On standard analyses, there is a tight connection between duties that are owed to particular people (‘directed duties’) and those people’s standing to claim performance of the duties, where what it is to have such a claim is not merely to be owed the duty but to be in a position to demand or insist upon performance. Some examples raise intuitive doubts about whether this connection holds in every case, however. I explore the possibility that some directed duties cannot be claimed in the relevant sense, reject the explanations of this possibility that others have proposed, and propose one of my own. The basic idea is that claiming involves assertion of authority, and in some cases the duties that one is owed aren’t there by one’s authority alone.

It looks as if most people accept the possibility of legitimate injustice—that is, the possibility of unjust policies or laws that are nevertheless legitimately pursued or imposed by the government or state, so that the government or state has a claim against interference as it goes about pursuing the policies/imposing the laws, for example. This would seem to be a puzzling instance of the ‘right to do wrong’. I argue that attempts to make sense of it that assume duty-based accounts of political authority fail. I propose a broadly political realist account instead, which explains the claim against interference that is essential to legitimacy as an aspect of the state’s title to act as the representative or agent of the political community, where such community, structured by relations of authority, is necessary for the pursuit of almost any human good at all.

I discuss recent work on aggregating competing claims and saving the greater number. I argue that contributors to debates about these topics aren’t sufficiently clear about key details of the examples they appeal to, so that there is good reason to doubt that what are taken to be our intuitive moral judgments about these examples really reflect ordinary moral commitments. I distinguish individual and collective contexts of choice, and defend a view according to which there is no requirement to save the greater number in the former (following Anscombe, Taurek, Munoz-Dardé, and Setiya). I then discuss some new puzzles that arise for this view, and argue that they can be solved.