Many experts believe that gasoline taxes shouldbe increased for a variety of reasons. Their argu-ments are unpersuasive. Oil is not disappearing,and when it becomes more expensive, marketagents will substitute away from gasoline to savemoney. The link between oil price shocks andrecessions, although real in the 1970s, has beenmuch more benign since 1985 because of the ter-mination of price controls. Market actors proper-ly account for energy costs in their purchasingdecisions absent government intervention.Pollution taxes, congestion fees, and automobileinsurance premiums more closely related to vehi-cle miles traveled are better remedies for the exter-nalities associated with automobile travel than asimple fuel tax. Gasoline consumption does notnecessarily distort American foreign policy,impose military commitments, or empowerIslamic terrorist organizations.

State and federal gasoline taxes should be abol-ished. Local governments should tax gasoline onlyto the extent necessary to pay for roads when usercharges are not feasible. If government feels com-pelled to more aggressively regulate vehicle tailpipeemissions or access to public roadways, pollutiontaxes and road user fees are better means of doingso than fuel taxes. Regardless, perfectly internaliz-ing motor vehicle externalities would likely makethe economy less efficient—not more—by inducingmotorists into even more (economically) ineffi-cient mass transit use.

The arguments advanced against increasinggasoline taxes are applicable to the broader dis-cussion about America’s reliance on oil generally.The case for policies designed to discourage oilconsumption is nearly as threadbare as the casefor increasing the gasoline tax—and for largelythe same reasons.

Don’t Increase Federal Gasoline Taxes—Abolish Them

by Jerry Taylor and Peter Van Doren

_____________________________________________________________________________________________________

Jerry Taylor and Peter Van Doren are senior fellows at the Cato Institute. Peter Van Doren is editor of Regulationmagazine and author of Politics, Markets, and Congressional Policy Choices (University of Michigan Press,1991). They are contributing authors to Energy and American Society—Thirteen Myths (Springer, 2007).

Executive Summary

No. 598 August 7, 2007�������

PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1

Introduction

Economists almost uniformly believe thatmarkets should be left alone by governmentunless market failures exist. They go on tocaution that government intervention willimprove efficiency if—and only if—theprospective intervention remedies one ormore of those market failures. And even ifmarket failures exist, actual government poli-cies may not improve market operations,because politicians rather than economistsdesign the policies.1

The economic case for a gasoline tax is rel-atively straightforward. Gasoline consump-tion imposes costs on third parties. If gasolineconsumers had to compensate third partiesfor those costs, the total cost of gasolinewould rise, demand would fall, injured partieswould be made whole, and gasoline con-sumption would be optimal. But becausethose who suffer damages find that the trans-action costs associated with securing com-pensation are high, gasoline consumers donot pay for the burden they impose on others.

Many economists believe that gasolinetaxes are too low relative to the external costsfuel consumption imposes on others and thatthe economy would be more efficient with asubstantial increase in the federal fuels tax.That argument is embraced by conservatives aswell as liberals. For example, Harvard professorGreg Mankiw, a prominent free-market econo-mist and former chairman of President GeorgeW. Bush’s Council of Economic Advisers, hasrecently formed “The Pigou Club,” which ismade up of prominent economists and publicintellectuals who support an increase in thefederal gasoline tax.2

We examine those arguments and findthem unpersuasive. Some arguments for fueltaxes—such as the need for society to facilitatethe inevitable transition away from an oil-based economy, encourage energy conserva-tion, or reduce foreign oil imports—fail toconvince because they are unlikely to improveupon resource allocations that would occurabsent government intervention. Other argu-

ments for fuel taxes are unpersuasive becausethey are second-, third-, or fourth-best reme-dies to problems—such as automobiletailpipe emissions and roadway congestion—that are best remedied by direct charges onoffending externalities.

In fact, we find no compelling reason for afederal gasoline tax at all and call for itsrepeal. Nor do we find any compelling casefor state gasoline taxes. The only circum-stance in which gasoline taxes might makesense are those in which the transaction costsassociated with road use charges are so highthat gasoline taxes are the only reasonableway to pay for road construction and mainte-nance. This implies that fuel taxes are at bestmatters of local governmental concern andthat they should only be a fraction of currentcharges on motorists.

This paper is primarily concerned withgasoline taxes, but the arguments we makeagainst the gasoline tax are applicable to thebroader policy discussion about oil’s place inAmerican society. Although liberals and con-servatives, Democrats and Republicans,appear to agree that government should “dosomething” to move the country away fromoil consumption, the case for governmentalintervention is little different from—and nobetter than—the case for raising gasoline taxes.

Energy Depletion andFuture Generations

Because fossil fuels are exhaustible, somegasoline tax advocates argue that we need toration production in order to save resourcesfor future generations.3 Future generationshave no say in energy markets, but their pref-erences regarding resource availability in thefuture should be considered. Markets willnot provide that consideration, so govern-ment must do so.

Another version of this argument doesnot emphasize the rights of future genera-tions. Instead, it paints a picture of inevitablefuture shortages as production declinesoccur. Fuel shortages will be accompanied by

2

Fuel taxes are atbest matters of

local governmentalconcern and

should only be afraction of

current chargeson motorists.

price hikes, recessions, and political struggle.Those unpleasant effects can be avoided onlyif government starts planning now. As arecent report for the U.S. Department ofEnergy put it, “Intervention by governmentswill be required, because the economic andsocial implications of oil peaking would beotherwise chaotic.”4

Oil depletion concerns, however, rest onshaky ground. First, they are primarily aboutthe future availability of conventional crudeoil. Unconventional crude oil deposits—suchas those found in heavy bitumen, tar sands,and shale rock—are extremely plentiful andonly lightly tapped at the moment because ofhigh extraction costs.5 Moreover, the tech-nology exists to convert coal and natural gasto synthetic petroleum liquids, which meansthat other, more plentiful, fossil fuels couldbe harnessed to produce vast amounts ofpetroleum if the economics are favorable.Second, concerns that conventional crude oilis becoming scarce in any meaningful sensehave not withstood close scrutiny.6

If petroleum depletion were to become agenuine problem, would intergenerationalequity demand conservation? We think not.The strongest normative argument againstconservation is that it transfers resourcesfrom the relatively poor to the relatively rich.7

That’s because today’s generation is almostcertainly much poorer than future genera-tions will be. For instance, if per capitaincome grows at 2 percent a year, people 100years from now will be approximately 7 timeswealthier than we are today. Those concernedabout intergenerational equity should worrymore about standards of living today thanabout standards of living tomorrow.

The strongest positive argument againstgovernment intervention is that markets aremore capable than government of reactingquickly and efficiently to declines in petrole-um production. True declines, rather thantemporary shocks, will permanently increaseoil prices, which will induce investments inalternative energy sources and conservation.

But what about temporary (albeit multi-year) price shocks? If low prices most of the

time and high prices some of the time are aproblem, is there a market solution? Indeedthere is. Long-term oil futures contracts areavailable to those who are worried aboutfuture price increases.

The fact that marketers have not tried tooffer long-term stable prices to consumers byarbitraging between the futures and retailmarkets suggests that most consumersbelieve that they benefit by accepting lowspot prices most of the time in return forunpleasantly high spot prices some of thetime. Said differently, we are “dependent” onoil exported from unstable countries ratherthan domestic oil or alternative sources ofenergy, and we don’t attempt to contract ourway out of that instability, because it ischeaper in present value terms to remain“dependent.”

The “solution” to oil price instability is toaccept higher prices most of the time inreturn for lower prices some of the time.There is nothing wrong with such a trade-offas long as it is achieved through contract.Thirty-year fixed rate mortgages, for exam-ple, allow consumers to shift to others therisk of varying daily spot rates for borrowing(whose mean is lower but accompanied byhigher variance) in return for higher meanand no variance (fixed) prices.

We don’t, however, see those sorts of con-tracts in energy markets. Instead, what we seeare proposals for European-style taxes ongasoline consumption, mandated or subsi-dized alternative energy production, and reg-ulations that require energy producers toretain excess production capacity.

Unlike contractual solutions, governmen-tal solutions have the dubious distinction ofbeing more expensive not just most of thetime, but all of the time. That is, the “alterna-tives” to fossil fuels are more expensive thanconventional fossil fuels, even when the latterprices are at peak, which is, of course, whysuch “alternatives” are not embraced withoutgovernment subsidy or coercion. For exam-ple, we have recently calculated that the fed-erally owned Strategic Petroleum Reserve hascost the taxpayer between $65 and $80 per

3

Markets are morecapable than government ofreacting quicklyand efficiently todeclines in petroleum production.

barrel (2004 dollars) to fill, which rivals thehighest spot market prices ever recorded.8

We believe that market actors are alsomore likely to work in the interests of futuregenerations than are governmental actors.That’s because democratically elected gov-ernments, and the regulatory agencies estab-lished by them, have a tendency to reflect theinterests of swing voters in swing voting dis-tricts. Accordingly, it’s unreasonable toexpect governments to be more interested inthe well-being of future generations thanswing voters in swing districts who haveshort time horizons and political prefer-ences. A single glance at America’s lavishcommitments to retirees in the form ofMedicare, Medicaid, and Social Securityshould disabuse everyone of the notion thatcurrent voters make major sacrifices forfuture generations—even when the case forsacrifice is mathematically indisputable.

The opposite, in fact, is the case; voters arehappy to rob future generations. EconomistJagadeesh Gokhale, for instance, calculatesthat the current Social Security benefit struc-ture taxes future generations for the benefit ofthose currently alive. Taxes for future genera-tions are more than $1 trillion greater than thebenefits they are scheduled to receive.9

Markets, on the other hand, can reflectlonger time horizons. In fact, because themarket value of assets is determined byexpectations about what others might payfor them in the future, speculators representfuture generations’ interests in today’s mar-kets more effectively than politicians whofollow swing voters—whose time horizonrarely exceeds the next election.

To summarize, there is no market failureassociated with oil depletion. If oil becomesmore scarce over time, prices will rise toreflect that scarcity and resources will be allo-cated efficiently. Nor is there a market failureassociated with the interests of future gener-ations. Market agents have more incentive toconsider the interests of the future than gov-ernment actors because asset values areaffected by estimates of future profitability.We recognize that markets do not take the

distant future into account because of dis-counting, but the government’s treatment ofcurrent versus future Social Security costsand benefits does not support the view thatgovernments are good stewards of the future.

