This is the summer break and I’m publishing old essays written when the audience of this newsletter was confidential. This post has been originally published March 16, 2022.
This morning, I almost fell off from my chair (fortunately, I was in my bed) when I read an op-ed (gated, in French) from Cécile Philippe, head of the libertarian French think tank Molinari Institute. Philippe – whose ideas I was familiar with back in the time I was reading libertarian fora, a long time ago so – is not the kind of person from whom you would expect that she regrets the time when the great Leviathan was curbing our liberties in the name of a fight against a deadly virus. But this is exactly what she is doing in this article (I translate):
“Paradoxically, I was ultimately freer when the circulation of the virus was relatively low, thanks to containing measures, than since it has been deciding to let it circulate widely… Can we really celebrate the return of individual freedom while we are still unable to eradicate the virus, a Damocles sword above our heads? A borrowed freedom implying my and others’ loss of months or even years of life in good health… As a natural phenomenon, this virus then looked less dangerous than the coercive power of the state which was fighting it. Rather than trying to defeat the virus by all the means at our disposal – including vaccines, the goal is for some to oppose the tyranny of the state… As a lover of liberty, I think that society should focus on eradicating the virus to finally create again the adequate circumstances for making exchanges free from the virus…”
Philippe goes on to argue that the collective decision regarding the trade-off between fighting the virus and loosening the formal restrictions on freedom should consider the nature of the risk at stake. If the risk is global, as is the case considering that the more it is circulating the more the likelihood of the emergence of a more dangerous variant is high, then the use of the precaution principle can be justified. The reasoning is indeed similar to the Hobbesian disaster-avoidance principle that, I have argued, should lead us to view the goal of vaccinating everyone in the world as the absolute priority. Once the global nature of the risk is acknowledged, Philippe points out that freedom claims cannot be made without taking responsibility considerations into account: it would be irresponsible to not try to prevent a catastrophic event affecting everyone in the name of freedom.
Beyond the surprise of discovering nuanced thinking from someone who belongs to an ideological community not known for that, I find this article interesting because it illustrates abstract considerations from economists and philosophers about the nature of individual rights and the best way to formalize them. I have especially in mind here the debates in social choice theory about the definition and the measure of freedom in terms of choice opportunities on the one hand, and about the idea that rights are spheres of individual sovereignty on the other. In both cases, a point of contention is whether the characterization of freedom and rights should depend on the individuals’ preferences or not. As I am more familiar with the literature on rights, I will focus on this one here.
In his famous Paretian-liberal paradox, Amartya Sen defines individual rights in terms of private spheres such that the individual preferences over alternatives in these spheres are always decisive. Less formally, the idea is simply that freedom corresponds to the fact that an individual can make choices that others must respect. If I want to read a philosophy book rather than watching TV, then whatever others’ preferences about what I do, the collective choice should respect my preference. That means that in the state of affairs that results from the collective choice, I must indeed be reading my book rather than watching TV. To be free is just to have the right to do what I judge best in my private sphere, irrespective of others’ judgments.
This is undoubtedly a very Millian conception of rights and freedom. In particular, it captures well the Millian idea that freedom builds on the existence of a protected private sphere in which one is free to entertain all the opinions she wants. As Mill famously argued in what is now known as the “harm principle”, interference with someone’s behavior in her private sphere is justified only in cases where this behavior itself causes harmful effects on others. With this definition in hand, Sen proceeds to show that there are circumstances where the social choice cannot respect individual rights while at the meantime guaranteeing that an alternative that is unanimously less preferred than another one is not chosen. More technically, the “minimal liberalism” and “Pareto” axioms are inconsistent.[1]
Sen’s paradox has triggered impressive literature. A significant part of it has focused on Sen’s characterization of individual rights in terms of protected private spheres. What has emerged as the controversial feature in Sen’s characterization is the role played by individual preferences. Basically, rights have an effect on the social choice only through individual preferences: a right is defined by the fact that one is a “dictator” – in the Arrowian sense of the word – in her private sphere. Her individual preferences must also be the social preferences. Critics have argued that this is not how we generally understand rights. Proponents of the so-called “game form approach to rights” rather entertain what can be called a Nozickian conception where “individual rights are co-possible… the existence of these rights fixes some feature of the world” (Nozick, Anarchy, State, and Utopia, p. 67). The idea is fairly simple: individual rights just correspond to the strategy sets of players in a game. By definition, in a game, a player only has the ability to choose a strategy among her strategy set. The outcome of a game depends on the strategy profile, which itself is the result of the independent choices made by the players. The crucial point is that preferences completely disappear from the picture: rights are defined ex-ante, irrespective of the preferences players happen to have. Depending on the case, the specification of rights together with players’ preferences may lead to Pareto-inferior outcomes (as in the prisoner’s dilemma), but this is not because rights impose individual preferences on social choice.
