This study considers how states change policy innovations as they diffuse. This process, known as reinvention, has been the subject of recent innovation diffusion research (Mooney and Lee 1995; Glick and Hays 1991). This paper examines the role of controversy in structuring the nature of reinvention, creating a more fully specified model of the causes of reinvention. This research is a comparative analysis of the evolution of provisions of three policy innovations: child abuse reporting laws, crime victim compensation law, and public campaign funding laws. The data show that varying levels of controversy affect the nature of policy reinvention and that reinvention results from a combination of social learning, political characteristics, and contextual factors.
Innovation diffusion is a dynamic process; later adopters modify innovations based on the experiences of earlier adopters. Everett Rogers, a prominent communications scholar and author of the Diffusion of Innovations, termed this learning process during diffusion reinvention (1983,1978; Rice and Rogers 1980). Rather than a simple diffusion process in which subsequent adopters follow the lead of an initial innovator, reinvention assumes a dynamic, constantly evolving process with adopters molding and shaping the innovation as it diffuses. Specifically, reinvention refers to purposeful changes made to innovations as they diffuse (Rogers 1983,1978; Rice and Rogers 1980; Glick and Hays 1991).
Reinvention challenges existing notions of the diffusion of innovations among states. Studies which have analyzed reinvention suggest that later adopters increase the scope of a law as it diffuses (Glick and Hays 1991; Mooney and Lee 1995; for a different assessment see Clark 1985). If this is true, it raises questions about the innovativeness of the earliest adopting states. Early adopters may be considered innovators in the sense that they set the stage for later adoption by other states. However, these innovators may in fact be adopting weak or ineffective versions of the innovative law, which is improved upon by the later adopting, laggard states.
Current research on reinvention has focused on this policy change over time (Mooney and Lee 1995; Glick and Hays 1991; Clark 1985). The content of a policy is measured along some dimension to assess its scope or permissiveness, and change over time is demonstrated by correlating a state's score with the date it adopted the law This method implies that time is in some sense an explanatory variable for reinvention. Theoretical support for this proposition comes from the notion of a social learning process (Rogers 1978; Mooney and Lee 1995) through which states learn from and improve upon the policy of earlier adopters. But what if there is little correlation between time and policy scope? Does this imply that the policy was not reinvented? Everett Rogers provides some guidance on this point by suggesting that adopters can reinvent an innovation for various reasons (1983). While a social learning process may play a role, reinvention occurs whenever an innovation is adapted in any way by subsequent adopters during diffusion (Rogers 1978). Thus, there is little a priori reason to connect the concept of reinvention exclusively to the date the law was adopted. While social learning may generally guide a trend toward increasing the scope of a law many other forces may also stimulate later adopting states to reinvent an innovation.
The purpose of this research is to model policy reinvention in a general explanatory framework, examining a number of causal factors hypothesized to lead a state to reinvent an innovation. In addition, I select three laws for analysis which differ in policy controversy, a crucial dimension affecting reinvention. The dependent variable consists of changes in the provisions of child abuse reporting laws, crime victim compensation laws, and public campaign funding laws. Specifically, the dimension of reinvention I analyze is the comprehensiveness of each adopting state's law based on an evaluation of the substantive effect of the provisions included in each initially adopted law. …