Tuesday, December 18, 2007

Gas taxes are much too low

So gas prices went up again, and the press relays complaints about how it is the taxes that keep prices high. These voices ask that tax be reduced to bring relief to hard working Americans. How wrong they are.

Gas is taxed for three reasons: 1) because it is easy revenue: the tax needs to be levied at only few resellers; 2) because it penalizes adverse behavior in the sense that using gas pollutes (it exerts a negative externality on others); 3) the tax is a proxy for paying the use of a public good: roads.

Let me be more specific.

Whenever you tax, there is going to be some dead weight. Setting up the appropriate administrative structure, collecting the taxes and chasing evaders all imply costs. Compared to income tax or sales tax, the gas tax is incredibly efficient, as you can levy the tax at the level of the refineries.

Whether you believe in global climate change or not, everyone will agree that using gas will exert some externality on others. That can be pollution, but also contributing to congestion on roads. The idea is to make the user pay for the externality so that he changes his behavior to factor in this externality. As demand for gas is quite inelastic (reacts little to price changes), this can be quite substantial.

Gas use can also be a proxy for the use of some public goods. Driving on public roads is the perfect example, but think also about the cost of diplomacy/foreign aid/wars to secure oil imports. Someone has to pay for all this, and the best is to make the user pay. Road use would be best paid according to some monitoring of mileage, but gas use is the best proxy we have. In this regard, road use in the United States is currently subsidized, as the gas tax does not cover all of the costs of road construction and maintenance.

You may criticize the high-tax ways of Europe, but here the latter is dead right: with taxes at $5-6 a gallon, they are about right in pricing the externalities. In the US, externalities may be lower, as there is less density and thus congestion, and more land to absorb pollution. A $3 tax per gallon should entirely reasonable. It can be introduced gradually, say 5 cents a month over a five year period. And do not tell me the voters are against such tax increases, they actually favor it (pdf), as long as the revenue displaces some other tax like the income tax. This makes perfect sense, as a gas tax discourages a bad, while income tax discourages a good.


Anonymous said...

You seem to be ignoring two important points: 1) there are huge transition costs as people adapt to higher gas prices (relocation, smaller houses and cars, building public transportation). 2) such a tax is regressive: it hurts the poor more than the rich. That is fine in European high tax countries, but not in low tax USA.

Anonymous said...

Gas taxes are not necessary regressive. The poorest ones do not own cars. And if increases in gas taxes go towards public transportation, they are the first to benefit.

Independent Accountant said...

A tax "replaces" another tax? Grow up. Now! Neal Boortz has been pushing his "fair tax" to "replace" the income tax. Very nice. I oppose any tax increases. Period. If you increase taxes, Uncle Sam increases spending. Period. The prophet, Milton Friedman once did a study on tax increases. He found increasing taxes increases what in the US is called the budget deficit, what you limeys call the PSBR. On this side of the pond we have what is called the "highway trust fund" (HTF) which is the net result of gas taxes paid and federal spending on roads. Guess what? The HTF has plenty of money in it and is used for other things like: supporting mass transit. I suggested a "tax" the US could adopt in lieu of the "CAFE" standards. Assume a car has a 150,000 mile life. Also at 25 mpg, will use 6,000 gallons over its life. A car which gets say 15 mpg will use 10,000 gallons. The tax is based on the "excess" gallons used or 4,000 gallons times $1 per gallon or $4,000 levied as an excise tax on car purchases. Those worried about "regressive" taxes should see anyone buying a small car won't be affected. It can also eliminate the US tax credit for hybrid vehicles. A 30 mpg car uses 5,000 gallons and gets a $1,000 credit on purchase. This is predicated on repealing CAFE.

Economic Logician said...

To independant accountant: I suppose your are against taxes for two reasons: 1) it gives the liberty to governments to waste tax revenue, 2) it provides adverse incentives on the taxed goods. Point 1 may depend on what a government should do. I agree with point 2, and this is why gas tax should replace income tax.

Also, the HTF does not pay for local roads and most state roads. It does not pay for the cost of pollution (air, noise, landscape) and congestion. The HTF accounting is all skrewed up from an economic point of view.

johnsharp9 said...

