The Impact of Political Factors on Local Government Decentralisation

ABSTRACT

The process of local decentralisation of public services delivery has attracted research attention because of the wide range of factors that induce governments to make this decision. Nonetheless, most such studies have focused on economic and financial aspects, ignoring the impact of political factors. The scarce previous research about political factors is the motivation for this study. We used panel data models based on 153 Spanish municipalities with populations over 50,000 for the period 1999–2007. The process of decentralisation is disaggregated according to the legal form of the entities: corporations and foundations. Our results show that more corporatisation processes are carried out in environments where political competition is lower and the party in power enjoyed greater support in the last elections. Additionally, decentralisation processes usually take place in the years immediately following elections, especially through foundations. Furthermore, both forms of decentralisation are more likely to be initiated by right-wing politicians.

Notes

*, **, and *** indicate significance at the 1%, 5%, and 10% level, respectively.

*, **, and *** indicate significance at the 1%, 5%, and 10% level, respectively.

*, **, and *** indicate significance at the 1%, 5%, and 10% level, respectively.

These three Anglo-Saxon countries, pioneers in NPM techniques, adhere to the “Whitehall Model” of government.

Although the precise nature of NPM has been extensively discussed (Dunleavy and Hood 1994), Hood (1991) aggregates its practices into seven doctrines that summarise the guiding principles of NPM: “hands-on and entrepreneurial management,” “explicit standards and measures of performance,” “output controls,” “disaggregation and decentralisation,” “competition in the provision of public services,” “stress on private sector styles of management,” and “discipline and parsimony in resource allocation.”

Metcalfe and Richards (1989) point out that this new model places more emphasis on responsibility than on control. To do so they distinguish between internal and external responsibility. Control mechanisms are related to responsibility from an external point of view, ensuring the legitimacy of public authorities, whereas responsibility viewed from an internal perspective refers to authority over the use of resources and getting results.

These practices would benefit citizens’ choice concerning the provision of public services, since they would be able to find the desirable combination between services and taxes (Tiebout 1956).

For instance, Denmark relies on a top-down approach; that is, the support service personnel are transferred to shared services centers. In addition, ministerial budgets are decreased for the corresponding amount of these resources. Finland and the Netherlands establish temporary cut-back targets for support services. Sweden and the United Kingdom support a temporary or permanent productivity cut, which is not specifically for support services (OECD 2010).

For more information about the privatization reforms in the UK, see Parker (2003; 2009; 2012), Yergin and Stanislaw (1998), Millward (2000), Heald (1989), and Hulsink (1999).

Although this process was not as significant as the one that took place in the UK, it represents a great effort, taking into account that it was carried out by a socialist party that ideologically defended public enterprise (Gamir 2005). However, due to the negative financial situation of public agencies in certain sectors (the automotive industry, steel industry, shipbuilding industry, etc.), it was forced to address privatisation, although in political terms there was talk of “disinvestment” or “alienation.”

Support services include human resources, information resources, ICT, organisation, accommodation and facilities, communications procedures, and finance and auditing procedures.

This is the first Coordinating for Cohesion in the Public Sector of the Future (COCOPS) working paper to present a state of the art of the literature assessing the impact of NPM reforms. Many of the ideas set out in sections 1 and 2 of this article have been obtained from this first working paper. More information about COCOPS is available in its web: http://www.cocops.eu/

Local Government Regulatory Law sets out a series of minimum services to be provided by local authorities, including public lighting, refuse collection, street cleaning, water provision, sewerage, access to population centers, road maintenance, food and beverage control, and cemetery management. In addition, municipalities with more than 5,000 inhabitants must provide at least one public park, a library, a market and waste-processing facilities; those with more than 20,000 inhabitants must provide civil protection services, social services, fire prevention and fire fighting services, and sports facilities; finally, those with more than 50,000 inhabitants must provide, in addition, collective transport and environmental protection services.

Progressive parties support social programmes (social benefits, health care, etc.), while conservative ones devote funds to other expenditures such as prisons, the police force, and the army.

Statistical descriptors for the “ELECTION YEARS” variable are meaningless due to the form in which the variable is estimated.

In particular, all of these models were estimated through fixed and random effects, checking the validity of both models using the Hausman test. The fixed effects model allows us to avoid the information related to latent fixed effects, which are highly correlated with the variables included in the model, and provides us with consistent estimators. The random effects model is more appropriate when there is no correlation between the fixed effects and the model variables, and it enables us to obtain more efficient coefficients. The Hausman test examines the null hypothesis of absence of correlation between the individual effects and the independent variables; when not rejected, the higher degree of efficiency in the estimation leads to the use of the random effects model.

An important assumption is required for this test: no existence of heterokedasticity. Thus, when the homoskedasticity assumption fails, a robust version of this test is required. There is no implementation of this in Stata, and so we used the method proposed by Wooldridge (2002).

Additional information

Notes on contributors

Beatriz Cuadrado Ballesteros

Beatriz Cuadrado Ballesteros ([email protected]) is an Assistant Professor of Accounting at the University of Salamanca, where she earned her PhD in business. Her current research is focused on the public sector, specifically on public services delivery and causes and consequences of decentralisation processes.

Isabel M. García Sánchez

Isabel M. García Sánchez ([email protected]) is a Lecturer of Accounting at the University of Salamanca, where she earned her PhD in business. Her research is focused on the activities of the decentralisation processes in the public sector and disclosure of information by private and public agencies.

José M. Prado Lorenzo

José M. Prado Lorenzo ([email protected]) is a Professor of Accounting at the University of Salamanca, where he earned his PhD in business. His current research examines the activities of public administrations, especially the decentralisation processes and disclosure of information by public and private agencies.