crisis itself. The overly parsimonious conception of crises as moments of Knightian
uncertainty may, in this respect, obscure more than it reveals.
This is perhaps suggestive of a broader, indeed somewhat characteristic, failing of
constructivist institutionalism to date—its tendency to fall back upon, or at least not to
close oVfully, the return to a rump materialism. Very often, as in this case, alternative
and more parsimonious accounts can be oVered of the very same data constructivist
institutionalists present that make little or no causal reference to the role of ideas.
A second set of concerns relates to the theoretical status of constructivist
institutionalist insights. Again, the issue is a more general one. For, like much
work within this development tradition, although constructed as a work of
explanatory/causal analysis, it is not always clear that Blyth does adequatelyexplain
the outcomes whose origins he details. Indeed, it would seem as though abstracted
redescription and explanation are frequently conXated. In other words, an
abstract and stylized sequence consistent with the empirical evidence is presented
as an explanation of speciWc outcomes in the context being considered. While crises
may well be what states make of them, it is not clear that constructivist institu-
tionalists have explained why states make of them what they do—indeed, it is
precisely in this ambiguity that the possibility of the return to a residual materi-
alism arises.
This brings us to a further, and closely related, issue—the epistemological status
of the claims Blyth makes about the US and Swedish cases, speciWcally, and those
made by constructivists about institutional change more generally. Understand-
ably, Blyth is keen to stress that his chosen constructivist brand of institutionalism
provides us with a ‘‘better understanding of political change’’ than more conven-
tional materialist modes of political analysis ( 2002 , ix; see also Abdelal, Blyth, and
Parsons 2005 ; Berman 1998 ). Yet it is not clear from the text why sceptics should
accept such a view—largely because no sustained consideration is given to how one
might adjudicate preferences between contending accounts (see, for instance, Bevir
and Rhodes 2003 ). Nor is it clear that constructivists can easily claim the kind of
epistemological self-conWdence required to pronounce the analytical superiority
of their perspective. Presumably, ‘‘better’’ here means more complex, more
nuanced, and more able to capture the rich texture of social, political, and
economic interaction—in short, the standard that Blyth seems to construct is
one of correspondence to an external reality. This is all very well, but external
realities, as most constructivists would concede, can be viewed diVerently.
Moreover, whilst complexity and correspondence can plausibly be defended as
providing the standards by which competing theories should be adjudicated,
parsimony, analytical purchase, and predictive capacity have arguably just as
much claim to provide such a standard. And by that standard, most constructivist
institutionalism is likely to be found wanting.
Constructivism has much to contribute to contemporary institutional analysis,
though its appeal is likely to be greatest for those who do not believe that a
constructivist institutionalism 71