FUNCTIONAL CORPORATE KNOWLEDGE.

• Author: Diamantis, Mihailis E.

TABLE OF CONTENTS INTRODUCTION 322 I. BACKGROUND TO KNOWLEDGE, CORPORATE AND CRIMINAL 334 II. CURRENT DOCTRINES FOR CORPORATE KNOWLEDGE 337 A Respondeat Superior 338 B. The Collective Knowledge Doctrine 344 III. EVALUATIVE FRAMEWORK 349 A Optimal Deterrence 352 B. Retribution 365 IV. A FUNCTIONAL ACCOUNT OF CORPORATE KNOWLEDGE 369 A. Insights from Collective Epistemology 371 B. Effort as a Relevant Variable 374 C. Obviousness as a Relevant Variable 376 D. Summary of the Account 378 V. EVALUATING THE FUNCTIONAL ACCOUNT 381 A. Optimal Deterrence 382 B. Retribution 389 CONCLUSION: BEYOND CRIMINAL DOCTRINE 392 INTRODUCTION

Suppose I hand you $1.61, $2.37, and $0.96. Do you have at least five dollars? If you are like most people, you do not yet know. With pen, paper, and a few seconds, you could add the numbers and find that you have only $4.94. You had the information you needed but did not at first know the answer.

Corporate law has yet to grapple with this intuitive distinction between information and knowledge. The incentives corporations have to implement the information-gathering compliance mechanisms they need to prevent misconduct turn on this distinction. So does the difference between a justly punished corporation and a casualty of prosecutorial or regulatory overreaching. As argued below, current doctrines either treat no information within the corporate structure as knowledge or risk treating it all as knowledge. Neither extreme serves the interests of corporate law or justice.

Knowledge plays an important role in corporate liability. (1) For simplicity, this Article focuses on criminal liability, though its arguments carry over with some modest modification to civil and regulatory contexts. Many of the most common white-collar crimes--things such as false claims, (2) mail and wire fraud, (3) securities fraud, (4)

money laundering, (5) and tax fraud (6)--require that the defendant acted knowingly. If corporate employees commit the proscribed acts with the relevant knowledge, that can spell trouble for them in their individual capacities. (7) Prosecutors wanted to find out which individuals knew about General Motor's faulty ignition switches (8) or Volkswagen's cheat devices (9) in order to bring those individuals to justice. But more than individual liability is at stake. Under long-established law, (10) the acts and knowledge of employees are the acts and knowledge of their corporate employers. (11) Corporations can be liable for the same crimes as individuals. (12) The prosecutors investigating individuals at General Motors and Volkswagen were also after the companies themselves. (13) The stakes for getting knowledge right in corporate criminal law are high. On the one hand, an overly restrictive definition will make it harder to prosecute corporate misconduct. (14) Some estimates already put the annual costs associated with white-collar crime in the United States at around half a trillion dollars (15) (just shy of Sweden's GDP), (16) which is twenty times the total economic costs associated with every other sort of crime in the United States. (17) On the other hand, an overly permissive definition of corporate knowledge risks choking the corporate engines of our economy (18) and discouraging corporations from taking advantage of socially beneficial economic opportunities. (19) In addition, under a broad definition of corporate knowledge, corporations would likely make wasteful investments in excessive compliance. (20) Corporate compliance costs already rival total municipal policing costs. (21) JPMorgan alone has three times more compliance officers than the FBI has agents. (22)

Corporations worried about criminal liability have a Janus-faced stake in knowledge. From one perspective, knowledge can be power and security. (23) Regular audits of employee performance can help corporations streamline operations, increase efficiencies, and boost profitability. (24) Audits also generate information critical for designing and implementing programs to prevent employees--and hence their corporate employers--from committing crimes and incurring criminal liabilities. (25) From another perspective, knowledge can be a great vulnerability. The line between criminal and innocent conduct frequently turns on what defendants knew. (26) As a result, monitoring employees can sometimes make corporations worse off. (27) The same internal corporate compliance and audit programs that gather information that is crucial to help corporations prevent crime can also produce knowledge that converts otherwise permissible corporate conduct into crime. (28)

