Suppressed Premises in the Rule of Law Argument for Defending the Guilty?

Hi everyone. It’s good to be here. For my inaugural con-op post (and possibly several others), I’d like to be selfish and take advantage of the epistemic advantages of diversity to toss a loosely-formed thought around, in order to sharpen my own knowledge by seeing what you all have to say about it.

Next semester, I’m teaching professional responsibility for the second time. And this time around, I plan to structure the course around a handful of the Standard Big Questions(TM) that plague legal ethics. And, of course, one of the Standard-est and Biggest is “how do lawyers justify representing criminals whom the lawyers know to be guilty?.” This question becomes especially salient when the crime in question is particularly evil, and if the lawyer succeeds, the criminal may pose a further danger to society: imagine the lawyer who defends a confessed terrorist or child molester. And so I’ve been thinking about it a bit.

One of the Big Standard Answers to the Big Standard Question is that lawyers may (indeed, ought to) defend the guilty in order to defend the rule of law. On this argument, without the work of lawyers defending even those who are known to be guilty, the government may have an incentive to shirk on its constitutional duties to punish only the guilty. Moreover, by forcing the government to comply with constraints like those embedded in the Fourth and Fifth Amendments even to punish the guilty, lawyers help preserve those rights for us all. For example, if the government knew that it could get away with warrantless searches, because nobody would wield the exclusionary rule in defense of the guilty, then it would have an incentive to illegally search everyone in order to catch and prosecute the guilty.

From the standpoint of lawyers as a class, this argument seems pretty good. But it breaks down when we stop asking “why should lawyers in general defend the guilty?” and start asking “why should I, a single lawyer, represent this particular guilty defendant?” For the rule of law pretty clearly can be preserved with something less than 100% of the guilty getting a defense. If every lawyer in the world decides that Khalid Sheikh Mohammed is so revolting that he’s not entitled to a defense, the rule of law isn’t going to collapse, so long as lawyers generally are still willing to put up a vigorous defense of the guilty in general. Moreover, even if any individual lawyer is unwilling to represent KSM, some other lawyer may still be willing to do so.

For representing the guilty, then, we have a double free-rider problem. Representation of the guilty is a public good, but each individual lawyer has an incentive to shirk KSM’s representation both because it may not be necessary to represent that particular client in order to get the rule of law, and because it may not be necessary for any given lawyer to represent him in order to bring it about that he is represented. Strikingly, this isn’t just your basic garden-variety self-interested free-rider problem; it’s what we might call a moral free rider problem. (I have an old conference paper floating around on this class of problem.) For a consequentialist, representing KSM comes at a moral cost: to the best of our knowledge, he’s patently guilty as well as incredibly dangerous, and doing anything that, however slightly, raises the probability that he will go free, represents a moral evil that the lawyer is obliged to avoid unless doing so is actually necessary to avoid the greater evil of the failure of the rule of law. It seems to follow that each individual lawyer is morally obliged not to defend KSM.

There are a variety of strategies we might take to avoid this problem. Briefly, at least two come to mind. 1) We can be rule-consequentialists rather than act-consequentialists: lawyers ought to follow the best rule, where that rule may be “always defend the guilty.” 2) We can add uncertainty about how much representation of the guilty is needed: if we don’t know how many guilty people can go unrepresented before the rule of law collapses, there may be a reason to defend each individual guilty person in virtue of the risk that this is the tipping-point case where it all goes down the drain.

If not one of those two ideas, it seems to me that we must have something to span the gap between the collective reason lawyers have to defend the guilty and any individual lawyer deciding whether to defend any individual guilty person. But what do you think? (Surely someone has written on this…)

You may also like...

29 Responses

  1. Andrew Ayers says:

    How exciting to see this post! I wrote an article on just this question, making two claims. First claim: the gap between collective reasons and individual decisions can only be bridged by a sense of intrinsic value — something about the act that is valuable/admirable even if it has bad consequences.

    This is true even if you’re a rule-consequentialist, because the rule-consequentialist has to say why it’s intrinsically valuable to follow their rules. (And even if you’re Kantian, because the Kantian has to say why it’s intrinsically valuable to follow universalizable maxims.)