Oil Shocks CauseRecessions and InflationEnergy supply and demand are relatively

inflexible in the short run. As a consequence,small changes in either have very large effectson prices.10 This is the underlying reality thatexplains why oil and gasoline prices are sovolatile.11 Over a longer time period, however,both supply and demand are more respon-sive to prices.12

The short-run inflexibility of producers orconsumers, and the oil price shocks thatresult from such inflexibility, are allegedlyresponsible for inflation and recessions. Themacroeconomic damage inflicted by oil pricevolatility is an external cost imposed on soci-ety by gasoline consumers. Analysts at theOak Ridge National Laboratory peg the mar-ginal external costs associated with oil priceshocks at somewhere between 0 and $8.30per barrel of oil, or up to about 20 cents pergallon of gasoline.13

Economists disagree about the macroeco-nomic impact of oil shocks. Federal ReserveBoard chairman Ben Bernanke and his col-leagues, for example, have argued that different(“better”) monetary policy would reduce therecessionary effect of oil shocks, while econo-mists James Hamilton and Anna Herrera areskeptical of that proposition.14 The current oilprice explosion that began in 2003 has causedfar less economic harm than conventional wis-dom predicted, which adds credence to thoseeconomists who have argued that the reces-sions that followed previous oil shocks werenot caused by energy price increases.15

Recent work in the field tends to confirmthe suspicion that past analyses overstatedthe macroeconomic damage caused by oilprice shocks. A rigorous econometric analysisby economists at the Federal Reserve Bank of

4

Recent work inthe field tends to

confirm the suspicion that

past analysesoverstated the

macroeconomicdamage caused by

oil price shocks.

Atlanta, for instance, suggests that oil shockshad significant effects on the macroeconomybefore 1985 but not after. They argue that thefederal price control regime of the 1970s isthe explanation.16 Similarly, David Walton, aneconomist at the Bank of England, arguesthat wage rigidities in the 1970s were the cul-prit.17 Economists at the Federal ReserveBank of Cleveland, on the other hand, arguethat oil price increases might be painful formany, but they never have and never willcause inflation. They calculate that a dou-bling of oil prices would lead to a one-timeincrease in commodity prices of about 3 per-cent.18 A common theme of these recentpapers is that policy-imposed rigidities in theeconomy were responsible for the bad eco-nomic outcomes associated with past oil priceshocks, and the more flexible economy wenow have allows us to cope more easily.

Even though severely negative macroeco-nomic consequences may not follow oilshocks, the lack of supply and demandresponse in the short run leads to large trans-fers of wealth from consumers to firms intimes of high prices (1979–85, 2004–07) andfirms to consumers in times of low prices(1991–99). While energy policy discussionsoften invoke macroeconomic or market fail-ure rationales for government action, themost likely source of constituent demandsfor intervention in energy markets is the dis-tributional concerns of firms and con-sumers. Both consumers and firms attemptto enlist the assistance of government to pre-vent those wealth transfers.

Energy market interventions, however,have failed to help consumers and donemuch to damage efficiency.19 The oil price-control system in the 1970s induced short-ages and increased reliance on imports at atime when America’s stated policy was toreduce import dependency. Consumers weremade worse off as a consequence.20

In summary, price volatility is not a mar-ket failure. Recent evidence suggests thatmajor macroeconomic damage is not causedby oil price shocks per se but instead by poli-cy-induced rigidities including price controls

and wage rigidities that impede marketadjustment.

Consumer Failure toConserve

Claims that consumers fail to invest asmuch as they should in energy efficiency are anoften-invoked rationale for energy taxes in gen-eral and gasoline taxes in particular.Explanations vary as to why consumers actirrationally, but common complaints includelack of information regarding prospective sav-ings, cultural hostility to energy conservation,excessively optimistic expectations aboutfuture energy prices, imperfect access to capi-tal, and the demand for irrationally high ratesof return.21 Appropriate energy taxes wouldencourage optimal conservation expenditures.

How irrational are consumers when theymake energy decisions? Empirical investiga-tions find that consumers act far more ratio-nally than many analysts believe. Clemsoneconomist Molly Espey, for example, closelyexamined sales data from 2001 model auto-mobiles and found that consumers actuallyover-valued the gains possible from buyingfuel-efficient vehicles.22 An earlier examina-tion by Mark Dreyfus and W. Kip Viscusifound that consumers discounted the sav-ings from fuel efficiency by 11–17 percentwhen buying automobiles, rates equivalentto returns demanded by investors from otherinvestments at the time.23

Thus economists undermine the argu-ment that consumers are unwilling to paymore for a car to reduce gasoline use duringthe operating life of the vehicle. The govern-ment has no basis for regulating the use ofgasoline by vehicles either through a tax orthrough Corporate Average Fuel EconomyStandards (CAFE) standards.24

Environmental Externalities

Gasoline tax advocates frequently arguethat energy use causes environmental and

5

Empirical investigationsfind that consumers act farmore rationallythan many analysts believe.

human health damages and that those costsare not reflected in energy prices. Economistsdescribe such costs as “externalities” becausethey impose costs on others that are externalto the prices that govern the transactionbetween buyer and seller. Economists’ reme-dy for externalities is a tax that would quan-tify the cost of the externalities associatedwith each energy source in dollar terms. Thetax would force consumers to pay the exter-nal cost of their energy use (which would“internalize the externality”).25

The underlying objective of energy taxesin this regard is to approximate the marketthat would arise if polluters had to compen-sate those harmed by pollution.26 An energytax that considers environmental impactsfrom energy consumption is thus an attemptto mimic the market that would arise if thirdparties could hold polluters liable for thedamages caused by their pollution.

The first problem with a gasoline tax as ameans of internalizing environmental exter-nalities, however, is that it taxes the wrongthing. If we want to tax pollution, we shouldtax the emission of pollutants, not the raw con-sumption of gasoline.27 The two are not identi-cal given the differences in automobile age andmaintenance. For example, 5 percent of thevehicles on the road today generate 53 percentof volatile organic compound (VOC) emis-sions, while 10 percent of the vehicles on theroad today generate 76 percent of the same.28

Given that VOC emissions are a major contrib-utor to urban smog—and that vehicles thatemit unusually high loads of VOCs are likewisemore likely to emit unusually high loads ofother pollutants—this illustrates the difficultyof regulating fuel consumption rather thanemissions. A uniform gasoline tax will overtaxsome drivers and undertax others.

The second problem is that an increase ingasoline taxes would have very little effect onaggregate tailpipe emissions. That’s because con-sumers will primarily respond to a fuel tax overthe long run by purchasing more fuel efficientvehicles, not by driving less.29 And for every incre-mental increase of automotive fuel efficiency, a 20percent increase in vehicle miles traveled follows,30

and this increase in driving will greatly reduce theemissions reductions that we might otherwise seein response to the tax.31 Economist J. DanielKhazoom, for instance, calculates that doublingthe gasoline tax under the current regulatoryregime would only reduce tailpipe emissions by 6percent over the long run.32

The third problem with a federal gasolinetax designed to internalize environmentalexternalities is that the environmental andhealth-related damages imposed by air emis-sions vary by location. Air sheds have variablecarrying capacities and the harms caused byemissions are largely determined by back-ground ambient concentration and the mar-ginal impact of additional loads. Accordingly,a given amount of hydrocarbon tailpipe emis-sions will have a greater negative impact inLos Angeles, California, than in Sioux City,Iowa. Uniform national environmental exter-nality taxes will be inefficient and wrong allthe time—too low in some areas and too highin others.

The fourth problem with environmentalexternality taxes is the difficulty associatedwith monetarizing the aggregated nationalhealth and environmental externalities asso-ciated with energy consumption in theUnited States.33 Parry and his colleaguesreport that the plausible estimates for con-ventional pollutants range from $.36 to$4.20 per gallon.34 Of course, auto emissionscontinue to decline from the 2000-era esti-mates used in those calculations,35 and stud-ies that rely on toxicological risk assessmentsand epidemiological studies to ascertaindamages may overstate human healthimpacts.36 Estimates regarding the climate-related costs associated with consuming aton of carbon likewise vary greatly; accordingto one survey of the literature, from $9 to$200 per ton of carbon in 2000 dollars.37

Experts also disagree about the dollar val-ues one should attach to human morbidity,mortality, and environmental harms. Forexample, the peer-reviewed literature suggeststhat employers have to pay employees any-where between $0.7 million and $16.3 millionto compensate for a statistical risk of death.38

6

If we want to taxpollution, we

should tax theemission of

pollutants, notthe raw

consumption ofgasoline.

What number should analysts use when mon-etarizing mortality in externality internaliza-tion exercises? One might try to dodge thatproblem by estimating the number of qualityadjusted life years (QALYs) lost through pollu-tion and then calculate what it would cost gov-ernment to provide for an equivalent numberof QALYs through improved health services.39

But that would require politicians to dedicatepollution taxes to health services programs.Hence, pollution taxes might prove quite inef-ficient, reflecting not the cost of pollution perse but the cost of socialized health care.

A more recent estimate is offered by econ-omists Ian Parry and Kenneth Small. Theirreview of the “best guesses” in the literaturesuggests that a national gasoline tax wouldinternalize environmental externalities byimposing a tax of 16 cents per gallon to payfor cost of conventional pollution and 5cents per gallon to pay for the costs of green-house gas emissions.40

To summarize, the environmental damagesimposed on third parties by driving motor vehi-cles are indeed a market failure; the costs associat-ed with those damages are not reflected in drivingcosts. Gasoline taxes, however, will have littleeffect on aggregate tailpipe emissions. The correctremedy to the problem—assuming we wish toaddress it—is an emissions charge that varies withemissions as well as the capacity of the air shed tohandle extra emissions rather than a tax on gaso-line. A national emissions tax would be inefficientbecause it would ignore the large geographic vari-ation in damages associated with pollution. Andeven though the literature provides estimates ofdamages that could be used to set local emissioncharges, the range is so large that it provides verylittle guidance to decisionmakers.

Accident and CongestionExternalities

Gasoline tax advocates also see the tax as ameans to discourage highway congestion andreduce accidents on the roadway. Drivers donot pay the marginal costs they impose on oth-ers when they crowd the roads—including the

increased probability of accidents. Those costsare not trivial. Parry and Small, for example,calculate congestion externalities at 29 centsper gallon (8 cents more than the environmen-tal externalities associated with motor vehicleuse)41 and accident externalities at 24 cents pergallon (3 cents more than the environmentalexternalities associated with motor vehicleuse). A recent paper by economists Aaron Edlinand Pinar Karaca-Mandic estimates that acci-dent externalities in California alone exceed$66 billion, more than current individual andcorporate income taxes collected in the state.42

But internalizing those externalities via agasoline tax is a very poor way of addressingthose problems. Better approaches includetolls that vary with congestion43 and the pro-motion of “Pay-As-You-Drive” insuranceunder which premiums would vary in directproportion to vehicle miles traveled and theinsured’s risk factor as determined by insur-ance companies.44 Gasoline taxes are an imper-fect means to address congestion or accidentcosts because such taxes don’t vary with thedensity of the setting in which driving occursor the extent to which a driver might be acci-dent-prone.45

The futility of taxing gasoline as a second-best policy to tackle congestion is well illustrat-ed by policy in London. Gasoline taxes in theUnited Kingdom are $2.80 per gallon,46 morethan seven times higher than they are in theUnited States (where they average 38 cents pergallon).47 Yet, high U.K. taxes have not alleviat-ed congestion in urban areas like London.When the municipal government in Londonimposed congestion-based tolls, however, tocharge drivers for using inner-city streets, con-gestion was greatly diminished.48 When con-gestion charges were imposed in Stockholm in2006, traffic likewise decreased 22 percent andexhaust emissions decreased by 14 percent.49

National SecurityExternalities

The most common rationale heard todayfor higher gasoline taxes is the complaint that

7

Gasoline taxesare an imperfectmeans to addresscongestion oraccident costs.

oil consumption harms national security.There are four distinct arguments. First, oilimports require the United States military tosecure foreign production facilities and ship-ping lanes. Second, good relations with oil pro-ducers are necessary to ensure that oil flowsinto U.S. ports, but good relations with pro-ducers can impose unacceptably large short-and long-term costs on the Treasury and con-tribute to anti-American sentiment, whichitself imposes costs. Third, oil profits fundIslamic extremists. Fourth, oil revenues areoften captured by international bad actors, andthe harm done by those regimes both withinand without their borders is to some extent“paid for” by U.S. motorists. None of thosecosts, however, are paid by those who consumegasoline. Hence, higher gasoline taxes wouldinternalize the externalities.

