Since 1999, the U.S. Department of Justice (“DOJ”) has had a formal policy detailing the criteria its lawyers will use for deciding whether to prosecute corporations for federal crimes.1 Three U.S. Deputy Attorneys General (Eric Holder, Larry Thompson, and Paul McNulty) in two administrations have authored and lent their names to different versions of the policy. The content of the policy has, however, largely remained the same. So have the criticisms of the policy from the corporate bar as well as some academics and members of Congress. These criticisms largely miss the mark—despite their constant repetition—for reasons that defenders of the DOJ policy have to date not clearly articulated. The critics seek to lay at the feet of the DOJ policy problems whose primary causes lie elsewhere, in places the critics may be reluctant to have us look. Thus, abolishing the objectionable parts of the policy, as Senator Arlen Specter’s recent bill seeks to do,2 will not likely have much effect. Taking seriously the problems raised by the critics will require more drastic change than they (or anyone) may be willing to undertake. My aim in this essay is not so much to defend the DOJ policy as to deflate the dominant criticisms and to refocus the debate.
Deputy Attorney General
Paul J. McNulty
The critics’ main target has been the DOJ’s consistent position that, in evaluating the degree of a corporation’s “cooperation” in the government’s investigation of corporate criminal activity, the DOJ may consider whether the corporation is willing to waive its attorney-client privilege and work product protection for information related to the alleged criminal activity. Although the stated conditions under which the government will take waiver into account have changed, the basic thrust of the policy has not. The government’s stated rationale for the waiver policy is that the disclosure of otherwise privileged information may sometimes be necessary for the corporation to effectively demonstrate its lack of involvement in criminal activity, or its change in behavior, or its willingness to assist in the prosecution of individual wrongdoers within the organization.
The principal criticism of the waiver policy stems from the premise that the government has sufficient “leverage” to “coerce” a suspect corporation into waiving its (meaning the entity’s) attorney-client privilege and work product protection whenever the government wants, even if the corporation (meaning its current management) does not “want” to.3 From this premise, the critics argue that the government has used its leverage essentially to treat waiver as a requirement for cooperation, thereby effectively abolishing (or significantly undermining) the corporate attorney-client privilege and work product doctrine in the criminal context. At this point in the critics’ argument, we get the obligatory citation to Upjohn Co. v. United States and its rationales that the corporate attorney-client privilege beneficially fosters communications from corporate employees (even those outside the “control group”) to corporate counsel and vice versa.4
We will return to Upjohn shortly. Let us start by examining the critics’ initial premise. All acknowledge that the government does not seek directly to compel corporations to supply information or documents, for example by using the subpoena power at issue in Upjohn. Rather, the alleged coercion—the (explicit or implicit) threat to prosecute the corporation if it does not sufficiently “cooperate”—is indirect. The difference between direct and indirect compulsion is important here. The government cannot compel the production of privileged information. If it wants that information, the government must bargain for it.
Moreover, none of the critics argues that corporations should be prohibited from waiving their privilege to the government so long as the waiver is “voluntary.” Before the DOJ ever adopted its policy, corporations under threat of prosecution waived privileges with some regularity and without major uproar.5 And presumably corporations, being in a bargaining position with the government, were not waiving without getting something in exchange, namely some kind of assurance of more lenient treatment, including the avoidance of criminal prosecution. Thus, it is hard to see how the mere threat of criminal prosecution against corporations by itself makes waiver “coercive,” or not “voluntary.”
The fact that the parties are in a bargaining relationship does not, of course, preclude the possibility that one side has substantial, perhaps overwhelming, “bargaining power.” But the law has always hesitated to conclude that the mere possession and exercise of bargaining power constitutes illegitimate “coercion” that invalidates the resulting bargain. Even in the case of unsophisticated, individual criminal defendants, the law generally recognizes and enforces plea bargains as “voluntary” contracts, despite the often significant bargaining advantage possessed by the government.6 To do otherwise would overwhelm the criminal justice system and would deprive both parties of something they prefer to the status quo, making both worse off.
