My own work diVered increasingly from Carl Friedrich’s in two respects. Theimpact of American political sociology directed my interest to elites, interest
groups, and trade unions ( 1980 ) which were undeveloped in European comparative
studies. In studies on Communism, Carl Friedrich emphasized totalitarianism with
a static bias. The neglect of interest groups was also detrimental to studies on Eastern
Europe. No internal conXict and development was possible. Even Friedrich’s co-
author, Z. Brzezinski, was no longer able to follow Friedrich and did not participate
in the second revised edition ofTotalitarian Dictatorship and Autocracy( 1965 ). I
came into a conXict of loyalty with my teacher because I was not willing to substitute
for Brzezinski. Since my studies in Moscow ( 1959 – 60 ) I was more able than the older
generation of Sovietologists and theoreticians of totalitarianism to discover modest
steps towards liberalization and the erosion of dictatorship. Moreover, in compara-
tive studies in both East and West, I was not interested in institutions per se,
but in combination with their impact on policies ( 1982 ). In that respect I was a
‘‘neoinstitutionalist’’ before the label has been invented.
The most interesting institution for older institutionalists, like Friedrich, was
federalism. Especially when they worked on the institutions of the budding
European Community they started from the normative assumption that federalism
was ‘‘progressive’’ per se. Doubts from the rational choice school in the work of
William Riker ( 1964 ) who calculated the costs of federalism by reluctant veto
groups in the decision-making process and especially in the implementation of
decisions at the national level, were widely ignored in Europe. In recent studies on
federalism I turned rather to comparisons of federalist and decentralized unitary
states. Only in the 1990 s did scholars from smaller European countries, like
Switzerland, Sweden, or the Netherlands (D. Braun, H. Keman), discover that
decentralized non-federal states in many respects had better performances than
federalist systems. The institutional economy studies discovered in addition that
the American model of a ‘‘competitive federalism’’—instead of a ‘‘federalism of
joint decision-making’’—does not prosper in federations with many small units
and that corruption spoils the decision-making process of federal institutions.
The new wave of institutional studies in economics proved to be fertile in
political science, enlarging the range of institutions to many quasi-governmental
institutions from the national banks to units which administer unemployment or
protection of environment. Comparative politics as a study of institutions will
certainly continue to develop in the direction of policy studies and include a greater
number of actors and veto groups than recognized in the older schools of institu-
tionalism, still largely thinking in terms of a global ‘‘checks-and-balances’’ theory.
Neoinstitutionalism will never develop back into the old institutionalism.
Even specialists of institutions who are inclined to accept the organization they
have chosen as an independent variable, can no longer prevent that non-
institutionalist approaches accept institutions only as one dependent variable
among others. Even a blatant nostalgia for the older institutionalism can lead
political institutions—old and new 755