In this section we examine each argument.50

Taxing for the “Oil Mission”Motorists do not pay for the costs associat-

ed with the safe and reliable delivery of foreignoil. Protecting oil tankers from harm, after all,is an explicit mission of the United States mili-tary. Protecting friendly oil-producing statesfrom attack is also thought by many to be animplicit U.S. military mission.51

Quantifying the national security costs asso-ciated with ensuring the safe and reliable deliv-ery of foreign oil is difficult. The CongressionalResearch Service estimated in 1997 that thosecosts may be anywhere between $0.5 billion and$65 billion, or 1.5 cents to 30 cents per gallonfor motor fuel from the Persian Gulf.52 Deeperanalysis by Mark Delucchi of the Institute ofTransportation Studies at the University ofCalifornia, Davis, and James Murphy at theUniversity of Massachusetts suggests that, if theUnited States did not import Persian Gulfcrude oil, military costs would be $11 billion—$42 billion less than they are today. If we did notuse oil at all in the motor transport sector,expenditures would be $3 billion—$31 billion ayear less than they are today.53

Agreement about the extent of the mili-tary’s “oil mission” is difficult to achievebecause military and foreign policy expendi-

tures are generally tasked with multiple mis-sions and objectives, and oil security is simplyone mission among many. Analysts disagreeabout how to divide those missions into bud-getary terms. Agreement about total expendi-tures is difficult because it’s very difficult toknow what Congress would appropriate invarious counterfactual scenarios.

Debate about the size of the U.S. military’soil mission and related foreign policy expens-es is not particularly relevant to gasolinetaxes. From an economic perspective, the keyquestion is whether an elimination of U.S.military and foreign aid expenditures dedicat-ed to “the oil mission” would result in anincrease in the price of oil, and, if so, howmuch? That is the true measure of the nation-al security externality if it exists. Measuringthe externality by the amount of money gov-ernment spends on the oil mission is at best ameasure of how much politicians believe theexternality might be. Political assessmentsmay or may not be accurate.

To be sure, if the termination of theAmerican “oil mission” implied the termina-tion of all military, police, and court servicesin the region, petroleum extraction invest-ments would become more risky, oil produc-tion would decrease, and prices wouldincrease. But remember that oil companies inthe region are creatures of government. Sothe question is really whether Middle Eastgovernments would produce less oil becausethe United States ended its oil-related mili-tary mission and foreign aid. Or would oilproducing states provide—or pay others toprovide—military services to replace thosepreviously provided by the United States?

We believe that a cessation of U.S. securityassistance would be replaced by security expen-ditures from other parties. First, oil producerswill provide for their own security needs aslong as the cost of doing so results in greaterprofits than equivalent investments couldyield. Because Middle Eastern governmentstypically have nothing of value to trade exceptoil, they must secure and sell oil to remainviable. Second, given that their economies areso heavily dependent on oil revenues, Middle

8

From an economic

perspective, the keyquestion is whether

an elimination ofU.S. military and

foreign aid expenditures

dedicated to “theoil mission” would

result in anincrease in the

price of oil.

Eastern governments have even more incentivethan we do to worry about the security of pro-duction facilities, ports, and sea lanes.54 Third,even if producing countries were to provideinadequate security in the eyes of consumingcountries, consuming countries could pay pro-ducers to augment it.

In short, whatever security our presence pro-vides (and many analysts think that our presenceactually reduces security)55 could be provided byother parties were the United States to withdraw.The fact that the Saudi Arabia and Kuwait paidfor 55 percent of the cost of Operation DesertStorm suggests that keeping the Straits ofHormuz free of trouble is certainly within theirmeans.56 The same argument applies to al Qaedathreats to oil production facilities.

If oil regimes paid for their own militaryprotection and the protection of their ownshipping lanes, would U.S. Middle-East mili-tary expenditures really go down? Theanswer might well be “no” for two very dif-ferent reasons. First, the U.S.–Middle Eastmilitary presence stems from our implicitcommitment to defend Israel as well as theregion from Islamic fundamentalism, andthose missions would not likely end simplybecause Arab oil regimes paid for their owneconomic security needs. Second, bureau-cratic and congressional inertia might leavemilitary expenditures constant regardless ofIsraeli or petroleum defense needs.57

Thus, U.S. Persian Gulf expendituresshould not be viewed as a subsidy that lowersoil prices below what they otherwise wouldbe. Instead, the expenditures are a taxpayer-financed gift to oil regimes and the Israeligovernment that has little, if any, effect on oilprices. One may support or oppose such agift but not on “market failure” grounds.

Foreign Policy Externalities Many foreign policy analysts think that U.S.

oil imports are dependent on friendly relation-ships with oil-producing states. The fear is thatunfriendly regimes might not sell us oil.Maintaining good relations with oil producers,however, interferes with other foreign policyobjectives and increases anti-American senti-

ment in producer states with unpopularregimes. While the costs associated with thisdistortion of foreign policy are difficult if notimpossible to quantify, that doesn’t makethem any less real. Because a higher gasolinetax would reduce consumption, many believethat high fuel taxes would give us more free-dom to shun odious oil-producing regimes.58

The problem with this argument, however, isthat its fundamental premise is incorrect.Friendly relations with producer states neitherenhance access to imported oil nor lower its price.

Selective embargoes by producer nations onsome consuming nations are unenforceableunless (1) all other nations on Earth refuse toship oil to the embargoed state, or (2) a navalblockade were to prevent oil shipments into theports of the embargoed state. Once oil leaves theterritory of a producer, market agents dictatewhere the oil goes, not agents of the producer,and anyone willing to pay the prevailing worldcrude oil price can have all he wants.59

The 1973 Arab oil embargo is a perfect casein point. U.S. crude oil imports actually in-creased from 1.7 million barrels per day (mbd)in 1971 to 2.2 mbd in 1972, 3.2 mbd in 1973,and 3.5 mbd in 1974.60 As MIT’s Thomas Lee,Ben Ball Jr., and Richard Tabors observe: “Itwas no more possible for OPEC to keep its oilout of U.S. supply lines than it was for theUnited States to keep its embargoed grain outof Soviet silos several years later. Simple rerout-ing through the international system circum-vented the embargo. The significance of theembargo lay in its symbolism.”61 Granted,“there were short term supply disruptions,”but “the only tangible effect of the embargowas to increase some transportation costsslightly, because of the diversions, reroutings,and transshipments necessitated.”62 MIT econ-omist M. A. Adelman agrees:

The “embargo” of 1973–4 was a sham.Diversion was not even necessary, itwas simply a swap of customers andsuppliers between Arab and non-Arabsources. . . . The good news is that theUnited States cannot be embargoed,leaving other countries undisturbed.63

9

Friendly relationswith producerstates neitherenhance access toimported oil norlower its price.

In short, it does not matter to consumersto whom the oil is initially sold. All that mat-ters to consumers is how much oil is pro-duced for world markets.

Do oil-producing nations allow their feel-ings toward oil-consuming nations to affecttheir production decisions? Historically, theanswer has been “no.” The record stronglyindicates that oil-producing states, regardlessof their feelings toward the industrializedWest, are rational economic actors. After adetailed survey of the world oil market sincethe rise of OPEC, M. A. Adelman concluded,“We look in vain for an example of a govern-ment that deliberately avoids a higher income.The self-serving declaration of an interestedparty is not evidence.”64 Prof. Philip Auerswaldof George Mason University agrees, “For thepast quarter century, the oil output decisionsof Islamic Iran have been no more menacing orunpredictable than Canada’s or Norway’s.”65

Although this is indeed the orthodox viewamong oil economists, there are examples ofcountries selling oil and natural gas to othersat below-market rates: Russia sold oil to Cubaat below-market prices during the Cold War;Russia continues to sell natural gas to Ukraineat below-market prices but has ended its sub-sidy to Georgia as relations have soured; andChina sells oil to North Korea at low rates andused this as leverage to induce North Korea tobargain over its nuclear weapons program.66

What should we learn from those cases?First, sellers have leverage in natural gas mar-kets that is not possible in oil markets becauseoil can be transported easily while natural gasis shipped through pipelines. Buyers have fewnear-term alternatives if natural gas sellersreduce shipments. As liquefied natural gasgains market share, however, natural gas mar-kets will look increasingly like world crude oilmarkets and the ability of Russia or otherstates to extract concessions from consumerswill dissipate. Second, the Russia–Cuba andChina–North Korea cases involve poor coun-tries receiving foreign aid in the form of low-priced oil. We are unaware of any wealthy west-ern countries receiving such in-kind aid fromoil-producing countries.

Thus far, our analysis has examined thebehaviors and incentive structure of incum-bent regimes. But if a radical new actor were toemerge on the global stage, how would itbehave? For example, if the House of Saudwere to fall and the new government consistedof Islamic extremists friendly to Osama binLaden, the new regime might reduce produc-tion and increase prices.67 But that scenario isby no means certain, given that Iran—despiteall its anti-western rhetoric—has not reducedoil output, because the Iranian economy andregime are dependent on oil revenue and theSaudis are even more dependent.68

Regardless, the departure of Saudi Arabiafrom world crude oil markets would probablyhave about the same effect on domestic oilprices as the departure of Iran from worldcrude oil markets in 1978. Iran accounted forjust shy of 10 percent of global oil productionbefore the Iranian Revolution virtually shutdown oil production, whereas Saudi Arabiaaccounts for about 13 percent of global oilproduction today.69 Oil prices increased dra-matically after the 1978 revolution, but thosehigher prices set in motion market supply-and-demand responses that undermined thesupply reduction and collapsed world priceseight years later. The short-term macroeco-nomic impacts of such a supply disruptionwould actually be less today than they werethen, given the absence of price controls onthe U.S. economy and our reduced reliance onoil as an input for each unit of GDP.70

So while it is possible that a radical oil-producing regime might play a game ofchicken with consuming countries, produc-ing countries are very dependent on oil rev-enue and have fewer degrees of freedom tomaneuver than consuming countries.Catastrophic supply disruptions wouldharm producers more than consumers,which is why they are extremely unlikely.