It would surely be odd to argue that corporate defendants deserve more protection against government bargaining power than individual defendants, especially since individual defendants entering into plea bargains are waiving constitutional rights, and not a statutorily created privilege. Yet the crux of the critics’ argument is that the government has too much “leverage,” leaving corporations no real choice but to give away any claim of privilege.7 This argument is implausible. Unlike many individual criminal defendants, large corporations have at their disposal top-notch legal help and substantial resources. They are perfectly capable of assessing the risks and weighing the costs and benefits of waiver. And they have long demonstrated ample ability to resist the government and vigorously defend themselves in court against allegations of criminal wrongdoing.8
The critics try to rescue their government leverage argument by contending that circumstances have changed, so that corporations can no longer “just say no” or otherwise tell the government what it can do with its waiver request. The critics’ story is that in the good old days, the government, respectful of the privilege, did not run around demanding privilege waivers all over the place. Now, however, the DOJ’s policy has created a “culture of waiver.”9 Although it is not at all clear that this story is correct—the evidence is largely anecdotal along with surveys of self-interested corporate counsel—let us suppose the number of waivers has increased and that this increase alone is cause for concern. Even so, the suggestion that the DOJ waiver policy is itself the sole or primary cause of any increase in waivers is flawed.
What, then, might have caused a numerical increase in waivers, assuming there has been such an increase? There are many possible reasons. Some commentators have pointed to changes in criminal law, such as an ever-expanding number of white collar crimes (leading some to suggest that at any given time, most corporations are guilty of something) and an increasingly sweeping law of criminal vicarious liability.10 If these changes mean that the government’s likelihood of prevailing against a corporation is higher (on average) than previously, or its cost of proving a violation is lower (on average), then the government might rationally exact, or even simply expect, a higher “price” of cooperation, such as more waivers. Aside from changes in the law, changes in factual circumstances could also lead to more waivers. The recent “wave” of scandals, including Enron and other large, well-known, public corporations, began only two years after the original Holder Memorandum. The government again might rationally demand more from corporations seeking to cooperate if, compared to earlier times, any of the following are true: more corporations are committing crimes (perhaps because of increased competitive pressures); the crimes corporations are committing are more likely to be higher stakes offenses; corporate crimes are more likely to involve the complicity or reckless ignorance of upper-level management than rogue lower-level employees; or corporate crimes are more likely to occur in industries, such as the accounting profession (Arthur Andersen and more recently KPMG), where indictment effectively means the death of the business. Another source of government leverage could be a change in the political climate. In response to the corporate scandals, President Bush formed a Corporate Task Force, Congress passed the Sarbanes-Oxley Act, and the press and public focused more attention on corporate scandals and demanded a stronger government response. Political changes like these could also contribute to the DOJ being more aggressive about prosecuting corporations, and consequently demanding more from corporations to demonstrate cooperation. Finally, as a result of the recent scandals corporations might adjust their internal calculus about fighting the government versus cooperating in order to avoid bad publicity and plunging share prices.
Any number of these (and other) factors could increase the “price” of corporate cooperation, enabling the government to exact more from corporations in exchange for leniency. The point is these factors are completely independent of any waiver policy the government might have. The changed conditions would lead corporations to waive more often “voluntarily” regardless of what the government’s policy was. Suppose, for example, the government decided to double the DOJ’s budget for fighting corporate crime. In that case, even if the DOJ had never adopted any waiver policy at all, corporations might feel more pressure to waive privileges than they had previously. There is simply no reason to believe that whatever increase in waivers has occurred is the result of the government’s adoption of a waiver policy as opposed to these other factors. This understanding may help to reconcile the apparent discrepancy in reporting on the actual implementation of the government’s waiver policy between government attorneys, who contend that they rarely ask for waiver,11 and corporate attorneys, who contend that government waiver demands are now routine.12 Any pressure to waive that corporate lawyers feel, therefore, exists independently of the express words that the government lawyers use, because the specific waiver policy is not the primary cause driving whatever increase in waivers exists.
