Oil Revenues, Government Spending Policy, and Growth

Article excerpt

ABSTRACT

This study argues that government spending composition determines how oil abundance ultimately impacts growth. Using dynamic panel-data GMM and PMG techniques on a panel of oil exporters, we find that the negative growth effect of oil price volatility is channeled through fiscal policy. In particular, revenue windfalls may impede growth through at least three channels: (i) weakening the domestic tax base, (ii) lowering the social return on new public capital, and (iii) intensifying political spending pressures resulting from the accumulation of surpluses. The main policy implication for oil-exporting countries is that it is imperative to use strict fiscal rules, backed by the appropriate political incentives, to insulate public spending from oil cycles. Investing the surplus (in sovereign wealth funds) or retiring public debt amid oil windfalls would alleviate competitive rent-seeking pressures and enhance the social gains from revenue booms.

(ProQuest: ... denotes formulae omitted.)

1. INTRODUCTION

Since the 1970s, oil price cycles have been highly unpredictable. Fiscal policy in oil-exporting economies plays a significant role in managing the highly volatile and uncertain oil revenues. A notable characteristic of public finance in these countries is the strong pro-cyclicality of government expendi-tures and the non-oil operational balance in relation to oil price fluctuations (Villafuerte et al., 2010; El Anshasy & Bradley, 2011). Government spending usually acts as a key transmission mechanism of oil price shocks to the macro-economy (Husain et al, 2008; Pieschacon, 2009).

Meanwhile, it is well documented that oil exporters have experienced, on average, poor growth performance in the past few decades, a phenomenon often referred to as the resource curse. 2 Since the seminal work of Sachs and Warner (1995), a growing body of literature has emerged to identify some channels through which this growth deficit may occur.3 This paper raises the following question: Would the peculiar nature of fiscal policy in oil-exporting countries contribute to their growth deficit?

In particular, we study the growth effect of different types of government spending financed through fluctuating revenues while controlling for the effect of oil price fluctuations in a panel of oil exporters. The study has two underly-ing premises that have been neglected in the context of oil-exporting countries. First, public-spending policy decisions are inseparable from financing deci-sions. Hence, emphasizing the growth effect of one side of the budget can be misleading since it fails to account for the fact that every increase in spending has to be financed by raising revenue or accumulating more debt, and vice versa (Bleaney et al, 2001). Second, the composition of fiscal adjustments matters to long-run growth (Alesina & Perotti, 1996). The conventional prac-tice of using aggregate spending measures misses important effects of allocat-ing public funds among different uses. This can be of greater concern, espe-cially in the context of developing resource-rich economies.

In this study, the growth effect of fiscal policy materialize not only through government size, but also through two more channels: (i) the compo-sition of public expenditures, which manifests the political elite's choice to allocate a significant proportion of the resource rent among different public uses; (ii) the method of financing expenditures, which reflects the extent to which the government is fiscally dependent on the natural resource. The main argument we advance is that fiscal policy choices may turn oil windfalls into a curse in some countries and a blessing in others. By the same token, govern-ment spending policy can cushion the economy amid negative oil shocks, thereby restoring growth.

We first use a dynamic panel-data system GMM technique, utilizing the pooled dimension of the data. Then we check the results from panel cointegra-tion techniques and estimate an error correction model applying a PMG esti-mator that accounts for panel heterogeneity and corrects for potential cross-sectional dependence. …