Oil Profits for TerroristsMoney spent on gasoline flows to oil pro-

ducers, and many of those producer states usethose revenues to directly or indirectly fundIslamic extremists. Private individuals who

10

The recordstrongly indicates

that oil-producing states,

regardless of theirfeelings toward

the industrializedWest, are rationaleconomic actors.

profit from the oil trade likewise contribute toIslamic extremists. Those extremists pose for-eign policy and national security problems.This suggests that reduction in oil revenueswould reduce Islamic extremist activities.

Before we go on, it’s worth noting thatonly 15.5 percent of the oil in the world mar-ket is produced from nation-states accusedof funding terrorism.71 Hence, the vastmajority of the dollars we spend on gasolinedo not end up on this purported economicconveyer belt to terrorist bank accounts.

Regardless, terrorism is a relatively low-costendeavor, and oil revenues are unnecessary forterrorist activity. The fact that a few hundredthousand dollars paid for the 9/11 attacks sug-gests that the limiting factor for terrorism isexpertise and manpower, not money.

What is the relationship between oil pricesand Islamic terrorist incidents? We estimatedtwo regressions using annual data from 1983to 2005: the first between fatalities resultingfrom Islamic terrorist attacks and Saudi oilprices and the second between the number ofIslamic terrorist incidents and Saudi oilprices. In neither regression was the estimat-ed coefficient on oil prices at all close tobeing significantly different from zero.72

That probably explains why there is nocorrelation between Persian Gulf oil revenuesand terrorist activity. Inflation-adjusted oilprices and profits during the 1990s were low.But the 1990s also witnessed the worldwidespread of Wahhabi fundamentalism, thebuild-up of Hezbollah, and al Qaeda’s com-ing of age. Note too that al Qaeda terroristsin the 1990s relied on help from state spon-sors such as Sudan, Afghanistan, andPakistan—nations that aren’t exactly knownfor their oil wealth or robust economies.

What terrorists need most is a recruitingpool from which to draw. If the United Stateswere to tax gasoline to such an extent thatglobal oil demand, prices, and profits for oilproducers declined, the oil states would havesmaller economies and less to distribute totheir underemployed youth. To the extentthat deteriorating economic conditionsbreed social discontent and political resent-

ment, taxing gasoline to reduce revenuesflowing to Islamic terrorists might wellincrease the recruitment pool for Islamic ter-rorists and make matters worse.

Reducing oil revenue to noxious regimesmight be a risk worth taking if billions werefinding their way from such regimes into alQaeda coffers, but that seems unlikely.Everything we know suggests that al Qaedaterrorist cells are “pay as you go” operationsthat primarily engage in garden-variety crimeto fund their activities, and Islamic charitiesare the primary sources for organizationalrevenue.73 Given that the governments ofSaudi Arabia, Kuwait, and others in theregion are slated for extinction should binLaden have his way, those governments haveno interest in facilitating the transfer of oilrevenues to some post office box in Pakistan.

Producer states do indeed use oil revenuesto fund ideological extremism, and Saudifinancing of madrassas and Iranian financing ofHezbollah are good examples. But given theimportance of those undertakings to the Saudiand Iranian governments, it’s unlikely thatthey would cease and desist simply becauseprofits were down. They certainly weren’tdeterred by meager oil profits in the 1990s.74

The futility of reducing oil consumption asa means of reducing terrorism is illustrated byan examination of revenues earned from oilsales. A recent paper from the publishers of theLundberg Letter notes that oil exports fromstates accused of funding terrorism earnedthose governments $290 billion in 2006. Evenif that sum were cut by 90 percent, it would stillleave $29 billion at their disposal—more thanenough to fund terrorism given the minimalfinancial needs of terrorists. “Even a price of$10 per barrel crude (an unlikely scenario evenunder massive subsidy programs for plug-inhybrid vehicles and biofuels market share man-dates) would likely not cut off the purportedcash flow to terror groups.”75

Rents to Bad ActorsWhen oil prices are high, so too are oil prof-

its for infra-marginal (low-cost) producers.Even if those profits do not find their way to

11

There is no correlationbetween PersianGulf oil revenuesand terroristactivity.

international terrorists, they serve to prop upmany regimes we find distasteful. Oil produc-ers in the Second and Third worlds, after all,often use their robust flow of petrodollars tosquelch human rights at home and to menaceneighbors abroad.76 Many analysts blame U.S.motorists for indirectly funding these interna-tional bad actors and argue that the actions ofthese bad actors impose costs on third partiesthat are not reflected in gasoline prices. Taxesto reduce demand would internalize this par-ticular externality, reduce the flow of money tobad actors and, presumably, weaken their gripon political power. At the very least, it mightwell reduce the threat they pose to others.

To our knowledge, no one has attempted toquantify this alleged externality, and it is unlikelythat anyone ever will. After all, putting a price tagon lost civil and economic liberties, not to men-tion regional instability and military tension,would be extremely difficult. Moreover, onewould need to estimate the baseline degree of“bad acting” associated with a particular regimeor nation-state before one could estimate to whatextent oil profits are responsible for observedinstances of “bad acting.” Accordingly, policy-makers have little concrete information at theirdisposal to inform tax policy.

Regardless, it is unclear to what extent oilprofits are associated with human rights abus-es or militaristic activity. There are plenty ofexamples, after all, of relatively long-livedregimes with terrible human rights records(such as North Korea) that have no oil revenuesto speak of, and this is the case even within thesame socioeconomic regions. Denuding Iranand Libya of oil revenues might produce a gov-ernment that looks a lot like Syria; denudingVenezuela of oil revenues might produce a gov-ernment that looks a lot like Cuba; and denud-ing Russia of oil revenues might produce a gov-ernment that looks a lot like Russia used to be.After all, all of these “bad-acting” petro-statesyielded unsavory regimes even when oil rev-enues were a third of what they are today.

The claim that oil revenues increase thethreat those regimes pose to their neighborsseems reasonable enough, but here again, it isunclear to what extent this is true. Pakistan is a

relatively poor country with no oil revenues tospeak of, but it has still managed to build anuclear arsenal and is constantly on theprecipice of war with India. Impoverished, oil-poor Egypt and Syria have at various timesbeen the most aggressive anti-Israeli states inthe Middle East. Russia launched its war withChechnya before oil revenues engorged its trea-sury. While we have no doubt that (all otherthings being a equal) a rich bad actor is moredangerous than a poor bad actor, the marginalimpact that oil revenues have on “bad acting”might well be rather small.

Regardless, the fact that unsavory petro-states have been fully capable of holding onto power, oppressing their people, and men-acing their neighbors during a decade associ-ated with the lowest inflation-adjusted oilprices in history (the 1990s) suggests thatnothing short of rendering oil nearly value-less will have any real effect on regime behav-ior. A one dollar hike in the federal gasolinetax, for example, would reduce world crudeoil prices by only 1–5 percent, or by$0.65–$3.25 given $65 crude—not enough tohave any appreciable effect on bad acting orterrorist funding. Getting prices back downto 1998 levels (the lowest inflation-adjustedprice in history) would require a gasoline taxof more than $20.00 per gallon.77

For the sake of argument, however, let’sassume that there is some incremental benefitassociated with reducing oil revenues to bad-acting oil producers—an assumption thatseems entirely reasonable. Unfortunately, wehave only very blunt and imperfect instru-ments at hand to achieve that end. A gasolinetax, for instance, would reduce oil demand—and, thus, reduce revenues—for all oil produc-ers, whether they are bad actors or not.Producers in the North Sea, Canada, Mexico,and the United States (which collectively sup-plied 20.1 million barrels of oil per day in 2006,or 24 percent of the world’s crude oil needsthat year) would be harmed just the same asproducers in Venezuela, Iran, Russia, and Libya(which collectively supplied 20.3 million bar-rels per day in 2006).78

Imposing oil taxes to reduce profits for

12

A one-dollar hikein the federal

gasoline taxwould reduce

world crude oilprices by only 1–5

percent—notenough to haveany appreciable

effect on bad acting or terrorist

funding.

unsavory petro-states is thus akin to takingout a shotgun to kill a mosquito at 50 yards.You may or may not kill the mosquito, butyou’re sure to hit a lot of unintended targetswhen you pull the trigger.

Given there was plenty of “bad acting” in1998, it’s unlikely that even astronomicalgasoline taxes would have much effect onbad acting. Accordingly, we doubt that theforeign policy benefits that might accruefrom gasoline tax increases would outweighthe very real costs that such a tax wouldimpose on both consumers and innocentproducers. We suspect that there are betterremedies available to the United States tocurtail bad behavior abroad.

To summarize, we find little reason tobelieve that America’s national security is jeop-ardized to any great extent by oil consumptionor that gasoline taxes could reduce whateverproblems may exist. U.S. taxpayers do pay forU.S. military activities in the Middle East,which are justified in part by the desire tosecure oil production and export facilities. Butthose expenditures are properly thought of aswealth transfers rather than externality-creat-ing payments because their termination wouldnot alter oil prices. Good relations with oil pro-ducers have no effect on the price or the avail-ability of oil in the world market. Oil revenuesare not necessary for terrorist activity, and thevariation in terrorist activity over time does notseem to be related to oil revenue. And while badinternational actors do indeed get rich off oilrevenues, gasoline taxes are unlikely to sub-stantially reduce the degree or the extent of badacting.

Gasoline Taxes: Better than the Alternative?

Many if not most of the economists whoembrace federal and state gasoline taxes con-cede the arguments above when pressed.They continue to support fuel taxes, howev-er, for two reasons. First, they believe that fueltaxes are more efficient means of raising rev-enue than other forms of taxation. Second,

they fear that first-best means of addressingexternalities (direct taxes on pollution, roaduse, etc.) are not politically feasible and thatgasoline taxes are a second-best remedy thatis preferable to the alternative, which is toleave externalities unaddressed.

We examine each of those arguments inturn.

Fuel Taxes and the “Double Dividend”Gasoline taxes change fuel consumption

habits less than other taxes affect the con-sumption of other goods or services. They sim-ply extract revenue. From an economist’s pointof view, that’s good. When taxes decrease con-sumption of that which is taxed, individualwelfare is reduced over and above the amountof the taxes actually paid. Those welfare lossesare thought to be quite substantial. MichiganState economist Charles Ballard, for instance,has calculated that each additional dollar oftax revenue imposes 20–30 cents of welfare lossabove and beyond the losses associated direct-ly with the tax payment.79

Hence, some observers have argued that ifgasoline taxes were increased and other taxesdecreased so that overall revenue remainedconstant, a gasoline tax hike would provide a“double dividend.” That is, it would reducethe negative externalities associated withgasoline consumption while also reducingthe welfare losses associated with taxation.Even if a gasoline tax created no net benefits,as long as the welfare losses associated with agasoline tax were smaller than the welfaregains associated with cuts in other more dis-tortionary taxes, a gasoline tax hike wouldmake economic sense.80

But if gasoline taxes produce such bene-fits, a tax on vehicle miles traveled would beeven better because the demand for vehiclemiles traveled is even more inelastic thandemand for fuel.81 Given that monitoringvehicle miles traveled is quite simple and notparticularly costly, analysts who embrace the“double dividend” argument have no goodreason to prefer fuel taxes over taxes on vehi-cle miles traveled.

