# Consequentialism, Agent-Relativity, and Moral Side-Constraints **URL:** **By** Eric Mack **Published:** October 1, 1987 --- “If the agent-relativist is correct, the more standard forms of consequentialism, especially the standard forms of utilitarianism, must be mistaken.” _Humane Studies Review_ continues to provide its readers with superb reviews and commentaries on current research across the disciplines. Tyler Cowen’s “Political Philosophy and Welfare Economics” (Summer 1987) fits securely within this pattern. But I wish to correct what seem to me to be some misleading remarks in Cowen’s lead section on “consequentialism.” My purpose is to clarify the relationships among certain central ideas and arguments within recent political philosophy. In this section Cowen contrasts consequentialist theories with “nonconsequentialist or agent-relative theories” and thereby erroneously equates nonconsequentialist with agent-relative theories. This false identification, in turn, threatens to blur the distinction between agent-relative (yet, nevertheless, consequentialist) objections to agent-neutral forms of consequentialism and anticonsequentialist objections to any form of consequentialism. It is important to highlight this distinction between arguments proceeding from an agent-relative conception of value and arguments proceeding from moral side-constraints on the permissible means for pursuing value. Clarification is complicated, however, by some apparent and, perhaps, even some real links between agent-relativist and moral side-constraint commitments. If a certain outcome, e.g., Jones’s happiness, is good, that outcome provides a reason for the performance of at least some actions that promote that outcome. But for whom is that outcome a reason for acting? There are two obvious and competing replies. The agent-neutralist replies that, _ceteris paribus_, everyone has a reason to promote Jones’s happiness. But, the agent-neutralist continues, given the differing opportunities of individuals for promoting various agent-neutral goods, Jones’s happiness should be promoted by that person whose pursuit of that good is part of the most efficient overall scheme for the production of agent-neutral goods. That is, if Jones’s happiness is a good, it must be a good for everyone. Its value will demand the efforts of anyone whose abilities and energies are not more extensively demanded by some other agent-neutral value. Each person should impartially be prepared to contribute to the greatest net yield of agent-neutral goods, such as Jones’s happiness. > If the agent-relativist is correct, the more standard forms of consequentialism, especially the standard forms of utilitarianism, must be mistaken. In contrast, the agent-relativist maintains that the value of Jones’s happiness provides Jones with a reason for promoting it. But though Smith may recognize that Jones has such a reason, Smith also recognizes that the end which provides Smith with a reason for acting is Smith’s happiness, not Jones’s. The status of neither Jones’s happiness nor Smith’s happiness as reason-providing goods implies, on the agent-relative conception, that each is an end which both Jones and Smith ought to promote. According to the agent-relativist, each agent is rational in being partial to his own happiness. The dispute between the agent-neutralist and agent-relativist is, then, a dispute between alternative ways of construing consequentialist propositions. The present example entails a debate about whether the value of someone’s happiness is somehow good as such (or good impersonally), or whether it is only a good, in the sense of providing a reason for action, for the person whose happiness it would be. Of course, if the agent-relativist is correct, the more standard forms of consequentialism, especially the standard forms of utilitarianism, must be mistaken: All those doctrines are built on an agent-neutral conception of value. It is this conception that, for example, lies behind J. S. Mill’s famous and otherwise astonishing inference from “each person’s happiness is a good to that person” to “the general happiness \[is\] a good to the aggregate of all persons.” > Some philosophers have contributed to the confusion by suggesting an agent-relative explanation of a person’s being subject to moral side-constraints—specifically, an explanation based on the agent-relative disvalue of certain actions. But to reject doctrines like standard utilitarianism in the name of the agent-relativity of value is hardly to accept an anticonsequentialist doctrine of moral side-constraints on the manner in which acknowledged values may be pursued. Imagine that Jones, on listening carefully to our agent-relativist, becomes convinced that Smith’s happiness does not as such provide Jones with any consequentialist reason for deciding how he (Jones) should act. Jones might then conclude that only unconstrained pursuit of his own happiness (in some non-self-defeating form) is rational. Here we would have a convinced agent-relativist who, for that very reason, was prepared to be thoroughly unconstrained in his behavior toward others. This highlights how mistaken it is to equate the idea of the agent-relativity of value with a belief in anticonsequentialist side-constraints on behavior. I think that Jones would be wrong to draw that exclusively consequentialist conclusion because I believe that there is a rational, albeit complex, link between an agent-relative theory of value and a doctrine of moral side-constraints. At the very least, given such an agent-relative conception of value, if there are _any_ objective, interpersonal moral rules, they must proscribe the use of certain means in the course of each agent’s pursuit of his personal ends. (By hypothesis, there could be no objective interpersonal rules derived from common ends, in contrast to means, to whose production each must contribute.) Of course, if we assume a complex link between an agent-relative theory of value and moral side-constraints, the reason a person would have to forgo certain means in the pursuit of his ends would not be directly explicable by the the agent-relative conception of value. Unfortunately, some philosophers, for example, Thomas Nagel in _The View from Nowhere_ (Oxford: Oxford University Press, 1986), have contributed to the confusion by suggesting an agent-relative explanation of a person’s being subject to moral side-constraints—specifically, an explanation based on the agent-relative _disvalue_ of certain actions. It is worth seeing what motivates this account and why it fails. Consider Brown, who believes it would be impermissible for him to kill innocent bystander 1 even though this will prevent the fiendish White from killing innocent bystanders 2 and 3. One might try to account for this impermissibility in the following way: For Brown, there is a special disvalue in _his_ killing. He views his killing of 1 as worse (for Brown) that White’s killing of 2 and 3 (for Brown). On this account, in choosing not to kill 1, even though this killing would prevent a greater number killings, Brown minimizes agent-relative disvalue. But this sort of attempt to analyze the moral side-constraint on Brown not to kill 1 by way of the agent-relativity of value is entirely implausible; while the disvalue for Brown of his killing may explain his squeamishness (and may help motivate his not killing 1), it provides no explanation at all of why Brown’s killing 1 would _wrong_ 1\. It gives no account of 1’s possession of a right against Brown that is not exhausted by the negative consequences (for Brown) of Brown’s killing 1. No side-constraint on Brown and no right of 1’s is explained when the only claim made about Brown’s killing 1 is that it would diminish _Brown’s_ personal good. To sum up: First, agent-relativity is a conception of value that competes with agent-neutrality on the consequentialist playing field, so agent-relativity should not be equated with nonconsequentialism. Second, the sort of attempt to explain moral side-constraints (with their anticonsequentialist connotations) found in works such as Nagel is unsuccessful. Yet, third, when agent-relativity of value sweeps agent-neutralist consequentialism off the field, a new ball game may arise regarding what constraining interpersonal moral rules it may be rational for individuals to acknowledge. _Eric Mack teaches in the department of philosophy and the Murphy Institute for Political Economy at Tulane University._