Of course, change in government bargaining power is not a one-way street. Circumstances could easily develop (and in fact may already have) in ways that decrease the number of waivers. For example, as the corporate scandals have played themselves out, corporations may have started to sense that the government does not really have the stomach to carry through on its threats to indict them, perhaps because of changing political attitudes or backlash against perceived government overreaching. As a result, corporations might once again feel emboldened to resist. The ferocity of the corporate reaction to the government policy may itself be an example of renewed corporate strength and receding government power. Moreover, if the government ever started to bring cases in which it lacked sufficient evidence of significant corporate wrongdoing or pursued overly aggressive legal theories, its threats to prosecute corporations upon noncooperation would become less credible. Critics of the government’s policy tend to ignore or downplay these factors that could dampen whatever pressure to waive might otherwise exist.
Moreover, the debate over “selective waiver” also tends to support the view that the DOJ’s bargaining power may in fact be weaker than the policy’s critics presume. The traditional rule, still followed by most courts, is that waiver with respect to one party waives with respect to the world. In the context of corporate wrongdoing, the traditional waiver rule means that if the corporation waives its privilege in dealing with the DOJ, private plaintiffs in subsequent civil suits will be able to obtain the information, thereby increasing the burden of the DOJ waiver policy on corporations.13 In response, some courts and lawmakers have proposed a doctrine of selective waiver, under which a corporation would be able to waive its privilege with respect to the government and not lose its privilege in subsequent suits.14 The main rationale for selective waiver is that it would lower the cost to corporations of waiver to the government and thereby encourage corporations to do it. That rationale, however, is inconsistent with the critics’ core premise, specifically, that the government’s policy coerces corporations into waiving. If the government really has sufficient leverage to coerce waiver routinely, the government should not need selective waiver. Yet the government has generally supported the proposal. Corporations, by contrast, originally supported the selective waiver doctrine but more recently have soured on it. Their stated concern is that selective waiver will just encourage the government to seek waiver more often. The reality behind corporations’ collective change of heart may be instead that, as a result of “class action reform,” the consequences of private litigation are less drastic than they used to be. Whatever the explanation, it is difficult to square the corporations’ changing position on selective waiver with their contention that the government’s policy currently coerces waiver.
Although we have disposed of the critics’ main objection to the government’s waiver policy, there is one last objection to consider. Again, let us suppose that the critics are correct that the number of waivers has in fact increased, whether this increase is due to the government’s waiver policy or not. Would such an increase be any cause for concern? The critics’ focus here shifts from the coercion of corporations to the harmful effects of waiver on the corporation’s employees. The two most noted effects are that employees will have fewer incentives to cooperate in corporate internal investigations and that the corporate employer’s protection of its employees’ rights will diminish.15 The critics significantly overestimate the size and likelihood of these effects, which are related in an important way that the critics underestimate or overlook entirely.
First, any increase in waiver of the corporate privilege will not likely have much effect on the employees’ incentive to cooperate. The argument to the contrary comes from the Court’s decision in Upjohn to extend the protections of the corporate privilege beyond lawyer communications with managers in the “control group.”16 The Court’s argument was (and remains) weak, however, for reasons directly connected to the great waiver debate. The main motivation for employees to cooperate with corporate investigations has always been the threat of being fired or incurring other job-related consequences, not the corporate privilege. The corporate attorney-client privilege simply does not protect the corporation’s employees, except for the executives who control the privilege. The reason is that the corporation (meaning the executives just mentioned) has always had the right to waive the privilege if it was in the corporation’s interest to do so, over the objection of employees and regardless of the effect on them.17 Long before the government’s waiver policy was ever dreamed up, corporations were exercising this right to waive the privilege and put the blame for illicit conduct (rightly or wrongly) on their employees.