However, academic researchers—even those

13

Academicresearchers—eventhose who support increasing thegasoline tax—have several reservationsabout the double-dividend claim.

who support increasing the gasoline tax—haveseveral reservations about the double-dividendclaim.82 First, while it’s true that a gasoline taxwill not change behavior very much in gasolinemarkets, they impose significant distortions inother markets. For instance, a gasoline tax willreduce after-tax wages in precisely the sameway as a direct tax on wages. Hence, a gasolinetax will introduce distortions in the labor mar-ket. It will also create distortions in other com-modity markets by reducing demand for somegoods while increasing the demand for others.Those distortions are at least as large as the dis-tortions introduced by other forms of labortaxation, and they tend to “exacerbate, ratherthan alleviate, preexisting tax distortions—evenif revenues are employed to cut preexisting dis-tortionary taxes.”83

But for reasons that are partially illustratedabove, a gasoline tax (and a tax on virtually anyother commodity, for that matter) is implicitlya tax on labor,84 and taxing labor creates lesswelfare loss than taxing capital.85 So even if agasoline tax hike leads to welfare losses in laborand other commodity markets, if the revenuesassociated with the tax were used to cut taxeson capital, it is possible that the welfare gainsassociated with the capital tax reductionswould exceed the welfare losses associated withincreasing the gasoline tax.

Despite this theoretical escape hatch, thegasoline market offers too narrow a tax base tosubstitute in any substantial way for taxes oncapital. To be revenue neutral, the gasoline taxwould have to be too high. And once the taxbecomes high enough, behavior will change(motorists will switch from gasoline to someother fuel—like ethanol), and behavior changeimplies welfare losses. Accordingly, the efficien-cy gains that might result from a tax swap willnot offset the efficiency losses caused by thegasoline tax increase.86

There are two other practical complica-tions. First, replacing income with gasolinetaxes decreases the efficiency of revenue collec-tion because it is cheaper to collect a givenamount of revenue from a broad tax base rela-tive to a narrower tax base.87 Second, the “dou-ble dividend” can only occur if the revenues

from the gasoline tax are used to offset cuts incapital taxes. If the revenues are rebated tolower-income Americans to offset the regres-sivity of the tax swap—which would almost cer-tainly happen to some extent in the currentpolitical climate—that would reduce the rev-enues available to “buy” cuts in capital taxesand, thus, further reduce or eliminate the effi-ciency gains that result from the tax swap.88

So while replacing capital taxes with labor(consumption) taxes is welfare-improving,the gains associated with that switch cannotbe secured absent generalized tax reformacross all sectors of the economy.89 As econo-mist Stephen Smith observed after surveyingthe literature:

Ecotaxes are likely to involve distor-tionary costs at least as high as thoseinvolved in raising equivalent revenuesthrough existing taxes. If the question isposed whether we would choose to useenergy taxes, in preference for existingtaxes on labour and other bases, in theabsence of any environmental benefits,then the answer is almost certainly thatwe would not. Energy taxes would belikely to involve just as much distortionof the labour market as income taxes,and at the same time distort the com-modity market. Only if there are expect-ed to be environmental gains can the useof environmental taxes be justified, andthe case for ecotax reform must be madeprimarily on the basis of the environ-mental gains that would result.90

First- vs. Second- vs. Third-Best Policy Are gasoline taxes worth embracing as a

“second-best” policy, given the widespreadbelief that “first-best” remedies, such as directtaxation of the externalities in question, are offthe table?91 We don’t think so.

First, as a factual matter, it’s unclearwhether alleged first-best remedies such astailpipe emission taxes and road use chargesare truly more difficult to pass in a legislaturethan fuel tax increases. Energy taxes, after all,

14

Although replacing capitaltaxes with labortaxes is welfare-improving, the

gains associatedwith that switch

cannot be securedabsent generalizedtax reform across

all sectors of theeconomy.

are among the most politically unpopulartaxes in America, as President Bill Clinton dis-covered when he attempted to impose a Btutax during the first year of his presidency.Pollution taxes, on the other hand, are some-what more “virtuous” in the public’s mind,and highway tolls are increasingly common. Ifone posits that gasoline taxes are unpopularbecause they are visible, unavoidable, andimposed on a commodity for which demand isrelatively unaffected by price, then pollutiontaxes and road-use charges would likely proveno more unpopular than gasoline taxes.

Second, economists who argue forincreased gasoline taxes rarely concede (tonon-economists, anyway) that those taxes aredeeply problematic and only worth embrac-ing because better policies are presumably offthe table. Instead, the case for higher gasolinetaxes is usually offered to the public with agreat deal of intellectual bravado that almostalways overstates the ability of gasoline taxesto solve identified problems.92 We believethat economists should argue for first-bestpolicies and let the political chips fall wherethey may. After all, if academics (who don’thave to worry about winning popularity con-tests at the ballot box) don’t make the casefor politically unpopular first-best economicpolicies, who will? Abandoning the case forideal policy in the public realm because itmay prove unpopular implicitly assumesthat good arguments do not persuade. It alsorequires economists to make judgmentsabout what is politically feasible and what isnot, and economists have no particularexpertise in that matter.

Third, it’s unclear whether gasoline taxeseven qualify as a “second-best” means ofaddressing air pollution or road congestion.That’s because the difference between theupper- and lower-bound externality cost esti-mates are larger than the marginal gainspromised by intervention. Hence, raising thegasoline tax too high could well make priceseven less, not more, reflective of total costs.

The only way to hedge against that risk isto support gasoline tax increases that fallwithin the lower bound of the aggregated

externality estimates, but that would pro-duce correspondingly little efficiency gaineven in theory. As economist Stephen Smithpoints out:

It is perhaps an over-generalization tosuggest that environmental taxes shouldbe large, or not imposed at all. However,the costs of complexity and the risk thatminor environmental taxes will simplybe ignored should both caution againsttoo much environmental fine-tuning ofthe fiscal system.93

That’s particularly the case given thatsmall, incremental tax increases do not guar-antee only small, incremental welfare losseswhen the preexisting tax system is inefficient.Relatively small carbon taxes, for example,yield disproportionately large gross costs intheoretical simulations replicating the exist-ing tax system.94

Fourth, and most important, even if agasoline tax increase were able to perfectlyprice gasoline’s externalities, that would notnecessarily lead to greater efficiency. In fact,the economy might become less efficient ifgasoline prices are corrected in isolation ofother prices.95 Transportation economists,including Mark Delucchi, a research scientistat the Institute of Transportation Studies atthe University of California, and CliffordWinston at the Brookings Institution, havedemonstrated that if all inefficiencies intransportation markets were corrected, therewould be more, not less, automobile use thanat present because mass transit “prices” (usercharges) are even more distorted than theprice of automobile travel.96 Hence, a perfectcorrection of gasoline externalities wouldlikely make the economy less, not more, effi-cient to the extent that even more inefficienttransit use increased.

Conclusion

Many economists argue that gasoline taxesshould be increased to internalize externalities

15

A perfect correction ofgasoline externalitieswould likelymake the economyless, not more,efficient to theextent that evenmore inefficienttransit useincreased.

associated with oil use. Some of the concernson which those arguments are based have no(or a very weak) foundation, including con-cerns about future generations, the macroeco-nomic consequences of oil shocks, the “failure”of consumers to conserve, and the four nation-al security arguments. The concerns about theenvironment, accidents, and congestion havemuch more validity.

Regardless, it is striking that the two fre-quently cited researchers in the literature per-taining to motor vehicle externalities—IanParry of Resources for the Future and MarkDelucchi of the Institute of TransportationStudies at the University of California, Davis—agree that internalizing those externalities via agasoline tax should be resisted.97 The greaterfamiliarity one has with the literature, the lessinclined one is to embrace gasoline taxes as aremedy for the externalities associated withmotor vehicle transport.

Gasoline taxes represent a “second-best”means of internalizing the externalities associat-ed with motor vehicle travel. Unfortu-nately,federal gasoline taxes, no matter how carefullyconstructed, always send the wrong signals tomotorists. When addressing road constructionand maintenance costs, for example, they over-charge motorists in low-maintenance, low-con-struction locations and undercharge those inhigh-maintenance, high-growth areas. Whenaddressing pollution costs, they overchargerural motorists and undercharge many urbanmotorists. When addressing congestion, theyovercharge non-peak road users and under-charge peak road use.

In an ideal world, there would be direct exter-nality charges rather than gasoline taxes. Butfirst-best charges levied directly on pollution (viacomputerized internal monitoring equipment),congestion (via tolls), and road use (via usercharges) in lieu of gasoline taxes would almostcertainly be counterproductive for two reasons.

First, the error bars associated with theexternality estimates are very large. Anyattempt to internalize those externalitieswith government-imposed charges risksreducing economic efficiency. That’s particu-larly the case given the poor track record that

legislative bodies have when it comes to“problem-solving” exercises like this.98

Second, correcting motor vehicle externali-ties would likely result in some incrementalincrease in mass transit use, and that wouldmake the economy less efficient because theeconomic distortions induced by mass transituse under the current subsidy regime aregreater than the economic distortions inducedby uninternalized motor vehicle externalities.

Notes1. For a review of the literature, see The Theory ofMarket Failure: A Critical Examination, ed. TylerCowen (Fairfax, VA: George Mason UniversityPress, 1988). See also Charles Wolf, Markets orGovernment: Choosing Between Imperfect Alternatives(Cambridge, MA: MIT Press, 1991).

2. A manifesto of sorts for the Pigou Club is N. GregoryMankiw, “Raise the Gas Tax,” Wall Street Journal,October 20, 2006, p. A12. A roster of inductees into thePigou Club is maintained on Mankiw’s website,http://gregmankiw.blogspot.com/2006/09/ rogoff-joins-pigou-club.html. Journalist Terence Corcoranhas countered by forming the “The NoPigou Club,”http://www.canada.com/nationalpost/features/nopigouclub/index.html.

3. For representative arguments, see Edith Weiss,In Fairness to Future Generations (Dobbs Ferry, NY:Transnational Publishers, 1989); and PaulBarresi, “Beyond Fairness to Future Generations:An Intergenerational Alternative to Intergenera-tional Equity in the International EnvironmentalArena,” Tulane Environmental Law Journal 11, no. 1(1997): 59–88.

4. Robert Hirsch, Roger Bezdek, and Robert Wendling,“Peaking of World Oil Production: Impacts, Mitiga-tion, and Risk Management,” Report Commissionedby the U.S. Department of Energy, February 2005, p. 5;http://www.hilltoplancers.org/stories/hirsch0502.pdf.

5. Recoverable oil deposits within heavy bitumen inthe Venezuelan Orinoco Belt may be nearly equal toSaudi proved reserves. Juan Forero, “For Venezuela,A Treasure in Oil Sludge,” New York Times, June 1,2006, p. C1. Shale rock in the United States is esti-mated to contain three times the amount of petro-leum found in proved Saudi reserves. James Bartiset al., “Oil Shale Development in the United States:Prospects and Policy Issues” prepared by the RandCorporation for the National Energy TechnologyLaboratory of the U.S. Department of Energy, 2005.For an overview of unconventional petroleum

16

It is striking thatthe two

frequently citedresearchers in the

literature pertaining to

motor vehicleexternalities agreethat internalizing

those externalitiesvia a gasoline tax

should be resisted.

resources, see Robert L. Bradley and RichardFulmer, Energy: The Master Resource (Dubuque, IA:Kendall/Hunt, 2004).

