Political competition impacts on government expenditure growth: An evidence from OECD countries.

• Author: Ozdemir, Ali Riza
• Position: Organisation for Economic Co-operation and Development - Report

1. Introduction

   The phenomenal growth of the size and scope of government expenditure in all the OECD countries during the latter half of the 20th century has become a well-established trend. On average, the size of government expenditure as a share of GDP in these countries increased from around 25 percent in 1960s to 50 percent in 1970s, reaching 60 percent in 1980s (GFS, 2000). There has also been tremendous cross-country variation among OECD countries with respect to this trend. Government expenditures in some OECD countries have grown much faster than others. For instance, over the 1960-1997 period, the average annual growth rate in real per capita central government total expenditures was around 5 percent for countries such as Denmark, Italy, and Norway, while around 3 percent for countries such as Canada, Australia, and the United Kingdom. (1) Overall, the rate of growth of government expenditures was faster in the 1960s and 1980s, fastest in 1970s, and slowest in 1990s. For instance, the average annual growth rate of real government expenditures in Italy was 5.5 percent in 1960s and 1980s, 10 percent in 1980s, but it was -2.5 percent for the period 1991-97 (GFS, 2000).

   Both high and varying government expenditure growth rates in OECD countries have been associated with such serious consequences such as high unemployment rates, large budget deficits and debt burden, and low economic growth (2) which are in turn linked to the many significant social, economic, and political problems. (3) In the current literature, many demand and supply oriented social and economic influences on government size have been tested extensively; (4) however, the role of political factors, such as political institutions, electoral rules, regime types, and the degree of competition among political parties are still open for deeper investigation. Among these, political competition draws particular attention as governing parties take rather expansionary actions which may result in budget deficit and larger public debt over the long lasting recessionary period starting 2008. There are basically two theories designed to explain the relationship between political competition and public expenditure growth. In the neoclassical competition theory, it is argued that competition among political parties to attract more votes would lead to the implementation of more efficient policies and a slower rate of government expenditure growth (Stigler, 1972 and Wittman, 1989 and 1995). (5) Under the public choice theory, however, a greater competition would lead to new spending programs and hence a higher rate of government expenditure growth (Buchanan and Wagner, 1977 and Peltzman, 1980). It then seems that how the rate of government expenditure growth across countries is influenced by political competition has more than one answer in the literature.

   The existing empirical literature that also seems to be supporting both of the hypotheses about the relationship between political competition and the rate of public sector growth is still unsettled and far from convincing. For instance, some studies find a significant positive relationship, while others suggest no relationship, or a negative relationship. The reasons for these controversies in the results are multifold. Firstly, the results that are produced in these studies using varies methods of econometrics. Some techniques are just simple correlations or regressions with no consideration of time series or cross section related problems, while some others do not control for multiple sources of influence. Secondly, especially in political science literature, what happens to wasteful size of government budget or revenue-expenditure package may not be viewed that importantly. They do not disentangle the effect of interactive factors such as economic and institutional variables.

   This article examines empirically which theory provides a better explanation by testing the relationship between political competition and government expenditure growth on data from OECD countries for the second half of 20th century. Using recent econometric techniques and panel data, this article demonstrates that there is a negative relationship between political competition and the rate of public sector growth. The relationship found to be robust to the inclusion of other critical explanatory variables such as per capita income, relative price ratio, coalitions, change of governing party and ideology. To the best of my knowledge, none of the existing studies used the relative price ratio as one major determinant and control for government growth. The second unique feature of this article is to control for the effect of change in government and change of ideology in government. The evidence presented here supports the notion that the integration of competitive parliamentary structure will be held accountable for allocating resources efficiently and suppressing the excessive growth of the public sector. The organization of this article is as follows: Section 2 reviews the results from previous studies. Section 3 describes the methodology, variables used, and data sources. Section 4 presents the empirical results, and Section 5 provides conclusions and policy implications.

2. Literature Review

   Political competition can be defined as competition among political parties in a parliament or/and senate to gain more electoral support in the next election (Downs, 1957). Why should residents care whether there is market-like competition within parliament, even if parliamentarians were originally chosen through an election process? The literature offers two lines of reasons: one is related to the deficiencies of monopoly (or advantages of competition) such as excessive prices and rent extraction, and restrained supply levels (or minimized prices and rent extraction), and the other is related to the principal-agent nature of political outcomes such as limited information and bureaucratic behaviour. Citizens care about competition because, in the absence of externalities or other market distortions, unrestricted competition allows resources to flow to their most productive uses, and maximizes the net welfare of the society. In an early article, Tullock (1965) extends this reasoning to the political process by viewing governments as a natural monopoly. Leaving the governing party without opposition means that voters face a despotic state. Democracy, with a competitive political party process, reduces this problem. By eliminating entry barriers into the political system, a competitive multiparty political decision-making process is created. In competing for votes, the governing party is induced to provide more public goods and services at any given cost and will be less able to extract monopoly rents from taxpayers. Stigler (1972) also argues that competition will force parties to satisfy the preferences of voters. In other words, legislation enacted by a more competitive system of government, with the agreements of different groups, may increase welfare by ensuring that a change in taxes and expenditures generates more gain than loss. This line of reasoning supports the analogy between traditional economic competition and political competition in a classical modern state. This analogous effect implies that increased competition in politics would most likely enhance efficiency of public good provision, minimize rent extraction and thus reduces excessive government expenditure growth (Stigler, 1972, Becker 1983 and Wittman 1989). (6)

   A second reason for the concern about political competition is its relation to bureaucratic behaviour. Following Niskanen (1971) and subsequent Leviathan models, the assumption that government officials can benefit personally from the budget that they control means that bureaucrats will pursue activities that increase the size of budgets under their control. For example, bureaucrats in a monopoly government, without competition, may offer two choices: a high level of spending or a very low level of spending. If the very low spending entails a considerable underprovision of public goods and services, voters may well accept the high level of spending. While some public resources are allocated to productive services benefiting voters, another portion goes to unproductive services (rents) benefiting public officials and to redistributive transfers benefiting a more narrow group of voters. In a competitive political environment though, the flow of resources to unproductive services such as excessive rent seeking is minimized, potentially increasing the efficiency of resource allocation and society's welfare, and thus reducing the excessive government expenditure growth (Wittman, 1995 and Besley et al, 2005).

   According to some political economists, however, political competition in relation with some other factors such as limited information and representative nature of political system may have increasing effect on government spending. Pommerehne and Schneider (1982) indicates that there is greater public spending in representative democracies compared to the direct ones due to weaker control of politicians (as agents) by voters (as principles) especially due to information asymmetry. Even though public spending is greater in representative democracies, as Mueller and Murrel (1986) argue, public goods and services are under-provided because some of this spending goes to powerful groups of voters rather than goods and services. This type of behaviour may be not merely opportunistic, due to information asymmetry, but politically rational. Politicians in real politics, may choose not to satisfy the preferences of every single voter instead, they please a subset of voters that is necessary to win the election. As a result, government growth is encouraged if parties compete to satisfy different voter subsets (Persson and Tabellini 1998). It is also argued that politicians may try attract new voters by...