Hard Determinism and Distributive Justice

I know this is going to seem weird, but here's a post from me with at least some philosophical content. This is mainly in response to Tamler's prior post, concerning the consequences of hard determinism, but I thought it was lengthy enough to merit a post, and besides, now we can claim that our blog includes posts on political philosophy, too!

Anyway, Tamler sez this: "Suppose for a moment that the hard determinist thesis is correct. Determinism is true and conseqently we do not have free will and no one is robustly morally responsible--deserving of blame or praise--for anything. Gardeners: what do you think follows from this state of affairs? How would it affect (a) your individual perspective on life, assuming the world stayed pretty much the same as it is now, and (b) social policy and legislation in a society that attempted to be consistent with hard determinism?"

With respect to (a): I sez "It wouldn't (much)," and with respect to (b) I sez "I don't know, but below are some thoughts."

First though, an entirely predictable explanation of my response to (a): I'm inclined to think that if incompatibilism and the thesis of determinism are both true, that we would have very good reasons to be revisionists. That is, I think that praise, blame, and maybe even desert (that's more complicated- see below) can be justified on grounds that are independent of a libertarian conception of agency. (Note: it is a further issue whether commonsesense is libertarian. Though I think most of the commentators on Tamler's original post were presuming that at least by hypothesis commonsense is libertarian, you don't have to think that. It is consistent with the assumptions as stated that (i) commonsense is compatibilist but (ii) the facts of the matter are hard determinist). All of this is not to say that post-hard determinist revisionism is free of consequences. If commonsense is libertarian, then one consequence is that commonsense will need to get re-arranged if it is going to reflect the appropriateness of revisionism. And, I think that - at least at the margins- there will be some differences in the our social/moral practices. But, I do think most of those will be largely at the margins, and that the bulk of our responsiiblity practices can survive on grounds independent of libertarianism, even if (1) incompatibilism is true and (2) the thesis of determinism is true. So, given that I'm already a revisionist, learning about the truth of a view that ought to lead to revisionism won't really require much change on my individual perspective on life. (Though I would be suprised to learn that determinism is true).

Now for the (maybe) more interesting part promised by the title: This afternoon I was talking with an excellent former student of mine (Katie Richardson) and our conversation lead me to think about something Sam Scheffler has written about: in contemporary political philosophy, a great many theories of political justice have shied away from relying on a notion of pre-institutional desert. However, there are some theories out there that seem to rely on a pretty strong notion of pre-institutional desert. Perhaps the most well known variety is, as Katie pointed out . . . libertarianism of the political sort.

Several things:
(1) First, is there some interesting historical connection between libertarianism (metaphysical) and libertarianism (political) that accounts for the names? Until this afternoon I've assume the answer was no, but now I wonder. Anyway, where the heck did libertarianism of the metaphysical sort get its name? As you all know, I am not much of a historian in this domain and would appreciate some edification by the more historically minded members of our blog.

(2) It strikes me that the truth of hard determinism might have some significance for social policy and legislation, on the assumption that varies pieces of legislation and social policy currently presume some degree of pre-institutional desert of the sort implicated by hard determinism. Unless we could find independent grounds for a commitment to pre-institutional desert (and maybe we can- I haven't tried, though) we'd have to chuck any policy or justifications that presuppose pre-institutional notions of desert. Minimally, it would mean some trouble for the already beleagured Libertarian Party. But it might have related implications that show up at least implicitly in some how people formulate policy about, for example, poverty and homelessness, not to mention debates about whether Shaq deserves his paycheck.

(3) However, it may also be relevent that a good many theories of political philosophy do not rely on pre-institutional notions of desert and responsibility. (As an aside, it may also be worth recognizing that many of these accounts are not consequentialist accounts- so for at least a large range of normative issues, the truth of hard determinism does not obviously entail purely consequentialist justifications for normative practices. I'd even argue that the praise and blame in a post-hard determinist sort of way can be justified on grounds that are compatible with both consequentialist and non-consequentialist theories . . . but that's an argument for another day.) Why does it matter that many contemporary accounts of distributive justice do not rely on a notion of desert? Well, what it might mean is that (at least in something like a well-ordered society), hard determinism wouldn't have an impact even for significant social policy. Of course, we don't live in one of those societies, but the difference between us (in the hard determinist world we're imagining) and the right kind of society isn't so much the difference made by the true metaphysics of agency, but merely the gap between the society we live in and the society that we ought to (politically speaking) live in.

