Welcome to our latest Ethics discussion forum, on Kieran Oberman’s “Freedom and Viruses,” kicked off with a precis by Ian Carter. Take it away, Ian!
In this excellent article, Kieran Oberman aims to contest the common view, encountered in much of political discourse, that the lockdown, in the recent pandemic, traded off freedom against other values, such as health or economic prosperity. He pursues this aim with admirable clarity and rigor, assuming more than one possible conception of freedom and demonstrating in each case the falsity of the claim that lockdowns sacrifice freedom for other goods.
Kieran’s first and most important point, which covers the first eight sections of the article, is about the effect of lockdowns on negative freedom. He rightly gives more space to this conception because it is the one implicitly assumed in claims about freedom being traded off.
Negative freedom is here conceived as the absence of external constraints imposed by others (intentionally or unintentionally). It is freedom understood as a social relation. Simplifying somewhat, these constraints make it impossible for the agent to do certain things, or sets of things, and might consist either in physical barriers or the threat of sanctions. On this conception of freedom, lockdowns certainly restrict freedom to an enormous extent.
Naturally caused diseases, on the other hand, make people unable to do certain things, but not unfree to do them, for it is only actions of other people that count as sources of negative unfreedom. So it looks as if viruses, like other naturally caused obstacles, don’t restrict negative freedom. And this explains why many people see the issue as a trade-off between health and freedom: lockdowns restrict negative freedom, viruses do not.
As Kieran points out, however, spreading a virus is an action performed by someone. And viruses disable people, making many actions impossible or costly; or they kill people, drastically restricting their action possibilities to those occurring within a shorter lifespan. If I start an avalanche, disabling or killing some rock climbers down the hill below me, I harness the forces of nature, intentionally or unintentionally, in a way that removes action possibilities on their part. Those forces of nature would not have such an effect if it weren’t for my action of starting the avalanche. The action of moving around in a crowded space when I have a potentially lethal virus is similar to my action of starting the avalanche.
Kieran discusses, and rebuts, a series of objections to his claim that ‘viruses restrict freedom’. Here are the ones I consider most important, or at least most helpful in further clarifying his position.
First, it might be objected that viruses are really internal constraints, not external ones, since they occur inside the agent’s body. The answer to this objection is that we need to look at sources of constraints, not just at their location. If I shoot you in the leg, making it impossible for you to walk, we would not deny that I restrict your freedom, even though the bullet that disables you is located inside your body. The location of the constraint is internal, but the source is external and consists in the action of another person. Spreading a virus is no different.
Another objection might be raised by appeal to a more refined analysis of the sources of interpersonal unfreedom. Some theorists of negative freedom claim that obstacle-causing actions restrict freedom only if someone can be held morally responsible for their obstructive effects. If we assume this moral-responsibility account of the sources of unfreedom, it might seem that we can only claim that virus-spreading actions amount to restrictions of others’ freedom if the spreaders realize exactly what they are doing. However, even on the moral-responsibility account, virus spreaders are generally morally responsible for the effects of their actions: if they do not know what they are risking, they ought to know. (There will be exceptions, as at the very beginning of the covid pandemic when people lacked the relevant knowledge.)
To be sure, potential victims are risking infection when they go out. And one might object on this basis that the victims’ disabilities are self-inflicted rather than brought about by other people: if they hadn’t gone out, they would not have become infected. But, as Kieran points out, while the potential victims take a risk in going out, they themselves do not create the risk. Imagine that the land around your house were full of dangerous holes previously dug by other people. You can’t go out without risking falling into one of the holes. Still, if you choose to go out, and fall into one of the holes, and break your leg, your inability to walk counts as an unfreedom created by those who dug the holes. The same point applies to the risk of infection when you go out.
Kieran concludes that, while lockdowns restrict freedom, so do viruses. One might contest the wording of this conclusion. The actions of virus-spreaders restrict freedom. Viruses themselves don’t restrict freedom. Kieran’s answer to this objection is that, if we can’t say that viruses restrict freedom, then we can’t say that prisons or lockdowns restrict freedom either. Well, I don’t think prisons restrict freedom; I think prison guards do. Lockdowns, by contrast, are sets of coercive actions. Viruses are not sets of actions, though they are spread by actions. As I said earlier, viral infections, like avalanches, are forces of nature that human actions might or might not harness. Naturally caused avalanches do not restrict negative freedom, and similarly a naturally caused virus does not restrict the negative freedom of patient zero. But this is a minor quibble on my part. The substance of Kieran’s arguments about the effect of virus spreading on negative freedom seems to me correct.
Assuming, with Kieran, that a lockdown effectively limits the spread of a disabling virus, its overall effect on freedom is somewhat complex. Lockdowns restrict freedom, but they also limit the restrictions of freedom created by viruses. Moreover, if lockdowns make it safer to move around when you do go out, then they preserve certain freedoms that would otherwise be restricted. So there is no simple trade-off. Kieran is surely right that we should not simply compare freedom under the lockdown to freedom before the pandemic. Instead, we should compare the sorry state of freedom under the lockdown with the sorry state of freedom under the pandemic unchecked by the lockdown. This is more tragic, but also more realistic.
