Incentives, taxes and policy feedback

by Henry Farrell on December 22, 2006

Via “Dan Drezner”:, Greg Mankiw “accuses Deval Patrick”: of not understanding basic economics.

The money quote from Patrick:

“If we’re trying to cultivate here in Massachusetts an energy-smart economy, then the notion of relying for additional revenues on something we’re trying to break our dependence on doesn’t seem to me to be a formula for long-term success.”

Mankiw’s response:

Okay, let me get this straight: It is important that we reduce our consumption of oil. Therefore, we should not tax it.

Hmmm….Governor Patrick was Harvard class of 1978, and it’s a good bet that he took ec 10. I wonder what they taught about the slope of demand curves back then.

There usually isn’t much need to remind well known right-of-center economists of the importance of incentives, but it looks to me as though Mankiw is misreading an on-the-face-of-it-quite-reasonable claim here. Patrick is suggesting, as far as I can see, that if we want to move away from a petroleum based economy, it may be a bad idea _ex ante_ to make gas into an important source of tax revenues for the government. If the Massachusetts state government comes to rely on a gas tax as a significant source of income, then it will have an incentive over the longer term not to want to lower tax revenues, say, by introducing non-tax regulations that make hybrid vehicles more attractive and gas-guzzlers less so. It will have created a long term constituency that favors the status quo, and that is likely to resist vigorously attempts to move away from it that would threaten funds going to this or that favoured project.

This kind of feedback loop, in which policy shifts create and provide resources for new constituencies which then push for the maintenance of that policy, has been explored at length by Paul Pierson and others (see “here”: for a nice and reasonably brief overview). Now it could be that this effect is swamped by the substitution effects and so on described in basic micro textbooks. But it also could be that it isn’t swamped – and I can’t think of any very good way empirically to test for which effect is likely to prevail under which circumstances. Mankiw and Patrick are on different sides of an argument about the appropriate instruments for changing incentives. But this in no sense means that Patrick is an economic illiterate.

{ 1 trackback }

Legislative Report » Charity in all things…
12.26.06 at 10:23 am



stuart 12.22.06 at 12:37 pm

There is a similar argument, after the fact in this case, about taxes on cigarettes in the UK, the argument runs that because the government has so much tax revenue from smokers they have a disincentive to discourage smoking because then they would have to raise taxes elsewhere (or cut spending). Not sure how valid it is in practise – they have continued to ban smoking in various places, stop advertising on TV both directly and indirectly, and so on.


Shelby 12.22.06 at 12:46 pm


The same applies in the US with regard to cigarettes. US states impose heavy taxes on them, and have further raided their coffers through various lawsuits on behalf of “the public,” but then end up working to ensure the companies remain viable and able to keep paying both taxes and what amounts to protection money. The states often justify these taxes and extortions as paying for anti-smoking programs, but rarely does more than 10% go for that purpose. Instead it becomes a general revenue source.


bi 12.22.06 at 1:10 pm

An interesting question. Is it meaningful to say that a state government has “incentives”? The way I see it, “incentives” belong more to individuals and private corporations.


P O'Neill 12.22.06 at 1:31 pm

I think a background issue is the effectiveness of higher taxes versus quantitative restrictions in achieving fuel economy goals. In the US context, higher gas tax versus stricter fuel efficiency standards. For the time being, the US is stuck with a very inefficient vehicle fleet, courtesy of the SUV madness, so a gas tax is much more likely to be revenue earner than a fuel economiser. Incidentally, it’ll be interesting to see Mankiw’s response, if any, to Krugman essentially announcing the death of low deficit political economy in this morning’s NYT.


Firebug 12.22.06 at 1:54 pm

All references to Econ 101 in political arguments should be banned. We need a simple formula similar to “Godwin’s Law” to do this. Econ 101 is such a gross oversimplification that it is simply not a serious guide to public policy. It is one of the few college classes that people come out of dumber than when they went in.


lemuel pitkin 12.22.06 at 2:46 pm

1. It’s fair to assume that anyone who makes policy arguments bases on “Econ 101” or equivalent either (a) has not thought seriously about the issue or (b) is bullying rather than arguing, or both.

2. Of course governments face incentives. They aren’t the full or often the main determinants of government behavior, but then neither are they for individuals or even businesses. I’m sure Mankiw in other contexts is happy to talk about rent-seeking, government failure, policy endogeneity, etc.

