Gas Tax Economics
Reviewing Mankiw's proposal

20 Oct 2006 in Regulatory Economics, Regulatory Policy

Harvard economics professor and former Bush administration chief economist Gregory Mankiw says in the Wall Street Journal (subscription required) that Congress should raise the gas tax.

With the midterm election around the corner, here's a wacky idea you won't often hear from our elected leaders: We should raise the tax on gasoline. Not quickly, but substantially. I would like to see Congress increase the gas tax by $1 per gallon, phased in gradually by 10 cents per year over the next decade.

He gives seven arguments for his proposed $1 per gallon increase in the gas tax: (1) carbon dioxide abatement, (2) reducing road congestion, (3) relief from counterproductive regulations, (4) balancing the federal budget, (5) burden-sharing with oil producers such as Saudi Arabia and Venezuela, (6) a preference for consumption over income taxes, and (7) enhanced national security.

Which of these arguments stands up to elementary economic scrutiny?

Herewith, a short economic analysis of Mankiw's gas tax proposal:

First, it's important to remember that gasoline already is taxed. The federal government levies a tax of 18.4 cents per gallon. State motor gasoline taxes vary from a low of 7.5 cents per gallon in Georgia to a high of 34.0 cents per gallon in Washington (21.3 cents per gallon is the state average). These figures are different for other motor fuels, exclude county taxes, local option taxes, and sales taxes where such taxes apply to gasoline. Gas taxes are typically justified as user fees for highway construction and maintenance, but revenues can be used for any purpose. (Money is fungible, so even highway "trust" funds that legally cannot be spent on other purposes can supplant appropriations from federal and state general funds.)

In the analysis which follows, we assume for simplicity that existing gasoline taxes are not intended to internalize any of the externalities identified by Mankiw. That is, we are looking at Mankiw's proposal strictly within the boundaries that he himself has set.

(1) CARBON DIOXIDE ABATEMENT

Mankiw writes:

The burning of gasoline emits several pollutants. These include carbon dioxide, a cause of global warming. Higher gasoline taxes, perhaps as part of a broader carbon tax, would be the most direct and least invasive policy to address environmental concerns.

Burning gasoline does emit "several pollutants:" including carbon monoxide and the ozone-forming volatile organic compounds and nitrogen oxides. But these are tightly regulated at both gasoline refining and the tailpipe, and residual damages from the remaining air pollution externally are likely to be trivial.

Mankiw isn't after these so-called criteria air pollutants but carbon dioxide, a non-air pollutant responsible for an unknown share of global warming. Burning a gallon of gasoline emits about 8.78*10^-3 metric ton of CO2. Carbon dioxide offsets cost no more than $10 per metric ton. Therefore, a gas tax of $.09 per gallon fully internalizes the CO2 externality. Mankiw's proposed $1 per gallon gas tax is ten times larger. See below for details.

If the purpose of Mankiw's gas tax is to get market prices in line with social cost, he overshoots by a substantial amount and causes consumers to buy too little gasoline, not too much.

Subsequently in his op-ed, Mankiw notes that even after his proposed tax US gas prices would "still be less than half the level in, say, Great Britain, which last I checked is still a democracy." This comparison is inapt if Mankiw's purpose is economic efficiency, for there is no basis for assuming that UK fuel policies are efficiency-based. The BBC says UK fuel taxes were "designed as a means both to raise money and discourage car use on environmental grounds." UK fuel duties rise at the rate of inflation plus 3%. UK also impose a value added tax [VAT] on both fuel and the fuel duty -- that is, a tax on a tax. Moreover, fuel taxes in Britain have been controversial, leading to blockades of refineries by truck drivers and farmers and an ongoing protest movement.

(2) REDUCING ROAD CONGESTION

Mankiw writes:

Every time I am stuck in traffic, I wish my fellow motorists would drive less, perhaps by living closer to where they work or by taking public transport. A higher gas tax would give all of us the incentive to do just that, reducing congestion on streets and highways.

Like global climate change, road congestion is indeed an externality resulting from motor vehicle use. But it is a fundamentally very different one because it's dependent on the time and place gasoline is burned. In urban areas, of course, road congestion is greatest at rush hour and absent in the middle of the night. But in rural areas, it simply does not exist. That makes a gas tax a very odd and clumsy tool for internalizing the externality of road congestion.

If internalizing road congestion is the objective, there are much better policy tools available such as tolls, congestion pricing, and HOT lanes. As policy tools for congestion management go, the gas tax is a sledgehammer. Policy makers could just as easily follow Will Rogers' advice and remove from the highways all cars that aren't paid for.

