Justice and Utility

Most of what utilitarianism asks of us feels right in everyday cases. Do not lie, because lying causes harm. Keep your promises, because broken promises damage trust. Help others when you can, because their well-being matters too. So far, so good.

The difficulty comes with justice. Consider some standard cases that test consequentialist theories.

A sheriff in a Southern town can prevent a riot by framing an innocent man for a murder he did not commit. The riot will kill many people; the frame-up will kill only the one. The utilitarian calculation seems straightforward: framing the innocent man produces more total happiness (more lives preserved, less suffering) than allowing the riot. But almost everyone thinks framing the innocent man is wrong. We do not think it is wrong because the math is unclear. We think it is wrong even granting the math.

A hospital has five patients who will die without organ transplants. A healthy person walks in for a routine check-up. If the doctor kills the healthy person and harvests their organs, five lives are saved at the cost of one. The aggregate happiness is much greater than the alternative. But almost no one thinks the doctor should do this.

A government is considering whether to allow some small number of innocent people to be tortured to extract information that will save many more lives. A consequentialist calculation seems to favor the torture. But there is something the calculation seems to miss.

In each case, what blocks the utilitarian conclusion is something like a sense of justice. Each person, we think, has some kind of standing that protects them from being used as a mere means to others' greater happiness. The aggregation that defines utilitarianism, the adding up of well-being across persons, seems to override exactly the protections justice provides.

Mill knows all of this. Chapter 5 is his attempt to show that utilitarianism does not have this problem after all, because justice is itself a special class of utility properly understood.

Mill's argument in Chapter 5 has three steps. Track them carefully; the chapter can feel like it is doing many things at once, but the spine of the argument is straightforward.

Step 1: Justice is not a unified concept. Mill begins by canvassing the many things people call "just" or "unjust." Respecting people's legal rights. Giving people what they deserve. Keeping faith. Treating equals equally. Punishing in proportion to guilt. He shows that there is no single feature that all and only just acts share. "Justice" is a family of related concerns, not a single principle.

Step 2: What justice has in common is the kind of utility it tracks. The cases of justice, Mill argues, are unified by what they protect: the most important interests human beings have. Security from harm, freedom from arbitrary interference, the conditions of cooperation. These interests are utility-relevant on a grand scale. We feel especially strongly about violations of justice not because justice is a separate domain of value but because the utilities protected by justice are uniquely important. A society in which lives, contracts, and basic security cannot be relied upon is not just inconvenient; it is one in which the most basic conditions for any human happiness have collapsed.

Step 3: This explains why justice feels different from other moral demands. Mill argues that what distinguishes the demands of justice psychologically is the intensity of the resentment we feel when justice is violated. Other moral failures provoke disapproval; injustice provokes outrage. This is not because injustice is a different kind of moral failure but because the utility at stake is so great. We feel injustice as personal because we know, at some level, that we depend on these protections ourselves. The intensity of feeling is the psychological correlate of the importance of what is at stake.

Put the three steps together and you have Mill's account. Justice is a special class of utility, marked off by the importance of what it protects and the intensity of feeling that protects it. It is not separate from utility; it is utility at its most urgent.

The most useful single passage in Chapter 5, for our purposes, is Mill's account of rights. He defines a right as something the possessor has a valid claim on society to protect. This is purely functional: a right is whatever it is socially useful to treat as inviolable. If society treating something as a right (life, liberty, property, freedom of expression) produces more happiness than not treating it that way, then it is in fact a right.

Notice what this account gives up. Rights are not absolute in the Kantian sense; they are not based on the inherent dignity of persons; they do not exist independently of human institutions and feelings. Rights are, in Mill's view, instrumentally justified: they protect important utilities, and that is why they exist.

Notice also what this account keeps. Within their proper scope, rights function exactly as the deontologist says they should. The whole point of treating something as a right is that we do not weigh it case by case against competing utilities. We treat it as inviolable in the run of cases precisely because treating it as inviolable produces better consequences over the long run than treating it as a case-by-case calculation would.

This is, in effect, a rule-utilitarian argument: the rule "treat life as a right" produces more total happiness than the rule "weigh life against competing utilities case by case." So we follow the first rule. The rule does not stand outside utility; it is justified by utility. But once justified, it operates as a constraint on individual case-by-case calculation.

Whether this account succeeds in capturing what we think about justice is a substantive question. Defenders of utilitarianism see Mill's strategy as showing that consequentialism has the resources to accommodate what is right in our intuitions about justice without abandoning its consequentialist foundation. Critics see it as a fudge: either the constraints really are absolute (in which case they are not really utilitarian) or they are not absolute (in which case the theory still permits frame-ups and forced organ harvests in cases where the math works out).

Once you have Mill's account of justice on the table, it is worth returning to the test cases.

The sheriff and the innocent scapegoat: Mill can resist framing the innocent man without abandoning utilitarianism. His argument: the rule "do not punish the innocent" is justified by utility because the social practice of punishing only the guilty is what makes the institution of law possible at all. A society in which sheriffs frame innocent people whenever the math works out is one in which no one can trust the justice system, and the long-run consequences of that loss of trust dwarf the short-run gains of the particular case. So the rule against framing the innocent should be treated as inviolable, even when it appears, locally, to produce worse outcomes.

The transplant case: Same argument. Doctors who would kill healthy patients to save sick ones make hospitals dangerous places to visit. A society in which we follow the rule "doctors do not kill their patients" produces more happiness in the long run than one in which doctors weigh patient lives case by case. So the doctor should not harvest the healthy patient's organs, even when the math seems to favor it.

The torture case: This is harder for Mill. The rule against torture clearly produces more happiness than its absence in almost all real-world situations. But in the genuine emergency case (the ticking time bomb, the captive who knows where the explosive is) the utilities of the particular case seem to swamp the long-run damage of breaking the rule. Mill would have to either bite the bullet (allow the torture in the genuine emergency case, while insisting that such cases are vanishingly rare in practice) or argue that the institution of torture cannot be confined to genuine emergencies and so should be ruled out altogether. Both responses are available; neither feels fully comfortable.

Whether these responses succeed is one of the most active debates in contemporary ethics. Your assessment of utilitarianism as a whole will depend largely on how convinced you are by them.

This lecture video covers Chapters 3, 4, and 5 together. Watch it after you have worked through the whole book and the three lecture pages. It will pull the strands together.