American Economic Association Does Britain or the United States Have the Right Gasoline Tax? Author(s): Ian W. H. Parry and Kenneth A. Small Source: The American Economic Review, Vol. 95, No. 4 (Sep., 2005), pp. 1276-1289 Published by: American Economic Association Stable URL: https://www.jstor.org/stable/4132715 Accessed: 11-02-2019 21:26 UTC JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The American Economic Review This content downloaded from 149.125.253.122 on Mon, 11 Feb 2019 21:26:24 UTC All use subject to https://about.jstor.org/terms Does Britain or the United States Have the Right Gasoline Tax? By IAN W. H. PARRY AND KENNETH A. SMALL* Changing fuel prices and new energy policy taxes. The fuel tax, by contrast, is administrainitiatives have heightened interest in the approtively simple and well established in principle, priate level of gasoline taxation. These taxes even at very high rates in many nations. Therevary dramatically across countries: Britain's tax fore it is entirely appropriate to consider how of 50 pence per liter in 2000 (about $2.80 per externalities that are not directly priced should be taken into account in an assessment of fuel U.S. gallon) is the highest among industrial taxes. countries, while the United States, where federal and state taxes averaged about 40 cents/gal,As for revenues, a well-developed publichas the lowest rate (International Energyfinance literature rigorously compares the effiAgency, 2000). ciency of different tax instruments for raising The British government has defended highrevenues. Recently, this literature has been exgasoline taxes on three main grounds. First, by tended to compare externality taxes with laborpenalizing gasoline consumption, such taxes rebased taxes such as the income tax (e.g., A. duce emissions of carbon dioxide and local air Lans Bovenberg and Lawrence H. Goulder, pollutants. Second, they raise the cost of driving 1996; Parry and Wallace E. Oates, 2000). It is and therefore reduce traffic congestion and now feasible to bring the insights of this literatraffic-related accidents. Third, motor fuel taxes ture to bear on a tax, such as the fuel tax, that is in the United Kingdom provide significant govpartially intended as an imperfect instrument for ernment revenue-nearly one-fourth as large as controlling externalities. that from personal income taxes (Lucy Chen- A number of previous studies attempt to nels et al., 2000)-and do so efficiently sincequantify the external costs of transportation; fuel has a relatively low price elasticity. typically costs are estimated on a per-mile basis, A counterargument to the externality ratio-and they sometimes are converted to a pernale is that, except for carbon dioxide, it would gallon equivalent by multiplying by average be better that a tax be placed on something othervehicle fuel efficiency or miles per gallon.' As than fuel: local emissions, peak-period conges-our formulation makes clear, however, it is crution, or miles driven, preferably with a rate thatcial to account for the endogeneity of fuel econvaries across people with different risks of causomy: to the extent that people respond to higher ing accidents. Nonetheless, ideal externality fuel taxes by purchasing more fuel-efficient vetaxes have not been widely implemented: theyhicles rather than driving them less, the contribution of distance-based externalities to the raise objections on equity grounds, they require administrative sophistication, and there is often optimal fuel tax is substantially diminished. stiff political opposition to introducing new In this paper we derive the second-best optimal gasoline tax, disaggregating it into components that reflect external costs of congestion, accidents, and air pollution (local and global), * Parry: Resources for the Future, 1616 P Street, NW, well as a "Ramsey tax" component that reWashington, DC 20036 (e-mail: parry@rff.org); Small:asDepartment of Economics, University of California, Irvine, flects the appropriate balance between excise CA 92697 (e-mail: ksmall@uci.edu). We are grateful to Klaus Conrad, Amihai Glazer, Howard Gruenspecht, Larry Goulder, Charles Lave, Don Pickrell, Richard Porter, Paul Portney, Stef Proost, Mike Toman, Kurt Van Dender, Sarah 1 For example, John Peirson et al. (1995), Inge Mayeres West, and David Wildasin for helpful comments andet sugal. (1996), European Conference of Ministers of Transgestions, and to Helen Wei for research assistance. Kenneth port (ECMT) (1998, ch. 3), Richard C. Porter (1999), Small thanks the University of California Energy Institute Werner Rothengatter (2000), and various papers in David L. Greene et al. (1997). for financial support. 1276 This content downloaded from 149.125.253.122 on Mon, 11 Feb 2019 21:26:24 UTC All use subject to https://about.jstor.org/terms VOL. 95 NO. 4 PARRY AND SMALL: DOES BRITAIN OR THE U.S. HAVE THE RIGHT GASOLINE TAX? 1277 nearly four times as large in the case of the United taxes and labor taxes in financing the governStates. Indeed, optimized VMT taxes are quite ment's budget. Based on a detailed assessment high,we equivalent to around $2.50/gal for the of evidence on underlying parameter values, United apply the formula to the United States andStates and $3/gal for the United Kingdom, leading United Kingdom, thereby illustrating why, andtherefore in both cases to a higher tax to what extent, the optimal tax may differburden acrosson motorists than currently exists.2 countries, and under what circumstances, ifOur any, analysis abstracts from many potentially relevant