Skip to content

McMahan’s Hazardous (and Irrelevant) Thought Experiment

Written by Professor Allen Buchanan and Professor Lance K. Stell  

 

This is a response to an earlier post, by Jeff McMahan, about the right to carry guns, https://blog.practicalethics.ox.ac.uk/2015/04/a-challenge-to-gun-rights.

Before we criticize McMahan’s argument, it is important to ascertain its implications: Assuming that, as McMahan thinks, there is no moral right to gun ownership, what follows, practically speaking? One might think that, given the number of gun deaths, it follows that there should be a legal ban on gun ownership. As we shall show, however, that conclusion does not follow.  Whether or not gun ownership should be banned is independent of whether there is a moral right to gun ownership. We will show that McMahan has not established that there is no moral right to gun ownership, but that even if he had, he would not have thereby shown that there should be a ban on gun ownership.

He claims that the case for gun ownership rests on two claims, one empirical, the other moral. The empirical claim is that we are safer the more guns there are in private hands.  The moral claim is that if someone suffers from an unjust attack under conditions in which the state has denied her access to guns, then the government has violated her right of self-defense.   But neither claim is needed to make the case in favor of gun ownership, as will become clear.  To make a sound case for gun ownership one needs only (1) a presumption that people should not be deprived of the means for exercising the right of self-defense, (2) a rebuttal of the claim that the presumption is defeated.

McMahan tries to refute the factual claim by recourse to an analogy between prisons and society.  He says that that allowing inmates to possess guns would make them all less safe and that this shows that by depriving them of access to guns the state does not violate their rights of self-defense.  The analogy is flawed.  First of all, one could argue that, by virtue of their crimes, inmates have forfeited some elements of the right of self-defense, including the right to access to firearms. (In fact, in U.S. law, inmates have no right of self-defense. Rowe v. DeBruyn, 17F.3d 1047 (1994)).  McMahan swiftly dismisses this possibility with the remark that no one can forfeit his right of self-defense against unjust attacks.  He gives no reason why this is so, but the question is not whether one can forfeit one’s right of self-defense against unjust attacks, but whether one can forfeit one’s right to access to guns.  Further, if it were true that one cannot forfeit one’s access to means for successful defense against unjust attacks, then it would follow that imprisonment massively violates the right of self-defense, a conclusion many would find counterintuitive and which undercuts McMahan’s prison analogy strategy.

Second, the situation of inmates is different from that of the general population: They are in a state of abject dependence on the authorities. The law recognizes this by acknowledging a duty of care and protection on the part of the authorities (Collingnon v. Milwaukee Co., 163 F.3d 982 (1998); DeShaney v Winnebago Co. Dept. of Social Serv., 489 U.S. 189 (1989)).  In contrast, the law explicitly denies that the police have a duty to protect individuals against lethal assault even in a courthouse (Zelig v County of Los Angeles, 27 Cal.4th 112 (2002)), presumably in recognition of the fact that the police cannot fully protect us.   Where the state has a duty to protect a group because of its special condition of abject dependence, it makes good sense to say that they need not have access to the means of self-defense; but that is compatible with saying that when individuals are not in a condition of dependency and where the police have no duty to protect them, they should have access to means of self-defense.

Third, the prison population is relevantly different from the population at large:  It contains a higher percentage of persons who have committed murder or assault.  McMahan ignores differences between inmates and the general population and, more importantly between perpetrators of gun violence and the general population, because he thinks that mere access to guns makes ordinary people dangerous—that we would frequently be overcome by passion and engage in gun violence if we had access to guns.  McMahan cites no evidence to support this prediction and the evidence speaks against it.  According to the Department of Justice, the U.S. homicide rate has declined by 50% since 1980 while gun ownership has skyrocketed.  Further, every state now allows concealed carry of firearms, with the number of ordinary people carrying guns increasing dramatically, yet there has been no increase in gun violence.  These facts show that most of us are not potential murderers, poised to explode in homicidal rage, thwarted only by our lack of access to guns.

