Can We Bargain our Rights?

A Coasean Analysis

Note: I have a new working paper titled “Living in Disagreement. Public Reason and Jurisdictional Rights” that I hope to present soon at some seminars and conferences. The article partially builds on reflections I’ve developed in several publications of this newsletter (see here and here). What follows below expands on these essays and the working paper.

In his famous article “The Problem of Social Cost,” Ronald Coase exposed one of the most beautiful – by its simplicity and generality – results of economic analysis.[1] Abusively called the “Coase’s theorem,” it states that when individuals’ activities have effects on other individuals outside voluntary market exchanges (i.e., what economists call “externalities”), these effects can be dealt with by creating and assigning additional property rights that individuals can freely bargain. Moreover, if there are no costs associated with bargaining, then it doesn’t matter how property rights are assigned, people will bargain until all opportunities for mutually advantageous exchanges have been exhausted. In economic jargon, irrespective of people’s initial endowments in those rights, the bargaining procedure will end somewhere on the contract curve, i.e., on a Pareto-optimal allocation of rights where you cannot improve a person’s situation without negatively affecting someone else’s situation.

As I generally tell my students, Coase’s contribution is not in proving the result itself. In a way, this result was commonly known before because it is nothing but a restatement of the two welfare theorems. You can easily picture those with an “Edgeworth box” in the case of a bilateral bargain (see below). If you start at endowment E where agent 1 has property rights on a quantity x of good X and a quantity y of good Y (with 1-x and 1-y for agent 2 respectively), then (if there is no transaction cost) agents will bargain until they reach the contract curve (the first theorem) where, by definition, there is trade that can move the allocation to a higher indifference curves for both agents.[2] However, you don’t know exactly where you will end up on the contract curve – the only thing that is certain is that it cannot be on a point that is located below the indifference curves that pass by point E (that would mean that an agent has voluntarily made an exchange that deteriorates her situation). Where you end depend on the exact bargaining procedure and on the initial endowments. The second theorem precisely says that, with perfectly competitive markets, you can literally “choose” the final optimal allocation by changing the initial endowments.  

Coase’s result does not prove more than that. Give people’s property rights and they will bargain until an efficient outcome is reached. How rights are distributed at the end of the bargaining procedure depends on how they are distributed initially, but the efficiency of the procedure is independent on the initial distribution. Imagine for instance that company A produces waste that pollute the environment of agents B and C. Let say that A makes 50000$ yearly profits and assume that B and C are respectively willing to pay 3000$ and 5000$ per year to have a clean environment. If A had to change its production process such as to eliminate waste, its yearly profits would fall to 43000$. The current situation is inefficient the social surplus is 50000$ - 3000$ - 5000$ = 42000$, while it would be 43000$ if A were to take the social cost of its activity into account. In this situation, Coase’s result just says that we should assign property rights over a clean environment. If the rights are given to B and C, A has the choices between eliminating waste (its profit falls to 43000$) or to buy to B and C their rights for a clean environment. Since it will cost at least 8000$ for A to do so,[3] A will change its production process. If, to the contrary, A is given the right to pollute, then B and C will be together willing to pay 8000$ to buy A’s right. Again, A will accept to change its production process because then it earns 43000$ + 8000$ = 51000$, against 50000$ if it refuses the offer. However, suppose that in this last case B and C have to coordinate before making an offer to A. If it costs them more than 1000$ to do so (the transaction costs, that I assume to be equally divided between them), they will be willing to offer less than 7000$ to A. In this case, A is better refusing their offer and keeping its right to pollute. Hence, when transaction costs are positive, it matters in terms of efficiency to whom rights are allocated initially (in this example, it’s better to give the rights to B and C).[4]

