theories of politics, traditional political institutions, such as the legislature, the legal
system and the state, as well as traditional economic institutions, such as theWrm, have
receded in importance from the position they held in earlier theories’’ (March and
Olsen 1989 , 1 ). TheWrst chapter is thus devoted to what are described as ‘‘Institutional
Perspectives on Politics,’’ but nowhere is the chapter concerned with the deWnitional
problem. In the last section of that chapter, entitled ‘‘The role of institutions in
politics,’’ the authors state: ‘‘In the remainder of the present book we wish to explore
some ways in which the institutions of politics, particularly administrative institutions,
provide order and inXuence change in politics’’ ( 1989 , 16 ). Thus March and Olsen are
not apparently more conscious than their predecessors of the fact that a
problem of deWnition arises with respect to institutions and that a distinction has to
be made between institutions and other ‘‘elements’’ which play a part in politics.
The question of the meaning of the concept of ‘‘institution’’ came to be raised only
a few years later, in the mid- 1990 ’s; yet this was done atWrst somewhat marginally by
Goodin who noticed the diVerent part which institutions play in the various social
science disciplines ( 1996 , 1 – 24 ). Probably theWrst full confrontation with the prob-
lem was in Lane and Ersson’s volume onThe New Institutional Politics( 2000 ). This
was something of a ‘‘volte-face’’ by these authors, as they still adopted, in the fourth
edition of theirPolitics and Society in Western Europe,publishedin 1999 , the kind of
‘‘unproblematic’’ language of March and Olsen. No attempt was then made, for
instance, in the section of the introductory chapter entitled ‘‘Social Structure versus
Political Institutions,’’ to deWne these expressions. The authors had stated, without
any further concern, that ‘‘[t]he focus on the variation between institutions of
political democracy and their sources in civil society as well as their consequences
for political outcomes creates a certain logical structure for the contents of the
volume’’ (Lane and Ersson 1999 , 14 ). Yet, one year later, they devoted a whole chapter
to the question ‘‘What is an institution’’ ( 2000 , 23 – 37 ) in which they referred, among
other problems, to the ‘‘ambiguity of institution’’ ( 24 – 7 ) and proceeded to discuss a
distinction which they made between what they described as ‘‘holistic’’ or ‘‘socio-
logical institutionalism’’ and ‘‘rational-choice institutionalism’’ ( 29 – 36 ). This was
real progress, but Lane and Ersson do not appear to consider that institutions have
a particular ‘‘resonance,’’ so to speak, in the political science context. ToWnd out why
there is such a ‘‘resonance,’’ we needWrst to turn to a general examination of the
diVerences among the various disciplines on the subject.
- 2 What Institutions are for Economists and Sociologists
and why Political Scientists have to DiVer
InThe Theory of Institutional Design, edited by R. E. Goodin ( 1996 ), the question of
diVerent kinds of ‘‘institutionalisms’’ is evoked, possibly for theWrst time, at least in
720 jean blondel