The fourth flaw in the prison analogy is that prisons are tightly controlled environments in which the authorities can prevent at least some of the bad unintended consequences of a ban on gun ownership in society at large.  Later we will explore some of these, but for now it is enough to appeal to another analogy:  Prohibition.  The ban on alcohol in society at large had quite different and much worse consequences than the ban on it in prisons.  So we shouldn’t assume that banning guns in society at large would be as beneficial as banning them in the tightly controlled prison environment.

McMahan thinks that banning guns in prisons is justifiable because doing so makes inmates safer than they would be if they were allowed to own guns.  Since he offers no qualification, he appears to be committed to the thesis that the state can justifiably deprive one of something whenever one would be safer without it.  That is repugnant, since it implies that the state may justifiably ban all sorts of items that we all think it shouldn’t, including skiis, cars, airplanes, power tools, swimming pools, squash raquettes, alcohol, and libidos.  All of these items are dangerous and some (especially cars) produce as many deaths as firearms.

As surprising as his failure to acknowledge any of these relevant differences between prisons and society is McMahan’s silence on the problem of unintended bad consequences of a ban on gun ownership.  These include a flourishing black market for guns, with predictable burgeoning of organized crime and corruption of law enforcement personnel (the analogy with the war on drugs is apt here).  In addition, it is likely that law-abiding citizens would become less safe than scoff-laws, which seems unfair.  McMahan ignores the Problem of the Second Best:  He wrongly assumes that if a fully effective ban (as can be achieved in a prison) would be a good thing, then the attempt to ban guns (in society) would also be a good thing.

What of McMahan’s criticism of the moral claim?  He thinks it’s obvious that if we would be safer without gun ownership, the state does not violate our rights of self-defense if it deprives us of access to guns in situations in which we suffer violence that we could have avoided if we had had access.  The problem is that he has not shown that we would be safer if there were a ban on gun ownership; at most he has shown that we would be safer if there were a fully effective ban combined with fully effective police protection against assault.   So he has not refuted the commonsense idea that if one has a right of self-defense, then no one should deprive one of the effective means for self-defense, if the government cannot fully protect one.  Even if we would all be safer if there were no guns in private hands (and assuming effective protection by the police), it is hard to see how that is relevant to whether we should attempt to ban private ownership, since any such attempt will not result in their being no guns in private hands but will result in some rather severely negative unintended consequences.  And given that, as the statistics show, more guns does not mean more violence, we can’t assume that fewer guns (as opposed to the unattainable no guns) will make us safer, if we take into account the threat to our safety posed by the black markets, organized crime, and corruption of law enforcement that would predictably result from a ban on gun ownership.

Share on

12 Comment on this post

  1. Two points in reply to Buchanan and Stell. One minor and one major.

    “These facts show that most of us are not potential murderers, poised to explode in homicidal rage, thwarted only by our lack of access to guns.”

    That is a straw man.

    “McMahan ignores the Problem of the Second Best …”

    Starting in that paragraph the authors write as if the United States is the only nation on earth and omit any comparison to different gun control policies in other nations. But in fact there are other nations that have much stronger gun control laws that disallow concealed carrying and that also have markedly lower numbers of gun related killings per capita compared to the US. If they can do it why cannot the US do it?

    1. Dear Ed;

      The question we addressed (and McMahan addressed) is not whether more gun control would lower gun violence rates, but whether a ban on private ownership would be justified.

      all the best,
      Allen

      1. Dear Allen,
        Your reply does not engage with the two points of objection I levelled at your argument, both of which I do think are on topic.
        1 regardless of what question you adress you should not engage in straw manning. Are you of another opinion?
        2 I am aware that you adress the question “whether a ban on private ownership would be justified”, but in doing so you make assertions about factual matters concerning guns och gun violence. E.g. this snippet from your text: “… given that, as the statistics show, more guns does not mean more violence, we can’t assume that fewer guns …”. If you wish to make such arguments, in the process of adressing the main question, then you should engage with international comparative empirical research on gun control.

