One of the goals of the Kantian tradition in moral and political philosophy has been to defeat the egoist by showing that morality is rationally required. Early in A Theory of Justice, John Rawls seemingly renounces this particular Kantian goal. Egoism, Rawls tells us, is logically consistent, and not irrational, and as such it is no part of his project in moral and political philosophy to defeat it.

Accordingly, although in TJ, Rawls initially includes three forms of egoism: first-person dictatorship egoism (Everyone is to serve my interests), free-rider egoism (Everyone is to act justly except myself if I choose not to), and general egoism (Everyone is permitted to advance his interest as he pleases) on his list of conceptions from which persons in the original position are to choose, he immediately tells us that these forms of egoism do not, strictly speaking, belong on his list. First-person dictatorship egoism and free-rider egoism obviously violate a generality condition, and general egoism violates an ordering requirement because if everyone is authorized to advance his or her aims, there is no overall ranking or ordering of competing claims. Given that Rawls assumes that any conception of justice must satisfy both a generality condition and an ordering requirement, these assumptions provide the basis for the quick elimination of these forms of egoism from Rawls’s list (TJ 107). All the other conceptions on Rawls’s list are clearly moral conceptions, and egoism, Rawls tells us, is “a challenge to any such conception.” This might lead us to conclude that Rawls was simply not interested in considering any kind of challenge that egoism might present to his work. That conclusion would certainly accord with Rawls’s previous work from “Outline of a Decision Procedure in Ethics” (1951) to TJ (1971).