    Second claim: a given agent’s sense of intrinsic value depends on a complex constellation of background values — background values that may vary widely across lawyers. One lawyer’s sense of the value of defending a guilty person might stem from her commitment to beliefs about structural injustices in the American legal system, which in turn might stem from her own experiences with legal institutions and her beliefs about what constitutes justice. Another lawyer’s sense of value might stem more from compassion for defendants as human beings in a uniquely vulnerable situation. That sense of value would depend on the lawyer’s own experiences of human relationships and beliefs about trust or human dignity.

    The complexity, particularity, and diversity of our background networks of values makes it impossible to reduce those values to propositions or maxims. A sense of intrinsic value has to be understood culturally and humanistically, in its context, before it can be usefully communicated or criticized.

    Or that’s the argument, anyway.

    Article is here:

  2. Paul Gowder says:

    Oh, wonderful! I knew someone had to have written on this—looking forward to reading your article.

  3. Orin Kerr says:

    I wonder if background legal rules address this problem in practice. For example, if the rule is that a person must be represented by counsel in order to be punished, then someone has to represent each person for them to be punished.

  4. Paul Gowder says:

    Hmm. Now I wonder if that’s true of the real world. What would we do with a Sixth Amendment claim by a defendant who said that there was no lawyer willing to represent him/her?

  5. pc says:

    For the benefit of any budding lawyers that might not know it, the argument for defending the guilty (someone you believe is guilty, I guess, unless you saw her do it) doesn’t “break down” at all. The author of this post hasn’t thought this through, not even close.

  6. pc says:

    um, in response to 4, we would not try the defendant. What “real world” are you talking about?

  7. Orin Kerr says:


    If a criminal defendant can’t afford a lawyer or find one willing to represent them, the judge will appoint counsel. I haven’t seen a lot of opinions on this, but my understanding of the ethics rules is that lawyers are not permitted to turn down a court appointment on the ground that they just don’t want to defend that particular person. Refusal to take the case is considered contempt of court. See, e.g.,

  8. Brett Bellmore says:

    Is it really that big of a problem? I’d expect there to be some segment of the defense bar willing to defend pretty much anybody, on the basis that their money spent just like anybody else’s. Perhaps this random appointment thing is actually meant to shield that group from exposure?

    “No, really, I don’t have any choice in the matter!”

    I mean, my limited exposure to lawyers has not given me the impression that they are uniformly the most fastidious group.

  9. Paul,

    Perhaps the more interesting question we might ask is why there does not seem to be, in practice (on the ground as it were), any significant free rider problem whatsoever when it comes to representing “guilty clients,” conspicuously, even those deemed especially morally reprehensible or evil: domestic and foreign terrorists, those accused of genocide, serial killers, and so forth. Perhaps this has something to do with the “legal and moral ecology” of criminal defense lawyering, the norms and duties of which appear to have been successfully internalized by most of its practitioners. And routine fulfillment of these obligations seem to bring in their wake enhanced personal and professional esteem, respect, or status among one’s peers as well as among those (as part of an overlapping set) outside the profession likewise concerned with rule of law values and constitutional and democratic principles and practices.

    There is of course a litany of publicly avowed reasons* that will be proffered by way of defense of such representation having to do with rule of law values generally (both substantive and procedural), constitutional protections and rights, and related democratic values and principles in particular. Criminal defense lawyers who defend “guilty clients,” notorious or not, could be said to be clear exemplars of the moral and legal norms and professional responsibilities of criminal defense lawyering, evidencing and modeling the moral, legal, and logical breadth and depth its members commit to in virtue of their professional role and duties of legal office. The necessary exercise of discretion and judgment will assure that some (in this case fairly small class of) individuals may studiously avoid recognizing such obligations, perhaps finding sufficient rationalization owing to a “personal ethics” that ill fits the predominant moral and legal ecology of criminal defense lawyering.

    The real problems of criminal defense lie elsewhere: recalcitrant access to justice problems and the unfulfilled promises of Gideon, for example. As W. Bradley Wendel succinctly puts it, “In general…there does not appear to be a pervasive problem of potential clients being turned away on account of their morally disagreeable qualities, as opposed to their inability to pay.”

