stable equilibria are observable outcomes of this complex game, any of which
would involve each agent taking different actions in the two ex ante identical
sub-economies. Matsuyama (2000, 2002, 2004) coined the term ”symmetry-
breaking” to refer to this heterogeneity-generating process.
The third approach originates in the business strategy literature, and is often
presented as part of a general critique of economic theory. With their traditional
emphasis on investigating the workings of firms as complex organizations, strat-
egy scholars have been particularly concerned with understanding the sources
and dynamics of inter-firm heterogeneity along various functions and charac-
teristics. In the dominant view, as articulated by Nelson and Winter (1982),
firms operate in such highly complex and ever-changing environments that they
entertain no hope of ever accumulating enough knowledge about their world
to view it as a strategic game or formulate a precise game-theoretic strategy
to guide their overall behavior. Rather, firms grope for economic performance
via a heuristic learning process of trial and error and the continual updating
of routines and rules of thumb eschewing optimization. In this evolutionary
vision, heterogeneity is simply an inevitable outcome of this groping behavior,
with firms ending up with different heuristic strategies and core capabilities to
implement them. These ”discretionary” differences can then be sustained over
extended periods of time due to the presence of barriers to successful imitation
generated by the differences in core competencies, and also by forces of path
dependence in the evolution of firms’ choices. This literature often criticizes
economic theory for not adequately accounting for inter-firm differences, other
than postulating them either as reflecting variations in initial conditions, or as
exogenous consequences of the luck of a draw in stochastic models. This failure
is attributed to the fact that economic theorists persist, as part of their excessive
reliance on complete rationality, in ”taking a firm’s choice sets as obvious to it
and the best choice similarly clear and obvious” (Nelson, 1991).1
1 An interesting development over the last two decades is a strand of literature straddling the
traditional boundaries between industrial organization and business strategy and addressing
issues of interest to both fields, making them increasingly related: See Shapiro (1989), Rumelt,