An Empirical Analysis of Energy Demand in Turkey: Estimating Price and Income Elasticities of Crude Oil With Different Data Frequencies

Abstract

The present study aims to examine crude oil import demand in Turkey. The country has quite limited oil production and is dependent on foreign supply to power its growing economy. We estimate the price and income elasticities of imported crude oil for Turkey exploiting three different data sets (i.e., monthly, quarterly, and annual) to test whether data frequency matters in estimation. We employ an Auto Regressive Distributed Lag (ARDL) model which shows both short-run and long-run effects of price and income changes on crude oil import demand. Results show that i) data frequency plays an important role in the estimation process as the findings of monthly and quarterly model show a long-run equilibrium while annual models do not and that ii) the estimated elasticities of monthly and quarterly data are in line with microeconomic theory and literature in that price elasticities are inelastic suggesting that consumers are less responsive to the price changes due to high dependency and lack of substitute for crude oil and that income elasticities are elastic suggesting the demand for crude oil increases as income level increases.