I am convinced that common-sense securities-litigation class-action reform is not only a necessary component for safeguarding our global competitiveness, but that it also will cut the costs for plaintiffs in legitimate suits without diminishing the quality of representation. And while it is encouraging to see New York�s Democratic Sen. Chuck Schumer write in the Wall Street Journal this week that �it may be time to revisit the best way to reduce frivolous lawsuits without eliminating meritorious ones,� I would be more confident in the prospect of bipartisan collaboration on this effort were it not for the hundreds of millions of dollars in campaign contributions the Democrats reap from these firms � Milberg Weiss among them.

I would like to thank the Manhattan Institute for inviting me to participate in this online exchange on "Criminalizing Corporate Conduct: How Far Is Too Far?" It's my job to open the dialog, so I will do so by answering the question directly. It is all too far. There is no justification for criminalizing any corporate conduct.

Let's be clear about what I am asserting. Plenty of criminal activity takes place within corporations and other business organizations. (Given the extremely broad and amorphous nature many federal criminal statutes, e.g., mail and wire fraud, money laundering, RICO, obstruction of justice, this is almost necessarily the case.) The individuals who perpetrate such crimes are subject to prosecution for any crime they commit. I am expressing no opinion about whether the federal government has overcriminalized such individual conduct (although I have one). What I am asserting is that when individuals engage in criminal conduct within a business, there is no justification for imposing criminal punishment on the business as a corporate entity. There should be no corporate criminal liability.

Criminal law is penal law. Its about punishment. It is not designed to compose disputes, provide compensation to wronged parties, or impose administrative sanctions. It is designed to punish. This means that criminal sanctions may be properly imposed only on those persons and entities that can be deserving of punishment, that is, to those capable of acting in a morally blameworthy way. That's why infants, the incompetent, and the legally insane are excluded from criminal punishment. Thus, moral responsibility is a necessary condition for the application of the criminal sanction.

But corporations are not morally responsible agents. In fact, corporations, as opposed to the individuals who comprise them, are not agents at all. There is no ghostly corporate entity hovering above and separate from the set of individuals that labor for the corporation. There is no corporate brain within which intentions that are distinct from the intentions of these individuals can form. This fact is implicit in the current respondeat superior standard of corporate criminal liability that must attribute the mental states of a business organization's employees to the corporate entity.

Corporate criminal liability is theoretically incoherent because it imposes punishment, which requires morally blameworthy action, on an entity that is not a moral agent.

Perhaps you don't like that argument. Maybe it's too philosophical. OK, ignore it. Assume for the sake of discussion that corporations can be morally responsible for their employees' conduct. There is still no justification for imposing criminal punishment on them. That is because punishing corporations does not serve any of the legitimate purposes of punishment.

Theorists perennially dispute whether the purpose of punishment is retribution, deterrence, rehabilitation, or some combination of these. There is no need for us to speculate about how this dispute should be resolved because punishing corporations does not serve any of these ends.

Retribution can justify punishment only for those who have acted in a blameworthy way. Retribution clearly justifies punishing those who personally commit an offense. But how can it justify punishing a corporation? Corporations, as collective entities, cannot be imprisoned; they can only be fined. When a corporation is fined, it is the owners, i.e., the shareholders, who pay the fine. But the defining characteristic of modern corporation is the separation of ownership and control. The shareholders, who own the corporation, have no control over the actions of the employees who commit the offense. Hence, inflicting punishment on a corporation's shareholders (and its other employees who had no hand in the wrongdoing but may nevertheless lose their jobs) is punishing the innocent. Punishing those who are innocent of wrongdoing cannot be justified on retributivist grounds.

How about deterrence? A major purpose of criminal punishment is to deter wrongdoing. But not by any means; not by punishing the innocent. Much of the crime attributable to teenagers could undoubtedly be deterred by punishing parents for their children's offenses. The Nazis sought to deter acts of resistance by punishing innocent members of the communities in which such acts occurred. Although such measures may be effective, they are not permitted under our system of law. Deterrence as a justification for criminal punishment refers to punishing those who engage in wrongdoing to deter others from similar activities. It does not refer to punishing the innocent to pressure them into suppressing the criminal activity of their fellow citizens. Threatening innocent shareholders (and employees) with punishment for the offenses of culpable corporate employees may be an effective means of reducing criminal activity within business organizations, but it does not constitute the type of deterrence that can justify criminal punishment in a liberal legal system.

Perhaps punishment can be justified for purposes of rehabilitation. I have heard prosecutors argue that corporate criminal liability can be justified on rehabilitative grounds because fear of corporate prosecution will make business people behave better. But consider the nature of this argument. For punishment to be justified on rehabilitative grounds, it must be designed to reform the character of the wrongdoer, or at least, reduce the tendency of the wrongdoer to engage in future wrongful acts. One can rehabilitate only wrongdoers. Threatening those who have not engaged in wrongful conduct with punishment in order to make them "behave better" is not rehabilitation. It is coercing them to act in the way the coercive agent believes they should. "Rehabilitating" the innocent is simply depriving them of their liberty. This form of rehabilitation was familiar in the Soviet Union and Mao's China in which those whose conduct was unacceptable to the government were sent to psychiatric hospitals and "re-education" camps. Threatening to punish shareholders (and innocent employees) in order to make corporate executives behave in ways that prosecutors believe that they should is not a form of rehabilitation that can be countenanced in a liberal system of justice.

I have encountered the argument that punishing corporations for the actions of their employees is not distinct from any other form of criminal liability. After all, criminal punishment always wreaks harm on the innocent. The families and dependents of convicted criminals are inevitably adversely affected by the incarceration or impoverishment of the offender, both materially and emotionally. Hence, criminal liability always punishes the innocent, and corporate criminal liability is no different.

This line of reasoning elides a crucial distinction, however. In the case of traditional criminal liability, punishment is directed solely toward the wrongdoer. The harm that results to innocent third parties is not the intended object of the punishment regime. Such harm is always viewed with regret as an unfortunate collateral effect of visiting punishment on the blameworthy that should be minimized as much as possible. In the case of corporate criminal liability, however, punishment is intentionally directed toward those who have not committed an offense. Punishing the innocent is not a regrettable side effect and is certainly not to be minimized, but is the very object of the punishment regime. It may be true that all criminal punishment wreaks incidental harm on innocent parties, but this fact cannot justify a form of vicarious criminal liability that is intended to punish those who have not themselves broken the law.

This is the one hundredth anniversary of New York Central & Hudson River R.R. Co. v. United States, the Supreme Court decision that created corporate criminal liability. New York Central was a mistake when it was decided, remains a mistake today, and should be explicitly overruled. I make this statement fully aware that advocating for the reversal of a century-old precedent that is the foundation for much of federal criminal jurisprudence and law enforcement policy is quixotic. Few precedents are so deeply entrenched and have been so repeatedly affirmed in subsequent decisions. As the recent decision in United States v. Ionia Management S.A., 555 F.3d 303 (2009) summarily dismissing a challenge to the New York Central standard of corporate criminal liability makes clear, the prospects of the judiciary seeing the light on this issue are not good. Nevertheless, the process by which society determines who or what should be subjected to criminal punishment should not be like jazz, which I once heard defined as the musical form in which one legitimizes a mistake by repeating it. After one hundred years, it is time we stopped repeating the New York Central court's mistake.

I, too, would like to thank the Manhattan Institute for inviting me to participate in this discussion with Professor Hasnas. Having read his first post, I can assure the readers about one thing: I disagree with just about everything John has said so far. So, I have no doubt that this will be an interesting exchange.

I don't intend in this first posting to respond to every argument John has made. I will pick a few of the most fallacious, and will add to my position as the week progresses.