6. For a review of the literature, see Robert Arnott,“Supply Side Aspects of Depletion,” Journal ofEnergy Literature 8, no.1 ( June 2002): 3–21. For ageneral critique regarding past and present worriesabout a near-term peak in global oil production, seeVaclav Smil, Energy at the Crossroads (Cambridge,MA: MIT Press, 2003), pp. 181–213. For a witheringcritique of current “peak oil” hypotheses, seeMichael Lynch, “Crop Circles in the Dessert: TheStrange Controversy over Saudi Oil Production,”International Research Center for Energy andEconomic Development, Occasional Paper 40,Boulder, Colorado, 2006; and “Crying Wolf: PeakOil Alarmism: Fact or Fiction?” Blue Book, CLSAAsia-Pacific Markets, March 2007.

7. Steven Landsburg, “Tax the Knickers off Your Grand-children,” Slate, March 7, 1997, http://www.slate.com/d/2036/.

8. Jerry Taylor and Peter Van Doren, “The Case againstthe Strategic Petroleum Reserve,” Cato Institute PolicyAnalysis no. 555, November 21, 2005.

9. Jagadeesh Gokhale, “Is the Fed Facilitating anUnpleasant Fiscal Arithmetic?” Cato Journal 27,no. 2 (forthcoming): Table 2.

10. From 2000–2006, the short run-price elasticityfor gasoline ranged from -0.034 to -0.077, that is, a10 percent increase in gasoline prices would reduceconsumption by only about three-tenths of 1 per-cent to about eight-tenths of 1 percent. This sug-gests that consumers have become less responsiveto gasoline price increases over time, because datafrom 1975–80 reveal that short-run price elasticityover that time ranged from -0.21 to -0.34. JonathanHughes, Christopher Knittel, and Daniel Sperling,“Evidence of a Shift in the Short-Run PriceElasticity of Gasoline Demand,” CSEM WP-159,Center for the Study of Energy Markets, Universityof California Energy Institute, September 2006.

11. Between March 1982 and January 2002, the stan-dard deviation in monthly oil prices was 29.5 percentof its mean. By means of comparison, the standarddeviation in monthly steel prices was 9.2 percent. Oilprices fluctuate more than, or at least as much as, themost volatile commodity prices in other sectors.Hillard Huntington, “Energy Disruptions, InterfirmPrice Effects, and the Aggregate Economy,” OP 51,September 2002, p. 4, http://www.stanford.edu/group/EMF/publications/doc/op51.pdf.

12. Estimates in the literature vary from -.3 to -.9. SeeRobert M. Ames, Anthony Corridore, and Paul W.MacAvoy, “National Defense, Oil Imports, and Bio-

Energy Technology,” Journal of Applied CorporateFinance 16, no. 1 (Winter 2004): 48; Ian W. H. Parryand Kenneth A. Small, “Does Britain or the UnitedStates Have the Right Gasoline Tax?” Resources forthe Future Discussion Paper 02-12, revisedSeptember 2004, p. 22; and Gilbert E. Metcalf,“Federal Tax Policy Towards Energy,” NBERWorking Paper 12568, October 2006, p. 30.

13. Paul Leiby, et al., “Oil Imports: An Assessmentof Benefits and Costs,” ORNL-6851, Oak RidgeNational Laboratory, 1997.

14. Ben Bernanke, Mark Gertler, and Mark Watson,“Systematic Monetary Policy and U.S. AggregateEconomic Activity,” Brookings Papers on EconomicActivity no.1 (1997): 91–142; and James Hamiltonand Anna Maria Herrera, “Oil Shocks andAggregate Macroeconomic Behavior,” Journal ofMoney, Credit, and Banking 36 (April 2004): 265–86.

15. Donald Jones, Paul Leiby, and Inja Paik, “OilShocks and the Macroeconomy: What Has BeenLearned Since 1996,” Energy Journal 25, no. 2(2004): 1–32; and Robert Barsky and Lutz Kilian,“Oil and the Macroeconomy since the 1970s,”National Bureau of Economic Research, WorkingPaper 10855, October 2004.

16. Rajeev Dhawan and Karsten Jeske, “HowResilient Is the Modern Economy to Energy PriceShocks?” Federal Reserve Bank of Atlanta EconomicReview 91, no.3 ( Third Quarter, 2006): 21–32. Anew paper by U. Michigan economist Lutz Kilianattempts to separate oil shocks into three cate-gories: changes in oil supply, changes in oildemand resulting from changes in overall eco-nomic activity, and changes in oil demand result-ing from precautionary inventory buildup (fearsof future disruption). He argues that since 1975all oil price increases have been result of increasesin overall economic activity or precautionaryinventory buildup. Actual reductions in crude oilsupply have played a small role. The price increasesince 2003 is the result of general economicgrowth rather than precautionary inventorydemand. Lutz Kilian, “Not All Oil Price ShocksAre Alike: Disentangling Demand and SupplyShocks in the Crude Oil Market,” Social ScienceResearch Network, February 2007, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=975262.

17. David Walton, “Has Oil Lost the Capacity toShock?” Bank of England Quarterly Bulletin 46, no. 1(Spring 2006): 105–114, http://www.bankofeng land.co.uk/publications/quarterlybulletin/qb060109.pdf.

18. Eric Fisher and Kathryn Marshall, “TheAnatomy of an Oil Price Shock,” EconomicCommentary, Federal Reserve Bank of Cleveland,November 2006.

17

19. See Joseph Kalt, The Economics and Politics of OilPrice Regulation (Cambridge: MIT Press 1981); PeterVan Doren, Politics, Markets, and Policy Choices (AnnArbor: University of Michigan Press, 1991); andJerry Taylor and Peter Van Doren, “EconomicAmnesia: The Case against Oil Price Controls andWindfall Profit Taxes,” Cato Institute PolicyAnalysis 561, January 12, 2006.

20. Taylor and Van Doren, “Economic Amnesia.”

21. For representative arguments, see David Greene,“Why CAFE Worked,” Energy Policy 26, no.8 (1998),pp. 595–614.

22. Molly Espey, “Do Consumers Value FuelEconomy?” Regulation 28, no. 4 (Winter 2005–06),pp. 8–10.

23. Mark Dreyfus and W. Kip Viscusi, “Rates ofTime Preference and Consumer Valuations ofAutomobile Safety and Fuel Efficiency,” Journal ofLaw & Economics 38, no. 1 (April 1995): 79–105.

24. For critiques of CAFE standards see Andrew N.Kleit, “CAFE Changes, By the Numbers,” Regulation25, no. 3 (Fall 2002): 32–35; and Paul Portney et al.,“The Economics of Fuel Economy Standards, Journalof Economic Perspectives 17, no. 4 (Fall 2003): 214–15.

25. This argument was first made by Arthur C.Pigou, The Economics of Welfare (London: Macmillan,1920).

26. Some economists have argued that if governmentwere to acknowledge and enforce private propertyrights over environmental resources, most environ-mental regulations would be unnecessary and thatthe resulting legal regime would be more efficientthan the current regulatory regime. MurrayRothbard, “Law, Property Rights, and Air Pollution,”Cato Journal 2, no 1 (Spring 1982): 55–100. But mostbelieve that the costs associated with policing privateenvironmental rights for mobile sources (“transac-tion costs” in economic parlance) would be (prohibi-tively?) large.

27. Passenger vehicles are so heavily computerizedtoday that the technology necessary to collect emis-sions data for tax purposes is already onboard alarge percentage of the U.S. auto fleet. If the trans-action costs associated with monitoring, collecting,and billing motorists for tailpipe emissions provetoo large, then a fuel tax might well be a superiormeans of internalizing the environmental external-ities associated with automobile travel. While thatmay have been true in the past, it is unlikely to betrue today.

28. Joel Schwartz, No Way Back: Why Air PollutionWill Continue to Decline (Washington: American

Enterprise Institute, 2003), p. 23.

29. Ian Parry, “The Uneasy Case for Higher GasolineTaxes,” Milken Institute Review 4 (2005): 43, contendsthat 60 percent of the reduction in gasolinedemand from any increase in gasoline taxes willcome from vehicle mileage improvements.

30. Fuel efficiency reduces the marginal costsassociated with driving, and reducing marginalcosts will induce some demand response. DavidGreene, James Kahn, and Robert Gibson, “FuelEconomy Rebound Effect for U.S. HouseholdVehicles,” Energy Journal 20, no. 3 (1999): 6–10.

31. Carolyn Fischer, Winston Harrington, and IanParry, “Economic Impacts of Tightening theCorporate Average Fuel Efficiency Standards,” reportprepared for the U.S. Environmental ProtectionAgency and the National Highway SafetyAdministration, 2005; J. Daniel Khazoom, “GasolineConservation Versus Pollution Control: UnintendedConsequences, Continued,” Journal of Policy Analysisand Management 7, no. 4 (1988): 710–14, and “TheImpact of a Gasoline Tax on Auto ExhaustEmissions,” Journal of Policy Analysis and Management10, no.3 (1991): 434–54.

32. If EPA regulated tailpipe emissions based onemissions per gallon rather than on emissions permile traveled, a 9 percent increase in the gasoline taxwould reduce emissions to the same extent that adoubling of the gasoline tax would have under thecurrent regulatory regime. Khazoom, p. 438.

33. For a review of the literature, see Ian Perry andKenneth Small, "Does Britain or the United StatesHave the Right Gasoline Tax?" American EconomicReview 95 (September 2005): 1276–89.

34. Ian Parry, Margaret Walls, and WinstonHarrington, “Automobile Externalities and Policies,”Resources for the Future Discussion Paper 06-26, June2006, p. 4, gives a range for local pollution of 1.6–18.6cents per mile based upon an analysis from the U.S.Federal Highway Administration. Assuming 22.6miles per gallon (Parry and Small 2005, Table 1), thattranslates into environmental costs of $.36 to $4.20per gallon. The “best estimate” from the U.S. FederalHighway Administration, however, is that local envi-ronmental costs total 2.2 cents per mile, a figure quiteclose to the 2.3 cents per mile embraced by KennethSmall and Camilla Kazimi, “On the Costs of AirPollution from Motor Vehicles,” Journal of TransportEconomics and Policy 29, (1995): 7–32. That suggests aplausible mean estimate of about 50 cents per gallon.

35. Schwartz, No Way Back. Tailpipe emissions willdecline rapidly with the new Tier-2 emissionsstandards. That, combined with the improveddurability of pollution control equipment, sug-

18

gests that this externality will decrease in impor-tance over time.

