In recent years, researchers have given greater attention to the potential importance of networks in many human processes, such as the reinforcement of personal lifestyle choices, the diffusion of policy innovations, and the delivery of social services. Although the connections that link together the units in a network generally represent information exchange, the content and influence of the information exchanged has not yet received as much attention. This is especially the case in studies of policy diffusion, where networks link multiple levels of government to nonprofit organizations with varying interests and resources. In How Information Matters, Kathleen Hale seeks to remedy this inattention through an empirically rich study of the organizational network that has been instrumental in the diffusion of drug courts in the United States.

Drug courts are intended to break the link between substance abuse and criminal behavior by means of therapeutic jurisprudence that imposes treatment while respecting due process. Although judges play a central role, drug courts require cooperation not only from the members of the conventional courtroom workgroup but also from probation officers and those who provide treatment services. The drug court model was first demonstrated in 1989 in Dade County, Florida. Subsequently, federal grants, favorable state policies, and information provided by a variety of nonprofit organizations contributed to the spread of drug courts to over 2,100 local jurisdictions by 2007.

Drawing on survey responses of state officials, interviews with staff of national nonprofit organizations, and statistical data, Hale maps the role of information within the national drug court network. By 2004, all state governments had public administrators whose policy portfolios included drug courts. Hale surveyed these administrators to identify their sources of information about drug courts and the content and usefulness of this information. She supplemented publically available information about the 19 relevant national nonprofit organizations with interviews of their staffs to determine the types of information they produce and the methods they used to disseminate it. She developed short case studies of drug court policy development in several states to help interpret the role of network information in particular contexts. In an effort to assess the impact of drug courts on crime, she collected data on program adoption and crime rates for each of the 50 states over 16 years.

This wealth of empirical evidence not only enables Hale to offer an authoritative assessment of the role of information in the diffusion of drug courts, but also supports her inductive generalizations about the operation of organizational networks in policy diffusion. One generalization is a typology of nonprofit organizations' information positions based on their preferences for the policy and their levels of engagement in promoting it. Champions strongly support the innovation and are highly engaged in helping diffuse it. Supporters strongly support the innovation but, because of broader missions, have lower levels of engagement. Challengers do not support the innovation, perhaps because it competes with policies they prefer more, but are highly engaged. Bystanders share information about the innovation without providing either strong support or high levels of engagement. Hale convincingly argues that interaction among these types of organizations leads to a testing and refinement of information that enhances its effectiveness and credibility.

In the case of drug courts, the National Association of Drug Court Professionals was the champion. In addition to participating in information exchange within the network, it encouraged the creation of similar but independent associations at the state level. Importantly, it set out general principles for implementing the drug court model. Further, as a nongovernmental organization not directly affected by changes in administration, it could remain committed to drug courts over a long period of time. Its interaction with other organizations interested in drug courts led to the development of mutual expectations that constituted a stable institutional arrangement for information exchange.

Another generalization offered by Hale concerns the value of synthesized information to potential adopters who want information about successful programs, best practices, and evaluations as resources for designing their own programs, applying for grants, and gaining necessary political support. Program templates facilitate policy borrowing and best practices, which often encapsulate the tacit knowledge of practitioners, and can provide guidance in how to tinker with the template to make it applicable to local circumstances.

Nonprofits play a central role in credibly synthesizing these sorts of information.

Hale reminds us of the many channels through which information can flow within the network and to potential adopters. Aside from common channels like newsletters and reports, rotations through advisory and governance boards establish personal links between organizational staff and their members that facilitate informal communications. Information exchange can even be done by creating opportunities for direct observation; for example, mentor courts that could be visited and observed by potential adopters were made available through coordination among network participants. This diversity of channels suggests that researchers should take considerable care in relying on readily observable but simple measures of connections to define, let along understand, networks.

The primary value of this research lies in providing a better understanding of the link between information developed within networks and policy adoption and implementation. However, it also attempts to relate policy to outcomes by estimating the effect of drug courts on changes in state-level crime rates. Although commendable, especially in light of efforts by the federal government during the era of the Law Enforcement Assistance Administration to induce the diffusion of questionable innovations in criminal justice, it is not clear how this state-level analysis could be more convincing than evaluations of specific drug courts. Indeed, analysts at the Washington State Institute for Public Policy identified 30 such evaluations of adequate quality to include in their meta-analysis, which reports statistically significant reductions in recidivism and positive net benefits from drug courts (Steve and Robert Barnoski, Washington State's Drug Courts for Adult Offenders: Outcome Evaluation and Cost–Benefit Analysis, 2003).

The pooled state-level time series analyses presented by Hale purport to show that the density of drug courts reduces crime. However, without fixed effects to take account of different average rates of crime across states, these results are suspect. More generally, the presentation of statistical results, such as reporting zero as the value of a statistically significant coefficient in Table 3.7, and alluding to but not presenting alternative statistical analyses, detracts somewhat from what is an otherwise well-crafted book.

Clearly, How Information Matters should be read by public administration scholars involved in the project of moving the unit of analysis from the agency to the network. Its focus on the content of information increases our general understanding of how networks create public value. It should also be read by political scientists who wish to understand that the policy diffusion focus on the roles of nonprofit organizations suggests a more complex organizational environment than is usually depicted in studies of policy diffusion across states.