Getting the Odds Right:
Casino Gaming Di usion, the Initiative Process and Expected Voter Support
Frederick J. Boehmke 1
California Institute of Technology
October 26, 1999
Paper prepared for presentation at the 71st Annual Meeting of the Southern
Political Science Association, November 3-6, 1999.
1boehmke@hss.caltech.edu. I would like to thank Mike Alvarez, Valentina Bali, Je Banks,
Garrett Glasgow, Jonathan Katz, Rod Kiewiet, Richard McKelvey, John Patty and the members
of the Caltech Social Sciences Graduate Student Seminar Series for their comments. Additional
thanks go to Rod Kiewiet for access to his data on state nances. The nancial support of the
Division of Humanities and Social Sciences at Caltech and the John Randolph Haynes and Dora
Haynes Foundation is greatly appreciated. Travel to this conference made possible by funding
provided by the Division of Humanitites and Social Sciences, Caltech.
Abstract
In this paper I develop and test a formal model of the role that the initiative process plays
in shaping policy outcomes. By introducing uncertainty over the median voter's ideal point,
I show that the possibility of proposing an initiative makes it cheaper for an interest group
to change policy, both through the initiative process and by using contributions to convince
the legislature to change policy. By also allowing groups to use information drawn from
neighboring states' adoptions to estimate the probability an initiative would pass, I show
that policy di usion should function primarily as information di usion between initiative
states. The predictions of the model are then tested through an event history analysis of state
casino gaming legalizations. The initiative process is found to be a positive and signi cant
predictor of adoption and its e ect is increased by voter liberalism. There is strong support
for the informational approach to di usion with a positive
ow between initiative states and
none from or to non-initiative states.
1 Introduction
Scholars have documented the
uctuations in th