8
clear evidence that measures associated with effective local networks increase the level of
inspections and decrease the level of violations. In the case of water pollution, at least, local
institutions and networks have worked as a complement to federal regulations, enhancing
bureaucratic outputs and outcomes alike. We argue that the observed complementarity arises
because of the specific (but not unique) conditions affecting water policy, where increasing
stress on the natural hydrological system from population growth and development has
dramatically increased the adverse impacts of fragmented authority. Independent federal, state,
and local agencies grapple with complex problems spanning their individual policy domains,
problems that they individually have neither the authority nor competence to resolve.
Fragmentation of authority creates opportunities for joint optimization and for
avoiding negative externalities across policy domains, which in turn provides a niche for the
development of local policy networks and institutions to informally coordinate policies locally
that would be very difficult to coordinate formally through centralized authorities (Bardach
1998; Lubell et al 2002). The creative drive behind these institutions and networks is to foster
benefits through mutual exchange. Their power lies not in the control of authoritative
punishments, but rather in threatened withdrawal of benefits from actors shunned by the
networks. These benefits are substantial enough to encourage even private permit holders to
forgo the short-term benefits from “optimal” violations in favor of long-term cooperation with
the broader water policy community—a policy community generally preferring cleaner water
that enhances other potential uses.
To extend this initial finding, we need to extend the empirical and theoretical
foundations for the study of policy institutions. Empirically, we need to develop more direct
measures of networks to replace the rough proxy measures in the previous study. Theoretically,
we need to identify the characteristics of networks that differentiate between more and less
efficient networks. For example, is the average number of contacts in a local policy network
sufficient to judge its efficiency, as is implicit in Schneider et al’s (2003) study of water policy
networks? Or is the structure of contacts—the weak versus strong ties— a more important
determinant of efficiency than the number of contacts. Are “learning networks” in which
policy activists mimic the actions of their most successful peers effective, or are “information
networks” that transmit a greater amount of current information necessary to make local
networks effective? The remainder of this presentation explores these specific questions.
Cooperation and network structure
The influence of local political networks in Wang and Scholz (2003) is based primarily
Convention