Privatization reduces the role of government by increasing the role of other organizations, such as non-profit organizations and private firms. As Savas puts it, privatization relies more on the private institutions of society and less on government to satisfy people’s needs (Savas, Emanuel S. 2000). The trend of using privatization has mushroomed all around the world since 1980s, from developed countries like the U.S., Japan, Great Britain, and France to less developed countries like China, Sri Lanka and Turkey. The practice has made governments more “commercial-enterprises-like” and encouraged the development of competitive market economies within procurement systems (Moe, 1987).

As the 2012 APPAM fall conference theme states, we are living in an age of scarcity; and public sectors face a tremendous pressure to “do more with less” and “to balance budgets in a setting with declining revenues”.  Although numerous scholars and practitioners believe that privatization could be a potential remedy for public sector problems, there still lacks empirical consensus in literature. The majority of the studies is based on case studies and on site investigations, which make the empirical results less generalizable or applicable to municipalities across the country. Empirical studies that assess national data are still considered scant; it is even rare to see privatization studies that incorporate diffusion mechanisms. It is the aim of this paper to fill part of this gap; we propose to investigate what are determinants of privatization and how privatization policies diffuse over time.

We analyze a panel data set for more 1200 American cities and countries between 1997 and 2007 using spatial econometric techniques. The privatization data are purchased from International City/County Management Association  (ICMA). ICMA conducts Alternative Service Delivery survey every 5 years at city and county level. The survey covers service delivery choices for the following areas: public works/transportation, public utilities, public safety, health and human services, parks and recreational activities, cultural arts, and support services. The survey data obtained are for the year 1997, 2002-2003 and 2007. In addition, city and county financial and demographic data are obtained from Census. A panel dataset is formed based on those years’ data.

The study explores the key determinants of adopting privatization and discovers whether there are any spatial interdependencies for privatization policies. The dependent variable is a privatization activity index that is created based on the survey response for the different service areas mentioned above. The explanatory variables are city/county financial and demographic data; namely population, debt, and revenue as well as other control variables. The inverse distance matrix will be also included as another regressor to explore the spatial pattern.

The study will contribute to the literature in two ways. First, the study examines privatization practices at the city/county level comprehensively. The dataset contains more than 1200 cities and counties. Second, the study incorporates spatial econometric techniques to assess how privatization practices diffuse over time. These features are rarely seen in privatization literature.