considerations. One of the most promcurrent rates might be justified. We summarize the results as follows. First, inent is dependence on oil imports. Adjustment costs during oil price disruptions may not be under our benchmark parameters the optimal fully taken into account by energy suppliers or gasoline tax in the United States is $1.01/gal consumers. However, a careful assessment for (more than twice the current rate) and in the the United States by Paul N. Leiby et al. (1997) United Kingdom is $1.34/gal (slightly less than the overall external costs from oil depenhalf the current rate). The higher optimal puts tax for dency at the equivalent of only a few cents per the United Kingdom mainly reflects a higher of gasoline. In the United States, monopassumed value for marginal congestiongallon costs. sony power in the world oil market could justify Significantly different values are obtained under fuelbut taxes as part of strategic trade policy; but reasonable alternative parameter scenarios, weunexpect that U.S. gasoline taxes have much a Monte Carlo analysis suggests that it is effect on world oil prices than does U.S. likely for either the optimal U.S. tax toless be as foreign low as its current value, or the optimal U.K. taxpolicy. Nonetheless, there remains room for legitimate debate about the role of energy to be as high as its current value. restraint Second, the congestion externality is the larg- in overall world politics, which is beyond our scope. est component of the optimal fuel tax. The We also ignore distributional concerns. HowRamsey component is the next most important, ever, at least when measures of lifetime income followed closely by accidents and local air pol- opposed to annual income) are used, gasolution. Global warming plays a relatively (as minor line taxes appear to be less regressive than is role-ironically, since it is the only component commonly thought (e.g., James M. Poterba, for which the fuel tax is (approximately) the right instrument. 1991). Furthermore, there is scope for using other Third, the optimal gasoline tax is greatly re-policies to offset any adverse distribu- duced by the fact that less than half of the tax-induced reduction in gasoline use is due to 2 A VMT tax has been advocated as a replacement for reduced driving, the rest coming from changes in average fleet fuel efficiency. If we hadOregon's incor-fuel tax by the Road User Fee Task Force (2003), established by the Oregon Legislative Assembly. In the rectly assumed that vehicle miles change in United Kingdom, a more far-reaching plan, endorsed by the proportion to changes in fuel consumption, we in June 2005, would introduce a nationwide government would have computed the optimal gasoline taxof variable VMT tax rates (Alistair Darling, 2005; system Department for Transport, 2004). in both nations to be much higher, wellU.K. over 3 Other considerations we do not address include the $3.00/gal in the case of the United Kingdom. industrial organization of the oil industry, tax favoritism for and consumer myopia. We expect the first two Fourth, when considered as part ofthethe industry, broader fiscal system, the optimal gasoline tax considerations to affect primarily the distribution of eco- nomic rents rather than marginal resource costs. Consumer is only moderately higher than the marginal myopia may create a case for regulation rather than pricing external cost of gasoline. The Ramsey compoif fuel economy is the primary goal (Greene, 1998), but is nent is only about 25 cents/gallon, and this is not particularly relevant to distance-related externalities. offset in part by the higher excess burden Thereof are a of narrow-based tax relative to a labor tax. course other external costs from motor vehi- cles, including road damage, noise, water pollution, vehicle Finally, we simulate a vehicle miles traveledand tire disposal, and policing needs. Estimates of these (VMT) tax, which more directly addresses the costs are small relative to those from congestion, accidents, and pollution-see, e.g., Mark A. Delucchi (1997), U.S. distance-related externalities. The potential welDepartment of Transportation (1997, pp. 111-12-23, and fare gains from this policy are considerably larger 2000, section entitled "Other Highway-Related Costs" and than those from optimizing gasoline-tax rates-Table 10). This content downloaded from 149.125.253.122 on Mon, 11 Feb 2019 21:26:24 UTC All use subject to https://about.jstor.org/terms 1278 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 2005 tional effects of fuel prices, H), through as arguably paying for is compute done in Western Europe.3 combustion or improved drive trai comfort or payload to drive I smal The paper is organized as follows. Section etc. model and the optimal describes our analytical gasoline tax formula. Section Driving time is determined II discusses as follows: parameter values. Section III presents the quanti- tative results. Section IV discusses model limitations and concludes. I. Analytical Framework (3) T = 7rM = wT(M)M where i7 is the inverse of the average travel speed and M is aggregate miles driven per cap- ita. An increase in aggregate VMT leads to A. Model Assumptions more congested roads, so ur' > 0. Agents take wr as fixed-they do not take account of their own Consider a static, closed economy model impact on congestion. where the representative agent has utility func- We distinguish two types of pollutants: cartion: bon dioxide (denoted PF), which is proportional (1) U= u(G(C, M, T, G), N) -
Purchase answer to see full attachment