Agency representatives were asked to list agencies with which regular contact was maintained in the process of dealing with the youth gang problem. Policy response was measured by the adoption of four policy options: (1) making staff training available; (2) having a policy for dealing with the youth gang problem; (3) having a policy in writing; and (4) attempting to influence legislation on youth gangs. Rasch modeling supported the scalability of the four policy options, and the Structure program was used to identify elements of network structure within each community. A generalized, linear model analysis of variance of the policy response scale demonstrated that structural equivalence and clique membership accounted for 54 percent of the variation in policy responses. An examination of clique means indicated that network structure may retard as well as enhance policy response adaptations. In policy terms, this finding suggested that network structure would be mobilized by those who wanted to develop a unified, national-level response to youth gangs at the local level. Policy implications of the analysis results are discussed in terms of agency linkages, community organization strategies, research on interagency relations, and quality of information available to local networks. 44 references and 6 tables