As I said, this result was not in itself new at the time Coase published the article. However, Coase’s contribution is to have shown how the welfare theorems could be relevant in the context of “missing markets” where the price system fails to account for all the effects of economic choices. In the same spirit, it may be worth exploring how the Coasean logic could apply to other kinds of jurisdictional rights than property rights, i.e., rights of privacy and rights of association. I’ve discussed in a recent essay how the problem of externalities presumably applies to activities and practices that are covered by rights and liberties such as free speech, religious freedom, or the right to watch pornography. What makes things maybe more complicated than in the standard economic case is that externalities in those domains are not or are less tangible. Pollution affects health in a way that can be more or less objectively ascertained. A person’s speech may certainly affect someone else, but the effect is also surely more difficult to measure and is anyway more subjective. Or, to take another illustration, while I may absolutely not care about how people are dressed and whether they signal their religious beliefs, other persons may be sincerely and deeply affected by it. The difficulty here is that individuals in general tend to perceive the world in many different ways. Even if we agree that, say, preserving human life is more important than protecting women’s bodily autonomy, we may nonetheless disagree on the moral status of (and the justification of a law permitting) abortion if some perceive a less than 12-week old fetus as a human being while others don’t.

When people’s perspectives significantly differ in such a way, it might be difficult if not impossible to design a system of rights that account for all possible “moral externalities” of this kind. Whatever the way you arrange rights, some persons will feel harmed or even disrespected by the behavior of others. In their view, rights should be arranged differently.[5] The standard liberal reaction is just to assert that people should be endowed with an extensive scheme of basic liberties and rights that is compatible with the liberties and rights of others.[6] The problem is that this can lead to a lot of disputes and conflicts. Depending on the scheme adopted, the outcome may be “sectarian,” in the sense that it some groups within the population may feel unfairly treated because their perspectives are not taken seriously. It is at this stage that it is tempting to introduce a Coasean analogy. Basically, whatever the system of rights you design, you will not be able to eliminate all moral externalities. Moreover, because perspectives are subjective, it is anyway difficult to anticipate the emergence of those externalities. So, in the spirit of Coase’s result, why not leave individuals bargain their rights between themselves?

The idea that rights could be bargained in such a way is explored in detail by Ryan Muldoon in his book Social Contract Theory for a Diverse World, though Muldoon does not explicitly appeal to Coase.[7] How that could work? Because they hold different perspectives and eventually disagree on what matters and what is good, we can imagine that individuals could be willing to trade their rights against others. Rights could be simply traded against money (i.e., property rights over monetary resources). If you’re really deeply affected by the fact that some people exhibits their religious convictions through how they dress, then you could just “bribe” those people (at least those that you see regularly) so that they dress differently. By offering money in exchange for some persons’ rights to dress as they wish, you’re revealing the “social cost” that the exercise of these rights has on you. Whether other persons will be willing to give up their rights indicate whether they value enough the fact of having those rights to internalize the externality that they impose on you. Note that in this context the fact that individuals’ have different perspectives that others cannot access or understand is irrelevant. What matters are the tradeoffs that individuals are willing to make, as reflected by the marginal rates of substitution in a market system, whatever the reasons. Also, there is no obligation to give a monetary value to rights – we can also imagine that rights are directly traded against each other.

The bargaining approach is interesting for at least two reasons. First, it helps individuals to understand others’ perspectives. In the same way as market prices transmit information about preferences and production costs by incentivizing economic agents to make tradeoffs based on those preferences and production costs, the exchange rates of rights display an information about the relative importance that individuals give to different rights. This importance is directly related to each person’s perspectives. Second, Coase’s result indicates that a bargaining procedure will help to reduce external effects and so to limit conflicts between rights and hence between individuals. Bargaining rights therefore favors moral reconciliation by promoting an ethos of understanding between persons holding different perspectives and by delineating individual spheres of sovereignty that minimally conflict. It therefore reduces the risk of sectarianism.

Of course, nothing is perfect, and the bargaining approach also raises some worries. I discuss the most important of them in detail in my working paper but here are some elements:

·       High transaction costs. There are good reasons to think that transaction costs would be prohibitively high. Contrary to property rights, specific privacy and association rights are likely to be difficult (and therefore costly) to enforce and monitor. Suppose you are willing to trade you’re right to hold some kind of speech against the right of a group of persons in your neighborhood to wear some kind of clothes. You’ve first to agree on what exactly counts as the kind of speech concerned, as well as the kind of clothes. This may be difficult when people do not hold the same perceptions. Moreover, it will be obviously costly to monitor the respect of rights has they have been traded.

