Sacrificing the Client to Save the Innocent Man

A few of us are debating the merits of the Alaska and Massachusetts versions of Rule 1.6 over at another blog called Public Square.  Have a look.

Comments

[john steele]

Andy,

What do we know empirically? Do any lawyers actually avail themselves of the option? Under what conditions? Are any (some? man? most?) criminal defendants aware that their lawyers have this option?

John

[Andrew Perlman]

John,

There is an important empirical question here. Massachusetts has had this exception for a number of years, and I am not aware of any lawyer's use of it. And to the extent that it has occurred, I have not heard reports from any criminal defense attorneys that defendants are reluctant to talk with them because of the existence of this exception.

Of course, lawyers have used a related exception involving death or substantial bodily harm, such as in the well-known Purcell case, which arose in Massachusetts. I have not heard that defendants have been any less willing to talk with their attorneys as a result of Jeff Purcell's unusual disclosure in that case. Given that some lawyers have made disclosures similar to Purcell's in other states and given that I have not heard of any reports regarding a consequent drop off in client communications, I am skeptical of the claim that an exception of the sort that exists in Alaska and Massachusetts would have any impact on the attorney-client relationship.

[Alice Woolley]

Andy and John,

Although I availed myself of the empirical point on confidentiality, Andy's objection to it is well taken. I know there are studies on the relationship between the exception and a client's willingness to talk, but I tend to like David Luban's non-empiricist justification for the exception.

For me the more interesting question here is with respect to the role of regulatory rules/codes of conduct. Specifically, rules which give broad, unfettered discretion with no guidance as to the basis for the exercise of discretion seem to have two (directionally incompatible) issues. On the one hand, a discretion with no guidance seems to create the possibility for bad exercise of it, which I would identify here as disclosing when disclosure has too high a price for the client whose confidences are being breached. On the other hand, when a broad discretion is given it seems to have extraordinarily little regulatory heft - there is no ability, with a rule like this, to say that a lawyer who fails to disclose has violated a rule which we, as a constitutional democracy, would have liked him to follow. My preference would be for a rule which was tailored more narrowly to avoid these two issues - giving more clarity around when it would be a bad idea to disclose, and more regulatory heft to discipline a lawyer who exercises the discretion wrongly.

The other issue with highly discretionary rules is that I think they tend to operate on the margins of the "real" rules, which are reflected (in this instance) by the actual practice of the defence bar, which is not to disclose confidential information, even where doing so would prevent the incarceration of a third party. That may not be a bad thing substantively, but I would rather the written rules aligned with the actual rules, either by not writing written rules which no one pays attention to, or writing rules which lead to a modification of actual practice.

Finally, I think that this is an instance where you have a problem of regulatory gaps, particularly in the relationship between confidentiality and privilege. The problem with a code of conduct is it has no power to, for example, change how the broader legal system treats information disclosed in a situation like this. It can respond to changes in the legal system, but it has a difficult time leading them, both because without the other changes the rule might be undesirable (which I think is the case here) and because, relatively speaking, the rules on lawyer conduct I don't think are especially weighty in the legal system as a whole (that's impressionistic, and I could be wrong about that).

How does this relate to the actual Rule 1.6 issue? I am not as comforted as Andy is by the discretionary aspect of it, nor by the fact that lawyers aren't actually disclosing information under the exception. I think it is a not especially well thought out piece of "statutory" drafting, giving lawyers too much discretion and with a limited likelihood of meaningful enforcement, one way or another. This perhaps reflects my overarching concern with legal ethics as a matter of professional regulation, and my dislike of "ought" or "can" regulation.

Alice

[john steele]

Thanks. The existence of an exception within a jurisdiction, coupled with the obvious empirical fact that the sky hasn't fallen in that jurisdiction, is often offered as proof that there's no downside to enacting the exception. I'm always curious to know if that is because (i) invoking the exception simply doesn't result in the parade of horribles that were predicted by the opponents of the exception, or (ii) the exception sits on the books but is virtually never invoked, so there is no upside or downside to enacting the exception--it's really all about expressive politics or about press relations for the lawyers, neither of which are utterly compelling to me.

I'm aware of Luban's non-consequentialist argument. But opponents of exceptions claim that overbroad exceptions cause real human suffering, and I'd like to get a handle on whether that's true. Neither side, alas, seems to have clear evidence on the point. In the case that we'll have to suffer through perpetual ignorance on that point, I'd favor exceptions that save human life. But if the proposal is to permit a criminal defense lawyer to reveal her client's confidence not to save a life but rather to free someone else from prison, and the revelation is against client wishes and detrimental to the client, I remain unconvinced (and in California, if that matters).

