Essays in public finance: theory and practice of decentralisation

Abstract

Functioning of the government is vital in economic development. This role can be played on cither centralised or decentralised basis. In recent years, due to developments in the European Union, suspicion of powerful centralised system of government, growing disenchantment with the functioning of the public sector, and cultural and ethnic diversity, decentralisation has become an attractive topic among both economists and political scientists. Thus, developed countries, developing nations and economies in transition have introduced decentralised systems of government. Sri Lanka belongs in this group. A discussion of decentralisation of government touches on theoretical work on provision of public goods, Ticbout model, club theory, and the Leviathan model. In respect of the first, Samuelson's discussion on determining of public goods, is widely accepted. However, not all public goods (i.e. judiciary, education etc.) can be explained in terms of Samuelson's criteria. The Leviathan model explains the relationship between the degree of decentralisation and size of govenmient. However, it discusses a situation after decentralisation and is not useful in discussing situations before decentralisation. The Tiebout model can be explained in terms of club theory. Consequently, club theory was selected to analyse the provision of public goods in this study. In the theoretical part of the study, first, Samuelson's case is reviewed and modified. Secondly, after reviewing the literature on club theory, that theory is presented with some modifications and new applications. These modifications of the club theory, include how club theory is used to analyse both 'representation of individuals' in their collective behaviour and 'social groups'. When social groups incur a new kind of transaction cost, namely security expenditure, a new formulation can be made with respect to social groups. In addition, the way club size is affected as a result of increase in consumer's income is presented in a new manner. Club theory has not been sufficiently improved to embrace the supply-side efficiency of the club. To fill this gap, the theory is applied to explain centralisation within decentralisation. In provision of some public goods, when consumers have to bear a transportation cost to reach the service centre, club theory violates its assumption that clubs are serving flying consumers. On such occasions, the way of establishing public service outlets is mathematically worked out. Not al! economists accept decentralisation as desirable. They argue that decentralised systems arc not more efficient than centralised ones. However, empirical studies on this issue are scarce and the hypothesis cannot be tested against evidence. In order to fill this gap partially, an empirical study was conducted with respect to the Provincial Council system introduced in 1987 in Sri Lanka. The empirical study consisted of two components, namely overall economic features and performances and school education. The study of the first component, consisted of descriptive analysis for the post decentralised period in light of quantitative data. The second component of the empirical analysis assessed the efficiency of school education under Provincial Councils, since education is a widespread public good of the country. Constraints on both theoretical and empirical analyses arc very strong. On the issues in both areas, there is a dearth of reading material in Sri Lanka. Further, decentralisation remains little studied by economists. . The conclusions reached in the study are the following. Samuelson's criteria for the determination of public goods have been improved and modified in this study. Club theory has been used in the analysis of decentralisation. It was used to explain 'representation by individuals', and social groups. The theory can also be used to explain a reason for decentralisation, since there is a negative relationship between consumer incomes and club size. Mathematical presentation of the distribution of service outlets demonstrated that a decentralised system is more efficient. There is no uniformly accepted definition of decentralisation among economists. Definitions vary from country to county and from person to person. Of all the functions of government, allocation, distribution and some elements of the constitution (judiciary) can be decentralised while others cannot. Even though decentralisation was introduced by the Provincial Councils Act 42 of 1987, the degree of decentralisation is very limited and the country still remains as a unitary state. However, there is high degree of decentralisation in education. Over 50 percent of its work was reported to have been decentralised. After decentralisation, the output of school education has quantitatively increased at provincial level, but efficiency has not. Even if the overall economy has grown considerably after the introduction of Provincial Councils, regional economies have not grown satisfactonly. Therefore, Provincial Councils have failed to ensure same balanced regional growth. There is no evidence to assess that the degree of decentralisation has made positive contributions on economic growth. Provincial Councils have some decision-making powers m respect of generating ahemative income sources and expenditure activities. Yet there are no electoral cycles to be seen in this country. In the proposed package for the devolution of power the country is considered to be a union of regions, where legislative, judicial and executive powers of government, will be devolved increasing the degree of decentralisation. After implementation of these proposals, further studies will be required to appraise the relationship between decentralisation and economic growth.