7
D I V E R S I T Y A N D S T R U C T U R E D I N T E R A C T I O N S
must self-consciously make is even larger than experienced in the past.
While the usefulness of a universal model of rational behavior is chal-
lenged in chapter 4, the assumption of a universal framework composed
of nested sets of components within components for explaining human
behavior is retained throughout the book.
Building a Framework
Game-theoretical analysis is drawn on and expanded in this book in sev-
eral ways. First, I do not confine analysis to those situations that are sim-
ple enough to be analyzed as formal games. The core concept of an action
situation (discussed in chapters 2 and 3) can be formalized as a mathemat-
ical game to represent many simple and important situations. Many other
significant situations—particularly where rules are the object of choice—
are too complex to be modeled as a simple game. (Agent-based models
and simulations of diverse types will provide the modeling tools we need
to capture patterns of interaction and outcomes in many of these more
complex settings [Janssen 2003].)
Second, I dig further to develop a consistent method for overtly analyz-
ing the deeper structures that constitute any particular action situation.
For some game theorists, this deeper structure is irrelevant once the struc-
ture of a game itself is made explicit. Third, the narrow model of human
behavior used in game theory is viewed as one end of a continuum of
models of human behavior appropriate for institutional analysis. The
three basic assumptions of that model are used as a foundation for speci-
fying the type of assumptions that a theorist needs to make when animat-
ing an institutional analysis.
The challenge for institutional theorists—as I discuss in chapter 4—is to
know enough about the structure of a situation to select the appropriate
assumptions about human behavior that fit the type of situation under
analysis. Thus, the approach presented here encompasses contemporary
game theory as one of the theories that is consistent with the IAD frame-
work. Also included, as discussed in chapter 4, are broader theories that
assume individuals are fallible learners trying to do the best they can in
the long term by using norms and heuristics in making their immediate
decisions.
As a scholar committed to understanding underlying universal compo-
nents of all social systems, I do not introduce complexity lightly. I view
scientific explanation as requiring just enough variables to enable one to
explain, understand, and predict outcomes in relevant settings. Thus, for
many questions of interest to social scientists, one does not need to dig
down through nested layers of rules that are examined in the last half of
this book. One can develop a good analysis of the situation (chapters 2
8
C H A P T E R O N E
and 3), decide what assumptions to make about participants (chapter 4),
and predict outcomes. If the predictions are supported empirically, that
may be all that is needed.
3
If the predictions are not supported, however, as is the case with much
contemporary work on social dilemmas and settings involving trust and
reciprocity, one has to dig under the surface to begin to understand why.
And if one wants to improve the outcomes achieved over time, one is
faced with the need to understand the deeper structure in the grammar of
institutions discussed in chapter 5 and the types of rules used to create
structure as discussed in chapters 7 and 8. This volume can be viewed as
presenting a series of nested conceptual maps of the explanatory space
that social scientists can use in trying to understand and explain the diver-
sity of human patterns of behavior. Learning to use a set of conceptual
maps and determining the right amount of detail to use is, however, itself
a skill that takes some time to acquire just as it does with geographic
maps (see Levi 1997b).
Frameworks and Conceptual Maps
For example, if I want to know the quickest route from Providence Bay
to Gore Bay on the Manitoulin Island, where Vincent Ostrom and I spend
summers writing at our cabin on the shores of Lake Huron, I need a very
detailed map of the interior of the island itself. If I want to explain where
the Manitoulin Island is to a colleague—who wants to know where we
spend our summers—I need a less detailed and larger map that shows its
location on the northern shores of Lake Huron, one of the Great Lakes of
the North American continent. If I try to use a map of the entire Western
Hemisphere, however, the Great Lakes are all so small that locating the
Manitoulin Island itself may be a challenge. I may only be able to point
to the Province of Ontario in Canada, where it is located, or to the entire
set of the Great Lakes. The advantage of a good set of geographic maps
is that after centuries of hard work, multiple levels of detailed maps of
most places are available and are nested in a consistent manner within
one another. Most of us recognize that there is not one optimal map
that can be used for all purposes. Each level of detail is useful for different
purposes.
The “map” that I will elucidate in this volume is a conceptual frame-
work called, as mentioned above, the Institutional Analysis and Develop-
ment (IAD) framework. The publication of “The Three Worlds of Action:
A Metatheoretical Synthesis of Institutional Approaches” (Kiser and Os-
trom 1982) represented the initial published attempt to describe the IAD
framework. Our goal was to help integrate work undertaken by political
scientists, economists, anthropologists, lawyers, sociologists, psycholo-