Unfortunately, no one has ever been able to substantiate the claims of massive negative gasoline externalities.

-Costs of foreign wars do not count because they are not necessary to secure that the oil is sold on the world market.
-If you are going to count the negative externalties of gas, you have to count the positive ones: the massive economic growth it enables, the option to use life-saving plastics in medical operations, etc.

Taxing an inelastic good has solid economic justification, but the externality argument never quite comes together on close inspection.

Bill Conerly said...

A gas tax to fight congestion? Not close enough. If I drive downtown mid-day or after 7 pm, I do not add to congestion, but I burn gas. If I drive out to the country at most hours, I don't cause congestion. You want to tax congestion, that's reasonable; use toll roads, HOT lanes, etc. But not the gas tax.

Anonymous said...

Ahh, the devil in the details of your last sentence.

Economists such as myself clearly see the sound logic of a gas tax. However, the reason I (and many others) simply will not support it is that politically we will never see the time when income or payroll or property taxes are reduced anywhere in proportion to the increase in gas taxes. So, what you are really proposing from a political economic standpoint is a huge tax increase.

jefff said...

That is a poor understanding of externality.

An externality is an effect that goes to someone who isn't part of the transaction.

Plastic isn't an externality. The precursors of plastic are sold by the refinery companies. The refinery company gets a benefit and whoever buys the plastic, including the ultimate consumer getting the heart surgery get the benefit. No benefit is given to a third party without payment. That high molecular weight oil goo would otherwise have to be disposed of in some other way at a cost to the parties of the gasoline transaction and possibly third parties (if, for example, it was burned into CO2, or dumped into the ocean). Some plastic does end up being dumped in the environment to form a negative externality. In particular medical waste is fairly likely to be incinerated and add to our CO2 problem, at no cost to anyone in the transaction, but medical waste does sometimes end up washing up on beaches. Your particular example is also absurd. Medical instruments cost enough that they could easily afford to use plastic when necessary even if it cost more because it was not part of a larger multi-party economic transaction. Only low cost uses of plastic, like our ubiquitous excessive packaging, are a significant part of the economic transactions involving gasoline, but again they pay for the plastic so it isn't an externality.

Economic activity due to gas use isn't an externality. The benefit of the driving goes to the parties of the transactions enabled. I don't get any secret unaccounted for benefit because some guy just drove past my window on the way to buy or sell something. He gets a benefit, and whoever he trades with gets a benefit. All I get is a woosh of noise and some air pollution.

Claiming externalites don't exist is quite ridiculous. They are all around us. There are some pretty obvious cases of extreemly negative balances of externalitie, though some of them have been dramatically alleviated by regulation in the wealthier countries (such as smog).

No magical principle causes positive and negative externalities to always balance.

dizmal said...

You may criticize the high-tax ways of Europe, but here the latter is dead right: with taxes at $5-6 a gallon, they are about right in pricing the externalities.

Do you have any evidence to support this claim?

Please share with us how you were able to derive what the price of gasoline "should be"...

jefff said...

I often wonder about these tax phobic nuts.

There have probably been thousands of tax changes in their lives. Given their certainty that taxes always go up in any tax change their tax rates must be well over 100% by now.

I used to live in Washington state. The arguments one could have with such people were entirely comical. Washington has no income tax, one of, I think, four states that do not. Washington pays a middling total state and local tax rate, but is on the upper end of total state, local, and national tax rates because of high income and poor federal deductibility (state income taxes are deductible, but if you don't have one you don't get the deduction). State and local taxes in Washington are also among the few most regressive in the country (it is mainly sales, property, and a gross sales business tax). One obvious way to fix that would be to impose an income tax. Washington would also get a few hundred million bonus dollars in federal deductions.

Every one of the tax phobic I ever talked to was certain that if an income tax was instituted total taxation would go way way up and continue to increase at some dramatic rate forever.

It is absurd. 46 states have those taxes now, and we were right in the middle without one. At least the vast majority of those states instituted an income tax at some point (usually in the 30's). None of them have far higher taxes than Washington (the range is fairly narrow). So in 100% of states that have instituted income taxes what they claim will happen has not happened, yet even when told this they insist it will happen.

