date primarily taken the form of growing interconnectedness at a regional rather
than a global level (the consolidation of the elements of the ‘Triad’ of Europe,
Japan and North America), does this open up scope for new forms of regulation
at a regional level (and notably, of course, by the EU)?
Other authors in this volume address a number of these questions. Here I
intend to focus more narrowly on three themes which underlie much of the
literature sceptical of the arguments for globalisation-induced convergence:
the interconnectedness of institutions; the logic of ‘path-dependence’; and the
relative autonomy of politics.
The first consideration is that within any society, the evolution of one set
of institutions is affected by that of others (see also Chapter 5). For example,
labour markets, education systems, welfare regimes and family structures are
interdependent in their operation. Hollingsworth and Boyer (1997: 2–3) use
the notion of ‘social systems of production’ to denote a multiplicity of regula-
tory mechanisms: ‘the industrial relations system; the system of training...; the
internal structure of corporate firms; the structured relationships among
firms...; the financial markets of a society; the conceptions of fairness and
justice held by capital and labor; the structure of the state and its policies; and
a society’s idiosyncratic customs and traditions as well as norms, moral princi-
ples, rules, laws and recipes for action’. These, they suggest, may be ‘tightly
coupled with each other’ and thus ‘coalesce into a complex social configura-
tion’. Similarly, Dore (2000: 45–7) writes of ‘institutional interlock’ as typical of
national economies and the relationship between the economy and broader
society. Of course, in no society are all institutions functionally integrated into
a stable whole: there are always tensions and contradictions which contribute
to institutional dynamism. Nevertheless, there are two obvious implications of
this interconnectedness. First, how an institution actually operates will depend
on national context; for example, the economic effects of German works coun-
cils are shaped by other features of German society and would not necessarily
be replicated in other countries where the institutional ensemble is different.
Second, an attempt to change one set of institutions (for example, ‘deregulat-
ing’ labour markets) may require alterations in ‘a complete architecture of gen-
erally interdependent institutions’ (Boyer, 1996: 55). Consequently, even were
it generally accepted that the Anglo-American model of capital performs best
under conditions of intensified global competition (and this is of course dis-
puted), ‘there can be no assumption that the evolution of national systems is
guided by a mechanism that selects more efficient arrangements’; ‘dysfunc-
tional institutional arrangements’ may prove resilient (Wade, 1996: 85).
This links to the thesis of path-dependence. Here, the argument is that past
choices constrain future options. Looking at the historical evolution of the
German and Japanese varieties of capitalism, Streeck (2001: 30) argues that ‘the
political economies of Germany and Japan were built by sequences of political
decisions, the outcomes of which were far from predetermined’. The
concept of path-dependence has also been used in a range of interpretations of
426 International Human Resource Management