3. My understanding is that gas demand is very inelastic, at least in the short to medium term. (Presumably demand for oil for electricity is more elastic, since there are much closer substitutes. but that’s not a big part of oil consumption in the US anymore, is it?) Gar Alperovitz has written a lot about this at MaxSpeak.

4. A gas tax, like most consumption taxes, is regressive in its incidence. I don’t know if Patrick mentioned this, but he should have.


dsquared 12.22.06 at 2:58 pm

Anyone mentioning “Economics 101” ought to be asked, sarcastically-but-haha-only-seriously, whether they took any other economics courses. Anyone who is British who utters the phrase (and for whom it is therefore meaningless as well as cliched) should simply be taken outside and shot.


spencer 12.22.06 at 3:20 pm

Mankiw has been campaigning and/or advocating a “Pigou Tax” on gasoline because he says it would cause gasoline consumption to fall and reduce our dependence on foreign oil.

So it looks like Deval has been reading Mankiw’s analysis and agrees with him that a higher gas tax might make revenues from the gas tax uncertain.

So Mankiw turns around and makes fun of Deval for agreeing with Mankiw’s analysis.

If this does not make Mankiw a political hack it clearly implies that he does not take his own analysis very seriously.


mpowell 12.22.06 at 3:36 pm

Well, one thing is clear from this: if you do institute a gas tax, don’t link the funds from the tax to anything specific. Just have them going into the general fund. Otherwise you will exacerbate your problem.

But at is, seeing as how the government is currently not incentivized to reduce gasoline consumption, its hard to see how creating a gas tax as a first step would be a bad idea.


Donald A. Coffin 12.22.06 at 3:36 pm

Mankiw’s gasoline tax proposal is, in its entirety, a revenue-neutral proposal–raise gasoline taxes and offset that with a cut in income taxes (or some other appropriate tax). (Note that this gasoline tax would have to rise over time as incomes rise, because consumption of gasoline rises with income–gasoline is what economists call a “normal” good.) In this case, the state (or the feds) could (underlined, in italics, etc.) gradually increase the income tax as gasoline tax revenues fall.

Unfortunately, many if not most of those proposing a gasoline tax like this are likely (as I take it Mankiw would be) to oppose increasing other taxes in the future. So while the gasoline tax might have the effect of reducing our use of gasoline, it’s not clear that it would, in the long-run, be revenue-neutral.


Jake 12.22.06 at 3:37 pm

Shelby raises a good point, but it’s even worse than he says. Most of the states that were awarded large amounts of money from the tobacco companies proceeded to sell bonds against the future revenue stream and spend the money so raised, to the point that if tobacco companies go bankrupt the states will be in serious financial shit. They realize this, and are now protecting the tobacco companies against future lawsuits or new competitors.

And at least here in California, Rob Reiner got a tax on cigarettes passed to fund some pre-school thing, and then proceeded to campaign vigorously against a proposal to tax cigarettes more to pay for emergency medical care on the grounds that it would imperil his preschools.

But all of this is easily avoided by redistributing the tax revenue equally to everyone who pays in (the fee-bate model).


Barry 12.22.06 at 4:28 pm

“Incidentally, it’ll be interesting to see Mankiw’s response, if any, to Krugman essentially announcing the death of low deficit political economy in this morning’s NYT.”
Posted by P O’Neill

The man was a quite biddable little wh*re for Bush; he’ll come up with something authoritative, which will be absolutely false as to recent political history.


harry b 12.22.06 at 4:49 pm

an innocent and perhaps off-topic question for Shelby and jake (and anyone else). I understand that the Feds subsidise tobacco growers. So are the States clawing back the subsidy the Feds supply? And how does this change the terms of the game (between two different governments and the manufacturers) relative to the UK (where there’s just one government, which seems to have behaved pretty well as stuart says)?


Walt 12.22.06 at 5:50 pm

God, the “Econ 101” argument. How embarassing for Mankiw.


luc 12.22.06 at 6:08 pm

Basically, are you’re saying that cigarettes should be subsidized because that would create a solid incentive for a government to reduce the amount of cigarettes sold?

It’s nice to note a countereffect, but is there any indication at all that that countereffect is larger than the standard effect according to that damned economy course we all had to sit through?


SamChevre 12.22.06 at 6:28 pm


I think I can speak knowledgeably to the “subsidizing tobacco growers” issue.

The Feds do not directly subsidize tobacco growing–never have, to my knowledge. What they did do (the system ended this year) was institute a quota system on tobacco sales. The quotas were changed to keep prices relatively stable and high (relative to world market prices); as smoking decreased, the quotas were steadily cut.