There is lots more to say about this, but I'm up past my bedtime.

Comments

[Dan Speak]

Manuel,

You have a bed time? I didn't even realize you slept.

In light of your reflections, here's a case that comes to mind. Nagel has this argument for the permissibility of affirmative action policies that turns on the failure of capitalism to track merit in the distribution of social goods. As I recall (and I'm not presently looking at the paper), it's a kind of natural lottery argument. The opponent of AA claims that it violates stringent rights based on merit. Nagel retorts that capitalism is inherently unjust because its distributions turn on irrelevant factors like being born Shaq-like or remarkably smart (aka Manuel-like). Apparently, relevant factors for the determination of distributions would have to connect up less accidentally to the agents efforts of will. This looks a lot like an expectation of free will in the assessment of "true" merit.

In the end (again, I'm trying to pull this out of the memory bank... frequently depleted), Nagel claims that it is just to reject an unjust system. Therefore, it is just to reject capitalism with its institutional conception of merit. Therefore, it is just to endorse policies of affirmative action. If I've recreated the argument correctly, it looks like one that would be put in jeopardy by hard determinism. That's because "luck in determination" would infect EVERY distribution of social goods. There would be no moral difference (in Nagel's sense) between distributions that tracked "real" merit and those that didn't. It would be the natural lottery "all the way down".

[Tamler Sommers]

Dan,

Nagel's argument would undermine the libertarian's attempt to justify inequality. But using a luck-neutralizing approach to ground a POSITIVE argument for affirmative action runs afoul of what S.L. Hurley calls “the egalitarian fallacy”—namely, the claim that if it's a matter of luck that two people are unequal, then it would not be a matter of luck if they were equal. Hurley points out that we have no more right to draw this inference than we would the reverse: from “it is a matter of luck that people are equal,” to “therefore it would not be a matter of luck if people were unequal.” (Shameless self promoting plug: see my review of her book Justice, Luck, and Knowledge in the May 7, 2004 TLS.) Hurley concludes, rightly I think, that egalitarian theories need to be established on other grounds.

Manuel,

That's interesting stuff and I agree that there's a lot more to be said. I had a student who worked on this question as well. He examined the extent to which state and federal law, as well as sentencing guidelines, relied on the notion of desert. His preliminary findings suggested that federal laws relied very little on this notion. States vary of course, but there he found more of a reliance on non-consequentialist considerations--both in the laws and the sentencing guidelines. States allow the judges more flexibility in setting the length of the prison terms according to the criminal's blameworthiness.

[Tamler Sommers]

I should rephrase the last sentence: States allow the judges more flexibility in setting the length of the prison terms according to (what the Judge DEEMS to be) the criminal's blameworthiness.

[Mark Smeltzer]

Tamler,

I think that's pretty interesting -- that judges can affect the sentencing based on the degree of blameworthiness they perceive in the culprit. On the HD model that kind of sentencing would be unjustified. So, does that mean that in order to "neutralize" the role of beliefs regarding MR within the sentencing process that only pragmatic concerns could be taken into account?

If so, would that mean judges would be justified in putting away (i.e. for life) repeat offenders for petty crimes if the judge discerned (rightly) that society would be "better off" without the offender? Or, on the flip side, if a valued member of society committed a heinous crime (such as rape, or murder), could the overall pragmatic "penalty" to society (caused by this person’s absence) justify a judge in acquitting this person or giving them a very mild sentence (such as community service, a fine, or a short period in jail)?

It seems to me that these kinds of scenarios would necessarily follow (to varying degrees) when the HD thesis is applied to the sentencing process.

[Tamler Sommers]

Mark,

I think you might get some fairly counterintuitive results from a hard determinist criminal justice system, but the ones you mention strike me as implausible. It can't be good (pragmatically speaking) to send the message that if you're a valued member of society, you can rape or murder whoever you want.

[Mark Smeltzer]

Okay, so let's assume that the judges will make their rulings with respect to a rule based pragmatism, instead of an act based pragmatism. Could you give some examples of fair sentences in the cases I mentioned above?

