Should contractualists drop the ‘individualist restriction’? (Q10b, 2025)

Related: notes · essay plans · Can contractualism explain why we should save the greater number? (essay)

No, contractualists should not drop the individualist restriction. Even though maintaining it comes with significant costs, their theory would lose much of its distinctiveness and attractive features if they abandoned the restriction and allowed for aggregation of complaints. In this essay, I define the contractualist criterion of rightness and individualist restriction, placing these in the context of the theory’s overall goals. I then give a motivating thought-experiment often used to justify the individualist restriction, and explain how that same restriction hampers contractualism’s ability to deliver what intuitive verdicts in cases involving numbers or risk. Finally, I argue that despite these costs, contractualists ought not to drop the individualist restriction, on pain of rendering their theory undistinctive.

Contractualism, as developed by Tim Scanlon, says that an action X is permissible iff and because it is licensed by principles to which nobody could reasonably reject. We say that a person S can reasonably reject to a principle P iff:

  1. S has a complaint C about P grounded in reasons which are:

    a. Personal [i.e., not about impersonal goods such as equality or beauty]

    b. Individual [i.e., about the effects of P on themself only; however these need not be exclusively welfare-based – for example, S could complain that principle P is unjust towards them]

    c. Generalisable [i.e., not objections which exist solely in virtue of S’s particular tastes or circumstances]

  2. And, there is some other principle P’ with a greatest complaint C’ which is no larger than C.

The structure of this theory is designed to capture what Scanlon viewed as the central feature of morality: that it is about relations of mutual respect and justifiability between humans, i.e. “What We Owe Each Other”. Deploying the Impersonalist Restriction and Individualist Restriction ensure that the reasons for rejection are of the sort that one person might genuinely (and justifiably) bring to another who was acting wrongly towards them, and this gives contractualism a closer connection to the phenomenology of moral motivation than other theories do.

Apart from fitting well with contractualism being a moral theory focused on what each human owes to each other, the individualist restriction accommodates commonly-held intuitions about the so-called separateness of persons which other theories like consequentialism deny. For example, consider the famous thought-experiment:

Transmitter Room. Jones is trapped under electrical equipment at the TV transmission office. He could be rescued, but this would require switching off the transmission of a hugely-popular World Cup game about to begin, preventing millions from watching it. Jones is in significant pain, and will continue to suffer throughout the transmission if he is not rescued.

Many people have a very strong intuition that it would be wrong to go ahead with the transmission, regardless of how many football fans would be disappointed by the cancellation of the broadcast. Moral theories which allow for aggregation of interests, however, would permit (or perhaps even require) the transmission to go ahead, for sufficiently many audience members. (This needn’t necessarily hold, because the theory might impose side-conditions on permissible actions, or be suffering-focused, and so on. But aggregating theories are not the focus of this essay, in any case, so I will set that aside from here.)

Contractualism, on the other hand, is easily able to explain the wrongness of continuing with the broadcast: any principle licensing it can be reasonably rejected by Jones, whose complaint (suffering in pain for 90 minutes) is much larger than any other audience member’s complaint to alternative principles requiring his rescue (and denying them the pleasure of watching a football match). So not only does the individualist restriction highlight the interpersonal character of morality, it also vindicates commonly-held intuitions about how the separateness of persons precludes us from requiring large sacrifices of one for a small benefit accruing to many others.

Given this, one might wonder: why would the contractualist even countenance dropping the individualist restriction? There are at least two weighty reasons in favour, both relating to how the individualist restriction seems to make contractualism blind to how quantitative considerations can have moral importance. First, consider “saving the greater number” (SGN) cases:

Dual Drowning. You are a coastguard confronted with two sinking boats, equidistant from your ship. You’re able to reach either boat, but not both of them. Boat A has one person on it, X; boat B has two people, Y and Z. The individuals on the boat you travel to will all be saved; any individuals on the other will drown.

Intuitively, given the choice between saving one person or two people, you surely ought to save two. But it’s not immediately obvious how contractualism can deliver this verdict, given the individualist restriction. Any principle P telling you to travel to boat B could be reasonably rejected by X, who objects that P leads to them losing their life for certain. Since this complaint is smaller than the objections raised by Y or Z to a competing principle P’ telling you to flip a coin to decide which boat to travel to (their objection being that they die with probability 0.5), it is not permissible to save those on boat B for certain.

For some contractualists, like Taurek, this is an attractive feature of the theory. He argues that the right decision genuinely is to flip a coin, as this is the only way to show appropriate respect for the weightiness that no matter what action you take, human lives are at stake, and every life is of immense value. But this does not seem like a sensible approach: consider a case where one boat had a single person on it and the other had one million – surely, it would be deeply irresponsible to flip a coin to decide who to save.

There are a number of potential resolutions to SGN cases, but the most compelling is from Hirose (2001), after Kamm. Hirose gives two non-aggregating principles (i.e., principles respecting the individualist restriction) which jointly imply one ought to save the greater number:

(1) Anonymity: Two outcomes are equally good if they differ only in respect of the identities of people.

(2) Pareto: If one outcome is strictly better for some people, and no worse for anybody, then it is better overall.

Since it would be better by (2) to save X and Z than to save X alone, and this is equally good by (1) as saving Y and Z, it is therefore better to go to boat B than boat A. So, contractualists seem able to maintain the individualist restriction as far as SGN cases are concerned (although some doubt whether Hirose’s principles are sound).

A more substantial problem for contractualists is what the individualist restriction implies about how to handle decisions involving risk. If they adopt ex post contractualism, where the size of complaints is based solely on the magnitude of harm which might occur, regardless of the probability of that harm obtaining, then the theory becomes almost absurdly risk-averse. As Ashford notes, it would imply, for example, that air travel for tourism ought to be prohibited, because the complaints of tourists who’d miss out on their holidays are certainly smaller than the ex post complaint of death that a remote villager in the Global South could raise, given the miniscule but nonzero risk that an airplane falls onto their home. Ex ante contractualism, which weights complaints by their probability of occurrence, has other problems. In particular, it implies that we should be willing to sacrifice any number of statistical lives for a single identifiable life – i.e., that if a drug could either be used to (i) save one person for sure, or (ii) save one million people from a 99.99999% chance of death, we ought to do (i). This is because the individualist restriction prevents us from aggregating any of the 99.99999%-chance-of-death complaints from the people in (ii), so that each is in isolation smaller than the certain-death complaint of the person in (i).

Since moral decisions are rife with uncertainty, it is a real cost that contractualism is not able to deliver more sensible verdicts in such cases. We would hope that a moral theory can provide judgements about how to act across a range of settings, including (for example) the ethics of public health campaigns that might involve imposing risks on certain individuals for benefits of others.

But in spite of this, contractualists should not drop the individualist restriction. Parfit famously advocated for a version of contractualism which abandoned both the impersonalist and individualist restrictions – but this would yield a theory that is hardly contractualist at all. Two of the most distinctive features of contractualism are the fact that it resists aggregation, and is able to explain morality as a relational practice between humans. Giving up on the individualist restriction means it foregoes both of these benefits, and one would have few remaining reasons to subscribe to it over a more naturally-aggregating theory like consequentialism.

Post-Claude & examiner report notes