    I should note that I’m not opposed in principle to attempting to theoretically discussing this along the lines of a hypothetical “double” free rider problem, but in so doing, it’s worth paying attention to the apparent fact—and in light of other more legally and morally pressing problems of our criminal justice system as well documented in a handful of recent books—that this is a fairly abstruse philosopher’s problem not reflective of an enduring or obvious problem intrinsic to criminal lawyering practice, at least in “our” adversary legal system.

    * Wendel provides one such list in his book, Lawyers and Fidelity to Law (Princeton University Press, 2010): 151.

  10. In other words, the collective/public good (synonymous terms for my purposes) in this instance is more or less distributively realized by criminal defense lawyers, the exceptional or anomalous cases requiring explanation: one that does not appear, to me at any rate, sufficiently (hence perhaps partly) captured in terms of a “free-rider” problem as such, but best viewed more simply (if not prosaically) along the lines of an ill-fit between the moral and legal lifeworld or ideology of the outlier and the prevailing moral and legal ecology that rules the minds of criminal defense practitioners in this country. This is a rather tolerable if not unavoidable problem we can live with….

  11. AYY says:

    I don’t see where the problem is. As someone said if you didn’t see him do it, then you probably don’t know if he’s guilty, and if you did see him do it, or even know anything about the case, then you’re disqualified.

    As a practical matter, you’re probably going to try to negotiate a plea. If he doesn’t agree to an offer there’s always the possibility that a difference of opinion on strategy can be a basis to claim conflict of interest, or for the defendant to try to get a different lawyer.

    If you have to go to trial, then you can always argue that the charges are piled on and he’s really only guilty of a lesser offense, or just a few of the charges, or he’s guilty but had an innocent motive, or there’s an ulterior motive for the prosecution, etc. etc.

  12. Paul Gowder says:

    Orin, I suspect the courts would treat a principled refusal to represent differently from a refusal because the lawyer no longer practices in juvenile court, etc.—if only because it’s generally accepted that if a lawyer is truly revolted by a client to the point that s/he can’t competently carry out the representation, there are strong ethical reasons not to go ahead anyway (e.g., Model Rule 6.2(c)).

    Ultimately, however, I imagine this is probably a counterfactual we’ll never see for the reasons Patrick and Brett give: either lawyers have internalized the conventional arguments for representing the guilty, or have enough interest in money not to care. So, Patrick, fair enough, this is mostly an abstruse philosopher’s question.

    AYY: Fair enough, but what if you’re really really confident he’s guilty? There’s never certain knowledge, but you can think, on the basis of all you know, that the probability of your client’s being innocent is vanishingly small.

  13. AYY says:

    “but what if you’re really really confident he’s guilty? There’s never certain knowledge, but you can think, on the basis of all you know, that the probability of your client’s being innocent is vanishingly small.”

    You might want to ask the DUI lawyers how many times they have clients with multiple DUI convictions, who have a blood alcohol level of .15 or higher, and admitted they had 4 or 5 beers and a few whiskeys before they were stopped while weaving on the freeway.

    Anyway, the lawyer probably isn’t going to find out what kind of a case the prosecution has until he assumes representation and gets discovery and interviews witnesses. At that point, the lawyer can’t just up and leave.

    If the case is so clear, the prosecution ought to be able to prove to it to a jury. But even aside from that, the defendant needs a lawyer to negotiate a plea, or to tell him whether he’d be better off taking a plea or going to trial.

    If it does go to trial all the defense lawyer is doing is making sure the prosecution plays by the rules. For example, if you have a client who is clearly in your mind guilty of possessing meth, shouldn’t the lawyer be able to bring out that the thingy the county uses to chemically analyze the meth hasn’t been recalibrated for 6 years when the manufacturer recommmends that it be recalibrated every 6 months. The jury might like to know that, but you won’t know that’s even a problem until you examine the records, and you won’t be able to examine the records unless you assume representation.

    If the case is so clear the prosecution ought to be able to easily prove it to the jury. If the jury doesn’t see it the prosecution’s way, then maybe they know something the defense lawyer didn’t.

    Besides even when the defendant is guilty, the question is whether he’s guilty of everything he’s being charged with. Even if he’s found guilty, the sentence isn’t automatic. Surely you’d agree that the defendant needs an attorney at that point.

    So that’s why I don’t see what the moral issue is in defending someone you know is guilty.