Let's start with the concept of whether corporations are "morally responsible agents." Corporations are, of course, legal constructs; we often refer to them as "legal persons." The primary reasons for their existence are twofold: (1) to allow individuals to pool their resources together to engage in economic activity on a scale that they could not achieve individually; and (2) to limit the liability of such individuals ("investors") should the operation go belly-up. Obviously, a corporation can only act through its "agents," i.e., its board of directors, officers, and employees. When these humans act in their agency role, however, they are not simply a bunch of people with titles doing whatever they want. They are acting on behalf of their principal, the corporation. They have a fiduciary duty to put the interests of the corporation ahead of their own.

As a result of their collective nature, corporations are extremely powerful actors in our economic system. They do a tremendous amount of good - adding trillions to our gross domestic product each year. For example, corporations do medical research, build skyscrapers, transport us by air and automobile, and supply us with endless means of entertainment. But corporations also do some things that are not so good. For example, they occasionally breach contracts and engage in tortuous activity. Just because a corporation acts through human agents, does it make any sense to say that it is not a "morally responsible agent" when its contract breach costs taxpayers millions of dollars or its poorly designed automobile blows up on impact and kills innocent people?

One could take the position that the corporation is not morally responsible for these acts; rather the responsible human or humans behind the corporate veil are. But this is not a satisfactory answer for two main reasons. First, the harm is often if not always an outgrowth of the collective nature of the enterprise itself. One person, for example, can't design, build, and mass-market an automobile. Thus, quite literally, the collective entity is at the root of the wrong - and it should take responsibility for it. Second, if we laid blame for harms caused by corporations solely at the doorstep of individuals, the result would be unworkable: no one person in the organization has the money or insurance to compensate for a contract breach or make victims of a tort whole. The corporation does.

Thus, corporations are undoubtedly morally (and legally) responsible for their contract breaches and tortuous behavior. So the question really is this: is there something different about criminal law that would lead us to a different result? The simple answer is no.

John is right that the aim of criminal law is to punish wrongdoers. We've established that a corporation can be a wrongdoer - it can, through its many agents, dump toxic chemicals or put adulterated food into the stream of commerce. Can a corporation be punished? Once again, I agree with John that the goals of criminal law are retribution, deterrence, rehabilitation, and (one I don't think he mentioned) incapacitation. I will explain later why all of these goals can, in fact, be achieved through corporate criminal liability. Here's a hint: the fact that a corporation cannot be put behind bars is irrelevant.

Corporations Are People, Too?

July 27, 2009 9:17 PM

Unlike Mike, I cannot say that I disagree with everything Mike says in his post. In fact, I agree with much of it. I do disagree with a few crucial points, however. Mike's entire first paragraph seems unobjectionable to me with the exception of one word. Corporate employees acting within their agency role are indeed acting on behalf of their principals. But their principal is not the corporation. Their principals are the shareholders who own the corporation. Their fiduciary duty is to shareholders, not to the corporation, which is a legal construction and not a real entity. This is not a semantic quibble on my part. As long ago as 1935, Felix Cohen was warning us about the danger of reifying abstract legal concepts in general and the concept of the corporation in particular. (See Felix Cohen, Transcendental Nonsense and the Functional Approach, 35 Colum. L. Rev. 809 (1935).) In my opinion, this tendency to treat corporations as real entities is responsible for much of the confusion about and support for corporate criminal liability.

I probably only disagree with one word of Mike's second paragraph as well, in this case, the word is 'morally.' Mike asks, "does it make any sense to say that it is not a "morally responsible agent" when its contract breach costs taxpayers millions of dollars or its poorly designed automobile blows up on impact and kills innocent people?" Yes, indeed it does. It may not make sense to ask whether a corporation is a legally responsible agent for purposes of such contract breaches or torts, but it certainly makes sense to ask whether the corporation is a morally responsible agent. Because it is not.

There is nothing logically offensive about the concept of vicarious legal liability. But there can be no such thing as vicarious moral responsibility. Moral responsibility applies only to intentionally acting agents who are causally responsible for producing certain consequences. Corporations are not morally responsible agents because 1) corporations are not agents, 2) corporations are not causally responsible for the actions of their employees, and 3) corporations do not act intentionally.

Corporations are not agents because they are not real entities. (This is the Transcendental Nonsense stuff again.) We speak as though corporations act rather than human beings because it is convenient and useful to do so. But we should not be misled by our semantic conventions into reifying the corporation. There is no "thing" there. There is only a complex set of relationships among a wide flung group of human beings.

Corporations are not causally responsible for the actions of their employees because corporations do not act, only people do. Corporations can exert causal influences on its employees in the sense that individuals in groups can get each other to behave in ways that none would if acting alone. But this effect is fully explainable in terms of the interior desires and beliefs of the individuals in the organization that lead them to behave as they do. There is no need to postulate "some kind of ghostly organizational spirit that is present in the organization and that somehow exerts external pressures and forces on its members." (See Manuel Velasquez, Debunking Corporate Moral Responsibility, 13 Bus. Ethics Q. 531, 544 (2003).)

Corporations can be said to act intentionally only in a metaphorical sense. We often attribute purposes or beliefs to groups on the basis of a pattern which the activities of its members exhibit. But this does not mean that the group has real intentions in a literal sense. We frequently speak of markets as "trying" to find a bottom, but that does not imply that the market is literally capable of intentional action. With no corporate brain in existence, a corporation is incapable of acting intentionally in a literal sense.

Why make a big deal over moral as opposed to legal responsibility? It makes no difference with regard to the assignment civil liability. As Mike points out, there are quite good reasons for holding corporations liable for breach of contract or tortuous actions. I can easily justify respondeat superior liability in tort. The reason lies in Mike's own question "Is there something different about criminal law that would lead us to a different result?" The simple answer to this is not no, but yes. Criminal responsibility is different in kind from civil liability. Criminal responsibility is concerned with punishment, not the proper adjustment of benefits and burdens among the members of society. And punishment requires blameworthy action. Human being can behave in a blameworthy way. Fictitious legal constructs cannot. Vicarious moral responsibility is an oxymoron. That is why there should be no corporate criminal liability despite the fact that there can be corporate civil liability.

In pursuing this line argument, I am probably doing a disservice to the reader by focusing on such a philosophical point. So, for purposes of our continued exchange, let me abjure further reliance upon it. Even if there were some sense in which moral responsibility could be attributed to corporations, criminal punishment would still be unjustified because it is improper to punish those who are innocent of wrongdoing, and punishing corporations as opposed to the individuals who commit the offenses does precisely that.

Before I get to the heart of the matter, let me briefly respond to John's latest claim that corporations, as mere legal fictions, cannot be agents (I assume he means "actors") because they are not "real entities." He's right, of course, in the sense that the corporate "being" exists on a piece of paper filed with the state. But he's being far too formalistic in his legal thinking. In reality, a "corporation" is a group of people who are organized in a certain way: there are shareholders, who are the owners; officers, who run the show; mid-level managers, who make decisions within their realm of authority; and employees, who carry out day-to-day tasks. If we decide to punish the corporation for wrongdoing, we're not punishing some ethereal being; rather, we're punishing in a collective fashion the people associated with the corporation. Viewed this way, the philosophical issue is really a red herring. (So, from this point forward, the reader should assume that I am using the term "corporation" as a shorthand reference for the human constituencies that comprise the overall entity.) The real question is whether punishment administered by the state against this collective is justified in deontological and utilitarian terms.

This question restates the one I posed earlier: does corporate criminal liability serve the fundamental purposes of punishment? The answer is yes.

One of the main purposes of punishment is deterrence - the prevention of future crime by the wrongdoer (specific deterrence) and others (general deterrence). One would not need corporate criminal liability if administrative fines and penalties were sufficient to keep corporations in line, but they are not. Corporations tend to treat fines as a cost of doing business; if the benefits of socially irresponsible behavior outweigh the potential costs (times the likelihood of getting caught), they will undertake it. The prospect of a criminal conviction, however, is different in kind. A corporation's reputation is one of its biggest assets, and a criminal conviction tarnishes that reputation in a serious and often unpredictable way. The corporation (really, its officers and managers) has an immense incentive to avoid this outcome. Stated differently, the prospect of criminal liability has a considerable additional deterrent effect over administrative remedies. Don't believe this? Look up the statistics on the number of corporations under investigation in recent years that decided to cooperate with law enforcement in exchange for a "deferred prosecution agreement" - that is, the promise that criminal charges would not be filed.

