« Brady “Materiality” and the Anemic Implementation of Prosecutors’ Disclosure Obligations | Main | New Constitutional Law Jot: "#Resistance, With Candor" »

Tuesday, May 15, 2018

Reply to Atiq's "What Unconditional Credence in Individual Desert Claims Does Retributivism Require?"

In Emad Atiq's response to my Punishment and Moral Risk, he states that he finds "negative retributivism" especially plausible and describes it as follows: "Punishing a person who does not deserve to be punished is morally impermissible." In the first sentence of his response, he writes, "Adam Kolber suggests that negative retributivism requires impossibly high degrees of credence in individual desert claims for punishment to be morally permissible." He goes on to argue that negative retributivism avoids my critique.

As a preliminary but important matter, I simply don't believe that negative retributivism falls under my critique in the first place. I write, for example, "I will focus on a pure deontological form of retributivism that takes desert to ordinarily provide a sufficient reason to punish without reliance on other possible goals of punishment like deterrence, incapacitation, and rehabilitation." Later in the paper, I have a section called "traditional hybrid theories" that explicitly discusses "limiting retributivism" which is the same or a close cousin to the "negative retributivism" Atiq favors. Atiq doesn't address this portion of my paper at all (and the arguments there are meant to be suggestive and not exhaustive in any event).

The reason I don't address Atiq's form of negative retributivism in my central argument about justification is that negative retributivism does not purport to justify punishment. As Atiq recognizes, it is simply a constraint on punishment. So the negative retributivist needn't assent to all nine of the propositions I discuss. For example, she might deny that "those who commit serious wrongs deserve to be punished (or to suffer) in response." After all, she merely views desert as a limitation on punishment, not something that necessarily provides an obligation or even a reason to punish (and the proposition probably implies at least a reason to punish).

Indeed, a person could be a serious consequentialist with the exception that she views negative retributivism as a limit on punishment. In my paper, I argue that consequentialism is less subject to my epistemic challenge than traditional forms of retributivism. Negative retributivism could require even less punishment than pure consequentialism; so it is on firmer justificatory ground than the retributivists I focus on. I do say things in the paper that might apply to negative retributivists, but it all depends on what their underlying justification of punishment is not their limitation on punishment.

In a footnote, Atiq writes that he takes negative retributivism to be among the views I target, even though "[t]he view that Kolber treats as paradigmatically retributivist is one that treats desert as a sufficient condition for punishment, not just a necessary condition." Still, he writes, "[w]hat I say in defense of negative retributivism applies with full force in the case of this alternative position, so long as it is consistent with ends like crime prevention being treated as valuable and as pro tanto reasons to punish (subject to the desert constraint). The "so long as" condition makes Atiq's version of negative retributivism sound a lot like consequentialism. So the short answer is, I think the portion of my paper Atiq focuses on has limited application to the retributivist view he finds most plausible. 

Atiq also describes a principle he calls "CERTAINTY:  Punishing a person without at least 90–95% credence that the person deserves to be punished is morally impermissible."  He points out, quite correctly in my view, that nothing about retributivism entails CERTAINTY.  Where I disagree is with his view that I support CERTAINTY (see, e.g., Atiq p.141 ("[Consider] Kolber's own reasons for thinking that the retributivist is committed to CERTAINTY"). I never suggest that retributivists need a justificatory standard of proof anywhere near as high as 90%. Most retributivists would readily concede that they need at least 50% confidence punishment is deserved (otherwise they would think it more likely that punishment is undeserved than deserved), and I suspect that the standard must be substantially higher than that to match the values that seem to underlie retributivist commitment to the beyond-a-reasonable-doubt (BARD) standard. But I don't expect retributivists to endorse a principle as demanding as CERTAINTY. 

Let's put these concerns aside for now and turn to what I think is the heart of Atiq's argument: Atiq argues that just because a person believes that juries should find that a defendant committed all the elements of a crime beyond a reasonable doubt does not mean the person believes defendants should only receive punishment when they deserve it beyond a reasonable doubt. So far so good. The BARD standard itself does not compel us to have a high overall justificatory standard. But I never claim that it does. Rather, as I repeatedly state, my focus is on the values that I attribute to retributivists who believe in the BARD standard. They typically defend the standard in Blackstonian-type terms: e.g.,  better ten guilty people go free than one innocent person be punished. The ratio takes a position on how to value the risks of error as to factual issues (or, perhaps more precisely, risks of error as to the satisfaction of the elements of a crime).

Once again, though, I don't claim that there is a logical entailment between belief in the Blackstone ratio and a similar hypothetical ratio for errors of deservingness more generally. Indeed, I acknowledge some reasons why one might differentially value these errors to some extent. But I do argue that it's hard to see why a retributivist's Blackstone-like ratio for errors of fact would be all that different than her Blackstone-like ratio for errors of desert. The reason is that the traditional retributivist justifies punishment in terms of desert. So it seems inconsistent to give radically different valuations to different ways of being undeserving. If punishment is underserved, it's not clear why it should matter all that much to retributivists the reason that it is undeserved.

