9
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
gists, and others interested in how institutions affect the incentives con-
fronting individuals and their resultant behavior.
4
During the time since
this publication, the framework has been developed further
5
and applied
to analyze a diversity of empirical settings. These include:
• the study of land boards in Botswana (Wynne 1989);
• the impact of institutions on creating effective monitoring and evaluations
in government development projects (Gordillo and Andersson 2004);
• the incentives of operators and state government regarding coal roads in
Kentucky (Oakerson 1981);
• the evolution of coffee cooperatives in Cameroon (Walker 1998);
• the causes and effects of property-right changes among the Maasai of Kenya
(Mwangi 2003);
• the performance of housing condominiums in Korea (J. Choe 1992);
• the regulation of the phone industry in the United States (Schaaf 1989);
• the effect of rules on the outcomes of common-pool resource settings
throughout the world (Oakerson 1992; Blomquist 1992; E. Ostrom 1990,
1992b; Agrawal 1999; Schlager 1994, 2004; Tang 1992; E. Ostrom, Gard-
ner, and Walker 1994; Lam 1998; de Castro 2000; Dolsˇak 2000; Futemma
2000; Yandle 2001; Gibson, McKean, and Ostrom 2000);
• a comparison of nonprofit, for profit, and government day-care centers (Bus-
house 1999);
• the impact of decentralization on forest governance in Bolivia (Andersson
2002, 2004);
• the evolution of banking reform in the United States (Polski 2003); and
• the effect of incentives on donor and recipient behavior related to interna-
tional aid (Gibson, Anderson et al. 2005).
Our confidence in the usefulness of the IAD framework has grown
steadily in light of the wide diversity of empirical settings where it has
helped colleagues identify the key variables to undertake a systematic
analysis of the structure of the situations that individuals faced and
how rules, the nature of the events involved, and community affected
these situations over time. What is certainly true is that the number of
specific variables involved in each of these empirical studies is very
large. The specific values of variables involved in any one study (or one
location in a study) differ from the specific values of variables involved in
another study.
The problem of many variables, and potentially few instances of any
one combination of these variables, has been recognized by other scholars
as one of the perplexing problems haunting systematic empirical testing
of social science theories. James Coleman (1964, 516–19) referred to the
development and testing of “sometimes true theories,” by which he meant
that explanations were likely to hold under specific conditions and not
10
C H A P T E R O N E
under others. If a small number of conditions were identified, sometimes
true theories would not present a major problem for the social sciences.
Rigorous analysis of many important questions, however, does eventu-
ally require examining a large number of variables. Viewing macropoliti-
cal orders in developed Western societies, for example, Fritz Scharpf
(1997, 22) points out that the national institutional settings “known to
affect policy processes can be described as being either unitary or federal,
parliamentary or presidential, have two- or multi-party systems in which
interactions are competitive or consociational, and with pluralist or neo-
corporatist systems of interest intermediation.” Each one of the five vari-
ables can exist in one or the other “setting” independently of the other
four variables. And, to make it worse, there may be variables related to
the particular policy area—such as banking, environmental policy, or edu-
cation—that may also change. “For comparative policy research, this
means that the potential number of different constellations of situational
and institutional factors will be extremely large—so large, in fact, that it
is rather unlikely that exactly the same factor combination will appear in
many empirical cases” (23). A similar level of complexity exists when
analyzing factors affecting the performance of city-county consolidation
efforts (Carr and Feiock 2004).
Hammond and Butler (2003) have illustrated this problem clearly in
their critique of the work of some institutional theorists who have made
overly strong claims for the overarching differences between parliamen-
tary and presidential systems. Presidential systems—according to Burns
(1963), Sundquist (1968), and Valenzuela (1993)—are thought to slow, if
not halt, policy change and lead to obstruction, frustration, and deadlock
interspersed with occasional bursts of change when a president faces both
houses of Congress dominated by his own party. Hammond and Butler
carefully analyze the interaction between rules and the preference profiles
that may exist in five variations of institutional rules. They conclude “that
considering institutional rules alone provides an inadequate guide to the
behavior of any system” (Hammond and Butler 2003, 183).
As Marwell and Oliver (1993, 25) put it, the “predictions that we can
validly generate must be complex, interactive and conditional.” And, we
can hope that some changes in a component are neutral—or have no im-
pact on outcomes—in at least some settings (as biologists are now learn-
ing about in regard to genotypes; see Gavrilets 2003). While verifying
the empirical warrantability of precise predictions has been the guiding
standard for much of the work in political economy, we may have to
be satisfied with an understanding of the complexity of structures and a
capacity to expect a broad pattern of outcomes from a structure rather
than a precise point prediction. An outcome consistent with a pattern