Let me be even more specific. In the absence of entity liability, corporate officers have every incentive to encourage criminal conduct by mid-managers and lower-level employees as long as the officers themselves are shielded from personal liability. They can accomplish this by, for example, setting productivity targets so high that they cannot be met through legal means, and then firing or demoting employees who fail to meet them. Employees quickly understand what they need to do to keep their jobs and get promoted, while high level management hides behind a veil of plausible deniability. Later, if criminal proceedings are initiated and lower-level employees get caught, management can (correctly) point to the fact that it never sanctioned criminal activity and was not "aware" of its existence. It can go even further and throw a couple of minor employees to the prosecution wolves, claiming that they were rogues and that their termination (and prosecution) has cured the problem. Meanwhile, the managers - the true rogues - will continue their way up the sleazy corporate ladder.

Entity prosecution significantly reverses this equation. If managers obliquely encourage widespread criminality and the entity gets caught, prosecution of the corporation means that the entity will pay a price. It will suffer a loss of reputation and likely lost revenues and market valuation. Harm to the corporation means harm to the officers. They may lose their jobs, or at least suffer monetary losses such as a reduction in the value of their stock options and portfolios, and perhaps the loss of future salary increases or bonuses. Certainly their professional reputations will be forever tainted. Given these prospects, preventing - as opposed to encouraging - criminality within the corporation now looks like the better path to choose.

Corporate liability also serves the second purpose of punishment: rehabilitation. This is far from a nonsensical notion. The people making up a corporate body create an "atmosphere" with many different characteristics, including whether criminality is encouraged, tolerated, or deplored. This atmosphere is self-perpetuating: agents of the corporation are likely to hire new agents with the same or similar proclivities, and these will be reinforced in an infinite number of ways within the organization over time. Punishing a few wrongdoers is not likely to change the atmosphere of a big organization, but collective entity liability will. By holding the corporation liable, prosecutors (and judges) can ensure that the corporation puts in place compliance programs with real teeth in them. In recent times, corporations have even agreed to place outside "watchdog" directors on their board to help with the oversight process. Over time, compliance programs and careful oversight can reform the organization.

One part of rehabilitation is the paying of restitution to the victims of one's crime. Often, white collar prosecutions involve millions - even hundreds of millions or billions - of dollars of fraud. Convicted individuals do not have at their disposal anything near the amount of money necessary to pay restitution to the victims. The corporate entity, however, does.

Corporate criminal liability can also serve the purpose of incapacitation, which is the disabling of the wrongdoer's ability to commit crime. For individuals, incapacitation is traditionally achieved by incarceration, obviously impossible in the case of a collective entity. However, one of the potential penalties for corporate criminal behavior involving government programs - which constitutes a huge percentage of health care fraud and all of defense procurement fraud - is debarment from that program, a form of incapacitation. Although this corporate form of "death penalty" should be used sparingly, it is undoubtedly warranted in cases in which the harm to taxpayers was extreme.

Finally, there is the question of retribution - punishment for punishment's sake. John's position is that imposing criminal liability on corporations is immoral because the "corporation" is nothing more than a group of wholly innocent shareholders. This position is overly simplistic: as noted at the beginning of this post, the "corporation" is made up of all of its human constituencies. When crime is committed by a set of individuals within the corporate structure, some members of each constituency are deserving of the "hurt" that corporate criminal liability imposes on them - and, admittedly, some are not.

Take the shareholders, for example. Some of them may not be so innocent: often, the major shareholders in a corporation are also its officers, who may very well have participated in the crime. Moreover, even some shareholders who are not direct participants should, indeed, help pay for it. And let's be clear: all they will do is pay monetarily through the reduction in the value of their shares. Their personal reputations will not be impugned; they will not serve any time in prison. Depending upon the timing of their purchases and sales, they may very well have benefitted monetarily from the criminality; if so, paying for it is not unfair. Even if a shareholder is truly innocent, one would expect that an efficient market would price in the risk that a company will get caught committing a crime, making the shareholder's loss in a particular case simply one of many perils associated with investing (and offset through diversification).

I'm not suggesting that in holding corporations criminally liable, some innocent people are not harmed. I concede that shareholders simply caught holding the bag are impacted solely as a result of bad timing, which is not particularly fair. In addition, to the extent that the corporation suffers monetarily because of the punishment and reduces in size (or in the rare instance, goes bankrupt) as a result, innocent employees will be hurt financially. Finally, if the corporation raises its prices to offset the costs of a criminal conviction, innocent consumers will literally pay the price, although market forces should act to keep this harm to a minimum.

Corporate criminal liability thus has some very significant benefits in deterring corporate crime and forcing corporations that commit crime to clean up their act. It punishes high- and mid-level managers who might otherwise mastermind immense frauds and get away scot-free. These benefits should not be underestimated - given the extent to which our economy is dominated by corporations, without such liability, white collar crime could very well run rampant throughout our business sector (even more so than it does today). On the other hand, corporate criminal liability does negatively impact some innocent people. Unlike John, I think these people are analogous to the innocent family members who are harmed by a loved one's incarceration. To me, the critical question is how to ensure that corporate criminal liability is imposed only when its benefits outweigh its costs. I will take up that question in a later post.

OK, this is getting interesting. From my perspective, real progress is being made. We now have agreement that corporations are not agents and that "[i]n reality a 'corporation' is a group of people who are organized in a certain way." (Although having cited legal realist Felix Cohen's brilliant attack on the legal formalism that gave rise to the reification of the "corporation" to make this point, I find it odd to be accused of being formalistic in my thinking.) We are also in agreement as to what it means to criminally punish a corporation. As Mike says, "If we decide to punish the corporation for wrongdoing, we're not punishing some ethereal being; rather, we're punishing in a collective fashion the people associated with the corporation." Thus, we are also in agreement that the essential question is whether this form of collective punishment is justified. Mike argues that it is. I will continue to argue that it is not. In fact, collective criminal punishment is never justified (unless all individuals are in fact culpable, but then it is unnecessary).

Mike's argument that collective corporate punishment is justified by its deterrent value is revealing. He begins by arguing that the financial penalties associated with civil liability and administrative sanctions provide insufficient deterrence because corporations treat them as a costs of doing business. But note that the penalties imposed on corporations convicted of criminal violations are financial penalties as well. By Mike's own hypothesis then, these must also be inadequate (especially since they are frequently considerably less than the amounts that corporations must pay out in civil judgments). As Mike's argument makes clear, the deterrent effect of collective corporate punishment is entirely due to the moral stigma that comes along with a criminal conviction. The essence of this form of deterrence is now made clear. It consists of stigmatizing a group of people to prevent individuals within that group from engaging in wrongdoing.

Mike wants to cite the statistics on DPA's support his claim that this form of punishment is an effective deterrent. There is no need to do so. There is no doubt that it is an effective deterrent. It's just not a morally legitimate form of deterrence.

Mike claims that in the absence of this form of collective punishment corrupt upper managers will try to create situations in which they encourage criminal activity in subordinates while insulating themselves from personal liability. I'm sure that this is true. Of course, this is not limited to corporate criminals. Criminals always try to accomplish their objectives in a way that makes it least likely that they will go to jail. No one said that prosecutors have easy jobs. It is very difficult to effectively enforce the law within the confines of a system that is designed to be highly protective of the innocent. You know, it's that presumption of innocence, proof beyond reasonable doubt, unanimous jury, right against self-incrimination, attorney-client privilege, Blackstonian "it is better that ten guilty persons escape than that one innocent suffer" thing.