There is also that few hundred million in deduction money that would appear as if out of nowhere (really it appears out of the federal budget). A reasonable expectation, given that these tax phobic people are a sizable voting block, is that the new income tax using tax system would have a net tax cut equal to part of that windfall, and that the rest would be used one some new spending goodies. It is the obvious political comprimise. They insist this won't happen.

If you are going to live in a fantasy land at least make it a happy one.

johnsharp9 said...


I think you misunderstood my argument.

you said: "[...] but again they pay for the plastic so it isn't an externality."

Read again what I wrote: "the option to use life-saving plastic"

Yes, the plastic that you buy is not an externality. I "get" that. However, the *option* to buy the plastic -- which itself has value, just ask any options trader -- is an externality. Without oil, I can't buy plastic at any price. With it, I can buy it at some price.

Likewise, any individual economic transaction is not an exterality. But the fact that a great economy surrounds me, with the option to buy many things I had no part in, is a positive externality to me.

Furthermore, I NEVER claimed there are no externalities. I simply pointed out that once you argue in favor of taxing because of the negative externalities, you have to go by the *net* externality, which means, yes, subtracting off the positives. Once you do that, it's far from clear that you have any remaining externality to tax.

I would like to add here that the road subsidy is not a gas subsidy for two reasons:

1) It subsidizes your vehicle, *whatever* the fuel source. If I fuel it with solar power, is that a solar subsidy?

2) There is extreme waste and misallocation in the government's handling of the funds for roads. Without the waste, I strongly suspect existing gas taxes cover the roads, which means gasoline users *on the demand side* see the full *real* impact of their wear on roads, even if it doesn't fill the actual cost with all the waste. (Plus, most damage comes from truckers, not ordinary drivers.)

jefff said...

"the *option* to buy the plastic -- which itself has value, just ask any options trader -- is an externality."

That is pretty thin gruel.

I think you are torturing the definition of 'externality' to death.

Hmm... ok an equally vague negative 'externality':
You have eliminated the option to buy the barrel of crude oil you refined for gasoline.

I find it implausible that, even if it were possible to get a vague estimate of such nebulous externalities value and it turned out to be positive, they could come close to balancing out the negative externalities that are obviously involved in gasoline transportation.

Imagine the effect it would have to have on the externality balance of transactions without major obvious negative externalities? Most transactions would have to have large positive externalities, and we would be underproducing everything!

An apt parable:
"And why beholdest thou the mote that is in thy brother's eye, but considerest not the beam that is in thine own eye?"

"I would like to add here that the road subsidy is not a gas subsidy for two reasons:

1) It subsidizes your vehicle, *whatever* the fuel source. If I fuel it with solar power, is that a solar subsidy?"

It's not a gas subsidy, it is a automobile transportation subsidy. This leads to more automobile transportation. At present this means more gas consumption, but it also means more consumption of everything else that goes into cars.

It is, however, a good point that if cars do move away from gasoline in a more dramatic way over time we may eventually need a mileage tax, or something like that to fund roads.

"2) There is extreme waste and misallocation in the government's handling of the funds for roads."

In other words:
In magical fantasy land where infrastructure was cheaper the current gas tax would cover them therefore road construction isn't subsidized now.

Huh? That's not a useful argument. How about this:

In magical fantasy land where highly skilled construction ponies grow on trees in such abundance that they are free for the taking and roads are thus cheaper to build the gas tax would cover it therefore road construction isn't subsidized.

Also in magical fantasy land public schools would make enough off their vending machines to be self funding, therefore schools aren't paid for by taxes today.

Eh, this isn't even fun anymore.

Economic Logician said...

In response to various comments: There is little doubt that gas has negative externalities, just think of the pollution its use generates. Other externalities are related to the use of gas, but could stem from other commodities. It just happens that gas is almost universally used for transportation, and thus any consequence of transportation can be fixed with gas tax (or subsidy).

There seems also to be some confusion between externalities and strategic complementarities. The latter happened for example when the US steel, coal and rail industries developed in parallel, each benefiting from the growth of the others.