Thus, the price support system didn’t in any way make cigarettes cheaper; it meant that manufacturers paid MORE for tobacco than they otherwise would have.

I think the best way to look at it is this:

There’s a certain amount of money to be made selling cigarettes to smokers; it gets divided up 3 ways–the government gets some, the manufacturers get some, the tobacco growers get some. The general anti-smoking position of the government reduces the available pool of money; higher taxes increase the government’s share relative to everyone else’s; the tobacco growers share was inflated by the quota system.

Most of the complaints about the state governments actions wrt tobacco taxes has been the way those taxes were implemented–as lawsuit settlements based on past conduct. However, State governments have aggressively tried to “spread the pain” to all cigarette-makers, even those new entrants not involved in the lawsuits. Thus, the small new companies pay much HIGHER direct taxes, to compensate for the fact that they aren’t payng lawsuit settlements.


Jake 12.22.06 at 7:42 pm

Harry, I can’t imagine that thinking of it as the states clawing back the federal subsidy on tobacco growth is at all accurate. For one, the feds had a quota, not a subsidy, so this just made the price of tobacco higher than it would otherwise be. For two, it looks like the quota was killed in 2005. For three, the quota was killed via a buyout, where the holders of quota rights agreed to give them up in exchange for $10B over 10 years. The size of the settlement in the lawsuit brought by the states was $200B over 20 years, so the values aren’t at all comparable.

I don’t know anything about how the situation is in the UK, and didn’t even think that they grew tobacco there.

I will say that I think the whole lawsuit and subsequent master settlement agreement are phenomenally poorly conceived and implemented with serious and predictable adverse results. Given that the end result was effectively a $2-a-pack tax on cigarettes for the next 20 years, the only people who are better off with the means of getting that result are the tobacco companies and the lawyers.


gmoke 12.22.06 at 10:31 pm

I remember a discussion around the lunch table of a conference on IT and Environmental Management. The EHS guys from a 3M plant said that in their area the state and county (as I recall) were taxing some pollutants and beginning to rely on those funds so much they were becoming somewhat hostile toward the possible elimination of those particular pollutants from the production process.


radek 12.22.06 at 11:03 pm

Well, in general Deval’s could be an argument against taxing ANYTHING harmful. Firms spewing pollutants into rivers? For god’s sake, don’t tax them or the government will become dependent on the tax revenue or the pollution fines!!!
Now one could finangle this so the story plays out – but even then it would just mean that a bit more thought needs to be put into where exactly the Pigouvian tax is set.
This is because of some HARD ECON 101 SCIENTIFIC FACTS:

1. The optimal amount of an externality is not zero. There’s a point beyond which YOU DON’T want to lower the consumption of gasoline, even with all the externalities.

2. An optimal Pigouvian tax is set equal to the magnitude of the MARGINAL externality (difference between Social Marginal Cost and Personal Marginal Cost). This sort of means that the size of the optimal per-unit tax is independent of the level of demand (it is if D and S are linear – not quite so if they’re not – this is where the “takes a bit more thought” bit comes in)

3. There’s a Laffer curve for gasoline. Before you scream at me please note that I’m not talking about a Laffer curve for supply of labor, which is empirically irrelevant. All I mean here is that it is quite possible that an increase in the tax rate can lower total tax revenue (this is probably true for long term TR from cigarette taxes at least in some states). So if we pass a Pigouvian tax on gas, and the gobemint gets addicted to the revenue there’s no guarantee that they will resist future tax increases in fear of losing revenue (this is assuming that politicians can think beyond “higher tax rate=more revenue”, which with guys like Deval might not be the case)(roughly again; the Deadweight Loss from setting a tax different then the optimal tax is U-shaped around the per unit level of the externality; tax revenue is reverse-U-shaped and its peak may be to the left or right of the optimal tax rate)

Also, relatedly;
4. The substitution towards other forms of transport, including more fuel efficient vehicles and alternative fuel sources is already accounted for in the slope of the demand curve.

So what have we learned? An important ECON101-type lesson about ECON101; There is some world out there in which Deval’s point would hold and be important. But there’s all kinds of reason why that world is not our world. No, at least in this case, ours is an ECON101 world where the basic principles hold and Mankiw is essentially right. A Pigouvian tax vis-a-vis alternative policies is a double whammy – less bad stuff consumed and tax revenue.


radek 12.22.06 at 11:12 pm

Ugh – Re 3 above is not clear (like the rest of it is)- if the optimal Pigou tax is to the left of the rate at which tax revenue is maximized and if in the future it becomes desirable to further decrease gas consumption by further raising the Pigou tax I don’t see how the government would object.