But the difficulty for prosecutors is intentionally built into the system. We want it to be difficult for the government to punish citizens. Our criminal law is based on the normative judgment that an unchecked government is a greater threat to liberty and individual well-being than are individual criminals. The last thing we want to do is to achieve law enforcement efficiency at the expense of the civil libertarian protections of the criminal justice system.

Collective punishment undoubtedly is an effective deterrent. But it is not the type of deterrence that justifies criminal punishment in a liberal legal regime. In a liberal legal regime, deterrence refers to inflicting punishment on a wrongdoer to discourage others from committing similar offenses. It does not refer to threatening to punish the innocent to pressure them into suppressing the criminal activity of their fellow citizens.

Consider Mike's own example of the advantages of corporate criminal liability for effective deterrence. Retain each step in his argument and description of how the punishment would work, but change its setting from the corporate business environment to, let's say, the threat posed by potential Japanese saboteurs during World War II, communist infiltration of the government during the 1950's, or Islamic terrorism today. There are many situations in which we can more effectively deter criminal activity by punishing entire groups or threatening group punishment to pressure innocent group members into suppressing the activities of others within the group. We just should not. I'm sure we could greatly deter teenage crime by threatening to punish the parents of all teenage offenders, who after all have more control over their children's behavior than shareholders have over corporate employees.

Threatening to employ the moral stigma of criminal conviction to damage the reputation of corporations, the job security of managers, and the financial well-being of investors, many if not most of whom may be innocent (think Arthur Andersen here), in order to more effectively deter individual criminal activity within the corporation is no different than threatening to employ the moral stigma of criminal conviction to damage the reputation, job security, and financial well-being of members of the communist party in order to more effectively deter subversive activity by individual communists. This form of deterrence is effective. It's just not justified.

What about rehabilitation? I have encountered arguments like the one Mike produces before. I have personally appeared on several panels with current or former prosecutors who argued that corporate criminal liability can be justified on rehabilitative grounds because fear of corporate prosecution can make business people behave better. And Mike is entirely correct that fear of prosecution can influence corporations to change their corporate culture and institute (government approved) compliance programs (which by the way are not "ethics" programs, see my book, Trapped: When Acting Ethically is Against the Law). The threat of collective corporate punishment can indeed cause corporate management to become deputy prosecutorial agents and to attempt to have a corporate culture that discourages individual criminal activity.

But this is not rehabilitation. Rehabilitation refers to imposing treatment on a wrongdoer designed to reform his or her character to ensure better behavior in the future. One cannot rehabilitate the innocent. Threatening those who have not engaged in wrongful conduct with punishment in order to make them "behave better" is not rehabilitation. It is coercing them to act in the way the coercive agent believes they should. "Rehabilitating" the innocent is simply depriving them of their liberty. This form of rehabilitation was familiar in the Soviet Union and Mao's China in which those whose conduct was unacceptable to the government were sent to psychiatric hospitals and "re-education" camps. Threatening collective punishment to cause corporate executives create what the government considers to be a positive corporate culture is not a form of rehabilitation that can be countenanced in a liberal system of justice.

In my opinion, restitution is a non sequitur. Restitution is not a purpose of punishment. Restitution is what the civil liability system is for. Corporate criminal liability merely interferes with this. For example, Arthur Andersen had negotiated a $750 million settlement with Enron's shareholders which fell through when the firm was destroyed by the federal criminal indictment.

Finally, retribution. On this point, there is nothing to answer. Mike states, and I agree, that "When crime is committed by a set of individuals within the corporate structure, some members of each constituency are deserving of the 'hurt' that corporate criminal liability imposes on them - and, admittedly, some are not." Mike continues by identifying situations in which shareholders may be personally implicated in criminal activity-"the major shareholders in a corporation are also its officers, who may very well have participated in the crime." The purpose of retribution is indeed served by imposing individual punishment on all such shareholders and any other individual associated with the corporation who is personally culpable.

But the rest of what Mike has to say is that to the extent that corporate criminal liability punishes those who are not personally culpable, the punishment will not be too severe; will consist only in financial losses, not prison time; may involve a fair redistribution of wealth; and may even be economically efficient. This is all very interesting, but it has nothing to do with retribution. Retribution refers to the process of requiting evil with evil in which harm is imposed on a wrongdoer in recompense for or in dissipation of the harm that he or she has done. Explaining why imposing punishment on those who have not done wrong is not so bad cannot justify the punishment on retributivist grounds.

Mike ends by admitting that in holding corporations criminally liable, some innocent people are harmed. He argues that the benefits to be gained are worth this cost. I disagree, and I have Blackstone on my side. It is never a good bargain to purchase law enforcement efficiency at the cost of empowering law enforcement agents to take shortcuts around the civil libertarian aspects built into our system of criminal law. As supporting evidence, I offer Department of Justice policy governing the indictment of corporations contained in the Holder/Thompson/McNulty/Filip Memorandum.

I'm going to respond to John latest arguments in my next post. Right now, though, I want to spend some time addressing one additional and critical purpose served by corporate criminal liability: obtaining cooperation from corporations in the investigation and prosecution of the individuals at the heart of white collar crime. Without the ability to threaten charges against the corporate body, such cooperation would be rare if non-existent. And without corporate cooperation, many serious white collar criminals would never be caught. Let me explain why.

For a number of reasons, the investigation and prosecution of white collar crime is extremely slow and resource-intensive. First, the crime itself is often very complex. Indeed, sophisticated white collar criminals do all they can to add to the complexity of their misdeeds by hiding them beneath layers of accounting tricks, false or fraudulent transactions, deleted records, and second sets of books. In a case of any significance, law enforcement might have to sort through hundreds of thousands or even millions of documents to unravel the criminal behavior. This work might take a team of investigative agents and multiple prosecutors years to carry out.

Second, white collar cases are often not open and shut. Many times, the key question is whether the individual defendants harbored the requisite criminal intent. The typical white collar defendant will plead good faith or advice of counsel. Specifically, he might claim that the action in question was, at worst, an innocent accounting mistake. Alternatively, the defendant might contend that he approved the questionable transactions, or the manner in which they were recorded on the company's books, only after carefully consulting with the corporation's accountants and legal counsel. In cases in which the prosecution alleges that the defendant stole from the company, he is very likely to claim that the money and benefits he "received" were merely "compensation" previously approved by the board.

These kinds of defenses are unique to white collar crimes; a bank robber, for instance, cannot plead ignorance of the law or claim that the bank approved of his (illicit) withdrawal. Moreover, they are difficult to overcome. At minimum, they require extensive interviews with the accountants, lawyers, and directors involved. If these individuals are not inclined to be cooperative, they must be subpoenaed to the grand jury and perhaps even granted immunity. In addition, countering these defenses often requires prosecutors to seek the advice of experts and prepare them for possible testimony at trial.

Third, major corporations and the individuals associated with them usually have the resources to hire excellent attorneys who specialize in white collar criminal defense. These attorneys have the ability to slow down an investigation to a considerable extent if they so choose. They can object to subpoenas duces tecum on a whole host of grounds, forcing repeated hearings relating to subpoena enforcement. They can claim attorney-client privilege and work-product protection of the documents subject to a subpoena, requiring the establishment of a system to filter the challenged documents to obtain a ruling from the court before government agents may see them. Defense counsel can advise their clients not to give voluntary statements to government investigators and to exercise their Fifth Amendment right not to be compelled to testify in the grand jury absent immunity. If they are coordinating their efforts through a joint defense agreement, counsel can ensure that this lack of cooperation is widespread, if not universal, forcing prosecutors to decide which potential witnesses to immunize in a situation of substantial uncertainty - something they are rightly hesitant to do. Unless it is fueled by a whistleblower or other inside information, these tactics - labeled the "delaying game" by criminologist James Coleman - can slow an investigation to a snail's pace, and perhaps even cause it to stall altogether.

