The Composition of Government Spending and Economic Growth in Developing Countries: The Case of Latin America

Abstract

This paper aims to identify the effects of the two economic components of government spending, namely, capital and current spending, on the per capita economic growth rate in a set of Latin American countries over the period 1975-2000. Within the neoclassical framework (Solow, 1956; Swan, 1956), government spending, and public policy in general, has no role in determining the long-run economic growth rate, since this is determined by the exogenous population growth and technological progress rates. On the other hand, in some endogenous growth models developed mainly since the early 1990s, such as Easterly (1990), Barro (1990), Barro and Sala-i-Martin (1992, and 2004), Cashin (1995), Bajo-Rubio (2001), and Milbourne et al. (2003), fiscal policy affects the long-term growth rate through decisions on either taxes or expenditures.

The empirical literature tends to reject the predictions of the neoclassical model, in the sense that according to this model, fiscal policy cannot affect growth in the long term. However, results are far from conclusive and it seems they depend on various aspects such as methods or techniques used, assumptions, country or set of countries analyzed, and so on. As long as theoretical models about the influence of public spending on growth is concerned, some of them such as Barro (1990), Cashin (1995), Bajo-Rubio (2000), and Milbourne et al. (2003) predict that a positive effect is expected to be found in countries where the size of government is smaller than a certain threshold, and a negative one in countries where the size of government is bigger than that. Therefore, since generally speaking, with few exceptions, one finds very large public sectors only in developed countries (DCs), studies evaluating the impact of public expenditure on growth should analyze DCs and less developed countries (LDCs) separately. In line with recent growth literature, the study uses a generalized method of moments as suggested by Arellano and Bond (1991) in order to obtain consistent and efficient estimates for a dynamic model, such as an economic growth model.

This paper’s findings suggest that neither government capital nor current expenditures have any impact on the per capita economic growth rate. The positive effect of government capital spending reported in some literature was not found here. Statistically insignificant estimated effects of these kinds of spending could be due to inefficiency. Perhaps they are vulnerable to rent seeking. In addition, inefficiency of government spending has widely been associated in the literature with poor governance and corruption, which are, typically, some characteristics of developing countries.

Keywords: Economic growth, generalized method of moments, government spending, Latin America