is premised. Actors cannot simply be assumed to have aWxed (and immutable)
preference set, to be blessed with extensive (often perfect) information and fore-
sight, or to be self-interested and self-serving utility maximizers. Rational choice
and historical institutionalism are, as Thelen and Steinmo note, ‘‘premised on
diVerent assumptions that in fact reXect quite diVerent approaches to the study of
politics’’ ( 1992 , 7 ).
Yet, if this would seem to imply a greater aYnity with normative/sociological
institutionalism, then further inspection reveals this not to be the case either. For,
to the extent that the latter assumes conventional and norm-driven behavior
thereby downplaying the signiWcance of agency, it is equally at odds with the
deWning statements of historical institutionalism. As Thelen and Steinmo again
suggest:
institutional analysis... allows us to examine the relationship between political actors as
objects andas agents of history. The institutions that are at the centre of historical
institutionalist analysis... can shape and constrain political strategies in important ways,
but they are themselves also the outcome (conscious or unintended) of deliberate political
strategies of political conXict and of choice. (Thelen and Steinmo 1992 , 10 ; emphasis added)
Set in this context, the social ontology of historical institutionalism is highly
distinctive, and indeed quite compatible with the constructivist institutionalism
which it now more consistently seems to inform. This brings us to a most important
point. Whether constructivist institutionalism is seen as a variant, further develop-
ment, or rejection of historical institutionalism depends crucially on what historical
institutionalism is taken to imply ontologically. If the latter is seen, as in Hall and
Taylor’s inXuential account, as aXexible combination of cultural and calculus
approaches to the institutionally-embedded subject, then it is considerably at
odds with constructivist institutionalism. Seen in this way, it is, moreover, incom-
patible with the attempt to develop an endogenous institutionalist account of the
mechanisms and determinants of complex institutional change. Yet, if it is seen, as
the above passages from Thelen and Steinmo might suggest, as an approach
predicated upon the dynamic interplay of structure and agent (institutional context
and institutional architect) and, indeed, material and ideational factors (see Hay
2002 , chs. 2 , 4 , and 6 ), then the diVerence between historical and constructivist
institutionalisms is at most one of emphasis.
Whilst the possibility still exists of a common historical and constructivist
institutionalist research agenda, it might seem unnecessarily divisive to refer to
constructivist institutionalism as a new addition to the family of institutionalisms.
Yet this can, I think, be justiWed. Indeed, sad though this may well be, the prospect
of such a common research agenda is perhaps not as great as the above comments
might suggest. That this is so is the product of a recent ‘‘hollowing-out’’ of
historical institutionalism. Animated, it seems, by the (laudable) desire to
build bridges, many of the most prominent contemporary advocates of historical
62 colin hay