·       Non-alienable rights. The previous difficulty implies that if rights can be traded at all, that will be sets of rights or rights defined in a broad way (e.g., the right of free speech, the right to dress as one wants). The problem is that the very justification of jurisdictional rights is largely grounded in the fact that individuals are moral agents with specific capacities that cannot realized without jurisdictional rights. Therefore, there are at least some jurisdictional rights that cannot be traded without jeopardizing the status of persons as moral agents.

·       The fair distribution of rights. The Coasean logic does not necessarily lead to a complete elimination of externalities, even when transaction costs are null. In the example above, if the profit of company A when it produces waste is 52000$ (instead of 50000$), then the efficient allocation is to give A the right to produce waste, even though it creates external effect on B and C. Hence, the Coasean logic rather militates for a minimization of external effects. The optimal scheme of jurisdictional rights may eventually adopt the same minimization criterion. But note the counterintuitive implication: the best way to reduce conflicts between rights is to reduce the number and the scope of rights. An alternative criterion is to aim for a distribution of rights such that the remaining externalities are equally distributed within the population. Because individuals and social groups are not all living in equal conditions with some more disadvantaged than others, the distribution of rights would then have to be unbalanced. For instance, more rights could be given to individuals who belong to social groups who have been historically unfairly treated and who, because of that, suffer disproportionally from typical moral externalities.[8] Again, this seems to have illiberal implications.

·       Nonexistent stable system of rights. Coase’s result doesn’t extend to the case of 3+ agents with two or more externalities operating simultaneously (see footnote 4). That would surely apply to a bargaining procedure of privacy and association rights, whether actual (people really bargain on their rights) or fictitious (we try to implement a system of rights that would result from a bargaining procedure). In turn, that makes the achievement of any criterion of fairness (minimization or equal distribution of moral externalities) unlikely.

·       Collectively harmful practices. In a previous essay, I discussed the challenge that the recognition of collective harmful practices poses to the definition of jurisdictional rights. What some persons may perceive as collectively harmful practices (e.g., watching pornography) are such that a single person’s action doesn’t cause any direct harm to any specific other person. The harm is at best indirect: the fact of taking part in a practice contributes to a collective harm that affects a group of individuals. Because jurisdictional rights are assigned to individuals and not to group, it is arguably unclear that any right is infringed. Still, if you argue that collectively harmful practices really exist and that they really harm people as members of a social group, then you may want to limit the rights of persons to participate in these practices. We end up close to the minimization criterion and its illiberal implications. If you discard the existence of collectively harmful practices, you’re implicitly favoring some perspectives over others without any obvious reason. In this case, you’re probably guilty of liberal sectarianism!

---

[1] R. H. Coase, “The Problem of Social Cost,” Journal of Law and Economics 3 (October 1, 1960): 1–44.

[2] This follows immediately from the fact that the contract curve is constituted by all the points where agents’ indifference curves are tangent.

[3] The assumption here is that an agent’s willingness to pay is equal to her willingness to accept. This assumption is natural but may be false in some circumstances, in light of what behavioral economists call the “endowment effect.” In this case, Coase’s result no longer holds because, even of transaction costs are null, some initial allocation of rights may not lead to an efficient outcome.

[4] There are other complications that are less often discussed. For instance, even without transaction cost, there may be no stable allocation of rights when two or more externalities are involved at the same time. In the example I give, suppose for instance that B is harming C by producing another externality. We can design configurations such that A, B, C will never be able to agree over an allocation of rights that internalize the social costs. For an illustration, see Dennis C. Mueller, Public Choice III (Cambridge ; New York: Cambridge University Press, 2003), pp. 30-2.

[5] A formal statement of this result is provided by Hun Chung and Brian Kogelmann, “Diversity and Rights: A Social Choice-Theoretic Analysis of the Possibility of Public Reason,” Synthese 197, no. 2 (2020): 839–65.

[6] As stated by Rawls’s first principle of justice. John Rawls, A Theory of Justice (Oxford University Press, 1971).

[7] Ryan Muldoon, Social Contract Theory for a Diverse World: Beyond Tolerance (Routledge, 2016).

[8] See Ryan Muldoon, “Exploring Tradeoffs in Accommodating Moral Diversity,” Philosophical Studies 174, no. 7 (July 1, 2017): 1871–83.