The first was the case of a person who greatly benefits the society and has been convicted of a serious crime like murder or rape (if it makes a difference you can pick one). The second was the case of a person who detracts from the society (let's suppose the person is a lazy drunkard who can't keep a job and leaches the welfare system) and has been convicted of a petty crimes like shop lifting.

[tnadelhoffer]

Mark,
You may be guilty of turning determinism into a bogey-man solely based on your own intuitions and beliefs about free will and moral responsibility. In any event, I see no reason why one could not opt for a system of criminal law that (a) lives up to the ideals of the rule of law (b) that is based entirely on strict liability. In such a system, it would not matter whether you were famous or not--everyone would get the same punishment. Moreover, the system would not depend on murky mental state--i.e., mens rea--language such as intent, intentional, knowingly, etc. Of course, you might find such a system to be unacceptable. But if you object to this type of system based on principles of desert and fairness, you may simply be begging the question by assuming that in some robust sense humans sometimes deserve--morally speaking--to be punished and that the purpose of the criminal law it to give people what they deserve. But even if determinism is true, people could still justifiably be punished because they are physically/causally responsible for breaking the law, but not because they are morally responsible. In this respect, people would be treated like shoppers--buyers beware, if you break you pay. In a system such as this, it wouldn't matter whether you meant to break the law only that you did break the law. Now without begging any questions concerning MR and desert, tell me why this sort of system is not compatible with determinism. As far as I can tell, a strict liability system would solve your concerns about punishing people of varying social importance differently for committing the same crimes. But I admittedly think that retributivists are often guilty of turning consequentialist theories of punishment into bogey-men as well. But that can wait until a later post.
Thomas

[Mark Smeltzer]

Thomas,

Given the following, "But even if determinism is true, people could still justifiably be punished because they are physically/causally responsible for breaking the law, but not because they are morally responsible," could you give me your input on the following thought experiment?

Let's suppose that a man is walking along the beach on the coast of Africa and a rock falls off the bluff and hits him on the head. He is good consequentialist and decides that since the rock might hurt someone else again in the future, he ought to do something with it.

Consider in one case that the rock is actually a valuable diamond, and in the other it is just a chunk of worthless sand stone. I think he would be justified in smashing the sand stone, but would be an idiot if he did the same thing with the diamond because the diamond is valuable despite the pain it caused him.

This example isn't exactly analogous with our questions about penalties in a just consequentialist society, but I think it is nonetheless telling.

You may be right that I'm chasing bogey men here, but I don't see how you've dispelled them yet. You said, "In a system such as this, it wouldn't matter whether you meant to break the law only that you did break the law." But unless you give up your primary concern for a consequentialist concept of the good, you'd be heading for a baseless deontological system.

If consequentially justified rules are meant to produce the highest overall benefit for a group, they must produce the highest overall good, or they be declared unjust.

In the first case I posed to Tamler I was not suggesting that the first person was famous. In fact, I think its better to presume that the person isn't famous -- at least not in the popular sense. Let's assume this person is a doctor who's close to a break through in cancer research, is best friends with the Emperor of China and is the only person who's able to keep them from invading the USA, etc.

All I said that the first person was valuable to society in a real and apparent way, whereas the second person is not.

Moreover, your suggestion of a "system of criminal law that ... is based entirely on strict liability" is subject to NUMEROUS straight forward counterexamples.

#1:

* Law: it is unlawful for any person to steal another person's property. The sentence for doing so is X.
* Situation: A thief wearing a mask has a man's wife hostage and orders you to steal something for him or he will kill your wife. The man is caught in the act and the thief flees the scene.
* Result on your system: the man is convicted, and given the standard sentence X; regardless that his action was compelled by the thief.

#2:

* Law: It is unlawful for any person to harm another person. The sentence for doing so is X.
* Situation: one person runs by and shows another person such that the second person knocks into an elderly person and causes this person to suffer a broken hip. The "shover" keeps running is not caught, but the "knocker-overer" is caught.
* Result according to your system: the "knocker-overer" is convicted and given sentence X; regardless of the fact that it "wasn't his fault".

