The Initiative Process and Policy Innovation in the American States

Abstract We utilize a new policy adoption database with over 500 policies to test whether the initiative process influences the timing of policy adoption. Prior studies have produced both supportive and null findings of the effect of the initiative, but typically examine policies one policy or a single composite score at a time. Theoretical accounts suggest that the initiative process should have heterogeneous effects on policy outcomes depending on the configuration of public and government preferences. By pooling hundreds of policies we are able to estimate the average effect of the initiative process on state policy adoption more systematically while also evaluating variation in its effect. We find via a pooled event history analysis that the initiative tends to increase innovativeness, but that this effect can be cancelled out by signature and distribution requirements. We find that this effect varies substantially across policies and is more consistently positive on average in states more liberal populations. We also find evidence that the initiative process moderates the effect of ideology on policy adoption, while making the adoption of non-ideological policies more likely on average.