برآورد عرضه و تقاضا ی بنزین با استفاده از داده مالیاتی

We estimate supply and demand functions for the U.S. gasoline market using information from excise tax returns provided by the IRS for the period 1990–2009. We find price and income elasticities of demand similar to those found using EIA data. We find a price elasticity of supply of 0.29, which differs from the common assumption of a perfectly inelastic short-run supply curve. By using a novel data source, the analysis provides a robustness check of aggregate studies of gasoline demand and a consistent, econometric estimate of the price elasticity of gasoline supply. The results are useful in guiding tax and regulatory policies regarding gasoline consumption.

Facing large and growing deficits for both the Highway Trust Fund and the overall federal budget, as well as increasing concern over curbing carbon emissions, there is a renewed interest in the taxation of gasoline and other fossil fuels. Indeed, taxes on fossil fuels have not kept pace with either government expenditures on transportation infrastructure or inflation. Table 1 shows federal taxes on gasoline and diesel have been stagnant (in nominal terms) since 1993. Since receipts from the excise tax on gasoline are the largest source of overall excise tax revenues, these taxes represent an important funding source for federal spending on transportation infrastructure. To understand how tax policies affect the behavior of consumers and producers, and impact government tax receipts, one must have estimates of the price elasticity of demand and supply for fuel. These elasticities are also necessary to uncover the incidence of such taxes. Yet, while there has been extensive study of the price elasticity of demand for gasoline (see Goodwin et al., (2004) for a recent survey), much less has been written about the price elasticity of supply. This study fills the gap by providing estimates of the price elasticities of both supply and demand for gasoline. Our analysis is novel in two ways. First, we use data from excise tax returns to provide quarterly observations on U.S. fuel consumption. Second, we are the first to estimate both supply and demand elasticities using the same data source and consistent models. We use data on fuel consumption from 1990 to 2009 to estimate the short-run price elasticities of supply and demand for gasoline. Simultaneous equations models, which correct for the endogeneity of prices and quantities in the market equilibrium, suggest the price elasticity of demand for gasoline to be − 0.07 and the price elasticity of supply for gasoline to be 0.29. These models yield an income elasticity of demand for gasoline of 0.41. The estimates of the price and income elasticities ofdemand fall within the range of estimates found in the vast literature on gasoline demand.2 While we are unable to find model based estimates regarding gasoline supply, an EIA report, cited by CBO, (2003), suggests a long-run price elasticity of supply of 2. We find a short-run elasticity significantly lower than 2. Thus our estimates support the view that the supply is more inelastic in the short-run and provide a precise estimate of this elasticity. A more inelastic short-run supply curve implies that policies such as a gasoline tax holiday would have less of an impact on consumption and that relatively more of the resulting surplus would go to producers. Much has been written on the price elasticity of demand for gasoline. Goodwin et al., (2004) provide a review of 69 published studies since 1990 that estimate price elasticities of demand for different countries, and present price elasticities ranging from 0.00 to − 1.81 depending upon the estimation technique and data used. More recently, Hughes et al., 2008 and Wadud et al., January 2010 estimate demand elasticities with more current gasoline consumption data and innovative techniques. Hughes et al., (2008) control for macroeconomic factors such as inflation, and use supply disruptions in an instrumental variables model to control for the endogeneity of prices and quantities. The paper finds short-run price elasticities ranging from − 0.03 to − 0.34 depending upon the model, estimation technique, and time period. Wadud et al., (2010) control for household characteristics such as urban/rural residence and find a median price elasticity of − 0.47. Less is written on supply elasticities. Austin and Dinan, (2005) and a Congressional Budget Office study CBO, (2003) of fuel taxes and fuel economy standards assume a price elasticity of supply of 2.0. However, this value is from an U.S. Energy Information Administration (EIA) report and not based on any econometric analysis of fuel supply. Often, such as in Davis and Kilian, (2010), it is assumed that supply is perfectly elastic (i.e. the marginal cost of refining fuels is independent of the quantity supplied) in the long-run and perfectly inelastic in the short-run. Others, such as Hsing, (1994) attempt to make inferences about the effects of tax policy on gasoline consumption without specifying a supply function. Providing a precise estimate of the elasticity of supply is an important contribution of the paper. The remainder of this paper is organized as follows. Section 2 describes the data. Section 3 outlines the econometric model and presents the main results. Section 4 discusses the results and the important strengths and limitations of the study. Section 5 concludes.

We estimate the supply and demand for gasoline using data provided by the IRS. We find demand elasticities (price and income) in-line with the existing literature. We find a price elasticity of supply of 0.29, which differs from the common assumption of a perfectly inelastic short-run supply curve. In doing so, we provide a robustness check on aggregate studies of gasoline demand and a consistent, econometric estimate of the price elasticity of gasoline supply. These results will be useful in light of current and future debate on tax and regulatory policies regarding gasoline consumption. In December of 2010, the Obama administration's National Commission on Fiscal Responsibility and Reform released a plan to achieve roughly $4 trillion in deficit reduction through 2020. Part of the plan includes gradually increasing the federal gas tax by 15 cents per gallon between 2013 and 2015 and dedicating the increase to the Transportation Trust Fund (which is proposed to replace the Highway Trust Fund). This increase, combined with proposed spending limits for the Fund, would in turn assist in deficit reduction by alleviating the need for continued bailouts of the Fund through deficit spending. The low elasticities of supply and demand for gasoline found in this paper suggest that, at least in the short-run, raising fuel taxes will generate significant amounts of revenue with relatively low efficiency costs. Given these results, and the historic recent deficits for both the Highway Trust Fund and the overall federal budget, policymakers should give careful consideration to the Commission's recommendation of increasing the federal gas tax. In addition to raising revenue and reducing deficits, many have made an argument for increasing fuel taxes for environmental purposes. Hsing, 1994 and Morrow et al., March 2010, and several others have found fuel taxes to be effective mechanism for reducing gasoline consumption and in turn carbon emissions. The elasticities found in this paper indicate that consumption of gasoline and resulting carbon emissions will be largely unaffected by marginal fuel tax increases, at least in the short-run. Finally, while not a micro-econometric study like West, (2004), (West and Williams, May, 2004), and (Wadud et al., (2010), by providing a short-run supply elasticity our study is able to provide greater insight into fuel tax incidence. While the entire statutory burden of fuel taxes fall on suppliers, the economic burden for consumers can be found by dividing the price elasticity of supply by the difference of the price elasticity of supply and the price elasticity of demand. Using our result of 0.07 for the price elasticity of demand and 0.28 for the price elasticity of supply, we find that in the short-run 80% of a fuel tax increase would fall on consumers.