The regressive gas tax (and why that should not be an argument against it)

I left a comment over at M1EK's Bake Sale of Bile awhile back. It was motivateed by the conclusion to The Gas Tax Isn't Regressive, Part Three:

This is timely because of a current thread on Environmental Economics on this very subject. Amazingly, I've now provided THREE links which are credible and contain supporting evidence for the claim that the gas tax isn't regressive across-the-board; for the most part blind assertion is still the only support for the 'regressive' position. Moral: Conventional Wisdom is hard to fight, even when it's wrong.

Here is my comment:

Here is a study that shows the gas tax is regressive: Estimates from a Consumer Demand System: Implications for the Incidence of Environmental Taxes, by West and Williams, Journal of Environmental Economics and Management, May 2004.

It finds a result similar to this:

"A tax on gasoline is not regressive across the lowest incomes but is regressive from middle to high incomes."

Those in the lowest incomes are not in the gas market because they don't have cars. Therefore, gas taxes are regressive for those in the market. That's one definition of regressivity.

A problem is that car ownership is one of those variables that helps low-income people increase their incomes. For example, if you don't have a car it is tougher to get to work at various places, tougher to find better employment, tougher to work and take your kids to daycare, etc.

So, excluding them from the definition of regressivity seems like circular logic.

Here is the response from M1EK:

Note that the typical way regressivity is measured is not across the market, but across all incomes. Thus, your argument that poor people who don't drive shouldn't count in the measure of the tax's regressivity is foolish. And, frankly, it doesn't surprise me that somebody who lives in such a car-dependent area views being without a car as an unmitigated negative - for you, it clearly would be. It would be nearly impossible to be economically productive in Boone without a car.

The same thing is definitely NOT true in major cities all over this country, though. In fact, freeing poor people from the necessity of owning a car in order to acquire employment is a huge POSITIVE in those cities, as seen by the various studies that show that in the most car-dependent cities, people actually spend as much or more on transportation as they do on housing.

Again, without redefining "regressive" in a way which seems inappropriate, you can't get away from the fact that even in this country, many poor people don't drive (don't HAVE to) and many more drive a lot less than your typical exurbanite SUV owner. That's why the gas tax is not regressive across the spectrum.

And a comment from Rob Williams, co-author of the JEEM paper:

John Whitehead mentioned my paper with Sarah West, but didn't mention one of the main points of the paper: the results depend a lot on how you use the gas tax revenue. By itself, the gas tax is (mildly) regressive. But as we show in the paper, you can easily use the revenue in ways that would make the gas tax highly progressive (by using that revenue to finance cuts in really regressive taxes, for example).

Right now, gas tax revenues aren't used that way. But there's no reason that has to be true. The additional revenue from a gas tax increase doesn't have to be used in exactly the same way that we currently use gas tax revenue.

Actually, I had mentioned this part of the argument in a comment to M1EK at ENV-ECON but I was rude about it so I deleted it, after apologizing to M1EK in an email (and consider this the public apology). Tim can back me up:

• I did mention the part about fixing the gas tax's regressivity
• I was rude

So, Rob, sorry that I didn't tell the whole story.

And this is why the regressive nature of a gas tax should not be an argument against it. An increase in the gas tax should be regarded as environmental policy. And environmental policy, just like trade policy, is a terribly inefficient place to try to right the wrongs of the income distribution. If you think the wrongs of the income distribution should be righted, head straight for the income tax.

Comments

[the oil drum (ianqui)]

Oh my. I can't believe it--I totally agree with this post!

Still, despite the fact that a gas tax is probably a good idea, middle and upper class Americans still get uppity about it--even supposedly forward thinking people. Just check out some of the comments to this post. (And don't let them fool you into thinking they're opposed to it because the government will abuse the profits. That might be true, but really, they just don't want to fork out either.) I, of course, can afford to be self-righteous about it since I live in NYC.

[odograph]

"A tax on gasoline is not regressive across the lowest incomes but is regressive from middle to high incomes."

