When Voters Make Laws: How Direct Democracy is Shaping American Cities
USC Gould School of Law
Mathew D. McCubbins
Duke University School of Law
December 1, 2007
USC CLEO Research Paper No. C07-14
USC Law Legal Studies Paper No. 07-13
Journal Public Works Policy and Management, Summer 2008
City, county and municipal ballot propositions are often used to raise funds for new infrastructure projects. We discuss three challenges facing the use the initiative and referendum process to fund local infrastructure projects; these problems suggest that we cannot be confident that voters have been able to make decisions that best serve their interests. We do not conclude that such difficulties are necessarily fatal to the use of direct democracy, but our analysis has consequences for the regulatory structure that shapes this process in cities, counties and special districts.
First, public agencies can make take-it-or-leave-it offers to voter. Voters do not have the ability to amend a bond proposal to move it closer to their preferences; they face a binary choice between either the status quo or the particular bond on the ballot. This type of decision making confers substantial power on the agenda setter and may result in an outcome that does not match the preferences of the median voter. This may be exacerbated by low voter turnout that characterizes many purely local elections where bond measures are decided.
Second, several common social choice problems can arise when sophisticated political actors manipulate the initiative agenda. One of the most pernicious of the problems they identify is that of sequential elimination agendas. The core problem with sequential elimination agendas is that they do not allow citizens to directly compare all of the alternatives and, therefore, do not allow them to make tradeoffs among their options. The typical problems created by sequential elimination agendas, suffered by state-level initiatives as well as local ballot propositions, are further compounded by two related aspects of local bond measures: limits on bond capacity created by state law, bond rating agencies or voters' (un)willingness to be taxed; and the presence of overlapping governmental authorities that can each propose bond measures to voters. The race for tax revenue can create a situation where bonds cannot be passed when they are needed; instead, funding is already allocated to unneeded or unwanted projects that were approved at an earlier election (i.e., earlier in the agenda).
Third, we find that voters can be faced with an information environment that precludes their making reasoned choices at the ballot box. For voters to make a reasoned choice that can improve their welfare, they must have correct beliefs about the consequences of their vote. Occasionally, voters may be able to identify the consequences of their vote on the basis of personal knowledge and experience, but they usually lack such information. In that case, voters must be able to learn from someone else - a knowledgeable, trustworthy endorser. They can then cast a reasoned vote, not depending on their own encyclopedic knowledge of a policy or their independent assessment of the merits of the argument, but using reliable cues provided by a knowledgeable, trusted endorser. In a series of case studies from Los Angeles and San Francisco, we find very mixed information environments for local bond measures, some which may allow voters to learn about the consequences of their decisions through trustworthy endorsements, and others where those important voting cues are few or nonexistent.
Number of Pages in PDF File: 59
Date posted: May 27, 2008 ; Last revised: November 16, 2013