Modeling Gasoline Demand
Modeling Gasoline Demand in the United States
Economics 375
DePaul University, Chicago, IL 60601
June 13, 2012
Abstract
This paper is an econometric approach to the estimation of price and income elasticities of gasoline demand in the United States from a translog model, and is based off of the most recent data available for use. This approach allows for variables to interact in a flexible yet instrumental way, providing for significant evidence that gasoline demand elasticities are construed by the income effect and price effect. In the linear model the variables are more easily defined and effects are more clearly understood but with a lesser accuracy due to the effects of random variables and skewed data. With the translog model, heterogeneity can be accounted for in the income and price elasticities and the different effects of price and income can be more accurately defined because the variables are functionally flexible. The results conclude that gasoline demand is significantly effected by both price and income elasticities, although to what degree is determined by what variables and instruments are set forth in play.
1. Introduction
Many factors can have an effect on gasoline demand – public transportation, fuel efficiency, market for automobiles, and something even as simple as cultural and regional differences (lifestyle). Governmental regulations and tax on gasoline have significant effects on the gasoline market but the most direct and measureable effects are price and income effect due to the wide variations across states, time, and households, and the availability of data for use.
With the constructed translog model, heterogeneity can be accounted for in the income and price elasticities and the different effects of price and income can be more easily defined because the variables are functionally flexible. In this model, the variables of income and price can interact with one another in the same function to determine
Economics 375
DePaul University, Chicago, IL 60601
June 13, 2012
Abstract
This paper is an econometric approach to the estimation of price and income elasticities of gasoline demand in the United States from a translog model, and is based off of the most recent data available for use. This approach allows for variables to interact in a flexible yet instrumental way, providing for significant evidence that gasoline demand elasticities are construed by the income effect and price effect. In the linear model the variables are more easily defined and effects are more clearly understood but with a lesser accuracy due to the effects of random variables and skewed data. With the translog model, heterogeneity can be accounted for in the income and price elasticities and the different effects of price and income can be more accurately defined because the variables are functionally flexible. The results conclude that gasoline demand is significantly effected by both price and income elasticities, although to what degree is determined by what variables and instruments are set forth in play.
1. Introduction
Many factors can have an effect on gasoline demand – public transportation, fuel efficiency, market for automobiles, and something even as simple as cultural and regional differences (lifestyle). Governmental regulations and tax on gasoline have significant effects on the gasoline market but the most direct and measureable effects are price and income effect due to the wide variations across states, time, and households, and the availability of data for use.
With the constructed translog model, heterogeneity can be accounted for in the income and price elasticities and the different effects of price and income can be more easily defined because the variables are functionally flexible. In this model, the variables of income and price can interact with one another in the same function to determine
Bibliography: Liu, Weiwei. Modelling Gasoline Demand in the United States: A Flexible Semiparametric Approach. State University of New York at Binghamton, 4 Aug. 2011. Web. 13 June 2012.