ABSTRACTHajer and Wagenaar (2003. Deliberative Policy Analysis: Understanding Governance in the Network Society. Cambridge: Cambridge University Press, xiv, 16) advanced a conception of policy analysis – “Deliberative Policy Analysis” – that “rests on three pillars: interpretation, practice and deliberation.” This form of policy analysis, they argued, supports “more direct, participatory forms of democracy” involving “democratic deliberation on concrete issues” (xv, 29). Since their writing, empirical research on such initiatives – “democratic innovations,” for short – has blossomed. However, while deliberative policy analysis is itself post-positivist in orientation, many researchers bring a (quasi-) positivist orientation to their work on democratic innovations. A key challenge for deliberative policy analysts is, then, how to participate in this field of inquiry while maintaining a post-positivist orientation. Pragmatist philosophy, I submit, can help them to meet this challenge. Pragmatism rejects a number of positivist assumptions about the nature of empirical inquiry. Relatedly, it supports the claim that policy analysis should be interpretive, practice-oriented, and deliberative. Indeed, it suggests that policy analysis cannot avoid being so. By way of illustration, I indicate how pragmatism points to an approach to case study research that rests on the three pillars.