[tnadelhoffer]

Mark,
First, you are correct in pointing out that your rock/diamond example is disanalagous to the issue of justified punishment in a deterministic world. Indeed, I would like you to say a little more about what exactly the example is "telling" us.

Second, If you adopt rule consequentialism (RC), then it is easy enough to explain why you the following rule would NOT maximize the good:

All other things being equal, doctors should be punished less severely than used car salesmen.

Minimally, this kind of rule would encourage doctors to worry less about obeying the law and it would encourage more ethically suspect individuals to practice medicine. The same can be said about nearly all legal rules that would violate the basic principles of the rule of law. For instance, given that all legal rules must be publicized (owing to purely consequentialist considerations), certain rules would do more public harm than good. Hence, a rule that permits intentionally punishing citizens who are known to be innocent would fail a RC cost benefit analysis. Of course, I think that if you opt for non-consequentialist theories of punishment you will have your own problems with punishing the innocent, but explaining why would take me too far afield for now.

Finally, simply claiming emphatically that there are "NUMEROUS" counter-examples to strict liability does not make them so. On the one hand, there are already strict liability crimes--e.g. statutory rape. Hence, if you have consensual sex with a 12 year old girl, simply insisting that she looked very old for her age, that she had what appeared to a perfectly valid state I.D., that she promised she was old enough, etc, will not get you out of trouble. And there are good reasons for this--namely, the fact that statutory rape is a strict liability crime gives men an extra incentive NOT to have sex with women whose age is even remotely in doubt. On the other hand, the "counter-examples" you discuss smuggle in the very same question-begging assumptions that I warned against. After all, your examples are supposed to provide us with instances where people are causally responsible for breaking a law (although in the "pushing down" case, it is unclear what law has been broken) but where they do not deserve to be punished. But my suggestion is that criminal law need not be grounded in desert at all--at least not moral or retributive desert. Hence, simply insisting that the people in these scenarios don't deserve to be punished, is to deny the very thing at issue. Nevertheless, I think that we can avoid some of worries that your counter-examples are aimed at getting at by (a) defining criminal acts very narrowly, and (b) instituting mandatory senteces for all people who commit these narrowly defined acts--regardless of how valuable they are to society.

[Mark Smeltzer]

Thomas,

I may just be naive, but I am skeptical that it would be possible to define criminal acts narrowly enough so as to avoid punishing people for acts they committed under compulision (such as the criminal forced the man to steal by taking his wife hostage) or external force (such as the man who hurt an elderly woman because someone else pushed him into her).

My second example wasn't meant to represent a real, nor even a plausible, law. The mechanics of the example were the important thing. We could easily substitute in the second example things like a person who fires a gun and kills someone because another person smashes into them then runs off, or a person who lies to another person and thus gets them to perform some criminal act under false pretenses, etc.

While I definitely agree that these examples import a notion of just-desert, that was the point. The hard determinist is quick to point out that such a thing does not exist, yet when we look at these examples it is plainly clear that it would be unjust to punish these people -- they are not criminals, they are victims.

That is the crucial element of these types of cases. No matter how narrow you get, unless things like an agent's intention, motive, knowledge, and volitional state are taken into account, the system will treat victims of unfortunate circumstances as real criminals.

I don't think the hard determinist has a live option to dismiss or equivocate here. The hard determinist must be able to construct a sense that makes sense of the unfortunate victim vs. criminal distinction in light of the truth hard determinism; one can't simply deny the distinction. And if particular hard determinist accounts can't make sense of this distinction, I think that is reason enough to dismiss them.

[Mark Smeltzer]

Thomas,

I forgot one thing in my last response. I can think of at least one case where statutory rape might be legally permissible: suppose a psychopath kidnaps a young girl and a young adult male and orders them to have sex with the threat of bodily harm and/or death. Eventually the girl and the boy agree that it would be best to just get it over with, so they have sex. Afterward, the psychopath releases them and when they are debriefed by police, the girl and the boy both describe what had happened.

First, while I'm no legal expert, I would imagine that according to current law (and law in a just legal system) an act like this would not be considered criminal and the boy would not be penalized. Second, I actually would be very hesistant about even calling this a statutory rape because of the element of external compulsion -- a distinction that may not hold up on your view.

Do you think your legal model, based on strict liability, can appropriately handle this case?