Lisa Hill's response to my critique of compulsory voting, like similar responses elsewhere,1 remind me how much a child of the 1970s I am, and how far my beliefs and intuitions about politics have been shaped by the electoral conflicts, social movements and violence of that period. But my perceptions of politics have also been profoundly shaped by my teachers, and fellow graduate students, at MIT. Theda Skocpol famously urged political scientists to ‘bring the state back in' to their analyses,2 and to recognize that political identities, interests and coalitions cannot be read off straightforwardly from people's socio-economic positions. In their different ways, this was the lesson that Suzanne Berger, Charles Sabel and Joshua Cohen tried to teach us, emphasizing that political participation and conflict, themselves, can change people's identities, their sense of what is desirable and possible, and their ability to make common cause with others.3 I do not take it as self-evident, therefore, that the poor and seemingly powerless should be politically apathetic, unwilling to vote, or incapable of imagining a political solution to at least some of the problems confronting them. Nor do I suppose that non-voters are all of a piece, and that their shared interests are, inevitably, more significant than those that divide them. Such assumptions seem mistaken in the case of voters, and I see no reason why they should be true of non-voters. The people we find in these categories are not predestined to be in one rather than the other; they do not always stay where they start off; and at an individual level, the reasons why people fall into one group, rather than another, are likely to be complex and sometimes unpredictable.