[Alice Woolley]

John,

How hard core does one need to be in thinking about these things empirically? I think if something is empirically falsifiable (e.g., character predicts moral conduct) then regulating on that basis is problematic. But if something is extraordinarily difficult to measure empirically, and there's a kind of logic to a position, how focused do we want to be on empirics?

I'm thinking of the two "imagine you" scenarios Andy and I put forward - that you are an innocent in jail or that you are a person scared about your criminal conduct and seeking the advice of a lawyer. Can one use that kind of analysis and nonetheless think meaningfully about what an exception like this should look like and, in particular, what norms it is that one as a society wants to adopt?

Simply put, the innocent convict is going to be fairly peeved (to put it mildly) to discover that a person out there knew of his innocence and could have shown it. I think without empirical evidence we can still be confident of that. And we (used loosely) as a society will likely also think it is a Bad Thing for people known to be innocent to be in jail. At the same time, I think we can also posit that a client who finds out that his lawyer spilled the beans is going to feel betrayed (it is harder to know whether that would prevent disclosure ex ante). And again I don't think a lot of empirical evidence is needed to believe that to be true. And also I think a society might not want people who seek legal advice when they do bad things to end up betrayed by their lawyers.

So then the question is simply this: what do we value more? Which principles and norms are more important to us, given that in this instance these conflict? And of course this is as often as not the basic dilemma around lawyer ethics, and one which may have been resolved whenever it was that the standard conception became standard, and confidentiality became the norm. Although even if it has been, that still leaves lots of room for debate around the margins about which norms should be giving in a particular instance, and to what extent.

But this also means that the debate around R1.6 is fundamentally normative, not empirical. I, like you (it seems), tend to value the ability of people to seek legal advice with only the most minimal risk of betrayal of their confidences by their lawyers. Can I pretend away the other norms this violates? No, but I'm not sure the empirical point sways me one way or another.

Alice

[Monroe Freedman]

With regard to empirical work on the importance of confidentiality, consider the following:

Conventional wisdom is often valid. We have quoted the conventional wisdom on the importance of confidentiality in the preceding section–the consensus of judges and lawyers that confidentiality is essential to ensure candid disclosure of embarrassing and potentially harmful truths from clients. This consensus is based upon the most extensive kind of empirical evidence–the day-by-day, year-by-year, century-by-century experience of judges and lawyers. It is not readily susceptible, however, to scientific empirical research. How would one determine scientifically how many clients in fact gave sensitive information to their lawyers which they would not have given but for assurance of confidentiality?
In the absence of scientific empirical data, some academics have questioned whether confidentiality is important to obtaining sensitive information from clients. The principal citation is to an article by Professor Fred C. Zacharias titled, Rethinking Confidentiality. Zacharias’ article, however, does not claim to be more than a proposal to rethink confidentiality; it does not purport to be an accomplished fact. As Zacharias candidly acknowledges about his own empirical study, which is reported in the article, its sample was limited and its methodology was “somewhat unscientific.” He concludes, “I would be the first to caution against overreliance on [this] study.”
There are, though, some studies that have been done by behavioral scientists, who are understandably concerned with the validity of responses to surveys about sensitive information. These are discussed in a book chapter by Robert F. Boruch titled, “On the Need to Assure Confidentiality and Privacy.” One study found respondents about five times more likely to admit to corporal punishment of their children when confidentiality was clear. Another study, which involved over three thousand servicemen, related to use of illicit drugs, racial attitudes, and racist behavior. Respondents were more than twice as likely to admit to undesirable behavior when the methodology assured confidentiality. In a Canadian study on the incidence of abortions, both legal and illegal, the authors concluded that less persuasive promises of confidentiality “could not obtain admissions of even all legal abortions.” Yet another study, of “leisure activities, including...intercourse, masturbation, and marijuana use,” found that “absolute assurance [of confidentiality] clearly increased the likelihood of response to sensitive, rather than innocuous, questions, relative to groups receiving qualified assurance or no assurance.”
There is reason to believe, therefore, that the conventional wisdom of the legal profession regarding confidentiality, based on the innumerable experiences of countless judges and lawyers, is indeed valid, and that it does not really need rethinking.
From ULE 139-140 (2004).
And then there's the Catholic Church's centuries of experience with the confessional.

