SIZING UP STATE POLICY INNOVATION RESEARCH

In the literature on state policy innovation, there are three major explanations for what causes a government to adopt a new policy. One is the internal determinants model, which posits that the main factors leading a state to innovate are internal political, social and economic characteristics of the state. The other two are diffusion models-the regional diffusion model, and the national interaction model-which see state policy adoptions as emulations of earlier adoptions by other states. Each of the three models has been associated with a distinct strategy for empirical testing. The regional diffusion model has been tested with factor analysis, the national interaction model with time-series regression, and the internal determinants model with sectional regression. In this paper, I explore the ability of these explanation'' methodologies to detect the true innovation process underlying state policy adoptions, by applying these methodologies to data generated from simulated innovation processes with known characteristics. I find that the methodologies often yield incorrect conclusions about the character of innovation. I conclude by presenting an agenda for refining a superior alternative methodology: the event history analysis approach to state policy innovation research introduced by Berry and Berry (1990).