Atiq tries to explain why different kinds of desert errors could be relevant to retributivists but does so by referencing consequentialist considerations that, as I discuss above, are supposed to be screened off. In a footnote, Atiq wonders whether his view of BARD involves a reinterpretation or rejection of the Blackstonian view. Either way, though, my argument concerns the values that retributivist commitment to the BARD standard seems to reflect and are not, generally speaking, consequentialist values. (Even if I hadn't screened off Atiq's flavor of negative retributivism in my piece, Atiq should argue not only that consequentialist considerations could lead to a divergence between the standard for factual guilt and for deservingness more generally but also that the consequentialist considerations he supports would still lead to a BARD standard at all. If his view doesn't lead to a BARD standard, then it is obviously outside the scope of my argument. And note that any consequentialist argument for the BARD standard is likely to be quite tentative since consequentialist commitment to the BARD standard depends on many hard-to-predict empirical considerations.)

After setting the BARD issue aside, Atiq asks whether the core retributivist prohibition on punishing the undeserving prevents retributivists from taking significant risks that they are punishing the undeserving. (Traditionally, the prohibition has been described in terms of punishing the innocent, but of course, I have no objection to treating it as a prohibition on punishing the undeserving.) Atiq argues, correctly in my view, that the traditional prohibition on purposeful and knowing inflictions of undeserved punishment does not directly apply to merely risking inflictions of undeserved punishment. That's why I don't locate the force that pushes retributivists to avoid excessive moral uncertainty in the deontological prohibition against punishing the undeserving.

As discussed, the force of my argument stems from the Blackstonian-type values that seem to underlie traditional retributivist support for BARD. Notice that even though the Blackstone ratio is typically described in terms that lack a mental state qualification such as "intentionally, knowingly, or foreseeably," the ratio is understood to refer to risk taking. That's why Blackstonian-style commitment to BARD is relevant to my argument. It reflects a tradeoff between errors of undeservingness. And, again, while Blackstonian errors refer more explicitly to the satisfaction of the elements of a crime, it is a very small further step to think that retributivist commitment to Blackstone-type ratios will lead them to similar ratios about errors of deservingness more generally.

Finally, though I don't think it bears on the argument in my paper, in conversation, I posed a challenge to Atiq's support for the deontological prohibition on purposely or knowingly inflicting undeserved punishment on a particular individual. Here's how Atiq describes it:

Kolber in conversation raises the following objection to my overall view:

Suppose that there’s a new public policy proposal that will save one million lives each year, but, unfortunately, it will require us to increase our punishment error rate a bit. Indeed, it will cause precisely 100 more innocent people to spend life in prison than would otherwise. These one hundred people weren’t targeted in anyway, and almost all people would prefer the substantial improvement in life expectancy from the plan relative to the rather tiny chance of being erroneously punished.

The negative retributivist might accept this deal on consequentialist goods. But now suppose that the policy makers ask the retributivist whether she would like to know who the 100 people would be. Kolber imagines my retributivist covering her ears, for identification would dramatically change the moral situation and prevent her from accepting the policy. Kolber is right that this reaction would be absurd. The reason it would be absurd is that in refusing to know the identities of the 100 innocents, the caricatured retributivist is imposing an artificial constraint on her epistemic situation. If members of the larger population were surveyed for their objection to the policy, they might reasonably ask: do we know who the innocents will be, and, more importantly, would I be one of the innocents? It is no response to this concern to say: we have the names written somewhere but we are not looking so we cannot tell you. The critical difference between this case and the standard case of undeserved punishment is that information concerning who the undeserving is [is] not available, either to the policy makers or the individuals being punished. The lack of knowledge concerning the identities of the undeservingly punished is not the result of a failure of enquiry. [Atiq drops a footnote here: "I recognize that here lie dragons: it is a hard question when lack of knowledge is attributable to the knower in the sense that the knower bears responsibility for the uncertainty and when it is not. We do not need a theory of this distinction for present purposes. The point is just that there is a distinction that is manifestly important from the moral point of view."] 

In the objection Atiq discusses, I question why it should matter whether we know the identity of the particular person being undeservedly punished. Atiq believes that he resolves my objection by saying that when one doesn't know a person's identity, we will treat it as knowledge of the identity if the lack of knowledge results from a mere failure of inquiry.

But as he foresees in a footnote, "here lie dragons." What constitutes a failure of inquiry? One method to determine when there is a failure of inquiry is to engage in a cost-benefit analysis that compares the value of the information we hope to obtain to the time and other resource costs of engaging in the inquiry. But why should the happenstance of costs of the inquiry bear on the moral permissibility of saving one million lives? Even if we don't use a cost-benefit analysis, retributivists will presumably identify some sort of duty to inquire lurking here. Shouldn't that duty be rather weak when a stronger duty would jeopardize one million lives? Atiq cannot respond to such important questions simply by adverting to a duty of inquiry and asserting that he identifies "a distinction that is manifestly important from the moral point of view."

To make matters even messier, imagine the following dialogue between a prosecutor and the sort of retributivist Atiq envisions:

Prosecutor: We can save an enormous number of lives this year, but it will require us to punish one innocent person.

