behavioralism was a reaction against the old institutionalism. This reaction comes
in three main guises, each rooted in one of the main social science disciplines. So,
political science gave us historical institutionalism, economics gave us rational
choice institutionalism, and sociology gave us sociological institutionalism (see
Goodin 1996 , 2 – 20 ; Hall and Taylor 1996 , 936 ). Approaches proliferate (Lowndes
2002 ; Peters 1999 ). The labels vary—sociological institutionalism begat ideational
institutionalism begat constructivism. The several proponents squabble. For
aWcionados of such debates, the several approaches, the key contributions, and
their diVerences are clearly set out in Chapters 1 – 5. A further summary is unne-
cessary.
There are important diVerences between the several approaches; for example,
between inductive and deductive methods. However, such diVerences are less
important than their common ground in a modernist-empiricist epistemology.
Thus, institutions such as legislatures, constitutions, and civil services are treated as
discrete objects that can be compared, measured, and classiWed. If American
concern with hypothesis testing and deductive methods raises the collective skep-
tical eyebrow of British political science, then Bryce’s claim ( 1929 , vol. 1 , 13 ) that
‘‘[I]t is Facts that are needed: Facts, Facts, Facts’’ would resonate with many. British
modernist empiricism has much in common with the positivism underpinning
mainstream American political science; both believe in comparison, measurement,
law-like generalization, and neutral evidence.
In so labeling the new institutionalism, I do not seek to criticize it, only to locate
it in a broader tradition. Adcock et al. ( 2006 ) do this job admirably. They explore
the diverse roots of the new institutionalism to dismiss the conventional narrative
of a shared rejection of behavioralism. They dispute there is a shared research
agenda or even the prospect of convergence. The new institutionalism is composed
of diverse strands, building on diVerent and probably incompatible intellectual
traditions, united only in the study of political institutions and their commitment
to modernist-empiricism. The new institutionalism may be a shared label but its
divergent roots in incommensurable traditions mean the several strands have
little else in common. When we move further aWeld, the divergence is even
more marked.
AtWrst glance, British political science took to historical institutionalism like a
duck to water. However, many British political scientists denied any novelty to the
new institutionalism. After all, in Britain, neither the behavioral revolution nor
rational choice had swept the study of institutions away. Also, the new institution-
alism is such a jumble of ideas and traditions that it can be raided for the bits that
easilyWt with other traditions. So, British political scientists could interpret the rise
of the new institutionalism in America as a vindication of British modernist
empiricism, with its skepticism toward both universal theory, and the scientism
characterizing American political science. Thus, Marshall ( 1999 , 284 – 5 ) observes we
do not need ‘‘more or deeper conceptual theories’’ because ‘‘we have already have
old institutionalisms 93