Fourth, the difficult nature of white collar cases means that they often must be prosecuted bit by bit, as prosecutors unravel the wrongdoing and work their way up the corporate ladder. Charges are first brought against the lower-level employees, who are much more likely to have been caught red-handed, with the hope that their indictment or conviction will lead to cooperation against mid-level management. If this succeeds, the mid-level managers are prosecuted with the hope that they will implicate the responsible high-level corporate officials. If so, prosecutors can finally attempt to bring these individuals to justice. Of course, because they have insulated themselves from the criminal activity, they undoubtedly have the best chance of either escaping conviction or having it overturned on appeal, despite being the most morally culpable of the bunch.

Corporate cooperation shifts the balance of power dramatically. No longer foes, the corporation and the government team up to unmask the individuals at the center of the criminal activity, thereby getting to the heart of the matter quickly and efficiently. This is because the corporation has likely completed an internal investigation long before law enforcement arrived on the scene. If so, the corporation has already isolated the critical documents and witnesses, compiled witness statements, and identified the key culprits. If it willingly turns this information over to the government, a huge amount of time can be shaved from the investigative phase of the criminal case. Indeed, with corporate cooperation, the successful completion of a complex white collar prosecution might very well be reduced from a matter of years to a matter of months. This is not mere speculation: in the post-Enron era, DOJ has successfully prosecuted an unprecedented number of very large corporate frauds by convincing companies to cooperate against the individuals involved.

Without the possibility of a criminal conviction hanging over its head, however, a corporation has zero incentive to cooperate. By circling the wagons, it has a good chance of shielding its own from prosecution. And even if this strategy eventually fails, so what? A few individuals will be prosecuted and the corporation will continue on its merry (and probably criminal) way.

John will probably respond that forgoing this enormous benefit is the price we must pay because corporate criminal liability is morally offensive in a liberal state. I consider myself a very strong proponent of morality, but I just don't think it's that simple. More on this later.

I admire Mike's latest post. It is a lucid, forthright, and honest statement of the purpose of corporate criminal liability. I have rarely seen a better explanation of the role played by corporate criminal liability in the contemporary criminal justice system.

Mike says he is describing an "additional and crucial purpose served by corporate criminal liability." I would insist on the removal of the word "additional." I think Mike has described the only purpose of corporate criminal liability, and it is a crucial one: coercing corporations into cooperating in the investigation and prosecution of their employees.

Mike asserts that "without corporate cooperation, many serious white collar criminals would never be caught." I entirely agree. Mike then gives four supporting reasons why this is true. They are impeccable. This is one of the best statements of the value of corporate criminal liability for the white collar prosecutor that I have seen.

Under our traditional criminal law, there is no duty to aid the government in one's own prosecution. The presumption of innocence places the burden on the government to prove every element of its case, and the reasonable doubt standard makes that very difficult to do. The privilege against self-incrimination guarantees that an accused does not have to help the government overcome this hurdle. Other civil libertarian protections built into the system increase the difficulty. The requirement to prove mens rea is a major impediment for prosecutors, and the accused's right to the advice of counsel and the protection of the attorney-client privilege further raises the bar. Yet, despite all these obstacles, prosecutors successfully prosecute and convict criminals and enforce the law every day.

This is because the traditional criminal law addresses actions that cause visible harm in the world. Murder, assault, robbery, rape, arson all have visible effects about which evidence can be gathered, and the nature of the conduct typically suggests the ill will or mens rea with which it is undertaken. Even traditional state level fraud (the offense of false pretenses) requires an outright misrepresentation, actual reliance upon it, and an actual loss of property. Such elements are susceptible of proof within the protective civil libertarian constraints built into the criminal law.

As Mike lucidly points out, white collar crime is different. Over the course of the 20th and 21st centuries, Congress has seen fit to create a broad array of amorphous and inchoate new offenses. The federal fraud statutes criminalize any scheme or artifice to defraud. This requires neither misrepresentation, reliance, nor loss, and can consist of any potentially deceptive conduct. The elements insider trading continue to confound courts and legal scholars. In addition, the federal government has enacted a myriad of arcane, malum prohibitum regulatory offenses. Boiled down to its essence, what we call the "white collar" criminal law consists of the effort to police the behavior of those engaged in business for compliance with regulatory requirements and general honest dealing.

Congress dumped the task of enforcing this body of law into the laps of federal prosecutors. But, as Mike points out, this task is virtually impossible within the civil libertarian confines of the traditional criminal law. In the first place, policing all of the business concerns in the United States not only for honest dealing, but for compliance with the myriad regulations that carry criminal penalties is a monumental task. These offenses typically consist of deceptive behavior, and have no corpus delicti or smoking gun to introduce into evidence. White collar crime is intentionally designed to be indistinguishable from non-criminal activity. As a result, considerable investigation may be required merely to establish that a crime been committed. Even then, a great deal of legal and/or accounting sophistication may be required to unravel the deception. No matter how large the Department of Justice's budget for white collar crime may be, it would still be insufficient to address anything beyond the tip of the iceberg of potential offenses if DOJ had to prove every element of its case beyond a reasonable doubt with its own resources.

Secondly, the organizational setting makes it extremely difficult to establish the mens rea of these offenses. The corporate form diffuses decision-making responsibility. Decisions made by one member of a firm may not be fully informed by what other members of the firm are doing or have decided, and corporations frequently take actions that were never explicitly known to or authorized by any identifiable individual or individuals within the firm.

And finally, the right against self-incrimination and the attorney-client privilege make it difficult for prosecutors to obtain the evidence they need to meet their burden of proof. Because white collar crime consists primarily in crimes of deception, the evidence necessary for a conviction will consist predominantly in the business records of the firm for which the defendant works and the testimony of co-workers. But to the extent that these records are in the personal possession of the defendant, contain communications between the defendant or other members of the firm and corporate counsel, or are the work product of corporate counsel, the right against self-incrimination and the attorney-client privilege render them unavailable to the prosecution. And to the extent that it consists of the testimony of others members of the business who may fear prosecution, the right against self-incrimination again renders it unavailable.

As a society, we face a choice. We can continue to pass criminal statutes that cannot be enforced consistently with the preservation of our civil liberties or we can draw back and let market forces and civil liability enforce those aspects of honest dealing that cannot be controlled criminally without severely impairing those liberties. But prosecutors have no such choice. They are charged with enforcing the law. And given the nature of the white collar criminal law described above, the only way to do so is to coerce business people into acting as deputy prosecutorial agents.

The purpose of corporate criminal liability is not to punish corporations. It is to force them to cooperate in the prosecution of their employees. As Pam Bucy has documented, see Pamela H. Bucy, Why Punish? Trends in Corporate Criminal Prosecutions, 44 AM. CRIM. L. REV. 1287 (2007), there is an ever-increasing number of federal criminal investigations of business organizations and an ever-decreasing number of corporate indictments and convictions. That is because if a federal prosecutor actually has to bring a corporation to trial, the prosecutor has already failed in his or her mission to get it to cooperate. Arthur Andersen was indicted and destroyed primarily because it failed to agree to waive its attorney-client privilege. See Julie R. O'Sullivan, Some Thoughts on Proposed Revisions to the Organizational Guidelines, 1 OHIO ST. J. CRIM. L. REV. 487, 496, n. 30 (2004) . This also accounts for the sharp increase in DPA's that Mike mentioned in an earlier posting. The DPA's are granted in return for the corporation's cooperation with the federal investigation.

As Mike points out, the ability to threaten the corporation with criminal indictment shifts the balance of power between prosecutor and defendant. And that is the purpose of corporate criminal liability. The reason why it does not advance any of the traditional purposes of punishment is that it is not designed to punish. It is designed to circumvent the pro-defendant, liberal bias inherent in our system of criminal law.

Mike has already spoken for me by indicting that my objection to this will be that it is morally offensive in liberal society. Prosecutors' jobs would be much easier if they could threaten to indict all those who might have knowledge relevant to their criminal investigations unless they aided in the prosecution of their fellow citizens. Generally, we do not permit this, and for good reason. It reminds us too much of the practices of the Nazi and Soviet regimes in which failure to inform on others was itself an offense. We don't want a society in which police agencies pursue their missions by adopting practices that turn citizens against each other.

