Most observers would call these dynamics real effects, to be found in the world,
not the model. I would call theminstitutional effects. When agents ‘see’ a world of
patterns and act on them, in the first instance by telling each other what they see,
they have begun a process of transforming patterns into institutions (models whose
formal properties are more or less fixed and publicly available as an ensembles of
rules). Both chapters detail what governments do: agents’ choices take effect in and
through institutions. In this process, agents have an effect on those institutions
through their choices, if only to reinforce institutional rules.
Waltz has never denied the causal significance of agents and institutions (recall
the first and second images). Indeed he has acknowledged that structures do no more
than ‘condition behaviors and outcomes’. If agents are moved by unseen forces, their
choices foreclosed by conditions they are not even aware of, then there is little point
in calling them agents at all. If agents do make choices with consequences, then we
need to ask, how do agents knowwhich structures constrain them in what ways? To
suggest that agents must have a relevant structural model in mind, or that institu-
tional rules must somehow accord with such a model, calls for theories about
structural effects as causes of institutional effects.
‘Structures select’
Speaking metaphorically (metaphors are simple models), Waltz drew a clear line
between theory and reality inTheory of International Politics, and proceeded to blur
it. The most striking example of this tendency arises in his well-known defence of
his structural theory of international anarchy ‘through analogy with microeconomic
theory’. Waltz advocated ‘reasoning by analogy’ when ‘different domains are
structurally similar’.^46 In this context, the term domainis ambiguous insofar as it
suggests a resemblance between different ‘parts’ of the real world – in this case, firms
and states, well-formed markets and international anarchy. Even if Waltz’s discussion
puts theory first, the analogy seems to operate on both sides of the theory/reality
divide.
‘Microeconomic theory describes how an order is spontaneously formed from
the self-interested acts and interactions of individual units – in this case, persons and
firms’. Key to this formulation is the phrase spontaneously formed. Markets are
‘individualist in origin, spontaneously generated, and unintended’. Not designed or
produced on any model, a ‘market is not an institution or an agent in any palpable
sense’. And yet ‘The market is a cause interposed between the economic actors and
the results they produce’.^47
Waltz appears to have it both ways. Markets do not really exist, but they operate
causally in producing real results. Appearances are deceiving. Either markets are
institutions, which seem to be causally implicated in market behaviour, or markets
are structural causes producing structural effects, as stipulated in a theoretical model.
It is easy enough to show – devise a model showing – that markets are always
institutions (‘hedged about’, as Waltz said himself), even if some markets form
spontaneously.^48 It is almost as easy to show that people respond to institutions as
96 Structure? What structure?