This dissertation consists of three essays on applied econometrics. An abstract for
each of the three essays follows.
Essay 1 (Chapter 1) investigates the demand for gasoline in Canada using annual household survey data from 1997 to 2009 and three of the widely used flexible functional forms. I find that the demand for gasoline is inelastic in Canada. Given that none of the studies in the existing literature checks if their estimates are consistent with economic theory, this essay addresses the importance of the theoretical consistency in the empirical work of demand analysis and is the first study in the literature that provides the elasticity estimates for the demand for gasoline that is consistent with the theory.
Essay 2 (Chapter 2) examines the effect of oil price uncertainty on gasoline prices
and investigates the relationship between gasoline and crude oil prices. The empirical
model is based on a structural vector autoregression model that is modified to accommodate multivariate GARCH-in-Mean errors. I use monthly data for the United
States from January 1976 to January 2013. I find that oil price uncertainty has positive effects on gasoline price changes and that there is an asymmetric relationship between crude oil and gasoline prices. My results are robust to alternative measures of oil prices and alternative model specifications.
Essay 3 (Chapter 3) investigates the demand for money and the degree of substitutability among monetary assets in the United States from 1967 to 2013 using four commonly used flexible functional forms, on three monetary subaggregates: M1, savings deposits, and time deposits. As in Essay 1, I pay explicit attention to economic regularity in order to produce inference consistent with neoclassical microeconomic theory. Moreover, I incorporate financial econometrics in the empirical flexible functional forms approach to money demand analysis by using a BEKK GARCH(1,1) specification to model the covariance matrix of the errors in the demand equations. I find significant evidence of ARCH effects in the estimations under the homoscedasticity assumption and that the demand for money is inelastic and that the three monetary assets taken as a whole are Morishima substitutes.