  2. Julian Savulescu

    In 2007, Norway had 31.32 firearms per 100 people, compared to the U.S. rate of 88.82 firearms per 100 people. There is roughly 3 times as many firearms per person in the US compared to Norway. Yet the murder rate in the US is around 40 times as high. Norway has one of the highest rates of gun ownership in the world and also one of the lowest rates of murder. I am not necessarily supporting gun ownership but it is a difficult topic to rationally debate and I would like to thank Jeff, Allen and Lance for contributing to such a debate.

  3. I think Julian Savulesco is right to draw attention to empirical issues and nation comparative statistics.

    One possibility consistent with that statistic is that the number of firearms per 100 people is a positive causal factor for the murder rate but that there is also (a) in the US additional causal factors that explain the comparatively extremely high murder rates there and (b) in Norway additional causal factors that explain the low murder rates in Norway compared to some other non-US countries with a similar frequency of firearms as that in Norway. One further possibility is that factors (a) and (b) are the same, Norway being a paradigmatic case of the high social equality nordic universal welfare state system and the US being the paradigmatic case of high social inequality thin welfare state capitalism.

    Another material of relevance for this topic is Professor David Hemenway surveying of “scientists who were publishing relevant articles were from the fields of criminology, economics, public policy, political science and public health” on gun policy topics. A popular overview to some of the results was recently published in LA Times, http://www.latimes.com/opinion/op-ed/la-oe-hemenway-guns-20150423-story.html

  4. Second, the situation of inmates is different from that of the general population: They are in a state of abject dependence on the authorities. The law recognizes this by acknowledging a duty of care and protection on the part of the authorities (Collingnon v. Milwaukee Co., 163 F.3d 982 (1998); DeShaney v Winnebago Co. Dept. of Social Serv., 489 U.S. 189 (1989)).

    I’d be interested to hear a further analysis of how the cited court cases help your point. Neither of them seems to be directly related to the authorities’ obligation to protect prisoners from assault by other prisoners; they and their citations seem mostly to deal with an obligation to provide prisoners with some medical care and reasonably safe conditions.

    But even if there is a principle in US law that the authorities are obliged to defend prisoners against attacks, it bears little on McMahan’s thought experiment about self-defense in prisons, because in fact the authorities notoriously do not defend prisoners against attacks. Rape and assault are common in US prisons. So the claim that “prisons are tightly controlled environments in which the authorities can prevent at least some of the bad unintended consequences of a ban on gun ownership in society at large” is, frankly, morally myopic–as if people outside prison need means of self-defense that prisoners do not because the prison guards do such a superb job of defending the prisoners! Aside from McMahan’s arguments, it does no one any good to minimize one of the great human rights crises in the US in this way.

    As for the conclusion of the post, in order to effectively make the case against gun control (as opposed to the case against McMahan’s argument), it would require argument that guns are an “effective means for self-defense,” and it certainly would require argument that “it is likely that law-abiding citizens would become less safe than scoff-laws.” We should also consider the negative effects of the widespread use of guns as a means of self-defense; unarmed people like Rodrigo Diaz, Yoshihioro Hattori, T.J. Darrisaw, and Renisha McBride were all shot by people who claimed to be acting in self-defense, and the culture of armed self-defense also contributes to the trigger-happy mindset that led to the killings of Trayvon Martin and Jordan Davis.

    1. Our point wasn’t that guards do a great job of defending prisoners. It was that because the prison environment is tightly controlled in some respects, some of the bad unintended consequences that are likely to occur with a ban on guns in society may not occur in the prison environment. In contemplating the wisdom of a ban on private ownership of guns in society at large it is important to avoid two mistakes: reasoning as if the ban would be effective in eliminating all or even most guns and would be accompanied by effective protection of citizens by police (as in McMahan’s example of a third party depriving two people of their weapons and then protecting them) and reasoning as if there are no predictable serious bad unintended consequences of the attempt to ban private ownership of guns in society at large. Our point about prisons being tightly controlled was addressed to the question of unintended bad consequences; we weren’t (stupidly) saying that prisons do a good job of protecting prisoners.

