On Small Externalities

Megan McArdle, one of my favorite economics writers, makes this rather odd statement on whether or not a small (say, $0.40) gasoline tax is worth the trouble:

It did take a lot of increase to change driving behavior, which is why in Europe, taxes can account for as much as 90% of the price of a liter of gas. There's a plausible argument that a 40 cent tax won't do much to change driving habits. But then, why have the 40 cent tax at all?

Is this really an issue? The reason to impose a Pigouvian tax is to correct an externality. I'm no expert on carbon taxes, so I'll outsource to wikipedia

Many estimates of aggregate net economic costs of damages and benefits from climate change across the globe, the social cost of carbon (SCC), expressed in terms of future net benefits and costs that are discounted to the present, are now available. Peer-reviewed estimates of the SCC for 2005 have an average value of US$43 per tonne of carbon (tC) (i.e., US$12 per tonne of carbon dioxide) but the range around this mean is large. For example, in a survey of 100 estimates, the values ran from US$–10 per tonne of carbon (US$–3 per tonne of carbon dioxide) up to US$350/tC (US$95 per tonne of carbon dioxide.)
[. . .]
A tax of $100 per ton of CO2 translates to a tax of $0.978 per gallon of motor gasoline.

Let's take the numbers here as given (since I know little about it). It would seem the maximum optimal tax is about $1/gallon (Greg Mankiw would approve). Megan is probably right: This wouldn't reduce driving all that much. But what I don't understand is why that's a problem. If the optimal tax is low, that means precisely that behavior shouldn't change after a tax is imposed, because the externality is not high. Now, maybe you disagree with the scientists and economists and think the external cost of carbon is higher. But if you agree with the calculations, well, they are what they are.

The failure of a hypothetical carbon tax to raise transportation costs to European levels or to slam Americans into higher density living isn't a failure of Pigouvian taxation. It's a recognition of the underlying economics; namely, from an economics point of view, even if you follow the IPCC recommendations (as did the Stern report), there's little justification to be found in global wamring for stratospheric energy prices.

Share this

Pigovian taxes

McArdle writes:

There's a plausible argument that a 40 cent tax won't do much to change driving habits. But then, why have the 40 cent tax at all?

To which you (curunir) respond:

Is this really an issue? The reason to impose a Pigouvian tax is to correct an externality.

Well, yes, but the point of correction is to optimize the behavior to maximize total social benefit (or however you want to call it). If the tax doesn't actually change behavior (much), then it is (close to) superfluous.

If the optimal tax is low, that means precisely that behavior shouldn't change after a tax is imposed, because the externality is not high.

But what we care about is the behavior. If the optimal behavior "shouldn't change", then you can achieve the same effect with no tax. Which brings us to McArdle's question: why have a tax at all?

I should have said

I should have said "shouldn't change much" there, because there is of course some marginal change. And I'm not sure what counts as small as a tax: a dollar a gallon increase would be roughly a 50% hike from what I pay now.

What I'm trying to say is that once you calculate the externality, then that's all there is to a Pigouvian tax. There's no point in further fretting that behavior didn't change as much as you thought it should.

I don't think that's it

What I'm trying to say is that once you calculate the externality, then that's all there is to a Pigouvian tax.

I think she agrees.

There's no point in further fretting that behavior didn't change as much as you thought it should.

I don't think that's what she's fretting about. She seems to be fine with behavior changing as little as it does. She's saying, given that it changes so little, why not just do away with the tax, since you get pretty much the same result.

Isnt it implied that the

Isnt it implied that the revenue of said tax is spent on mitigating the effect of the externality?

It might not alter driving behavior much, but 50c per gallon across the US is a lot of money. How does it compare to the budget of the DOE/EPA? Not sure, but its probably significant.

Apparently not

Isnt it implied that the revenue of said tax is spent on mitigating the effect of the externality?

Going by the descriptions I found at Wikipedia and Investopedia, no. Nor is how it is spent part of McArdle's discussion, either explicitly or implicitly.

Public finance

Isn't it implied that the revenue of said tax is spent on mitigating the effect of the externality?

I didn't draw that inference, but I think it's relevant to the larger discussion. Different taxes have different purposes. A Pigouvian excise tax is designed to modify a price in order to send more appropriate price signals, thereby influencing people's behavior. If no behavior is modified (except as a result of the de minimis wealth effect), does the tax do any good? Perhaps, depending on the use to which the tax proceeds are dedicated.

1. As suggested above, the proceeds might be used to finance a Pigouvian "subsidy."

Most households pay for electricity based on the number of kilowatt-hours (kWh) they use, regardless of when they use them. But the utility incurs higher cost to provide a kWh during the day that at night. Relative to the status quo, a change to time-of-day rates for electricity operates much like a tax on daytime usage and a subsidy for nighttime usage.