36. For an overview of the debate and the uncertaintiestherein, see Donald McCubbin and Mark Delucci,“The Social Cost of the Health Effects of Motor-Vehicle Air Pollution,” Report 11 in the series “TheAnnualized Social Cost of Motor Vehicle Use in theUnited States, based on 1990–91 Data,” UCD-ITS-RR-96-3(11), Institute of Transportation Studies,University of California, Davis, August 1996, pp.22–80, particularly pp. 22–26, http://www.its.ucdavis.edu/publications/1996/UCD-ITS-RR-96-03(11).pdf.For a review of the epidemiological problems associat-ed with mortality estimates from particulate matter—the most important of the environmental externalitiesassociated with vehicle emissions—see RobertCrandall, Frederick Reuter, and Wilbur Steger,“Clearing the Air: EPA’s Self Assessment of Clean AirPolicy,” Regulation 19, no.4 (Fall 1996): 35–46; andRandall Lutter and Richard Belzer, “EPA Pats Itself onthe Back,” Regulation 23, no. 3 (Fall 2000): 23–28; andJoel Schwartz, “Comments on EPA’s Proposed Rule,National Ambient Air Quality Standards forParticulate Matter,” Docket ID EPA-HQ-OAR-2001–0017, April 17, 2006, http://www.aei.org/ publi-cations/filter.all,pubID.24240/pub_detail.asp. For asummary of other health-related air pollution contro-versies, see Joel Schwartz, “Air Pollution and Health:Do Popular Portrayals Reflect the ScientificEvidence?” Environmental Policy Outlook 2 (2006).

37. R. Clarkson and K. Deyes, “Estimating theSocial Cost of Carbon Emissions,” EnvironmentalProtection Economics Division, Department ofEnvironment, Food and Rural Affairs, London,August 2001, cited in David Newberry, “Why TaxEnergy? Towards a More Rationale Policy,” TheEnergy Journal 26, no. 3 (2005): 21.

38. Cass Sunstein, Risk and Reason: Safety, Law, andthe Environment (New York: Cambridge UniversityPress, 2002), Table 7.5, p. 174.

39. David Newberry, “Fair Payment from Road Users:A Review of the Evidence on Social andEnvironmental Costs,” Automobile Association,February 1998, cited in Newberry, “Why Tax Energy?Towards a More Rationale Policy,” The Energy Journal26, no. 3 (2005): 24.

40. Parry and Small 2005, p. 1283, Table 1. A recentstudy by economist Nicholas Stern suggests that anoptimal climate change externalities tax would be$85 per metric ton of carbon equivalent, which worksout to about a 16 cent per gallon tax on gasoline.Nicholas Stern, The Economics of Climate Change(London: Cambridge University Press, January 2007,forthcoming), http://www.hm-treasury.gov.uk/independent_reviews/stern_review_economics_climate_change/stern_review_report.cfm. Parry and Small’s

review of the literature, however, reveals that Stern’sestimate is at the high end of the literature. See, forinstance, William Nordhaus and Joseph Boyer,Warming the World: Economic Models of Global Warming(Cambridge, MA: MIT Press, 2000), which estimatesthat marginal damages associated with global warm-ing work out to about $15 per metric ton of carbon,or 4 cents per gallon of gasoline.

41. Parry and Small 2005, p. 1283, Table 1.

42. Aaron S. Edlin and Pinar Karaca-Mandic,“The Accident Externality from Driving,” Journalof Political Economy 114, no.5 (October 2006): 951.

43. Variable toll congestion pricing now exists instate route 91 in Orange County California, I-15 inSan Diego, I-394 in Minneapolis, and I-25 inDenver. See Timothy Egan, “Paying on the Highwayto Get Out of First Gear,” New York Times, April 28,2005, p. A1. For a discussion of optimal congestionpricing see Kenneth Small, Clifford Winston, andCarol Evans, Road Work (Washington: Brookings,1989); and Kenneth Small, Clifford Winston, andJia Yan, “Differentiated Road Pricing, ExpressLanes, and Carpools: Exploiting HeterogenousPreferences in Policy Design,” AEI-Brookings JointCenter Working Paper 06-06-02, March 2006.

44. Ian Parry, in fact, believes that such innovationsmay emerge in the market with limited governmenthelp. Parry, Walls, and Harrington, pp. 29–30.

45. Although some economists have argued that atax on vehicle miles traveled (VMT) alone wouldinternalize the accident externality (see, forinstance, Todd Litman, “Distance-Based VehicleInsurance as a TDM Strategy,” TransportationQuarterly 51 [Summer 1997]: 119–37; and Edlinand Karaca-Mandic, pp. 952–53), statistical analysisfinds little correlation between VMT and automo-bile accidents. Oddly enough, a VMT tax mighthave the perverse effect of actually increasing the per-centage of bad drivers on the road because it woulddisproportionately affect good drivers. If accidentsare more closely correlated with bad driving thanwith VMT, a tax on the latter would likely increaseautomobile accidents. Clifford Winston andVikram Maheshri, “Towards an Efficient Policy forReducing Automobile Accidents,” Working Paper,April, 2007 (available from authors).

46. Parry and Small 2005, p. 1276.

47. The figure is a weighted average of existing statetaxes plus the federal fuels tax (U.S. Department ofTransportation, Federal Highway Administration,Office of Highway Policy Information, 2005).

48. As a consequence of road use charges, 60, 000(16 percent) fewer vehicles a day entered central

19

London, resulting in 30 percent less congestion.National Commission on Energy Policy, “Conges-tion Charging: Solutions for the Escalating Prob-lem of Vehicle Miles Traveled,” NCEP BackgroundPaper, http://www.energycommission.org/files/finalReport/III.5.a%20-%20Congestion%20Charging-VMT.pdf, pp. 1–2.

49. Leila Abboud and Jenny Clevstrom, “Stockholm’sSyndrome,” Wall Street Journal, August 29, 2006, p. B1.

50. Many economists doubt that there are significantnational security externalities associated with gaso-line consumption. See Douglas Bohi and MichaelToman, The Economics of Energy Security (Norwell, MA:Kluwer Academic Publishers, 1996).

51. Andrew Bacevich, “The Real World War IV,”Wilson Quarterly 29, no. 1 (Winter 2005).

52. Congressional Research Service, “Oil Imports:An Overview and Update of Economic and SecurityEffects,” CRS Report for Congress, 98-1, December 12,1997, Table A-1.

53. Mark Delucchi and James Murphy, “U.S.Military Expenditures to Protect the Use of Persian-Gulf Oil for Motor Vehicles,” Report 15 in the seriesThe Annualized Social Cost of Motor-Vehicle Use in theUnited States, based on 1990–1991 Data, UCD-ITS-RR-96-3(15) rev. 2, April 1996, rev. October 2006.

54. J. Robinson West, “Saudi Arabia, Iraq, and theGulf,” in Energy Security, ed. Jan Kalicki and DavidGoldwyn (Washington: Woodrow Wilson CenterPress, 2005), pp. 197–218.

55. Robert Jervis, “Why the Bush Doctrine CannotBe Sustained,” Political Science Quarterly 120, no.3(Fall 2005): 35177.

56. Saudi Arabia and Kuwait paid approximately$33 billion (55 percent) toward the total cost ofDesert Storm and Desert Shield, which was $60billion. The U.S. share was only $6 billion (10 per-cent). U.S. Department of Defense, news release125-M, May 5, 1992.

57. The claim that military costs are fixed ratherthan oil related is discussed in Portney et al.

58. Jonathan Rauch, “A Higher Gas Tax Is theAnswer. Who’ll Ask the Question?” National Journal,February 9, 2002.

59. This is such an obvious point that energy econ-omists rarely bother to explore the issue in detail.To understand how the world crude oil marketworks is to understand that embargoes are unen-forceable. See Philip Verleger, Adjusting to VolatileEnergy Prices (Washington: Institute for Internation-

al Economics, 1993) and M. A. Adelman, The Genieout of the Bottle: World Oil Since 1970 (Cambridge, MA:MIT Press, 1995).

60. Energy Information Administration, AnnualEnergy Review 2004, Table 5.3.

61. Thomas Lee, Ben Ball Jr,. and Richard Tabors,Energy Aftermath (Boston: Harvard BusinessSchool, 1990), p. 17.

62. Ibid., p. 30. See also Edward Fried, “Oil Security: AnEconomic Phenomenon,” in Oil and America’s Securi-ty, ed. Edward Fried and Nanette Blandin (Washing-ton: Brookings Institution, 1988), pp. 56–59.

63. Cited in Robert L. Bradley Jr., The Mirage of OilProtection (Lanham, MD: University Press ofAmerica, 1989), p. 140. For similar arguments, seeFrancisco Parra, Oil Politics: A Modern History ofPetroleum (New York: I. B. Tauris, 2004), pp. 184–85.

64. Adelman, p. 31. Former OPEC Secretary-GeneralFrancisco Parra makes the same point in Parra.

65. Philip Auerswald, “The Irrelevance of the MiddleEast,” American Interest, May/June 2007, p. 22.

66. See Steven Lee Meyers, “Russian Gas CompanyPlans Steep Price Increase for Georgia,” New YorkTimes, November 3, 2006, p. A12; and Joseph Kahn,“China May Be Using Oil to Press North Korea,”New York Times, October 31, 2006, p. A12.

67. Bin Laden has said on many occasions that hethinks the Saudi monarchy keeps oil prices belowtrue market value in order to maintain friendlyrelations with the West.

68. Oil revenues are 40–50 percent of Iranian gov-ernment revenues and 70–80 percent of Saudi gov-ernment revenues. See Energy Information Admin-istration, “Country Analysis Briefs,” http://www.eia.doe.gov/emeu/cabs/contents.html. Iran’s oil out-put increased steadily from 3.7 mbd in 2003 to 4.1mbd in 2005. Energy Information Administration,International Petroleum Monthly, Table 4.1a

69. Data on Iranian production in 1978 and Saudiproduction in 2006 from the Energy InformationAdministration; http://tonto.eia.doe.gov/merquery/mer_data.asp?table=T11.01a and http://tonto.eia.doe.gov/merquery/mer_data.asp?table=T11. 01b.

70. In 1978 the U.S. used 15,950 BTUs per ($2000)dollar of GDP but only 8,970 BTUs per ($2000) dol-lar of GDP in 2005, a reduction of 43.8 percent. Andthe BTUs used in 2005 came less from petroleumthan in 1978 (47.5 percent of 1978 energy consump-tion was petroleum versus only 40.5 percent in 2005).Energy Information Administration, Annual Energy

20

Review 2005, Tables 1.3 and 1.5, pp. 9 and 13.

71. Calculation from Lundberg Survey, “FatallyFlawed Premise: Why Anti-Oil Weapon in War onTerror Won’t Work,” Energy Détente 27, no. 11(November 30, 2006).

72. Data on international Islamic terrorism incidentsand fatalities were taken from the Memorial Institutefor the Prevention of Terrorism, Terrorism Know-ledge Base, an interactive website maintained by theMIPT, http://www.tkb.org/. Data on that websitecomes from the RAND Terrorism Chronology andRAND-MIPT Terrorism Incident databases; theTerrorism Indictment database; and DFI Inter-national’s research on terrorist organizations.Nominal Saudi oil prices were obtained from EnergyInformation Administration, Annual Energy Review2005, p. 169, Table 5.19, “Landed Costs of CrudeImports From Selected Countries,” and deflatedwith the GDP deflator. Unit root tests suggested thatfatalities and Saudi oil prices had unit roots but ter-rorist incidents did not, so the former were first dif-ferenced before the regressions. Even after first differ-encing, auto correlation existed so autoregressiveterms were added to each regression, which furtherweakened the insignificant relationships.