Corporate criminal liability is the exception to this, and it is an unfortunate one. For it turns employers into the adversaries of their own employees whenever those employees are suspected of committing a crime. This may not seem so harmful in cases in which the employees have truly acted purposely and malevolently. But given the amorphous nature of the federal criminal law and its myriad provisions that can be violated without one's moral sense indicting that one is doing anything wrong (try teaching white collar crime to business students and watch their eyes widen in fear), this becomes a very harmful and destructive practice indeed. As knowledge of the incentives placed on corporations by corporate criminal liability slowly become more widely known by those involved in business, its poisonous effect on trust and loyalty spreads. Years ago, my students used to believe that if one was loyal to his or her employer and did the right thing, he or she was entitled to expect loyalty and support in return. These days most of them have learned enough about how businesses respond to the incentives of corporate criminal liability to know when asked about any suspicious behavior in the corporation, rule #1 is: Calm up and get your own personal attorney.

What makes law so interesting is that, on the tough questions, there is no clearly right answer. This is because, once we are removed from the trivial and commonplace, legal decisions are based on policy considerations and empirical determinations. The problem is that, inevitably, competing policies judgments pull in opposite directions, and - as good as science is today - the facts most critical to decision-making are usually unprovable assumptions about which reasonable persons can disagree. While society is in the midst of making a tough legal decision, lawyers, judges, legislators, columnists, and citizens muster the best arguments they can to persuade others as to the rectitude of their position. For some period of time, the right answer seems beyond reach. Eventually, though, one side's arguments begin to sound more persuasive to a majority of the populous. Ultimately, a consensus is reached on the correct position for society to adopt. There is now a "right" answer to the question. It's critical to understand, however, that this answer is only "right" because of the consensus, and not the other way around. (This is a pragmatic view of the law; if the reader is interested in a more in-depth philosophical discussion, see Michael L. Seigel, A Pragmatic Critique of Modern Evidence Scholarship, 88 Northwestern University Law Review 995 (1994).)

So, at one time, the consensus in the United States was that "separate but equal" schools satisfied the Fourteenth Amendment's due process clause; today, it would be hard to find anyone who endorsed that previously "right" answer. At present, a debate rages about whether gays should be permitted to marry as a matter of the law's guarantee of "equal protection"; a generation from now, there will be a right answer to this question as well.

Given my pragmatic philosophy, John's position that collective entity criminal liability is "wrong" in some absolute sense simply does not compute. In part, he bases his conclusion on the proposition that criminal liability is "different" than civil or administrative liability. He claims that a liberal government can under no circumstances impose corporate criminal liability because of the collateral harm it causes to innocent people, which is categorically immoral. (I love how he resorts to the Soviet Union, Maoist China, and the German Nazis to support his arguments every chance he gets. You can't get any more immoral than genocidal communism and fascism!)

John bristled at my labeling him a formalist. But formalism is exactly what his argument boils down to. For, as I understand him correctly, if we simply relabeled criminal punishment and called it an "administrative sanction," he would have no problem leveling it at a corporation that has violated the law. (In his first post, he wrote, "Criminal law is penal law. It's about punishment. It is not designed to compose disputes, provide compensation to wronged parties, or impose administrative sanctions.") This is, however, a mere semantic or "formalistic" distinction. Administrative and criminal penalties are different only by degree. If administrative penalties are high enough, they will produce the same result as a criminal conviction: a loss by the company of a considerable amount of money. If so, they would have all of the beneficial effects associated with criminal punishment - retribution (though technically we wouldn't be "allowed" to call it that), deterrence, rehabilitation, and incapacitation - as well as the negative effect of harming some innocent shareholders, employees, and consumers. And threatening such penalties would encourage the corporation to cooperate against the individuals directly responsible for the wrongdoing.

It just so happens that, rather than creating a system of truly effective administrative remedies to counter white collar crime, our society has chosen instead to enact and enforce criminal statutes. One could argue about the wisdom and efficiency of this choice (it shifts regulatory responsibility and power from administrative specialists to generalist prosecutors), but the moral issue does not change. Only the words.

Perhaps John would argue that super-large administrative remedies would be as immoral as criminal sanctions because of the collateral damage they would do to innocents. This would, at least, be a consistent position to take. And it leads to this fundamental question: is it ever moral to harm innocents to achieve a great public good?

As much as we might not like it, the honest answer is "sometimes." War is a very patent example of when we make this painful calculus. Innocents - eighteen-year-old citizens whose otherwise routine and productive lives are interrupted by a draft - are sent to the front lines and sacrificed for the protection of the homeland. If we did not act this way, Nazis would now be running our government. (My turn!) Here's another difficult example. Let's say an innocent five year old, visiting his dad at the nuclear power station where he works, is about to push a button that will cause a meltdown and result in a million deaths. Shouting "stop" isn't working and no one is in reach of the kid. Nuclear disaster is imminent. An armed guard on the scene can stop the child only by shooting to kill. (Shooting to injure is too risky, especially at a small, moving target.) Should the guard fire?

There is no absolute "right" answer to this question. Some would say it is never right to kill an innocent, regardless of consequences - it is usurping the will of G-d. Others would say that to let a million persons die unnecessarily is the true immoral act. The co-called right answer lies only in consensus.

So, back to the prevention and punishment of white collar crime. Can we justify harming some innocents for the greater good of protecting the public from the havoc that would be wreaked by unchecked criminal behavior in the powerful corporate sector? One hundred years ago, the Supreme Court answered that question in the affirmative when it decided the New York Central case. It seems to me that society has reached the consensus that the Supreme Court was right in balancing the competing concerns.

If John took a step back, he'd realize that even he does not see these things in black and white. As he stated earlier, "[c]riminal punishment always wreaks harm on the innocent. The families and dependents of convicted criminals are inevitably adversely affected by the incarceration or impoverishment of the offender, both materially and emotionally. Hence, criminal liability always punishes the innocent . . . ." The difference between that harm and the harm of corporate criminal liability is, according to John, that in the former case it is "an unfortunate collateral effect of visiting punishment on the blameworthy that should be minimized as much as possible," while in the latter case latter case the "punishment is intentionally directed toward those who have not committed an offense."

This is not a legitimate distinction. First, in both cases it is not the government's intention to injure the blameless; rather, such injury is a foreseeable but unpreventable consequence of punishing those who are to blame. (Remember that John does not disagree with my contention that, without corporate criminal liability, many egregious offenders would escape detection and prosecution.) Moreover, even if one accepts this distinction, so what? Again, to use the war analogy, if the government enacts a draft, it is directly inflicting harm on those who are forced into battle against our enemies. We believe, however, that their "ultimate sacrifice" is justified by what is at stake. In fighting white collar crime, the stakes are less significant, but so is the harm to the innocents. And the government does take steps to keep this harm to a minimum.

Up to this point, John and I have been treating all corporate crime generically, debating in general terms whether corporate criminal liability can be justified in light of our values and its costs and benefits to society. I'm ready to move beyond that subject to take on the question of under what circumstances an alleged corporate criminal should be pursued.

Just when I thought our positions were moving closer together, Mike's last posting greatly widens the gulf. As the week comes to an end, it appears our conversation is more likely to end in discord than harmony.

Mike begins by getting down to fundamental philosophical starting points. He declares himself to be a pragmatist, with the right answer (or the "right" answer, as he expresses it) emerging out of slowing forming and ever changing consensus of opinion. I confess that I am not a pragmatist, either of the traditional James/Deweyian variety or the post-modern Rorty type. In the current idiom, I am not one who believes that it is turtles all the way down.