I'm happy to see my intuition confirmed.

[Tim Haab]

Yes John, you mentioned the part about fixing the gas tax regressivity (is that a word?) and you were rude. Can I go back to sleep now?

[mike]

Expenditure on gasoline and motor oil as percent of household income before taxes falls consistently as income is higher, according to the most recent Survey of Consumer Expenditures. By the ordinary standards of tax policy, it is going to take a tortured redefinition of regressivity to make motor fuel taxes anything other than regressive with that data. That doesn't mean that an extremely high motor fuel tax isn't good policy, just that it will be regressive according to normal ability to pay standards.

[M1EK]

ftp://ftp.bls.gov/pub/special.requests/ce/standard/2003/income.txt

Interestingly, the "0-5000" group spends more on gas than the 5000-9999 group. Wonder if this is mainly because of college students?

Looks to me like if you compare the 5K-9.9K group with the 15K-19.9K group, you don't see any regressivity - more like a flat tax. From where did you conclude that it was regressive across the board?

[joshua corning]

"middle and upper class Americans still get uppity about it--even supposedly forward thinking people."

Well, just don't forget us red-neck backward people as well...we dumb and hate them gas taxes.
Anyway it comes as no suprise to me that those who endorse the idiology of Mo, Hitler and Stalin name themselves forward thinking. Hey by the way how is that war on poverty going for ya?

"I, of course, can afford to be self-righteous about it since I live in NYC"

yeah cuz your whole city is heated, lit and run on oil...one of the only cities in the world to still do that by the way.

[John Whitehead]

"Anyway it comes as no suprise to me that those who endorse the idiology of Mo, Hitler and Stalin name themselves forward thinking. Hey by the way how is that war on poverty going for ya?"

Please, everybody, no more Hitler comments here.

[joshua corning]

"Please, everybody, no more Hitler comments here."

So I take it Stalin and Mao comments are OK? :)

Anyway I concure with the oil drum ...that gas taxes are regresive and that should not be the argument against it...the argument against a gas tax is "it will achive nothing towards a non-existant problem" there is nothing wrong with cheap oil..it is enviornmentaly friendly compared to most cost effective alternatives, it is cheap, and it is safe to transport. One day oil will lose its importance in the energy market...but just like the stone age didn't end for the lack of stones, the oil age will not end for the lack of oil.

Hey here is a question that will piss everyone off...why is there oil on Titan when there is no signs of life on Titan?

and no i am not hinting that there is life on titan or hinting that if we run out of oil here we can always get oil from titan...i am hinting that terrestrial oil is not produced through biological means and that it exists in near inexaustable quantities.

[M1EK]

"there is nothing wrong with cheap oil"

How about being addicted to cheap oil? That's where we are in this country. Imagine what happens to suburbia if gas hits $10/gallon.

[Robert Schwartz]

Whether or not a tax is "regressive" is simply unknowable ex ante by inspection of its administration. In an economy where prices and wages are mostly set by individuals operating in a market environment, the cost of any tax will be passed through the market system in myriad ways.

A tax on labor may increase wages. A tax on capital may increase its return at the expense of labor. A tax on consumption, such as gas tax may, increase the cost of labor and depress the return to capital.

Given the above, the better criteria for asseing a tax include its cost of administration and its intrusiveness in the economy.

[joshua corning]

"How about being addicted to cheap oil? That's where we are in this country. Imagine what happens to suburbia if gas hits $10/gallon."

your arguments get weirder and weirder...so what you are saying that high gas prices will destroy our civlization so we should raise taxes on gas?

Anyway I agree that gasoline has helped build our civilization...and in your remark about suburbia is some sort of critisism of sprawl of some such non-sense i would like to point to all the ghost towns of the west that are no longer there...they existed before gas powered cars and went away after the invention of cars....the heart of america has depopulated since the invention of the car...not the other way around.

sprawl is caused by bad regulation and the erotion of private property rights....if you want i can explain to you how zoning regulations promote sprawl.