Dynamic-opportunistic behaviour in local government contracting-out decisions during the electoral cycle

ABSTRACT

A major question for public managers is whether municipal services should be rendered in-house or contracted out. In view of the negative perceptions often aroused by contracting out, this political decision might be framed within a theoretical model that we term ‘dynamic-opportunistic behaviour’. According to this model, the probability of municipal services being contracted out is greater in the years immediately following elections; moreover, during this period the decision is taken more quickly. In this theoretical model, not all factors (budgetary, economic, political, service characteristics and socio-economic) have an equal impact on the contracting-out decision during each year of the electoral cycle. The model was applied to a sample of 2,274 Spanish municipalities, with respect to a broad time horizon (2002–2014), and the results obtained confirm our hypotheses regarding dynamic-opportunistic behaviour in the contracting out of local public services.

KEYWORDS:

• Opportunistic theory
• dynamic analysis
• local public services
• contracting out
• political cycles

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

1. In the Spanish regulatory system, mayors are paramount in the decision-making process, and the figure of professional manager found in countries like the USA does not exist. Therefore, in the present study, this term is used, at all times, in reference to the mayor.

2. However, this form of municipal management differs from that found in other countries such as the USA, where decisions are taken by a public manager who is employed by the local entity for this purpose.

3. We used variables obtained from budgetary and balance sheet information. However, although local authorities are obliged to provide budgetary information to the Directorate General for Financial Coordination with Regional and Local Authorities (DGCFCAEL, part of the Ministry of Finance and Public Administration), a number of local authorities do not comply with this requirement, and this further limits our sample size. Furthermore, in line with Bastida, Benito, and Guillamón (2009) and Prior et al. (2019) our analysis of the municipal situation in Spain is limited to municipalities with more than 1,000 inhabitants. As the latter authors point out, the reliability of financial data is doubtful for small municipalities and, moreover, the data for certain variables were only available for municipalities with more than 1,000 inhabitants.

4. The sample was reduced to 2,274 municipalities for two main reasons: first, a filtering process was carried out, in which municipalities of fewer than 1,000 inhabitants were excluded; second, municipalities for which financial information was not available for one or more of the years of the study period (2002–2014) were also excluded.

5. In discrete time models, the discrete time risk function is the probability of transition in a discrete time tj, j = 1,2,…, given survival until the time tj (Máñez, Rochina and Sanchís 2008; López-Hernández et al. 2017). In the present study, this method represents the probability of a municipality contracting a private company to provide a public service over a period of time t during the electoral cycle, when the company has not provided such a service prior to the period in question. Taking into account these considerations, and the data obtained, we implemented clog-log models to construct the hazard rates of the municipalities examined and to relate them to a set of time-varying covariates (Prentice and Gloeckler 1978) for each of the models representing the electoral cycle.

6. A hazard ratio is often reported as ‘the increase/reduction in the risk’ (in the present case, that of contracting out). For example, if the hazard rate is 0.9, this represents a 10% decrease in the risk (the reduction is calculated as 1 minus the hazard rate). If the hazard rate is 1.1, this represents a 10% increase in the risk (the increase is calculated as the hazard rate minus 1).

Additional information

Funding

The authors are grateful for the financial support received from Ministerio de Educación y Ciencia (Spain; ECO2016-76578-R) and Beca del Ministerio de Educación, Cultura y Deporte Formación Profesorado Universitario (FPU14/01403).

Notes on contributors

Emilio J. de la Higuera-Molina

Emilio J. de La Higuera-Molina graduated in finance and accounting, has a master’s degree in management and public management and is a Ph.D. student at the University of Granada.

Ana M. Plata-Díaz

Ana M. Plata-Díaz is an assistant professor at the University of Granada. She has published in various journals.

Antonio M. López-Hernández

Antonio M. López-Hernández is a professor at the University of Granada and President of the Chamber of Accounts of the Andalusian Government. He has published recently in several journals.

José L. Zafra-Gómez

José L. Zafra-Gómez is an associate professor in accounting at the University of Granada, Spain. His research interests focus on local government. He has recently published articles in journals including the Journal of Public Administration Research and Theory, Public Administration, Regional Studies, Urban Studies, the European Journal of Operational Research and OMEGA.