If I have to assign myself a philosophical label, I would call myself a legal realist of the Felix Cohen stripe. Like Cohen, I eschew the formalism of transcendental nonsense that reifies abstract legal concepts and attempts to deduce legal conclusions from such concepts rather than the actual effects that the decisions have on the world. And like Cohen, I am not an ethical relativist. Cohen wanted to peel away the legal word play and identify the empirical effects of our laws and legal decisions in order to make a proper ethical assessment of it. Like Cohen, I believe that if we allow ourselves to see what we are really doing, we have a better chance of doing the right thing, where there are no quote marks around the word 'right.'

Mike declares that legal decisions are based on policy considerations and empirical determinations. I do not agree. Law is a normative pursuit. It is a prescriptive enterprise. And legal argument, like all normative arguments, must have both a normative and an empirical premise. Getting the empirical or descriptive premise correct is crucially important. That's why it is essential to eliminate the legal transcendental nonsense. But it is not enough. There must be some normative value at work as well to reach a normative conclusion. This is basic Hume.

These philosophical differences may mean that Mike and I are unlikely to agree on fundamental issues. But they do not mean that there cannot be agreement on intermediary points. So let's consider some of the points Mike makes in his last posting.

To begin with, Mike somewhat overstates the position with which I began our dialog. I did begin with the contention that criminal liability is different in kind from civil liability. I did not assert that "a liberal government can under no circumstances impose corporate criminal liability because of the collateral harm it causes to innocent people, which is categorically immoral." I did assert that a liberal government should not employ a form of criminal liability that is intended to threaten the innocent with punishment to force them to help suppress the criminal activities of the their fellow citizens. I assert this because this form of criminal liability is utterly incompatible with the inherent liberal bias built into our criminal law that is designed to protect us against an over-reaching government-something that is always a greater threat than any number of individual criminals.

Mike seems a bit put out by my use of the Soviet, Maoist, and Nazi regimes in my examples. I apologize for this. But I cite those examples not because there is any equivalence between the evil of these regimes and our employment of corporate criminal liability, but because those are the regimes that employed the same principles of collective punishment and "rehabilitation" that are inherent in corporate criminal liability. I'd be glad to use the Inquisition in my examples more frequently, but the church did not employ such principles. The Inquisitors simply used torture. I use the examples to point out what makes a liberal regime different from a totalitarian one. In a liberal regime, citizens accept a less efficient police apparatus in return for a larger sphere for their personal liberty.

Mike accuses me of formalism again for insisting on the distinction between criminal and administrative sanctions. He then makes an odd argument claiming that there is no essential difference between them because "administrative and criminal penalties are different only by degree. If administrative penalties are high enough, they will produce the same result as a criminal conviction: a loss by the company of a considerable amount of money." The reason why I call this argument odd is because earlier in our exchange, Mike was at pains to distinguish criminal from civil and administrative penalties on the ground that only criminal penalties carry sufficient moral stigma to effectively deter criminal conduct. Remember his point about businesses simply regarding non-criminal sanctions as a cost of doing business? Now, either criminal penalties are or are not different in kind from administrative sanctions. If they are different because of the moral stigma associated with them, then we have something to talk about. If they are not different, then criminal sanctions are unnecessary and Mike has conceded the point about corporate criminal liability.

Having said this, let me concede that I believe Mike has a good substantive point is showing that administrative and corporate criminal sanctions are similar in effect. In opening our discussion on Monday, I distinguished between criminal liability on the one hand and civil liability and administrative sanction on the other. I did this because our assigned topic was the extent to which corporate criminal liability was justified and I wanted to keep the focus on that. However, if asked to address the more general topic, I would be willing to argue that market forces and civil liability are sufficient deterrents to fraudulent business practices, and hence, that administrative sanctions can be dispensed with along with corporate criminal liability. Further, I would argue that this would be a good thing for precisely the reason Mike points out-absent the moral stigma, such sanctions still function as a form of collective punishment. However, for those readers who do not accept this, and believe that administrative sanctions are necessary, Mike's last posting provides an argument demonstrating that corporate criminal liability is unnecessary.

Finally, Mike passes on to a more fundamental moral question, "is it ever moral to harm innocents to achieve a great public good?" Mike's answer is sometimes. I agree with him. He provides examples of cases such as war and the need to avert a nuclear disaster which illustrate this. Notice, however, how extreme these examples are. The question we are examining is whether the need to combat business crime is similar enough to these cases to justify harming the innocent. I am tempted to say that he answer must be: obviously not. Mike asks: "Can we justify harming some innocents for the greater good of protecting the public from the havoc that would be wreaked by unchecked criminal behavior in the powerful corporate sector?" But this is a non sequitur. For there is no threat of havoc being wreaked by unchecked criminal behavior in the powerful corporate sector. If we eliminate corporate criminal liability, business crime will not run wild. Prosecutors can still go after the culpable individuals directly. Corporations as collective entities will still be subject to the financial sanctions of both the market and civil liability judgments, and if you favor that sort of thing, administrative sanctions. If we eliminate corporate criminal liability we lose only one tool in the fight against business criminality, the illegitimate tool of collective punishment.

In stating that "In fighting white collar crime, the stakes are less significant, but so is the harm to the innocents," Mike greatly underestimates the harm of corporate criminal liability. And here I am not referring merely to examples like the losses suffered by the innocent partners at Arthur Andersen. As I think we previously established, obtaining convictions against corporations is not what corporate criminal liability is really about. It is about coercing corporate cooperation with law enforcement efforts. And the harm this does is great. It is too late in our exchange for me to go into detail about the untoward effects of placing employers and employees in adversarial roles. But if you want an inkling of what it is like, consider the lives that were ruined (and the Constitutional rights that were violated) by KPMG's behavior in attempting to "cooperate" with federal prosecutors sufficiently to be granted a DPA. Government-approved compliance programs are not ethics programs, and what the government considers cooperation is not consistent with what is called organizational procedural justice. There is presently over thirty years of organizational behavior research that demonstrates systems of command and control sanctions such as those DOJ mandates to avoid corporate indictment are less effective at reducing wrongdoing within business organizations than the trust-enhancing systems of procedural justice that DOJ considers evidence of an uncooperative corporation. And the reason for this is that the command and control systems tend to destroy the trust and alignment of interests and values between employee and firm that not only make businesses more successful financially, but also tend to produce ethical "corporate cultures."

Mike wants to discuss the circumstances under which an alleged corporate criminal should be pursued. The Holder/Thompson/McNulty/Filip Memorandum identifies the tactics DOJ will use in such a pursuit. That is enough for me to identify those circumstances with the null set.

It's taken us a week and thousands of words, but I think John and I have identified our fundamental differences. Our opposing positions are now crystal clear.

Let me start by clarifying a few things I said earlier that John seems to misapprehend. When I said that legal decisions are based on policy considerations and empirical determinations, I was not in any way suggesting that law is something other than a normative pursuit. That's what I meant by "policy considerations" - fashioning law in a manner that furthers a set of normative values. The only difference between John and me on this front is his certainty in the rectitude of particular normative propositions, and my reluctant but realistic acceptance of the view that, as William James said, "truth is what works."

John also appears to misunderstand my statement that "[a]dministrative and criminal penalties are different only by degree." My next sentence makes clear that I meant this only in the context of corporate criminal liability. For individuals, the difference is quite real - a criminal conviction brings with it the possibility of time in prison and a negative social stigma that may never be overcome. This stigma may impact the individual's family life, ability to get a job, mental status, and even whether he gets to vote in public elections. For corporations, however, prison is not an option and the stigma associated with conviction affects only the value of the good will attached to the corporate name. So, in the context of corporate criminal liability, a criminal conviction translates solely into a matter of dollars and cents.

I am not, however, conceding that such liability is unnecessary because it could, in theory, be replaced by a regime of aggressive administrative penalties and fines. I'm an academic, true, but I live in the real world. If we eliminated corporate criminal sanctions in the U.S., they would never be replaced by administrative remedies with real teeth in them because of the revolving door that exists between industry, on the one hand, and regulatory agencies and Congress, on the other. Federal prosecutors are wholly insulated from this symbiotic relationship; when pursuing a conviction they are not thinking about taking a job in the industry under investigation. So, for very pragmatic reasons, I believe that criminal prosecution of corporations is a must in the battle against white collar crime.

