Cities choose to privatize public services for several reasons. Elected chief executives emphasize efficiency and quality of public service provision as the primary benefits of privatization. Mayors resort to privatization in an attempt to balance the budget coping with deteriorating tax base during recession. Previous privatization literature approaches privatization mainly from the economic and managerial perspectives.

This paper examines the political perspective of privatization by analyzing the initial decision stage of privatization by politicians. Political dimensions of privatization in the decision making process have been less studied compared to other facets of privatization such as transaction costs (asset specificity and measurability of outcomes), monitoring costs, and incongruent goals across private and public agents, leading to “hollow” government propositions.  The smaller number of research that tested political factors examined the demand of the public for smaller government, public employees’ opposition, political ideology, and forms of government (i.e., council-manager or mayor-council). Their empirical results often proved that such political factors exerted little influence on privatization decisions. This paper introduces a new political factor in the empirical analysis format. Based on the rational choice assumption that Mayhew proposed that the ultimate goal of politicians is to get elected to office, this paper tests the electoral incentives of local politicians and its relationship to privatization decisions. This paper’s preliminary regression results show that electoral variables are statistically significant controlling for other factors mentioned in this proposal.

This research studies American municipalities with population of 100,000 or more. The authors built the data particularly regarding political variables. For the base, the authors use the Local Government Service Delivery Choices survey data for two time periods, years 2002 and 2007, purchased from the International City/County Management Association. The Census data was used for some of the control variables. For political factors, regression models test if politicians and/or citizens were involved in the decision making process. With the premise that privatization decision is primarily political, electoral incentives are tested by incorporating electoral cycles into analysis. Forms of government and other political factors suggested in the previous literature were also tested as control variables. Other control variables include local finance data and other socio-economic factors of residents that vary across municipalities such as income, education and unemployment rates.