Some studies suggest the difference between daytime and nighttime rates would not be enough to induce much change in behavior. Nevertheless, the change to time-of-day rates would permit both daytime people and nighttime people to better bear the true cost of their own decisions. In other words, Pigovian taxes and subsidies are, when appropriately understood, merely mechanisms for eliminating the hidden tax and subsidy structures embedded in the status quo.

2. Alternatively, the proceeds might be used to offset less efficient taxes.

One problem with taxation is it's tendency to distort price signals and therefore induce inefficient changes in behavior. Income taxes discourage income-generating activity. Sales taxes discourage mutually beneficial exchange. Payroll taxes (FICA) discourage employment. So if we find that a carbon tax doesn't actually change people's behavior, I say HALLELUJAH! Let's use it to displace some more pernicious taxes.

One problem with taxation is

One problem with taxation is it's tendency to distort price signals and therefore induce inefficient changes in behavior. Income taxes discourage income-generating activity. Sales taxes discourage mutually beneficial exchange. Payroll taxes (FICA) discourage employment. So if we find that a carbon tax doesn't actually change people's behavior, I say HALLELUJAH! Let's use it to displace some more pernicious taxes.

That principle is called Ramsey taxation, by the way. Tax goods with the most inelastic demand, because the resulting market distortions will be smallest. It's always struck me as morally problematic though. What makes the demand for something inelastic? Often it's precisely the necessity of the good. The demand for, say, Tamiflu in an pandemic is extremely inelastic, and taxing it to lower payroll and income taxes would be quite efficient. And yet . . .

Not that I'm equating medicine and gasoline, I hasten to add. Just that "efficient" taxation can open up all sorts of problems.

Great point

I think that's a great point. In light of your remark, I do think that a fetish has been made of not affecting behavior, without regard to the importance of that behavior. If a tax greatly affects some behavior, that may be because the behavior wasn't that critical and so no big deal that it's been modified.

Economists may have taken this into account in their effort to minimize deadweight loss. But I'm not so sure that their analysis goes far enough. Admittedly, I am in need of a refresher course, so I'm not one to judge.

Ramsey pricing/taxation

That principle is called Ramsey taxation, by the way. Tax goods with the most inelastic demand, because the resulting market distortions will be smallest. It's always struck me as morally problematic though. What makes the demand for something inelastic? Often it's precisely the necessity of the good. The demand for, say, Tamiflu in an pandemic is extremely inelastic, and taxing it to lower payroll and income taxes would be quite efficient. And yet . . .

Not that I'm equating medicine and gasoline, I hasten to add. Just that "efficient" taxation can open up all sorts of problems.

All sorts of DELIGHTFUL problems!

Ramsey pricing/taxation tells us that when we need to depart from marginal cost pricing (that is, to recover non-marginal costs – overhead costs, including the cost of most government services) we should recover them from people with the least elastic demand. Translation: Stick it to the guy with the fewest options!

1. Progressives remonstrate that capital gains taxes tend to be lower than income taxes, a very regressive situation. But capital is famously fluid. If taxes are too high, capital will flee to a jurisdiction where taxes are more attractive. Laborers are not nearly that mobile. Thus, if you think that labor taxes are too high, you might consider working less, but taxes would have to be damned high before you’ll consider emigrating. Thus, Ramsey pricing justifies taxing labor more highly than capital.

2. When it came time to integrate public schools, government started with the schools of poor whites. Pandering to the wealthy, right? Perhaps. But it’s also true that rich white kids would simply drop out and go to private schools. If you’re gonna integrate a school against people’s wishes, you’d better find a school where people are too poor to flee. The Little Rock school was not the only all-white high school in Little Rock; it was just the high school serving the white families from the poorer part of town. (South Boston famously frustrated the plans of integrationists. Impoverished white Southies somehow scrapped together tuition money and fled to Catholic schools in droves rather than attend integrated public schools.)

3. Imagine that Rhett Butler and his faithful slave die in a valiant attempt to hold off the Union forces, and today historians accuse Southerners of racist hypocrisy for heaping praise on Rhett while ignoring the slave that engaged in the exact same behavior. Yet was it the same behavior? Rhett was a man of means; he had many alternatives rather than remaining to fight. His slave was not; the slave had the choice to remain with Rhett or be shot as a deserter. In any number of situations in which people make a sacrifice, Ramsey pricing tells us that the richer people were the more noble, because they almost certainly had more alternatives than the poorer ones did to avoid making the sacrifice.

THERE’s a moral that’s sure to make you popular around any campfire....

Difference between the US and Europe is

that Europe had their cities laid out and rail installed before cars became cheap. Our cities and suburbs were designed around cheap cars and cheap fuel.