John reveals the fundamental difference between our world views when he states, "If we eliminate corporate criminal liability, business crime will not run wild." My reaction is this: What planet is he living on? The history of the past twenty years, starting with the S&L crisis, through the crackdown on healthcare fraud, then into and out of the Enron era, and now with the revelation of the rampant bank and mortgage fraud that nearly led to the ruination of our economy, I don't understand how he can make this argument with a straight face. Even with the existence of corporate criminal liability (which was not aggressively pursued by DOJ until the beginning of this century), white collar crime has run amok. If we've learned anything from the collapse of Bear Sterns et al., it's that market forces will assuredly do nothing to police bad actors with access to other people's money. I agree with the present Administration that more regulation is needed to prevent wrongdoing in the first place, but when that fails, the government needs every legitimate tool at its disposal to catch the wrongdoers and bring them to justice.

I am not advocating the pursuit of corporate criminal liability in every instance. The legal threshold for charging a corporation is quite low--the entity may be indicted for any crime committed by any of its agents whose activity fell within the scope of his employment and was undertaken, at least in part, for the benefit of the company. This means that, in theory, if one cashier at a national retail chain is purposely overcharging customers and putting at least part of the ill-gotten gain into the cash drawer, the national chain could be indicted for theft. This is, of course, ridiculous - and it doesn't happen. Instead, we have a system that relies heavily on prosecutor discretion; DOJ determines when pursuit of corporate liability is warranted.

This determination is based on a document referenced twice by John: the Holder/Thomson/McNulty/Filip Memorandum (named for the succession of Deputy Attorneys General who authored and revised it), found at Chapter 9-28 of the United States Attorney's Manual. The memo requires prosecutors to consider a number of factors when making the charging decision, including: (1) the pervasiveness of wrongdoing within the corporation, including the complicity in, or the condoning of, the wrongdoing by corporate management; (2) the corporation's history of similar misconduct, including prior criminal, civil, and regulatory enforcement actions against it; (3) the corporation's timely and voluntary disclosure of wrongdoing and its willingness to cooperate in the investigation of its agents; (4) the existence and effectiveness of the corporation's pre-existing compliance program; (5) 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; (6) collateral consequences, including whether there is disproportionate harm to shareholders, pension holders, employees, and others not proven personally culpable, as well as impact on the public arising from the prosecution; (7) the adequacy of the prosecution of individuals responsible for the corporation's malfeasance; and (8) the adequacy of remedies such as civil or regulatory enforcement actions.

The existence of these guidelines is evidence that, in the prosecution of corporate crime, DOJ is not a government agency run amok. Rather, the guidelines are carefully calibrated to ensure that charges are brought only when warranted by the magnitude of criminality within the organization and when the harm to innocent third parties does not outweigh the benefits of prosecution. They form the basis not only for decision-making at the level of the individual prosecutor, but at the supervisory level as well. Indeed, in a significant case, defense attorneys may get hearing on whether corporate prosecution is justified from the Deputy Attorney General him- or herself.

Of course, I cannot guarantee that DOJ's prosecutive power will never be abused. That's why we need the constant vigilance of individual defense attorneys, organizations like the National Association of Criminal Defense Lawyers and the National Organization of Corporate Counsel, the press, and the public. But the risk of a rogue prosecutor abusing his authority does not justify crippling the government's ability to fight white collar crime just when it is starting to get a handle on it.

Twice, John has referred to the Arthur Andersen case as a "Exhibit A" of government overreaching. I'm not surprised, because ever since Andersen's indictment on obstruction of justice charges in connection with the Enron debacle led to the accounting firm's dissolution, opponents of entity liability have used it as their rallying call. The protests got even louder when Andersen's conviction was overturned on appeal by the Supreme Court. Critics of DOJ claimed that the entire prosecution was ill-founded and that Andersen was "hanged" without cause.

This is simply not true. First, the collapse of the firm after indictment - obviously causing serious collateral damage to a variety of innocents - was a very unusual event. The best evidence of this is the huge number of corporations that have been criminally charged (or have settled such charges) over the years that have lived on to produce their widgets for another day. Andersen's situation was unique because, as a firm specializing in public accounting, it faced the loss of its ability to conduct public audits upon conviction. As a result, its indictment led clients to believe that the firm no longer had the credibility necessary to do its job - even if it was eventually exonerated - and so they jumped ship. This extreme reaction by a company's clientele is simply not present in the run-of-the-mill corporate case (and it was probably not anticipated by the Andersen prosecutors and defense lawyers who failed to broker a plea deal to avoid it).

In addition, Andersen suffered because it was a multiple recidivist: it had recently settled a number of administrative cases with the government in connection with prior claims of wrong-doing. The Enron debacle was the final straw; it was evidence that the firm had a deeply entrenched culture of misconduct that simply could not be ignored. Finally, the contention that the firm was exonerated on appeal is incorrect. The Supreme Court held that the trial judge's jury instructions on the criminal intent required for conviction were erroneous, and it remanded the case for a new trial. It did not declare the company innocent. Presumably, prosecutors did not retry the case because by that time the firm was more or less defunct.

On the other hand, the KPMG case is a legitimate example of government overreaching. In that case, prosecutors investigating illegal tax shelters pressured the accounting firm to cooperate early. As part of its cooperation, KPMG agreed not only to make its employees available to prosecutors for questioning, but also to sanction those employees who refused to comply. An employee's invocation of his Fifth Amendment privilege against self incrimination was considered lack of cooperation. Sanctions included the denial of previously promised attorneys' fees and termination of employment. Eventually, some of these employees were indicted.

Judge Lewis A. Kaplan found that the arrangement between prosecutors and KPMG violated the defendants' Fifth Amendment right to due process and their Sixth Amendment right to counsel of choice. In effect, the government and the firm had teamed up against individual employees in an illegal manner. Under Garrity v. New Jersey, there is no question that the government could not have used direct threats of employment termination to force KPMG employees to incriminate themselves. Accordingly, it should not have enlisted the aid of the company to do its bidding under the guise of cooperation. Chapter 9.28 of the USAM has been re-written to take this decision into account.

John points out, and I agree, that the net effect of corporate criminal liability when it results in cooperation is the deputation of corporate counsel and other corporate agents in the fight against white collar crime. Unlike me, however, he thinks this is a negative development. He alludes to the "untoward effects of placing employers and employees in adversarial roles, and claims that the "trust-enhancing systems of procedural justice" that corporations develop on their own are demonstratively better than the "command and control" compliance systems put in place as part of plea deals. Left alone, he says, firms will not only be "more successful financially" (which I don't doubt), but will also "tend to produce ethical 'corporate cultures.'"

Simply put, I don't buy it. Without doubt, aggressive corporate criminal enforcement, which leads in turn to the prospect of corporate cooperation, potentially pits employee against employee within the same organization. It is likely to make employees more distrustful of each other, of management, and of corporate counsel, and result in the workplace becoming a less pleasant environment in which to function. But is this really a bad thing? By many accounts, the present corporate environment is one in which fraudulent practices and other criminal activities abound. When an individual is caught committing a crime red-handed, his response often is that "everyone else was doing it, too." The "trust" relationships that exist are those in which employee A can count on employee B not to report his wrongdoing. Despite the personal hardships involved, shaking up the complacency of the corporate environment is in the public's best interest. We need to be more vigilant in our fight against white collar crime, not less.

Mike and my debate must come to an end at some point, and I think it only fair to give him the final word since I got to open the discussion on Monday. I want to thank Mike for an engaging conversation and the Manhattan Institute for hosting us. It has been a stimulating experience for me, and I hope one that may prove useful to the readers.




Rafael Mangual
Project Manager,
Legal Policy

Manhattan Institute

Published by the Manhattan Institute

The Manhattan Insitute's Center for Legal Policy.