Mao, Qi (2017) The elasticity of substitution for US energy price changes between 1947 and 2010. (PhD thesis), Kingston University, .

Abstract

Since energy price changes have been studies by much literature, this thesis tries to discuss it through the elasticity of substitution. More importantly, this thesis finds that the substitution effect itself cannot completely interpret the phenomenon of energy price changes. This is based on the results of the estimation of the AES, the MES and the CES, as well as some previous studies' results of negative substitution elasticities. This thesis adopts VECM, ADL and panel data as methodologies. Different data is also colelcted and analyzed. Most of the AES estimates are negative and the CES estimation shows many negative elasticities as well. The estimated elasticities from the MES and the Panel dta CES are positive. Out positive elasticities from the MES may are consistent with the previous literature that finds the MES estimation method is better than AES. The negative elasticities of susbtitution as the substitution effects are usually followed by income effects. Based on literature view, if income effects are involved in the energy price changes between energy exporting and importing countries, it may lead to new policy making and application. In addition, there are some other findings: (1) an application approach is used to test the cointegration when variables include I(0), I(1) and share variables. This approach is different from Pesaran, Shin and Smith's (2001) ARDL method which is involved I(0) and I(1) variables. However, in the application, the data being used covers share variables. Share variables sum to unity so they have collinearity. New procedures (ADL) are adopted to solve the problem of collinearity in econometric application. This then allows to analyze I(0), I(1) and share variables together in the model. This is one of our contributions in econometric VECM application. (2) By the literature review in Chapter 4, it's found that the elasticities of substitution vary from economies. Since some economies do not share free flow of capital, labor or energy, their substitution elasticities cannot be estimated as a whole. The different sustitution elasticities in each country show that, when selecting the data, we may refer to a region or a country with a common free flow market for capital, labor and energy. (3) Different data is used in the application. Previous literature uses the data of the UK or other countries to analyze the Allen's elasticity of substitution (AES), or the US data in the period different from that in this study, or uses the familiar data but in technology forecasting. Importantly, the empirical analysis uses the US data to test the Allen's elasticity of substitution (AES) and the constant elasticity of substitution (CES), in order to explore negative estimates of the substitution elasticity.

Actions (Repository Editors)

Item Control Page