role in establishing a new trajectory of institutional evolution. They are, in other
words, not reducible to the condition they seek to describe and explain.
The implications of this are clear—if we are to understand path-shaping institu-
tional change we must acknowledge the independent causal and constitutive role
of ideas, since the developmental trajectory of a given regime or policy paradigm
cannot be derived from the exhibited or latent contradictions of the old regime
or policy paradigm. It is, instead, contingent upon the ideational contestation
unleashed in the moment of crisis itself. Though this is not an inference that Blyth
himself draws, there is, then, no hope of a predictive science of crisis resolution, capable
of pointing prior to the onset of crisis to the path of institutional change—for the
causal chain is incomplete until such time as the crisis has been successfully narrated.
This is an important intervention and it provides a series of correspondingly
signiWcant insights into the developmental trajectories of Swedish and US capital-
ism in the twentieth century. In particular, it draws attention to the role of business
in proselytizing and sponsoring new and/or alternative economic theories and in
setting the discursive parameters within which inXuential crisis narratives are likely
to be framed, and to the crucial relationship between business, think tanks,
and professional economists. It also reminds us, usefully, that in order to prove
inXuential, (economic) ideas need not bear much relationship to the reality
they purportedly represent. In a classically constructivist institutionalist vein, it
demonstrates that, if believed and acted upon, economic ideas have a tendency to
become self-fulWlling prophecies (see also Hay and Rosamond 2002 ).
Yet its limitations also show that constructivist institutionalism is still very much
a work in progress. Blyth raises just as many theoretical, methodological, and,
indeed, empirical questions as he answers. Moreover, the text is characterized
by some signiWcant and by no means unrepresentative tensions, contradictions,
and silences. None of these are insurmountable impediments to the development
of a more consistently constructivist institutionalism. Yet they do perhaps serve to
indicate the work still required if the profound challenge that constructivism poses
to more conventional approaches to institutional analysis, and the insights it oVers,
are both to be more widely appreciated.
In the context of contemporary neoinstitutionalism, it is Blyth’s comments
on the relationship between ideas and interests that are likely to prove most
controversial. It is in these comments that the distinctiveness of the constructivist
variant of institutionalism resides. His core claim is, in essence, that actors’ conduct is
not a (direct) reXection of their material interests but, rather, a reXection of particu-
larperceptionsof their material interests (see also Wendt 1999 , 113 – 35 ). Our material
circumstances do not directly determine our behavior, though our perceptions of
such circumstances (and, indeed, of our stake in various conceivable outcomes),
may. 12 In his own terms, it is ideas that render interests ‘‘actionable’’ (Blyth 2002 , 39 ).
12 The parentheses are important here. There is something of a tendency in the existing literature to
treat the issue of interest-formation and representation as a question solely of the accuracy of the
information actors have about their external environment. If there is a disparity between an actor’s
68 colin hay