Whereas corporations have an ambivalent relationship with knowledge, the criminal justice system's interest in corporate knowledge is one-sided: the more corporations know, the better off the criminal justice system tends to be. Congress, (29) prosecutors, (30) and judges (31) are all committed to reforming corporations and promoting effective corporate compliance programs. A large part of what this means is ensuring that corporations have adequate procedures for generating and transmitting information through appropriate channels to personnel who can respond when problems are uncovered. (32) The more corporations know about themselves, the better able they are to anticipate and prevent misconduct. Should something still go awry, as it inevitably will, (33) corporate knowledge is crucial to ensuring accountability. Prosecutors are in a better position to bring charges against corporations that acted knowingly.

In light of the criminal justice system's strong interests in corporate knowledge, and corporations' mixed incentives to know things, we might expect criminal law to have nuanced doctrines about corporate knowledge. It does not. (34) Most circuits (35) simply apply the centuries-old, mechanistic doctrine familiar from civil law (36)--respondeat superior--and thereby treat corporate knowledge just like any other corporate mental state. (37) That doctrine allows prosecutors to attribute mens rea from employees to their corporate employers but requires prosecutors to find one single employee with all the knowledge needed for conviction. (38) Whatever merits respondeat superior has with respect to some mental states, (39) it performs poorly when it comes to others--such as knowledge--that can be dispersed over many people. (40) Corporations can manipulate such a mental state by partitioning it across employees so that no one employee has it in its entirety. Today's corporate behemoths, characterized by complex operations that require a diffusion of responsibility and authority, do not even have to try to spread knowledge thinly. (41) If anything, they must fight the entropic dispersion of information that is a natural product of large-scale operations. (42) Despite these conflicting incentives, respondeat superior decidedly pushes corporations away from allowing employees to acquire too much knowledge. The more individual employees know, the more likely it is that the corporation can be prosecuted when bad things happen.

Some jurisdictions have supplemented respondeat superior with the collective knowledge doctrine. This ham-handed approach frees prosecutors from the need to find one employee with all of the culpable knowledge by allowing them to aggregate the knowledge of all corporate employees for attribution to the corporation. (43) In so doing, the collective knowledge doctrine flips corporate incentives concerning knowledge on their head. If corporations are on the hook for every scrap of information known by every employee, they must have mechanisms to gather and process it all. While this may be a win for prosecutors--because corporations are more likely to have inculpating knowledge when they misbehave--these mechanisms come at a great cost. As one court put it, "[T]he imputation of every bit of knowledge known to each individual employee--from the Chief Executive Officer to the most recently hired recruit--would likely paralyze a corporation." (44) The marginal social costs for corporations to collect and process increasing quantities of information must eventually exceed the marginal social benefits from crime prevention. (45) We may approach that point soon, if we have not already in some industries, and social waste lies beyond it. (46)

The heart of the problem is the law's failure to recognize and grapple with the distinction between knowledge and information. Under respondeat superior, corporations have a strong incentive to turn knowledge into mere information dispersed through the corporation and to prevent dispersed information from becoming centralized knowledge. (47) Under the collective knowledge doctrine, corporations have strong incentives to implement inefficiently complex compliance systems that convert all information into knowledge by centralizing and processing it. (48) On the doctrinal line between knowledge and mere information rests the incentives criminal law gives corporations to scale their audit and compliance programs up or down. If the law draws the line under- or overinclusively, it will under- or over-incentivize corporations to collect and process information. Without a considered approach, criminal law is missing an opportunity to fine-tune the incentives corporations have to balance risk, compliance, and cost. (49)

More than incentive setting hangs in the balance. The criminal justice system is also a system of justice. Corporations must face the punishment they deserve, and the rule of law (50) and due process (51) demand that they do so on terms antecedently set in statute by the legislature. To convict corporations justly of knowledge-based crimes, courts must have an adequate theory of corporate knowledge. Characterizing corporate knowledge too narrowly, as respondeat superior does, will fail to hold corporations accountable when they engage in clear criminal wrongdoing. (52) This is unfair to the victims of that...