  14. Hi Paul,

    Nice topic. I’m interested in this claim: “For a consequentialist, representing KSM comes at a moral cost: to the best of our knowledge, he’s patently guilty as well as incredibly dangerous, and doing anything that, however slightly, raises the probability that he will go free, represents a moral evil that the lawyer is obliged to avoid unless doing so is actually necessary to avoid the greater evil of the failure of the rule of law.”

    First, consequentialism here is underspecified (what consequences matter?), but most versions are false, so I’m not sure how helpful it is to frame the problem in these terms.

    Second, there is an unargued for assumption that representing KSM will even slightly increase the probability of him getting off. But that depends on our theory of legal ethics and what it is permissible for a lawyer to do in the course of representation. More on that in a minute.

    Third, I think we have more particular reasons (more particular than just general maintaining the ‘rule of law’ reasons) to want each person to have a lawyer. The thought is that there are worries about political legitimacy that arise both in the creation of law, and in the application of law to particular cases. In the creation process, we specify certain ‘triggering conditions’ for State action against particular individuals and we specify certain rules for attaching consequences to individuals who satisfy those triggering conditions. As a result, on the application side, we have a powerful interest in making sure that in every particular case in which the State acts against an individual, those triggering conditions are actually satisfied and only legally specified consequences attach. That remains true in the case of KSM, and that remains true if we are certain (somehow) that KSM has satisfied those triggering conditions, because there may be mitigating circumstances that affect the appropriateness of attaching certain consequences, or there may be a range of consequences that are open, and some question of which should attach.

    If one goes in for this, then one might defend a principle regarding the decision to represent an individual threatened with State action against that individual that goes something like this:

    SVI Representation Principle: An Individual Lawyer may permissibly decide to represent a client only if, at the
    time at which she is making the decision, she reasonably
    believes that by representing the client she will make it more likely that (a) the State will take action against the client only if the relevant triggering conditions actually obtain; and (b) if the triggering conditions obtain, the client will receive an outcome which is legally appropriate, given the client’s situation (including possible excusing or extenuating circumstances), than if the client were unrepresented.

    Anyway, I defend that view (and the view that legal ethics should vary depending on whether one represents and individual against the State as opposed to an individual against another private individual) in “Lawyers, Context, and Legitimacy: A New Theory of Legal Ethics,” GJLE (2012), which you can read here:

  15. Monroe Freedman says:

    The book is “How Can Your Represent Those People?” edited by Professor Abbe Smith and me. It was published on August 12 and sold out the first day. I understand that copies are available again on Amazon at a discount.

    It consists of 15 essays by leading criminal defense lawyers, including Michael Tigar, Barbara Babcock, Paul Butler, and Alan Dershowitz. The essayists are a mix of age (25 to 85), ethnicity, gender, and areas of practice.

    There will be a Symposium at Hofstra Law School, featuring 15 of the authors, on October 9, 12-2:00. Other book signings are scheduled throughout the country.

  16. Paul Gowder says:

    Alex, I can’t wait to read this paper—thanks!

    Even before having time to read the paper, one quick question: where does the moral importance of legal accuracy in an individual prosecution come from? The moral importance of legal accuracy in prosecutions in general comes, I think, from the importance of ensuring that state officials only coerce people in accordance with the law, and thence (see Yours Truly, “The Rule of Law and Equality” in the current Law & Philosophy) from the moral/political value of equality. But the rule of law is a property of states and legal systems as a whole, and can persist in the face of isolated deviations.

    One thing you might say is that the moral importance of legal accuracy in an individual case is a matter of justice, but I’m not sure that’s true. The set of just punishments for KSM and the set of legally-permissible punishments for KSM are independent (unless one is a proceduralist about justice?), guaranteeing the latter need not guarantee the former.

    AYY, I think hidden in your last comment is another important point: it might be that in any individual case, the lawyer defending the guilty will reveal some government misconduct that’s also present in other cases (like the uncalibrated chemical thingy); that does seem to suggest the possibility of indirect contributions to good ends even in defending KSM.

  17. Paul,

    I might have to read your paper to answer this question, but I don’t know why you think what we care about is the “rule of law” as a property of states and legal systems as a whole, such that “it” can persist in the face of isolated deviations. I don’t care about the “rule of law” per se, nor do I care about whether there is some sense in which it is true that the legal system under which I live instantiates the property of being governed by the rule of law. (Or, if I do care about this, I only care about it derivatively, in the way I might care about being a good person, where what I really do and should care about is acting in particular good ways.)