Retributivist: Do we know who it is? Because if so, I disapprove.

Prosecutor: Well, we've engaged in a quite searching inquiry, far beyond the minimum required of us. The person looks like this [a photo is shown to the retributivist].

Retributivist: Then I disapprove of punishment. If we know who will be undeservedly punishment, there is a firm deontological prohibition against proceeding.

Prosecutor: All I said is that he looks like this. This is a photo of John Robinson. He might be the one who gets undeservedly punished but it might also be his twin brother, Steve Robinson, who looks virtually identical. So, you see, we really don't know who will be undeservedly punished.

I doubt Atiq's retributivist would now assent to the undeserved punishment, but why not? Is the pool of two people too small? What if there were ten or one hundred clones and we don't know which of the ten or the one hundred it would be? What if the prosecutor came to the retributivist not with a photo but with a physical human being and said we've got the person but we don't the person's identity. It could be one of one hundred different people as far as we know. Can the retributivist proceed to undeservedly punish now? What if we have the severed finger of the person who will be undeservedly punished but we do not yet know to whom it belongs?

I have trouble seeing the moral relevance of the possible distinction at play here that is supposed to be manifestly important. My own inclination is that morality should be pushing us to worry quite a bit about lives unseen for we are much more likely to neglect statistical lives than identified lives. Here lie dragons indeed. There may be dragons on my side as well, but I think there are too many on Atiq's side to confidently assert that the distinction he relies on is manifestly morally important.

While I believe Atiq largely responds to arguments that I don't make, working through his claims reveals a remarkable level of agreement between us. For example, we both seem to think that traditional retributivist beliefs entail little about moral uncertainty and that consequentialist views are less vulnerable to my epistemic challenge than pure retributivist views. The main source of disagreement, I think, concerns the nature of the retributivist values underlying the Blackstone ratio. Atiq seems to think that the Blackstone ratio entails nothing about the relative values of other errors of deservingness, while I say that, entailment is not at issue--it's still a small step from the retributivist values underlying the Blackstone ratio to the claims I make. Atiq's response helpfully encourages exploration of Blackstonian values and pushes me to explore the boundaries of the kinds of retributivism that likely adopt such values. 

I am grateful to Atiq both for his thoughtful response paper and for our conversations surrounding it, and I hope that we continue the conversation in the future.

Posted by Adam Kolber on May 15, 2018 at 04:22 PM | Permalink


I appreciate you saying that, Peter!

And, yes, you anticipated the direction of the response I would have given. We don't want police to set up essentially harmless sting operations (e.g., attempted theft of inexpensive baby powder made to seem like cocaine) just to be able to arrest and punish people. The hybrid retributivists you describe may have answers to this sort of thing: e.g., giving just desert is a "conditional good" meaning that it's good so long as a crime has been committed). I don't know if anyone has worked this out in detail, but it's possible that there are workarounds like this. But they make the whole theory less simple and less elegant, and those are certainly strengths, I think, of purer forms of consequentialism.

Posted by: Adam Kolber | May 16, 2018 4:58:59 PM

Thanks, those are very good points. I agree they wouldn't so self-describe but they do seem to (in practice) weigh the gains of retribution off against other consequentialist goods but that's only something you need to worry about if the resulting theory would be persuasive and your argument here that it wouldn't be seems pretty persuasive.

Though I suppose you could insist that the benefit from retributive punishment is never as great as the harm caused by the crime...but then this raises other problems (most retributivists presumably believe an attempted murder that fails unnoticed five minutes before the target dies in a freak lightning strike deserves substantial punishment even though it would create very little harm).

I have to admit that initially I was pretty skeptical that an argument with a number of parts and relying on assumptions about the values underlying BARD could be that persuasive but your explanations here and in the paper have brought me around to see it as a really interesting and powerful argument.

Posted by: Peter Gerdes | May 16, 2018 4:42:03 PM

Thanks, Peter! You're right that I'd classify it as primarily a consequentialist theory (hence, not a target of the main argument in the paper). Quite apart from the paper, I have some skepticism about such approaches. For one thing, they have to give an account of why it's bad to, say, tempt people into crimes and then arrest them so we can get the benefit of giving them what they deserve. (There are some potential responses, of course, but this gives you a flavor of some of the problems they might have).

As for whether the view you describe is closer to the view of most retributivists, my own sense is that most people who identify as retributivists don't purport to have that view. Now, maybe that view better reflects their actual beliefs or practices. But at least in terms of how they self-describe, I'd be surprised if it has a large following among self-proclaimed retributivists. Thanks again.

Posted by: Adam Kolber | May 15, 2018 7:37:09 PM

Ahh, I understand your argument much better now and it seems more appealing now that I do.

However, how would you classify a consequentialist view which accepted that, among the consequences to be considered, is the positive value of punishing those who deserve punishment. From what you say in sections III and IV of your paper I'm guessing this would evade the critique as a consequentialist theory even though it stipulates a fundamental value to punishing the guilty.

Yet, I would suggest this is much closer to the view most retributivists have than any pure deontic theory.

Posted by: Peter Gerdes | May 15, 2018 6:56:38 PM

The comments to this entry are closed.