One of the most problematic aspects of the 'Harvard School' of liberal international theory is its failure to fulfil its own methodological ideals. Although Harvard School liberals subscribe to a nomothetic model of explanation, in practice they employ their theories as heuristic resources. Given this practice, we should expect them neither to develop candidate causal generalizations nor to be value-neutral: their explanatory insights are underpinned by valueladen choices about which questions to address and what concepts to employ. A key question for liberal theorists, therefore, is how a theory may be simultaneously explanatory and value-oriented. The difficulties inherent in resolving this problem are manifested in Ikenberry's writing: whilst his work on constitutionalism in international politics partially fulfils the requirements of a more satisfactory liberal explanatory theory, his recent attempts to develop prescriptions for US foreign policy reproduce, in a new form, key failings of Harvard School realism.In 1995, David Long called for the closure of what he termed the 'Harvard School of Liberal International Theory'. He coined this term to denote efforts by scholars such as Robert Keohane and Andrew Moravcsik, then based at Harvard, to develop a liberal theory that would subsume or supersede (neo)realism. 1 Long (1995: 489-91, 501) welcomed attempts to 'frame liberalism' as an 'explanatory theory', but criticized Harvard School liberals for borrowing 'a number of normative, ontological, and methodological premises' from realism. He argued that this stifled the prospects for a renewal of liberalism's