[Monroe Freedman]

A critical question raised by the foregoing post is whether lawyers will have all that exonerating information without credible assurances of confidentiality. For example, the innocent man freed after 26 years (when the guilty client had died) might never have gotten out at all if the guilty client hadn't been willing to confide in his lawyers that he was in fact guilty.

[Andrew Perlman]

Monroe,

Those studies are helpful, at least to some extent. They're not exactly analogous, however, because I am willing to bet that the studies involved explicit statements to subjects about the scope of confidentiality. In contrast, in day-to-day law practice, it is relatively rare for lawyers to explicitly state the scope of the duty of confidentiality.

For example, lawyers in Model Rules states do not regularly say to the their clients, "Everything you tell me is confidential, unless of course, you tell me that you are about to go kill someone or cause them substantial bodily harm or unless you end up using my services to commit a fraud that will cause somebody substantial financial injury or if you fail to pay my bill or if I am sued as a result of my representation of you."

If lawyers gave such disclaimers, then those studies would be right on point. Otherwise, unless we have some other indication that clients are aware of these narrow exceptions, I don't think the studies give us too much empirical knowledge about how clients are likely to behave.

Moreover, even if we assume for the sake of argument that clients will be less likely to reveal their true guilt to their lawyers in Massachusetts out of fear that they'll be turned in if someone else is convicted, how much harm has that client actually suffered? Given that Rule 3.3 prohibits a lawyer from putting on knowingly false testimony, one could argue that a client's awareness of the Massachusetts exception might even do the client some good. Specifically, the client might refrain from revealing information that would hinder the lawyer's defense of the client. (Obviously, there's a vast literature about whether criminal defense attorneys should avoid learning certain information for this reason, and that literature is to some extent relevant here.)

Finally, I'm skeptical about your reliance on conventional wisdom over the centuries. An awful lot of the profession's conventional wisdom from 100 years ago has turned out to be quite wrong, ranging from the role of women and minorities in the profession, the effect of marketing, and (to some extent) the evils of contingency fees. The very strict views on confidentiality may be conventional, but I'm not convinced that they are always wise.

[Monroe Freedman]

ABA Standards Relating to Criminal Justice, Std. 4-3.2(b) Interviewing the Client

Defense counsel should not instruct the client or intimate to the client in any way that the client should not be candid in revealing facts so as to afford defense counsel free rein to take action which would be precluded by counsel's knowing of such facts.

* * * * *

Also, as shown in a number of sources, including "Getting Honest About Client Perjury," the lawyers who blow the whistle on client perjury are court-appointed lawyers representing indigent members of minority groups. Other clients assume, correctly, that they can rely on confidentiality as a traditional lawyers' ethic.

[Andrew Perlman]

Monroe,

I'm not suggesting that lawyers should instruct their clients not to be candid. I was merely speculating about how confidentiality exceptions might affect client behavior. (You and Alice had both voiced the concern that clients would be less vocal in states that adopt the Massachusetts/Alaska rule, so I think you both think this is a possibility.)

[Monroe Freedman]

Andy,

As you know, but others might not, I fought, alone, for over 20 years to get the ABA to recognize an exception to confidentiality in order to permit a lawyer to save innocent human life.

I chose to draw the line there for two reasons. First, in American constitutional jurisprudence (as well as my own moral view), death is different, for both non-consequentialist and consequentialist reasons.

Second, and also important here, the instances in which that exception would come into play are so rare that I do not think there would be a systemic threat to confidentiality. That is, few if any clients would ever become aware of that exception or care about it.

On the other hand, exceptions for client fraud, or unjustified imprisonment – if acted upon – would certainly become known to many clients. Frankly, I don’t think those exceptions are going to be acted upon, so the risk of systemic threat is probably very slight, if non-existent. However, I also don’t like writing rules that we expect to be ignored – call it a bow to legal realism or to simple honesty, plus the potential ill-effect of saying, “Some of our ethical disciplinary rules don’t really mean what they say.”

In short, I support the life-and-death exception, but not those for unjust imprisonment or for client fraud (except to allow the lawyer to withdraw her own, previously unaware, misrepresentation that is still subject to being relied upon by a third party).

[Andrew Perlman]

That's a very reasonable line to draw, Monroe. You and Abbe have a great exchange about this issue in your book, and as you might guess, I agree with your conclusion.

The only point that I would add is that I don't think the Massachusetts/Alaska exception will be "ignored." After all, it is discretionary, so I don't think the existence of the exception (and its rare use) poses a threat to the way people perceive the bite of other disciplinary rules.