  5. I’ll reiterate a point I made in response to Prof. McMahan’s original post: We should read his argument narrowly, as a negative claim against one particular line defending gun rights, rather than a positive claim establishing that we should ban/heavily regulate guns. This narrow moral argument infers from the existence of the right to self-defence (e.g., cops shouldn’t bat away the gun you’re using to defend yourself) to the existence of the right to possess firearms. (there’s an analogous empirical argument, but I’ll focus on the moral since it’s more easily evaluable from the armchair) The success of McMahan’s moral argument rests on whether he convincingly defuses that line of thought, not whether he establishes that we should ban/regulate guns. (I’m sure McMahan supports such a ban, but the full argument for that position is outside the scope of his short blog post)

    In arguing against prisoners’ rights to self-defence, Profs. Buchanan and Stell bring up Rowe v. DeBruyn, which denied prisoners a constitutional right to self-defence. But that’s not a very good example in case law for gun rights activists, as that ruling denies that the right of self-defence is implied for *anyone* by the US constitution: “we find no precedent establishing a constitutional right of self-defence in the criminal law context.” The case is further complicated by not being a criminal case anyway, but a prison discipline case, so the same decision would’ve been reached even if Rowe had a complete right to self-defence.

    Moreover, I presume McMahan’s response to that and the further case law Buchanan and Stell cite would be that those cases to not bear on the *moral* existence of a prisoner’s right to self-defence (nor the moral duty to protect that officers have). And the case for the existence of that right is meant to be as intuitive as the general existence of the “common sense” right. Even in prison, Rowe (e.g.) was perfectly within his *moral* rights (though apparently not his legal rights) to smack the unjust assailant, Evans, with the hotpot. And this is important, because the argument McMahan responds to derives gun rights from this more general right to self-defence against unjust attacks that Rowe and other prisoners intuitively retain.

    Furthermore, considering the principle “if one has a right of self-defence, then no one should deprive one of the effective means for self-defence, if the government cannot fully protect one”: McMahan’s prison example has (moral) force insofar as it fulfils the antecedents, thereby implying the (absurd) consequent. Prisoners retain the (moral) right of self-defence, as Rowe was morally permitted to hit Evans with the hotpot; and officials cannot fully protect prisoners, evidenced by instances of prison violence like Rowe’s. Therefore, by the principle, they would retain the right to possess firearms, an effective means of self-defence. This is I take it meant to be reductio of the principle – though perhaps modified versions of the principle (esp. the ‘fully protect’ clause) can avoid this.

  6. Some responses to the contribution that Lance Stell and I made wrongly assume that we are arguing against gun control. We are only arguing that McMahan has not made the case for a ban on private gun ownership.

  7. “We are only arguing that McMahan has not made the case for a ban on private gun ownership.” Philosophers dancing on the head of a pin, then – and also somewhat disingenuous. If you’re so concerned to limit the scope of what your argument is trying to do, then you ought to be equally scrupulous in recognising what McMahan was trying to do.

    He wrote:
    “The case for gun rights rests primarily on two claims…The claim about fact is that members of society as a whole are safer when more of them have guns…The claim about principle is that each person has a right of self-defense and that this right entails a further right not to be deprived of, or prevented from having, the most effective means of self-defense.”

    The word “ban” doesn’t appear in there. If your argument is intended to show that “McMahan has not made the case for a ban”, then you have directed it extremely poorly. Personally, I thought McMahan’s argument was pretty weak, but the gun advocates seem to always be able to go one step lower.

  8. I should have quoted one more line from McMahan to make clear what he said he was doing: “I will, however, present a…challenge to the central claims of advocates.”

  9. I agree that McMahan didn’t argue explicitly for a ban on private gun ownership. We wanted to make it clear that his argument doesn’t support such a ban and that, in order to argue for continuing private gun ownership one needn’t rely on either the empirical claim or the moral claim he identifies.

Comments are closed.