    What I care about are instances of coercion, or action backed by coercion, and whether each of those instances is morally permissible. I care about this full stop, but I also care about it particularly much because this State coercion is, at least in a democratic polity, partly authorized by me; it is democratically authorized by all of us. And (I would have thought) whether each instance of coercion is morally permissible is a function of, among other things, whether the State official acting coercively is democratically authorized to do so–acting under legitimately enacted law, but also in a situation in which the factual triggering conditions obtain.

    So, to my mind, that’s where the moral importance of legal accuracy in an individual prosecution comes from. By paying taxes, I’m supporting coercion. Ordinarily, coercion is morally impermissible. Whether I’m supporting something morally impermissible or not turns on whether the full justificatory story obtains, including that part of the story that involves getting the details right in this particular case. Each instance of legal inaccuracy is an instance of my having supported morally unjustified coercion. And that seems really bad. And it is much worse when the errors are systematic and predictable, as they are when we so woefully underfund indigent civil and criminal legal defense.

    But some of your tone suggests a much more consequentialist picture, so that maybe you will be unsympathetic with these worries about our support of morally impermissible coercion (you might think the badness of that can appropriately be ‘outweighed’ by the various goods that the system brings about). I’d certainly urge against that: one of the clearest counterexamples to consequentialism remains the obvious impermissibility of punishing an innocent person to quell an angry mob.

  18. Also, as a sort of meta-point, I really do agree with Patrick on this:

    “The real problems of criminal defense lie elsewhere: recalcitrant access to justice problems and the unfulfilled promises of Gideon, for example. As W. Bradley Wendel succinctly puts it, “In general… there does not appear to be a pervasive problem of potential clients being turned away on account of their morally disagreeable qualities, as opposed to their inability to pay.” ”

    As a philosopher, I can appreciate the interest in the question you ask in the original post. But I do worry that in a legal ethics class, it’s one of the last chances one has to focus people on thinking critically about the system they are going to be a part of.

    In addition to the huge issue Patrick raises, there is also interesting work on the ethics of plea bargaining (Richard Lippke has a recent comprehensive monograph on the topic), for example, that often doesn’t get covered in standard legal ethics classes, and little in the MRPC is on point. I could also see covering related issues having to do with collateral consequences, mass imprisonment, and the racial character of our criminal justice system.

    Anyway, it’s great that you’re doing this. I’m sure it will be a great class! I’ve only taught PR once, but the students really seemed to appreciate getting into a bit more theory behind (or next to) the rules.

  19. Paul Gowder says:


    Going backwards: first, I agree with your last comment; these are great ideas for things to cover, and I intend to deal with a number of them (certainly mass imprisonment and race, and I really like the others as well, though I personally know almost nothing about plea bargaining…).

    On the philosophical questions: I don’t particularly see why the legitimacy of each individual instance of the state’s coercion should matter to me, qua democratic citizen. For while it’s true that the state acts in our names and under our authorization, it seems like a big leap to think that every individual official state act is attributable to each citizen. It’s more intuitive to think that the general arrangement of the state, its laws and the norms that ordinarily regulate how its business is carried out are attributable to each citizen.

    More broadly, I’m not sure one needs to be a consequentialist to be skeptical of the complaint “I’m guilty, but you prosecuted me illegally.” Perhaps this is why I’ve never really found the arguments of Nozickian libertarians and the like terribly compelling. Should we care *for its own sake* if someone is illegally coercively punished, when they deserve punishment anyway?

    Of course, by “for its own sake,” I mean to bracket all the other reasons why we might care about all or almost all individual instances of illegal coercive punishment, including, among others, these systemic rule of law considerations I like to talk about, moral concerns about the particular ways the state might illegally punish someone (e.g., moral concerns about the torture used to extract a confession), moral concerns derived from epistemic concerns about the reliability of the procedure by which the punished’s deservingness is ascertained, character vices like arrogance which we might attribute to the person who carries out illegal punishing, and all the rest. Once we get rid of all those things, I’m not sure right now how much is left of the moral objection to illegal punishment of the guilty.

