Posted 5:32 PM by Robert

Some Andersen Considerations

The Man Without Qualities is always pleased to hear from Max Power, if only because his comments and criticisms are always so thoughtful and erudite.

In this case, Max seems to side at least provisionally with Judge Harmon's key ruling in the Andersen trial that the jury didn't have to settle on a particular "corrupt agent". But I don’t think Max understands how bad and dangerous that ruling really is. Automatic “guilty agent” criminal liability amounts to nothing more than the irrefutable assumption that the employer authorized the agent’s actions - thereby relieving the prosecution of any need to show "authorization". The same principle underlies the respondiat superior doctrine in civil law - a doctine that imputes civil liability to the employor for employee acts committed in the course of employment.

But suppose the employer is an individual and the prosecutors are able to prove four agents had all been expressly authorized to commit a crime for the benefit of the employer - say a man authorizes four “hit men” to kill his wife, who is then murdered. Suppose the employer and all four hit men are made codefendants in the murder trial and each member of the jury settles on his or her own “hit man” as having committed the murder – but the jury cannot agree that one of them did it. By assumption there is no question of “authorization” – so the situation is now the same as the Andersen case: if the prosecutor can show one of the agents committed the murder, then the employer-husband is guilty, too. The reasoning of Judge Harmon’s ruling would allow the conviction and execution of the individual employer where each “hit man” gets the benefit of a hung jury. Does that sound right? It’s by no means the most peculiar consequence of Judge Harmon’s thinking. Judge Harmon was right to call her ruling “terrifying.” Of course, it is no answer to say that Judge Harmon's ruling was limited to corporate employees, simply because the principles of criminal law she is employing are much more general than that. In fact, the ongoing and dangerous "backwash" from corporate crime to individual crime has been noted by legal scholars with increasing alarm - and Judge Harmon's ruling should be a wake up call:

"[F]irst, the general principles of criminal liability have become much more stringent so as to ensure that unscrupulous corporations are convicted; second, constitutional rights generally have been diluted to ensure that corporations are convicted. Both of these consequences in turn affect individual criminal liability because of the ... equation for legal purposes of corporations and individuals. That is to say that the pressure to convict corporations has a discernible impact on individuals, both in terms of the principles of criminal liability under which they are convicted and the strength of individual constitutional rights."

It is especially strange that Judge Harmon felt the need to effect her sweeping ex post facto expansion of corporate criminal liability in the face of considerable scholarly opinion to the effect that corporate criminal liability is already overbroad and has pernicious social effects. [See also: V.S. Khanna, “Corporate Criminal Liability: What Purpose Does it Serve?” (1996), 109 Harvard Law Review 1477 (arguing that corporate civil liability can achieve all of the goals of corporate criminal liability, without the attendant procedural protections and stigmatic effects of the latter)]. More of an already bad thing is not something judges are hired to produce.

Max also says that the Man Without Qualities believes that the Andersen attorneys should have anticipated this issue in the original jury instructions “and caught the prosecution unaware”.

But I didn’t suggest the government would or should have been caught unaware. Rather, I wrote:

"[I]f the Andersen attorney[s] had fixed the problem by moving to amend the jury instruction before the jury started deliberating, they probably would have succeeded, since neither the judge nor the prosecutors would have been motivated to insist on a likely erroneous construction of the law. Indeed, at that point, the prosecutors thought they had an easy case and would probably have favored a correct construction of the law, since the incorrect construction on which they are now forced to rely - and the judge has now adopted - invites reversal."

So - knowing full well what Andersen was doing - the prosecutors would likely have accepted revisions to the jury instructions that did not rely on what Judge Harmon has now characterized as “terrifying” new legal grounds. But once it became clear to the prosecutors and the judge that their choice was either a hung jury or reliance on “terrifying” new legal grounds, they dug in their heels. The issue is not esoteric and was not hard to see coming, despite the patronizing statements from the attorneys about how the jury should be working for law professors and the like. In fact, before the jury began deliberations, the Man Without Qualities noted that the jury instructions needed clarification in exactly this fashion. Contrary to Max's belief, in all likelihood, no fancy caselaw would have been necessary to secure the prosecution’s assent early on, simply because the prosecution should have wanted to forestall the possibility that following an easy conviction Andersen might cite to the ambiguity in the jury instructions as reason for reversal. Max may not agree with me, but he should first correctly represent and address what I wrote.

Max also disagrees with the assertion made here in a prior post that Duncan's attorney's advice has been demonstrated to be facially bad. Max says that Duncan was “faced with a classic prisoner's dilemma, and one with three co-defendants no less.”

Max is clever and thoughtful, as always, but quite wrong in this case. As the first sentence of his own cite demonstrates: “Tanya and Cinque have been arrested for robbing the Hibernia Savings Bank and placed in separate isolation cells.” There’s a reason the sentence ends that way: The classic prisoner’s dilemma relies on the two prisoners not being able to communicate, agree or cooperate. Obviously, that was not a feature of the Duncan negotiations, the tip-off being that none of the other three highly rational agents cooperated or (as far as we know) attempted to cooperate with the prosecution – a result quite unlikely in Max’s model, but not at all surprising given the real situation. If Duncan had joined with the other agents and Andersen in a common defense, Duncan would be vastly better off today – just as the other three agents are now. The other three agents and Andersen all have lawyers who understood that. Perhaps Max thinks the other three agents and Andersen should all fire their lawyers – since that would follow from the lawyers not having recognized the “classic prisoner’s dilemma” and advise their clients to cooperate with the prosecution. But somehow I doubt that the other three agents are regretting not racing Duncan to the prosecutor’s office right now.

No, from a technical game theory perspective the Duncan negotiations were anything but a “classic prisoner’s dilemma.” In fact, they are a fair example of what is known as an “asynchronous’ game – one quite ruled out by the “classic prisoner’s dilemma.” As Max’s cite states it , for example:

“It has often been argued that rational self-interested players can obtain the cooperative outcome by making their moves conditional on the moves of the other player. [One expert] favors a strategy of reciprocal cooperation: if the other player would cooperate if you cooperate and would defect if you don’t, then cooperate, but otherwise defect. Conditional strategies like this are ruled out in the versions of the game described above, but they may be possible in versions that more accurately model real world situations.”

Of course, if the prisoners are not kept in isolation many other forms of cooperation among them are possible which short circuit the effects of the “classic prisoners' dilemma.”