The effect of political information on direct democracy strategies and outcomes

Author: Lupia, Arthur William

Year: 1991

Degree: Dissertation (Ph.D.)

Advisors: Ordeshook, Peter C.; McKelvey, Richard D.

Committee Members: Ordeshook, Peter C.; McKelvey, Richard D.; Kiewiet, D. Roderick; Ledyard, John O.

Option: Social Science

DOI: 10.7907/rxpx-pg64

Abstract

The intent of the dissertation is to detail the effects of political information on participant strategies and outcomes in an electoral environment called "direct democracy." Direct democracy is a decision-making institution in which an agenda setter chooses an alternative to a pre-determined Status Quo and voters vote for either the Status Quo or the agenda setter's alternative. Through the use of a spatial election model, a survey of California insurance reform voters, and a series of laboratory experiments, I show how the direct democracy outcome corresponds to the underlying preferences of a majority of the electorate. The spatial model is used to establish that under conditions of incomplete information, the direct democracy outcome corresponds to the (full information) wishes of a majority of the electorate only when there are sufficient opportunities to cue off of the actions of other, credible, electoral participants. The empirical tools and experiments are used to examine electoral environments where different types of information are available. It is established that voters do not require full information in order to vote for their full information preferred alternative. It is also established that, in the absence of certain types of information, rational voters can cast votes for alternatives that lead to their least preferred outcome.

That voters do not require full information in order to vote for their full information preferred alternative suggests that voters do not necessarily need to understand an issue to vote in their own best interest. That rational voters can cast "ex post mistaken" votes under conditions of incomplete information implies that direct democracy outcomes can be manipulated by well-endowed interests. The dissertation details the conditions under which each of these outcomes is likely to occur.

Files