Dynamic representation(s): Federal criminal justice policy and an alternative dimension of public mood

Does public policy respond to public opinion? Previous research suggests dynamic representation occurs in the aggregate. Yet, most of the evidence for policy response is limited to the policy intentions of elected officials on issues related to more or less government spending. We examine policy response to an alternative dimension of public mood, public preferences for more or less punitive criminal justice policies, using multiple indicators of policy from various stages of the policy-making process. Criminal justice policy should be responsive to public preferences given the public's concern about crime and the negative social construction of criminals. Thus, there is an electoral incentive for public officials to respond to public preferences along this alternative dimension of public sentiment regarding criminal justice policy. We estimate a DYMIMIC model of federal criminal justice policy as a function of the multiple dimensions of public policy mood using Kalman filtering. The results indicate that criminal justice policy responds to the second, not the first, dimension of public mood. We find evidence that policy-makers at multiple stages of the policy process are able to differentiate among multiple signals from the public and respond appropriately. The results present a more sophisticated portrait of democratic responsiveness.