In any case, whatever protection the corporate attorney-client privilege provides is relatively weak. Apart from waiver by current corporate management, there are many ways the government or private plaintiffs can overcome the corporate attorney-client privilege. Waiver can occur if new management comes in, if the corporation files for bankruptcy and a court appoints a bankruptcy trustee,18 or if the corporation makes disclosures to “unnecessary” third parties, including D&O insurers, auditors, creditors, analysts, or even shareholders.19 Communications that have a business rather than a legal purpose are not protected by the privilege at all.20 Neither are facts, which employees can be compelled to disclose to the government absent a Fifth Amendment privilege or other protection. The crime-fraud exception may apply to communications that predate or are contemporaneous with the criminal activity.21 Shareholders who file derivative suits may be able to get access to privileged material.22 In light of these limitations, employees would be foolish to put much stock in the hope that the corporation’s attorney-client privilege will protect them.
If employees do not understand these limitations, the fault lies with the corporate bar, not the government’s waiver policy. Corporate lawyers owe no ethical obligation to corporate employees to protect them by warning them of the limitations on the corporate privilege, let alone the corporation’s waiver policy or the likelihood that the corporation would waive the privilege (in fact, it is hard to imagine a competent corporate lawyer risking an estimate of that sort in communicating to employees).23 Even if corporate lawyers do warn employees about waiver at all (which of course, makes it less likely that employees will be fully forthcoming), lawyers do not generally claim that they warn employees of any of the other ways the privilege could be overcome, or the likelihood of these events. To the extent that lawyers are in fact warning employees more about waiver than they used to in light of the government policy (again something on which evidence is lacking), any resulting decrease in cooperation is likely smaller than what would occur if lawyers had been providing full information about the privilege to employees all along.
For the employee incentive argument to have any plausibility, employees would have to believe that their corporate employers will look out for their interests and not waive the privilege and hang them out to dry when the government comes knocking. But corporations have always hung errant employees out to dry. And that is a good thing. The plain fact is that corporate employers generally owe no loyalty to their employees, at least as a legal matter. Nor should corporations owe their employees any loyalty if the employees have themselves shown disloyalty to the corporate entity by engaging in criminal activity.
Indeed, it is the very lack of corporate loyalty that gives the corporate employer the ability to bargain with the government for cooperation. The ability to help the government nail suspected employee “bad guys,” which includes the ability to waive the privilege, is precisely the “cooperation” that the corporation has to “sell” to the government. This ability existed long before the government decided to adopt a waiver policy. Moreover, to the extent that whatever corporate loyalty to employees exists as a matter of custom and culture has declined over time (a decline in loyalty shared by employees ready to jump to a competitor at a moment’s notice), that decline could be another factor that lowers the corporation’s cost of cooperating and so makes corporations more willing to waive. And again, a corporation would have to decide whether to exercise its right to “sell out” its employees to the government regardless of what the government’s waiver policy is or whether the government had ever adopted any waiver policy at all. Thus, the professed concern over employees’ “rights” is a smokescreen, and, when voiced by corporations and their lawyers, a disingenuous one. Corporations want to preserve the right to sell out their employees; they just want to do it on their own terms, not the government’s.
The vehemence of corporate opposition to the government’s waiver policy may be a matter of whose ox is being gored. It is one thing when the corporation on its own wants to finger some low-level employee and label him a “bad egg” acting contrary to company policy so that the corporation can avoid prosecution. It is quite another when internal investigations turn up evidence of misbehavior at the highest levels and diffused throughout the organization. But that is what the recent corporate scandals are all about. The “coercion” that corporations claim to suffer may in fact be the discomfort that upper-level executives feel when they have to choose between waiving the privilege for the good of the company and saving their own necks. If so, then criticism that has been dressed up as a noble stand in defending a venerable privilege against government abuse is in reality just the corporate bar’s age-old attempt to protect upper-level corporate management rather than the entity client that corporate lawyers are supposed to serve. That would not be a surprise. The surprise is that people have been taken in for so long.
Preferred citation: George M. Cohen, Of Coerced Waiver, Government Leverage, and Corporate Loyalty: The Holder, Thompson, and McNulty Memos and Their Critics, 93 Va. L. Rev. In Brief 153 (2007), http://www.virginialawreview.org/inbrief/2007/07/23/cohen.pdf.
[Back to this issue]