I agree a gas tax is not the best way to deal with congestion. The ideal would be to equip every car with GPS and tax optimally by current traffic. But you see the privacy issues. Gas tax is next best thing in the absence of tolls.

Telnar said...

On the theme of strategic complements, there are also economies of scale which are encouraged by inexpensive gasoline and generate a lot of consumer surplus. For example retail home delivery costs quite little compared to the value of the consumer time it saves to the busiest consumers, but I suspect that it has a minimum effecient scale which requires pricing attractive to people who place a fairly low marginal value on their time. Less expensive gasoline makes it easier to reach that scale.

dizmal said...

In response to various comments: There is little doubt that gas has negative externalities, just think of the pollution its use generates.

First, I'm willing to grant you that gasoline has negative externalities.

Congestion, however, is far more a function of "driving a vehicle at a specific time and place" than "using gasoline". A person driving a Prius at rush hour in Manhattan contributes to congestion. A person driving a Hummer on Saturday morning in Montana does not. There are far more direct and efficient ways of taxing congestion than taxing gasoline. For example, sell a special red license plate that allows the driver to be on certain roads during rush hour. Or, have electronic toll tags with rates that fluctuate higher during rush hour.

Second, I'm willing to grant you the idea that pigouvian taxes on things with negative externalities can make the world better.

Theory is good, but specific statements like "gasoline taxes should be $5 or $6" require support.

Can you offer anything at all to back this statement up?

How did you determine what the price of gasoline "should be"?

johnsharp9 said...

jeff, All I can see in response to my point about positive externalities is that you think they're small. Without any hard way to quantify this, there's nothing left to debate about. However, keep in mind the positive externality I'm referring to is the option to travel to lots of places, to have significant use of comparative advantage, to consume goods I would never have a change to otherwise touch. I can't see how that's "nebulous". I'm all for tabulating that externality (which you could measure by e.g. the amount people pay to move to such places), but don't be surprised if you don't like what you see.

As for roads:

1) You've conceded that it's not a subsidy for oil, just "whatever happens to be best". I'm waiting for your accusations of it subsidizing solar technology when people drive solar cars.

2) I think you miss my point there. I was refuting the claim that drivers aren't paying the "real" cost of road use. To answer that, you only have to look at whether the demand sides *feels* the true cost through prices. If someone's taxes fail to cover the useless, pork-laden bridge to nowhere, I just don't see that as a case of him "not paying the full cost of his road usage". You see?

My position is robust across the public education example you give. If someone's taxes don't cover inefficient schools, but would cover efficient schools, I consider him as paying the true cost of education -- since it's not his fault the government is so inefficient at using his money to satisfy that need.

Econ Chris said...

"All I can see in response to my point about positive externalities is that you think they're small. Without any hard way to quantify this, there's nothing left to debate about."

This seems all-to-convenient for you, while you hand-wave a point about how gasoline is somehow a positive externality (???) using, let's say a... very "particular" definition of what an externality is. In general your point seems to be carefully constructed around some highly questionable assumptions-- like what a "positive externality" is or the fact that governments must always and everywhere be full of irrational bumbling fools-- even though we are talking about something that has nothing to do with competitive pressure (the usual justification for assuming governments are all morons as you seem to), but something very simple like levying a tax.

Perhaps your assumption about governments being inefficient does make sense in a world of competitive pressures that cause private enterprise to be more efficient. But taxing is not competitive. Your point is now almost entirely an article of faith.

As for the blog post, I like. When I think of taxing gas, I think of it from the perspective of a tragedy of the commons. It's completely rational for one person to get in a car powered by fossil fuels and pollute, but when everyone does it, strong and irreversible environmental costs become imposed over time. A gasoline tax discourages this race-to-pollute, through the usual channels.

As for congestion, while I agree that a gas tax is a weapon against it, I also agree with several comments pointing out that it's far from perfect for that. dizmal's post has some fascinating ideas on how to effectively attack congestion regardless of the vehicle fuel.

Though, I think for covering the cost of a public good, tolls would be better. We know-- and you posted on this even-- that pollution taxes work best in the long-run when offsetting labor taxation.