The Impacts of Oil Rents on Fiscal Balance

Data used in the regression model are described in this section. The dependent variable is the fiscal balance measured in constant dollars, from the International Monetary Fund (Fiscal Monitor) 2016. A set of control variables are also included to alleviate the effect of omitted variables bias; these include Debt to GDP ratio, Real GDP per Capita, Government Expenditure (Size), Interest rate, unemployment rate, inflation rate, and control of corruption index. Real GDP and Debt to GDP are specified in Natural logs. Data of 20 selected oil dependent economies, and 61 net oil exporting economies (all listed in the appendix), as classified by the World Bank; Kuwait, Libya, Saudi Arabia, Iraq, Angola, Oman, Azerbaijan, Venezuela, Chad, Brunei, Kazakhstan, Iran, UAE, Ecuador, Algeria, Nigeria, Russia, Qatar, Norway and Mexico within 2000 – 2015, and 1991 – 2015 respectively. All have annual data for the period under consideration, and all the variables are taken from International Monetary Fund, except Real GDP per capita constant and Oil rents, which are taken from WDI dataset 2015.

本节将对回归模型所使用的数据进行说明。被解释变量为以不变价美元计价的财政收支余额，数据取自2016年国际货币基金组织（International Monetary Fund, IMF）发布的《财政监测报告》（Fiscal Monitor）。为缓解遗漏变量偏误对回归结果的干扰，本文纳入一系列控制变量，具体包括债务与GDP之比、人均实际国内生产总值、政府支出（规模）、利率、失业率、通货膨胀率以及腐败控制指数。实际GDP与债务与GDP之比均进行了自然对数变换。样本涵盖世界银行（World Bank）分类的20个石油依赖型经济体与61个石油净出口经济体（所有经济体名单详见附录），二者的样本期分别为2000–2015年与1991–2015年；具体样本经济体包括科威特、利比亚、沙特阿拉伯、伊拉克、安哥拉、阿曼、阿塞拜疆、委内瑞拉、乍得、文莱、哈萨克斯坦、伊朗、阿联酋、厄瓜多尔、阿尔及利亚、尼日利亚、俄罗斯、卡塔尔、挪威与墨西哥。所有经济体均提供研究期内的年度数据；除不变价人均实际GDP与石油租金数据取自2015年世界发展指标（World Development Indicators, WDI）数据集外，其余变量均来源于国际货币基金组织。