This last characterization of rights is truly a libertarian one. The idea is that, in social life, freedom is a “side-constraint” that must be respected irrespective of the consequences. This idea will however strike most as problematic. As we see with Philippe’s op-ed, even libertarians themselves are not keen to defend it in all circumstances. The libertarian conception partly builds on another, old but controversial idea that there are “natural rights” directly grounded in our existence as human beings. Whatever we may think about the natural rights theory, it tends to be oblivious to the fact that rights are, if not socially created, at least socially enforced. That means that systems of individual rights come into existence through social practices that lead individuals to de facto recognize the existence of private spheres in which interference is not permitted without a form or another of consent. In modern societies, the existence of these systems of individual rights is tightly related to the coercive power of the state, but this is more a contingent than a necessary feature – we may imagine “libertarian” societies with non-state enforced systems of rights. Anyway, the point is that systems of rights are not given, but socially designed. Now, I would claim that one cannot understand properly the role of a system of rights and the specific shape it takes in a given society without considering the issue of externalities. This is something economists are well aware of in the case of property rights. Externalities result from an inadequate specification of the system of rights and, as far as externalities impede the proper functioning of the market mechanism, the “failure” must be dealt with by redesigning the rights system.[2] Consider otherwise the example discussed by Sen in The Idea of Justice (p. 315-6) of an “invasive action” such as the smoke of a cigarette blown in the face of a non-smoker. We may admit that the latter has a right of being free from having smoke blown in her face. Note first that it is difficult to rationalize this right without considering the corresponding detrimental outcome it protects from. Based on this, a system of rights will specify the kind of restrictions that determine what each can or cannot do. These restrictions can obviously be formalized in terms of strategy sets in a game form approach. But,
(i) It is impossible to understand the (social) choice of a set of restrictions rather than another without looking at the resulting outcomes.
(ii) It is impossible to assess the possibility that the system of rights can be implemented through social practices without looking at the individuals’ interests, desires, motivations, and beliefs.
In other words, at least in cases of externalities, it does not make sense of reflecting on individual rights without looking at individuals’ preferences. This is well put by Sen: “the choice of a game form is guided by their effectiveness in bringing about the social realization that is aimed for the sake of liberty” (The Idea of Justice, p. 316).
This is exactly the point made in the article I have started with. Liberty, in the Berlinian negative sense of freedom of interference, is a core value of liberal societies. But (as Berlin pointed out) not only negative freedom is not the ultimate value to be pursued at all costs. Its very definition and the assessment of its implementation cannot be made without any consideration for the outcomes it leads to.
[1] More precisely, they are inconsistent when combined with an axiom of “unrestricted domain” which requests that a social choice be made for any profile of individual preferences.
[2] Implicitly, implementing a tax on, say, carbon emissions consists in reducing an agent’s private sphere, the state claiming property on part of her output.
Isn’t everything anyone cares about an externality? If I am pleased by what someone else does, that is a positive externality. If I am displeased, that makes it a negative externality. If so, then pointing out their status as externalities isn’t what matters; what matters is deciding which externalities we collectively care about enough to seek recourse, and which we do not.
“negative freedom is not the ultimate value to be pursued at all costs. “
No, it is the value that enables pursuit of all other values, the value that can be shared when others are contentious.
“Its very definition and the assessment of its implementation cannot be made without any consideration for the outcomes it leads to.”
But we inhabit a sort of inversion of Rawls' initial position. We all have an idea of what our interests are, but do not share an understanding of the consequences of the various alternative policies available. From a more humble view of epistemology, liberty allows us to investigate what we think is best.
The answer in the abstract is Nozick's utopia, the thought experiment that allows any person to exit from an unsatisfactory world and create a new one, restricted only in that they cannot prevent the inhabitants from also exiting for other worlds. In the static vision of most political philosophy, that would mean that participants would converge onto something like an optimum. In a more dynamic scenario, it would mean that experimentation might continue to improve our social knowledge indefinitely. How does this apply in reality? It is easy to say that we should consider allowing more experimentation via subsidiarity or economic zones. Is it really so simple?
Liberty is often thought of in terms of a static set of principles or institutions. But can liberated persons experiment with such principles without losing liberty? If they can’t, do they really have liberty? Perhaps they can use contracts to create new social restrictions they deem worthwhile. But can they use contracts to amend or repeal their basic norms?