Abstract

Crude oil demand function has been estimated using Panel Data Method in the present paper through utilizing the annual data of a period including the years 1980-2007. Mexico, Argentina, Brazil, Iran, Saudi Arabia, China, India, Indonesia, Taiwan, South Africa, Singapore, and Thailand have been surveyed as the developing countries’ representatives throughout the research. For model estimation using the Hausman test, the fixed effects model has been selected for this function and the explanatory variables of the function include crude oil price, economical growth of the selected countries and the lagged demand of the crude oil. The stationary testing of the model variables has been the IPS test¹, indicating all the variables in non-stationary level. Therefore the Pedroni cointegration test was used to ensure non-existence of the spurious regression and investigating the long term relationship among the variables. The results of the test showed a long term relationship among the model variables. Also the results of the model estimation shows that the crude oil demand in each of the 12 industrial countries has been inelastic from the price and income point of view, but the income elasticity of the crude oil has been by far greater than its price elasticity. Hence, it is observable that the economic growth changes in positive and negative sides are much more influential than the similar changes in crude oil price. On the other hand, because of the low price elasticity of the crude oil demand in the selected group of countries, such variables are unable to have substantial effects on the crude oil demand in short term.

Keywords: oil demand function, pool data, stationary and cointegration tests