(3) RELIEF FROM COUNTERPRODUCTIVE REGULATIONS

Mankiw writes:

Congress has tried to reduce energy dependence with corporate average fuel economy standards. These CAFE rules are heavy-handed government regulations replete with unintended consequences: They are partly responsible for the growth of SUVs, because light trucks have laxer standards than cars. In addition, by making the car fleet more fuel-efficient, the regulations encourage people to drive more, offsetting some of the conservation benefits and exacerbating road congestion. A higher gas tax would accomplish everything CAFE standards do, but without the adverse side effects.

Mankiw is correct that CAFE created the market for SUVs; that's only part of the story. CAFE made the family station wagon a fuel-efficiency pariah. Automakers invented minivans and SUVs to simultaneously escape CAFE and provide the family-friendly cars consumers still wanted but their elected representatives said they shouldn't have. It says something about the effectiveness and efficiency of regulation -- and the degree of genuine public support for it -- when both industry and consumers gladly conspire to evade it.

Mankiw also blames CAFE for encouraging people to drive more, but this complaint has limited appeal. In whatever manner greater fuel efficiency is achieved -- whether by government regulation, secular technological change, or market forces -- it always will encourage more driving. If the current generation of hybrids really do provide significant fuel economy gains, they should be expected to encourage owners to drive more.

If regulatory relief is a genuine purpose Mankiw's gas tax proposal, what's missing is an explicit statement linking the enactment of the tax to CAFE repeal. He cannot claim the banner of regulatory relief for his gas tax if is destined to be layered on top of CAFE instead of replacing it.

(4) BALANCING THE FEDERAL BUDGET

Mankiw writes:

Everyone who has studied the numbers knows that the federal budget is on an unsustainable path. When baby-boomers retire and become eligible for Social Security and Medicare, either benefits for the elderly will have to be cut or taxes raised. The most likely political compromise will include some of each. A $1 per gallon hike in gas tax would bring in $100 billion a year in government revenue and make a dent in the looming fiscal gap.

About 60 million gallons of motor gasoline are consumed in the US each day. That's 22 billion gallons per year. Mankiw's $100 billion in annual tax revenue requires 100 billion gallons to be sold, or five times the level of current gasoline consumption. Either he erred by a lot in his revenue estimate or he actually intends (but did not disclose) that his "gas" tax would apply to a wide variety of liquid fuels, About 330 billion gallons of liquid fuels of all types are sold in the US each year, including jet fuel (16 billion gallons), diesel (6 billion gallons), residual fuel oil (4 billion gallons), and consumer grade propane (0.5 billion gallons).

Whatever the answer, revenues can't be predicted based on current consumption. Mankiw would be first to acknowledge that consumption would decline once his tax was imposed. Indeed, that's the express purpose of his tax. But that means the current base for Mankiw's $1 per gallon tax must be larger than 100 billion gallons.

What's more curious is that the source of Mankiw's federal budget gap is Social Security and Medicare. Gas taxes might be useful to internalize a carbon dioxide externality, and they are a crude instrument for incentivizing road use to deter congestion, but they have no conceivable targeting purpose with respect to retirement and health care. All that can be said in favor of a gas tax is that it would not encourage baby-boomers from growing old or demanding more hip and knee replacements.

(5) BURDEN-SHARING WITH OIL PRODUCERS SUCH AS SAUDI ARABIA AND VENEZUELA

Mankiw writes:

A basic principle of tax analysis -- taught in most freshman economics courses -- is that the burden of a tax is shared by consumer and producer. In this case, as a higher gas tax discouraged oil consumption, the price of oil would fall in world markets. As a result, the price of gas to consumers would rise by less than the increase in the tax. Some of the tax would in effect be paid by Saudi Arabia and Venezuela.

It's an interesting empirical question what effect Mankiw's gas tax would have on the world price of crude oil, but to think about this we assume that Mankiw intends his "gas" tax to be applied to all petroleum-based liquid fuels, even though he didn't say so.

The US consumes 21 percent of world oil production, about one-third of which is produced domestically and two-thirds are imported. Initially, US prices would rise to cover the tax but not by the full amount, depending on the relative elasticities of supply and demand. From the perspective of both US and foreign oil producers, Mankiw's tax provides an incentive to sell outside the US. That would increase the supply of oil in overseas markets and thereby reduce overseas prices. But the change in prices would be small because the tax would be a small faction of the world market price of crude oil. See below for details.