73. Mark Basile, “Going to the Source: Why AlQaeda’s Financial Network Is Likely to Withstandthe Current War on Terrorist Financing,” Studiesin Conflict and Terrorism 27 (2004): 169–85.

74. Although little is known about funding trendsassociated with Iranian support for Hezbollah, theIranian government probably spends no more than$25–50 million on Hezbollah a year. AnthonyCordesman, “Iran’s Support for Hezbolla inLebanon,” Center for Strategic and InternationalStudies, July 15, 2006, p. 3. Even less is known aboutSaudi contributions to Islamic extremism. SeeAlfred Prados and Christopher Blanchard, “SaudiArabia: Terrorist Financing Issues,” RL32499, CRSReport for Congress, Congressional ResearchService, Updated December 8, 2004.

75. Lundberg Survey, p. 8.

76. For a brief review of the academic literature onthis subject, which is somewhat mixed, see PaulStevens, “Resource Impact: Curse or Blessing? ALiterature Survey,” The Journal of Energy Literature9, no. 1 (June 2003): 22–24.

77. This is a crude calculation based on Ian Parry’sestimate that a $20 per barrel oil tax would reducelong-term world crude oil prices by 1–5 percent. SeeIan Parry, “The Case for a Pay-by-the-Barrel OilTax,” Resources (Fall 2006 27). A $1 hike in the gastax is the equivalent of imposing a tax on crude of$42 per barrel if all the oil were used for gasoline,

but because only 44 percent of the oil consumed inthe United States becomes gasoline (http://www.eia.doe.gov/neic/infosheets/petroleumproducts.html), the tax is equivalent to $18.50, or close toParry’s original estimates of $20, so we used the 1 to5 percent estimate and $65 a barrel crude. In 1998,the price of oil was $13 per barrel in constant (2000)dollars. Average refiner acquisition costs in 2006were about $60 per barrel, or $52 in constant (2000)dollars. To reduce prices from $52 to $13 is a reduc-tion of 75 percent, which would require a 30-foldincrease relative to Parry’s calculation if the tax wereon crude oil (30 * 2.5 percent, the middle of the 1–5range, is 75 percent), or 60-fold on gasoline or about$28 per gallon.

78. Energy Information Administration, InternationalPetroleum Monthly, May 8, 2007, http://www.eia.doe.gov/emeu/ipsr/t22.xls.

79. Charles Ballard, J. Shoven, and J. Whalley,“General Equilibrium Computations of the Margin-al Welfare Costs of Taxes in the United States,”American Economic Review 75 (1985): 128–38.

80. This is known as the “strong” double-dividendclaim. Weaker versions of the double-dividendclaim exist, but they are not relevant to the case fora major swap of more distortionary taxes in favorof less distortionary energy taxes. For a discussionof weaker double-dividend claims, see LawrenceGoulder, “Environmental Taxation and the‘Double Dividend’: A Reader’s Guide,” NationalBureau of Economic Research Working Paper4896, October 1994, pp. 4–8.

81. Parry and Small 2005, p. 1286.

82. Stanford economist Lawrence Goulder, forinstance, has written one of the seminal papers crit-icizing the strong double-dividend claim despitethe fact that he supports raising the gasoline taxand is a member of Greg Mankiw’s “Pigou Club.”Weaker versions of the double-dividend claim,however, are on much firmer ground. See Goulder.

83. A. Lans Bovenberg and Ruud de Mooij,“Environmental Levies and Distortionary Taxation,”American Economic Review, September, 1994, pp.1085–1089. Bovenberg and de Mooij conclude thatthe double dividend argument holds only if theuncompensated wage elasticity of labor supply isnegative, but empirical investigations of the labormarket find that this is not the case.

84. Alan Auerbach and Laurence Kotlikoff,Dynamic Fiscal Policy (Cambridge, England:Cambridge University Press, 1987).

85. The fact that taxing capital creates more wel-fare loss than taxing labor explains why many

21

economists support a switch from income taxesto consumption taxes. Dale Jorgenson and Kun-Young Yun, Tax Policy and the Cost of Capital(Oxford: Oxford University Press, 1990).

86. Goulder, pp. 15-16; Bovenberg and de Mooij;and A. Lans Bovenberg and Laurence Goulder,“Optimal Environmental Taxation in the Presenceof Other Taxes: An Applied Equilibrium Analysis,”Stanford University Working Paper, May, 1994.

87. Wallace Oates and Ian Parry, “Policy Analysisin the Presence of Distorting Taxes,” Journal ofPolicy Analysis and Management 19 (Fall 2000)603–14.

88. Stephen Smith, “Environmental and PublicFinance Aspects of the Taxation of Energy,” OxfordReview of Economic Policy 14, no. 4 (1998): 80–81.

89. See for instance Laurence Kotlikoff, “TheEconomic Impact of Replacing Federal IncomeTaxes with a Sales Tax,” Cato Institute Policy Analysis193, April 15, 1993.

90. Smith, p. 82.

91. For explicit concessions regarding the second-best nature of gasoline taxes, see Parry, “The UneasyCase for Higher Gasoline Taxes,” pp. 36–45;Newberry, “Why Tax Energy?” pp. 23–26, 29–31,and 35; and Metcalf.

92. See, for instance, Mankiw, p. A12.

93. Smith, p. 69.

94. Bovenberg and Goulder, and Ian Parry,“Pollution Taxes and Revenue Recycling,” WorkingPaper, Economic Research Service, U.S. Departmentof Agriculture, April 1994, cited in Goulder, p. 30.

95. This is a well established and uncontroversialconcept known in economics as “the general theo-ry of the second best.” See R. G. Lipsey and KelvinLancaster, “The General Theory of the SecondBest,” Review of Economic Studies 24, no. 1 (1956):11–33; and O. Davis and A. Whinston, “WelfareEconomics and the Theory of the Second Best,”Review of Economic Studies 32 (1965): 1–14. For acontemporary summary of the theory, see Jean-Jacques Laffont, Fundamentals of Public Economics(Cambridge, MA: MIT Press, 1990), p. 167.

96. Mark Delucchi, “Should We Try to Get thePrices Right?” Access Magazine 16 (Spring 2000): 17.Clifford Winston and Chad Shirley, Alternate Route(Washington: Brookings Institution, 1998), p. 58.

97. Parry, Walls, and Harrington; and Delucchi,pp. 14–21.

98. David Mayhew, “Congress as Problem Solver,”in Promoting the General Welfare: New Perspectives onGovernment Performance, ed. Alan Gerber and EricPatashnik (Washington: Brookings Institution,2006), pp. 219–36. Clifford Winston, GovernmentFailure versus Market Failure (Washington: BrookingsInstitution, 2006).

22

OTHER STUDIES IN THE POLICY ANALYSIS SERIES

597. Medicaid’s Soaring Cost: Time to Step on the Brakes by JagadeeshGokhale (July 19, 2007)

596. Debunking Portland: The City That Doesn’t Work by Randal O’Toole(July 9, 2007)

595. The Massachusetts Health Plan: The Good, the Bad, and the Ugly by David A.Hyman (June 28, 2007)

594. The Myth of the Rational Voter: Why Democracies Choose Bad Policiesby Bryan Caplan (May 29, 2007)

593. Federal Aid to the States: Historical Cause of Government Growth andBureaucracy by Chris Edwards (May 22, 2007)

592. The Corporate Welfare State: How the Federal Government Subsidizes U.S.Businesses by Stephen Slivinski (May 14, 2007)

591. The Perfect Firestorm: Bringing Forest Service Wildfire Costs under Control by Randal O’Toole (April 30, 2007)

590. In Pursuit of Happiness Research: Is It Reliable? What Does It Imply for Policy? by Will Wilkinson (April 11, 2007)

589. Energy Alarmism: The Myths That Make Americans Worry about Oil by Eugene Gholz and Daryl G. Press (April 5, 2007)

588. Escaping the Trap: Why the United States Must Leave Iraq by Ted Galen Carpenter (February 14, 2007)

587. Why We Fight: How Public Schools Cause Social Conflict by Neal McCluskey (January 23, 2007)

586. Has U.S. Income Inequality Really Increased? by Alan Reynolds (January 8,2007)

585. The Cato Education Market Index by Andrew J. Coulson with advisersJames Gwartney, Neal McCluskey, John Merrifield, David Salisbury, andRichard Vedder (December 14, 2006)

584. Effective Counterterrorism and the Limited Role of Predictive Data Mining by Jeff Jonas and Jim Harper (December 11, 2006)

583. The Bottom Line on Iran: The Costs and Benefits of Preventive Warversus Deterrence by Justin Logan (December 4, 2006)

582. Suicide Terrorism and Democracy: What We’ve Learned Since 9/11 by Robert A. Pape (November 1, 2006)

581. Fiscal Policy Report Card on America’s Governors: 2006 by Stephen Slivinski (October 24, 2006)

580. The Libertarian Vote by David Boaz and David Kirby (October 18, 2006)

579. Giving Kids the Chaff: How to Find and Keep the Teachers We Needby Marie Gryphon (September 25, 2006)

578. Iran’s Nuclear Program: America’s Policy Options by Ted Galen Carpenter(September 20, 2006)

577. The American Way of War: Cultural Barriers to Successful Counterinsurgency by Jeffrey Record (September 1, 2006)

576. Is the Sky Really Falling? A Review of Recent Global Warming Scare Stories by Patrick J. Michaels (August 23, 2006)

575. Toward Property Rights in Spectrum: The Difficult Policy ChoicesAhead by Dale Hatfield and Phil Weiser (August 17, 2006)

574. Budgeting in Neverland: Irrational Policymaking in the U.S. Congress and What Can Be Done about It by James L. Payne (July 26, 2006)

573. Flirting with Disaster: The Inherent Problems with FEMA by Russell S. Sobel and Peter T. Leeson (July 19, 2006)

572. Vertical Integration and the Restructuring of the U.S. ElectricityIndustry by Robert J. Michaels (July 13, 2006)

571. Reappraising Nuclear Security Strategy by Rensselaer Lee (June 14, 2006)

570. The Federal Marriage Amendment: Unnecessary, Anti-Federalist, and Anti-Democratic by Dale Carpenter (June 1, 2006)

576. Is the Sky Really Falling? A Review of Recent Global Warming Scare Stories by Patrick J. Michaels (August 23, 2006)

561. Economic Amnesia: The Case against Oil Price Controls and Windfall Profit Taxes by Jerry Taylor and Peter Van Doren (January 12, 2006)

555. The Case against the Strategic Petroleum Reserve by Jerry Taylor and Peter Van Doren (November 21, 2005)

Untitled-2 2 2/7/06 4:35:00 PM