All of the above discussion concerns nonnormative freedom, that is, freedom understood as a fact about the physical prevention of actions. Perhaps, Kieran wonders, the freedom-based objection to lockdowns can instead be formulated in terms of normative freedom, by which he means, more specifically, moral freedom. Moral freedom is the presence of moral permissions, or the absence of moral prohibitions. The idea, then, is that lockdowns restrict people’s moral freedom by creating, through an exercise of moral authority, moral obligations to stay at home. To be sure, people already have moral obligations to engage in social distancing before the exercise of this authority takes place, but lockdowns can create new duties to ensure coordination, protect the most vulnerable, and so on. So lockdowns do seem to create moral unfreedoms without creating any moral freedoms.
Kieran has some sympathy with this argument, as he considers moral freedom to be important in itself. He also notes, correctly, that he differs from most theorists of freedom in this sense. Most political philosophers who theorize about freedom – myself included – are interested primarily in nonnormative freedom. The reason for this is that we consider nonnormative freedom to be a value on the basis of which we can assess certain structures of normative claims, liberties, powers and immunities – in short, certain structures of rights, whether moral or legal. Nonnormative freedom is more basic, normatively speaking. I don’t really see a challenge to this view of the primacy of nonnormative freedom in Kieran’s argument, despite his professed belief that moral freedom is important in itself.
In any case, Kieran thinks that even the objection to lockdowns based on moral freedom ultimately fails. The reason is that he thinks viruses, no less than lockdowns, restrict moral freedom. Viruses impose all the social-distancing duties, mentioned above, that preexisted the lockdown. Moreover, the lockdown ‘liberates’ us from some of those duties: by flattening the curve, it introduces the prospect of a return to more normal, less socially distanced relations.
Despite caring less than Kieran does about moral freedom in itself, I am not entirely convinced by his argument that viruses restrict moral freedom in no less serious a way than government-imposed lockdowns do. Kieran holds that it is not worse to have one’s moral freedom restricted by a virus than by an exercise of authority. I am not sure why he thinks this. I would have thought that, just as in the case of nonnormative freedom, we resent restrictions created by other people in a way that we don’t resent restrictions created by nature. This distinction forms the very basis of the negative conception of nonnormative freedom understood as a social relation. Why should normative freedom be different?
Finally, Kieran asks whether one can complain that lockdowns produce an overall reduction in freedom if, instead of the negative conception, one assumes the republican conception of freedom as nondomination. On the republican conception, in order to be free, one must not merely enjoy noninterference on the part of others, or a high probability of such non-interference; that noninterference must be guaranteed, by suitable constraints on possible interferers.
Far from generating a complaint about lockdowns, Kieran argues, the republican theory supports lockdowns. But there is more: it in fact takes us too far, as the theory supports lockdowns even when they seem to be unjustified – as, for example, where voluntary social distancing would be sufficient to limit the spread of the virus. On the republican theory, before the lockdown each infectious person is “dominating” those who might be infected. On the negative conception of freedom, those infectious people are not restricting others’ freedom if they voluntary stay at home or take other measures to avoid infecting others. But for the republican, this behavior is insufficient to guarantee freedom as non-domination: the infectious, indeed all possibly infectious people, need to be constrained against infecting others. This is an interesting implication, and a helpful contribution to the so-called ‘liberal-republican’ debate.
The implication amounts to a critique of the republican conception – a critique I’m very sympathetic to, having published arguments against the coherence of the ideal of ‘freedom as non-domination’ that point in a similar direction. The requirement that non-interference be not only highly probable, but also guaranteed by people’s incapacity to interfere, is pointlessly overdemanding and in many ways counterintuitive. ‘Republicans should feel uncomfortable about supporting lockdowns that are unnecessary to combat viruses’ – a nice example of British understatement.
The article contains a brief final section on an empirical question that the above discussion ignores: lockdowns are just one of an extensive set of government measures that have restricted freedom over the last two decades. The argument here is similar to that advanced by some critics of the security measures that followed the terrorist attacks at the beginning of this century: once governments acquire new powers, they are very reluctant to given them up, and are indeed happy to expand them further. Moreover, lockdowns have ‘normalized’ restrictions of freedom that were unthinkable before the pandemic. As a result, people have become more tolerant of such restrictions. I see this change among my students in Pavia when I ask them to vote, as I have done over the years, on the acceptability of restrictions of Rawlsian basic liberties if such restrictions would permit the pursuit of certain other social goals: their tolerance of unfreedom has increased notably over the last three years.
While he wisely avoids pronouncing on this empirical question, Kieran seems mildly optimistic about the long-term noxious effects of lockdowns compared to more targeted measures such as selective quarantining or surveillance. Lockdowns are at least visible and ‘create a large constituency with an interest in their removal’. Fair enough. But I remain less optimistic, because of their normalizing effect.
More generally, I think this slippery-slope empirical argument is the one that has most potential as a freedom-based complaint against lockdowns. It assumes the negative conception with which we started, and it survives all the purely conceptual objections that Kieran carefully and correctly rebuts. It also suggests taking very seriously the hypothesis, touched on only incidentally in the section on republican freedom, that in terms of the long-term overall effect on freedom, non-coercive guidelines might sometimes be preferable even where the immediate balance in terms of freedom seems to favor a lockdown.