Whatever the magnitude of this price change, why Mankiw picks on Saudi Arabia and Venezuela is easy to explain politically -- for different reasons each is are perceived as a "bad actor" -- but hard to explain economically. They are, respectively, the third and fourth largest countries of origin for imported crude oil. Canada and Mexico are the top two, and their combined share of US imports is 31%. Thus, to the extent that the incidence of Mankiw's gas tax falls on countries exporting oil to the US, it would effect them all in proportion to their exports.. So while it's true that "[s]ome of the tax would in effect be paid by Saudi Arabia and Venezuela," more of it would be paid by our closest trading partners. This may explain why self-described "policy wonk" economists like Mankiw are often excluded from political discussions.

(6) A PREFERENCE FOR CONSUMPTION OVER INCOME TAXES

Mankiw writes:

Public finance experts have long preached that consumption taxes are better than income taxes for long-run economic growth, because income taxes discourage saving and investment. Gas is a component of consumption. An increased reliance on gas taxes over income taxes would make the tax code more favorable to growth. It would also encourage firms to devote more R&D spending to the search for gasoline substitutes.

Many economists favor consumption taxes over income taxes. Taxes on income discourage work (if levied on labor) and investment (if levied on capital). It's generally undesirable to discourage either one. But taxes on consumption discourage consumption and make saving more attractive. A case can be made that increased savings is a good idea.

Yet gasoline taxes are not free of the same kinds of undesirable effects that accompany income taxes. In a 2000 article timed to coincide with British protests over high fuel taxes, Ian Perry of Resources for the Future wrote:

[G]asoline taxes produce similar types of economic costs; for example, by raising the cost to firms of transporting products, they can reduce the overall level of economic activity and employment. These costs appear to be a lot larger than the economic costs of raising more revenue from the income tax.

Mankiw's gas tax would surely discourage the consumption of gas (or whatever collection of fuels he really intends to cover). But gasoline, fuel oil, and other energy goods comprise a small fraction of personal consumption expenditures -- just 3% in 2005. See details below.

It is strange to argue in favor of a tax because it's consumption- rather than income-related, then exclude 97% of all consumption in the base. When Mankiw's proposed $1 per gallon tax is added to the existing $0.40 per gallon average motor gasoline tax, i's not hard to imagine the combined tax rate at about 50%. Why does it make sense as a matter of tax policy to tax 3% of personal consumption at a 50% rate and leave the other 97% untaxed?

One final note on this point: As was true in the case of CAFE (see [2] above] Mankiw's proposal lacks an explicit statement that income taxes should be reduced by an amount equal to the revenue gained from the gas tax. That's missing, of course, because separately he advocates the gas tax because it would increase federal revenues and thereby help pay for Social Security and Medicare.

(7) ENHANCED NATIONAL SECURITY

Mankiw writes:

Alan Greenspan called for higher gas taxes recently. "It's a national security issue," he said. It is hard to judge how much high oil consumption drives U.S. involvement in Middle Eastern politics. But Mr. Greenspan may well be right that the gas tax is an economic policy with positive spillovers to foreign affairs.

Unlike his erstwhile boss, Mankiw apparently believes that the Iraq war is all about oil and not about combating terror. But if that's his real reason for promoting a gas tax, then Mankiw should drop each of his previous justifications. His proposal is overblown for the problem (climate change), grossly targeted for the problem (road congestion), not targeted at all at the problem (budget policy), or targeted on innocent bystanders (ostensibly shifting the burden to Saudi Arabia and Venezuela). A gas tax might be better than CAFE, but Mankiw doesn't propose to repeal CAFE in trade for the tax. And his paean to consumption taxes is belied by the gas tax's narrow base.

Furthermore, if national security is really the issue, then Mankiw's tax clearly must be applied to all liquid fuels and not just gasoline, and if possible, on imports from those countries of origin believed to pose a threat to national security. Presumably, neither Canada nor Mexico pose such a threat. It's not at all clear that Canada and Mexico would welcome the opportunity to share the burden of Mankiw's gas tax.

Mankiw thinks it's possible for "policy wonks" such as himself to prevail over campaign consultants, who he says would never advise their candidates to embrace policies such as dramatically increased gas tax. The gas tax is, he says, "a wacky idea you won't often hear from our elected leaders." On that point, he may be right. Campaign consultants may have better economic intuition than Harvard professors.

Showing our work:

(1) CARBON DIOXIDE ABATEMENT

1. CO2 conversion from the U.S. Climate Technology Cooperation Gateway:
2. CO2 offset from CarbonCounter.Org

8.78*10^-3 metric tons CO₂/gallon gasoline * $10/metric ton CO₂ = $0.09/gallon gasoline

(5) BURDEN-SHARING WITH OIL PRODUCERS SUCH AS SAUDI ARABIA AND VENEZUELA

(6) A PREFERENCE FOR CONSUMPTION OVER INCOME TAXES

[add a comment]

Comments on Gas Tax Economics
Reviewing Mankiw's proposal

Add a Comment