  20. Paul Gowder says:

    (Of course, in a democracy, there might be an objection to illegal punishment based on the moral importance of popular sovereignty, but, contra the enlightenment/libertarian tradition, we need not see that moral importance as having much to do with sanitizing otherwise-impermissible coercion.)

  21. shg says:

    To criminal defense lawyers, this is called “the cocktail party question.” I twitted this post to the criminal defense community, and they were all going to comment as soon as they stopped laughing.

    As an aside, while I haven’t read Freedman’s and Smith’s book, I did read Abbe Smith’s WaPo op-ed and was not at all impressed. In fact, I was sufficiently disturbed that I wrote about it.

  22. Hi Paul,

    I’m not sure what you mean by “I’m guilty, but you prosecuted me illegally.” Similarly, I don’t know what you have in mind when you say someone is “illegally coercively punished.” I was thinking of cases in which even if guilt is clear, there may still be other relevant issues involving the details of the consequences that attach to guilt. So I don’t think we can just bracket “moral concerns derived from epistemic concerns about the reliability of the procedure by which the punished’s deservingness is ascertained.” I see this as one of the main reasons to have a lawyer even in cases in which technical guilt (satisfying the elements of the crime) is clear. So much happens after that.

    (Also, just to put it on the record, I think that it is very rare when it is clear ex ante that technical guilt is satisfied, given mens rea elements and the various ways in which those can be negated by mental health issues, which makes this even more of a philosopher’s question…)

    I’m a skeptic about moral desert claims in the realm of punishment, but even if I weren’t, I think almost all the issues here have to do with imperfect information, detection, and enforcement (there are also issues regarding disagreement about punishment, and the need to come up with some societally palatable way of responding to bad conduct in the face of this disagreement). If one believed in desert, and if one thought there was just some simple fact about what conduct morally deserved what kind of punishment, it would certainly be a better world if as soon as a person engaged in that conduct, the punishment was automatically enacted. And there’d be no moral objection on procedural ground, certainly. But that is not our world…

    Finally, it’s not that “every individual official state act is attributable to each citizen.” It’s that we should care that every individual official state act is morally justifiable. One reason we should care about that, in particular, is that we are partly responsible for the State existing and acting, but it’s not the only reason, and it certainly doesn’t require something like direct attributability. This is something we should care about even if we were in no way responsible for anything the State or state officials did. That we are just adds an extra reason.

  23. pc says:

    Why is this post being taken seriously? It’s major premise is absurd. I would ask the author what he thinks constitutes proof to a lawyer that his client is guilty, but I’m afraid of what he might answer (confession? head-desk). Moreover, as someone noted, there’s just a bit more to criminal defense than trying to convince a jury/judge your client is completely innocent – prosecutors often overcharge, and then there are the sentencing rules. This post does not have even one foot in reality.

  24. Paul need only only assume or stipulate that the lawyer believes his client to be guilty, whatever the specific reasons that motivate that belief: from the sound and persuasive to the highly probable (the belief may turn out to be false, yet that would change nothing for the purposes of illuminating the problem he imagines). And surely Paul appreciates that criminal defense is more than attempting to prove one’s client is innocent: he is free to circumscribe the specific nature and scope of the hypothetical problem or stylized scenario of a possible real world problem if only for heuristic purposes. This need not entail the burden of being true to all the real-world descriptive features of criminal defense lawyering.

  25. Monroe Freedman says:

    shg: Richard Bernstein said in The New York Times: “[W]hat the internet and its cult of anonymity do is provide a blanket sort of immunity for anybody who wants to say anything bad about anybody else, and it would be difficult in this sense to think of a more morally deformed exploitation of the concept of free speech.”

    What that means is that people say things anonymously that they don’t have the guts to say otherwise.

  26. Monroe,

    Presumably you did not know this, but you can discover who “shg” is if you click on the initials with your browser (, which links to his blog and name. You can also read the blog post in question with an “Abbe Smith” search on the blog.

    Warm regards & best wishes,

  27. I should have mentioned I like the quote from Bernstein.

  28. shg says:

    I like the Bernstein quote as well, Patrick, though I can think of quite a few more morally deformed exploitations of free speech than mere anonymous criticism. Then again, I hang out with a tougher crowd.

    Too bad Monroe didn’t save it for a more appropriate opportunity.