Columbia’s CCP: A Case Study
Limitations and Cautions
We should be cautious not to over-interpret or draw causal inferences about the observed networks for several reasons. First, these analyses and graphic presentations provide a one-time snapshot of interactions between individu- als early in the CCP project. Consequently, these data will not allow us to tease out any pre-existing relationships and networks that may be operating. Thus, whether these networks are CCP-induced or reflect pre-existing rela- tionships is unknown. A longitudinal look at these networks is currently in progress to see how these linkages change during the course of the CCP funding. Combined with careful fieldwork, this should give us a stronger as- sessment of CCP’s contribution. Second, these analyses are limited to inter- actions between individuals, which may or may not reflect the nature and ex- tent of partnerships between agencies. To capture interagency contacts, our unit of analysis for the longitudinal analysis will be the organization/agency rather than the individual (This analysis strategy also avoids the individual- level attrition problem that is always present in longitudinal data). Finally, the present analysis is limited by the nature of the original sample. Who ends up in the sample can have a large influence on the outcome of network analy- sis. While we are satisfied that this problem has been minimized by allowing sites to self-define a comprehensive list of CCP participants, nevertheless, we suspect that some individuals and groups have been overlooked at each site. Generally speaking, one might characterize this network analysis as a study of “elites” — in this case, community, city and agency leaders. Networks that may exist among street-level employees and community volunteers are un- der-represented (although not completely absent) from this analysis.
Despite these limitations, network analysis provides an important empirical tool for examining the nature and extent of community-based partnerships and coalitions. While it is easy to talk about “interagency cooperation” in grant proposals or in personal interviews, it is not so easy to create the illu- sion of a network (for the benefit of researchers and others) when members of that network are asked, individually, about their frequency of interaction with one another. The results here suggest that the number and density of networks varies by site and that resultant patterns of contact are generally consistent with our field observations.
Heinz, J. P., & Manikas, P. M. (1992).
“Networks Among Elites in a Crimi-
nal Justice System.”
Law and Society Review, 26: 831-861.
BOTEC Analysis Corporation