And come on. A bit of healthy cynicism clearly suggests that Deval is just trying to avoid the politically unpopular, but pretty effective, policy of an increase in gas taxes and cover it up with pretty talk of “energy-smart economy” (subsidized by taxes that are harder to point to and blame)


radek 12.22.06 at 11:20 pm

Ugh^2 (I really should proofread before clickin’ that Post button) – on #2 – optimal tax is independent of the level of demand if the marginal externality is constant not what I wrote above (brain addled up). If the externality increases with quantity consumed (the likely other case) then if demand decreases (say due to better alternative energy vehicles becoming available) then the optimal tax rate actually goes down (since marginal externality gets smaller). So what happens to tax revenue again depends on which side of that gas-Laffer curve you’re at.


Eric 12.22.06 at 11:47 pm

Mankiw perhaps took (and taught) ec 101 but he seems to have failed Intro to Reading/Thinking.

Deval is not making a static argument in which you move up/down an unmoving demand curve but, rather, a more sophisticated point (perhaps one about shifting demand curves/legislative incentives). But Mankiw– being a mainstream economist–can only respond by going into automatic pilot and insulting people.

Deval seems to be saying that if Mass comes to rely on oil tax revenue then they possibly create a disincentive for state legislators to kill the goose the lays the golden tax revenue (oil) and/or that if Mass comes to depend on oil tax what happens when oil becomes less important (in the long run)?

Perhaps it was only a half completed thought but it is not a thought that can be instantly rejected on the basis of ec 101.

Mankiw’s inability to look at the statement as anything other then through the lens provided by the world of ec 101 (“assume a can opener!) shows that he just can’t think right.


radek 12.23.06 at 12:21 am

If the Massachusetts state government comes to rely on a gas tax as a significant source of income, then it will have an incentive over the longer term not to want to lower tax revenues, say, by introducing non-tax regulations that make hybrid vehicles more attractive and gas-guzzlers less so.

And this bit is a bit strange. Presumably we want more fuel efficient cars in order to lower the consumption of gasoline and not because we want more fuel efficient cars for their own sake. If a gas tax reduces the consumption of gasoline down to the lower level we want it at (partly through reduction in overall driving and partly through a switch to hybrids etc.) then why would we want to make hybrids even more attractive?
(another HARD SCIENTIFIC ECON101 FACT – if you wanna correct an externality, usually the best way to do it is at its source; tax gasoline, not stuff that uses gasoline. In fact short to medium term more fuel efficient cars can mean higher gas consumption)

Or is this about something else. Like you really just want politicians picking winners, subsidizing the “good” folks who drive and develop the hybrid cars (i.e. the Demand for Smug) and generally monkeying around with things, whereas a direct tax just don’t let you monkey around as much?


bi 12.23.06 at 12:42 am

radek: To quote dsquared,

“Anyone mentioning ‘Economics 101’ ought to be asked, sarcastically-but-haha-only-seriously, whether they took any other economics courses. Anyone who is British who utters the phrase (and for whom it is therefore meaningless as well as cliched) should simply be taken outside and shot.”

lemuel pitkin says, “Of course governments face incentives. They aren’t the full or often the main determinants of government behavior, but then neither are they for individuals or even businesses.” Hmm… in which case it’ll be interesting to see how all these determinants in governments and private people interact.


Henry 12.23.06 at 1:13 am

shorter radek – we can prove that Econ 101 works using Econ 101, after pulling ourselves up by our own bootstraps.

Seriously, this is rather silly. Basic economic claims about supply and demand work quite well for lots of things – but the proof of the pudding is in the empirical testing. There’s plenty of evidence to suggest that policy feedbacks are pervasive – and no very good reason to believe _ex ante_ that they are necessarily going to be swamped by standard substitution effects. What you’re engaged in is an arid species of scholasticism, which seems to be invoking economic theory more as a kind of high blown rhetoric than as a serious, testable argument. You can do better than this surely. (moreover if you really want to combine the terms “hard scientific fact” and “Econ 101” I’d recommend you do it while muttering to yourself in the privacy of your bathroom rather than shouting it in ALL CAPS in a public forum).


radek 12.23.06 at 1:41 am

bi; Yes. And I’m not British.

Henry; There’s a good bit of reason to trust Econ101 on this as deviations from it in this case get quite convoluted quite quickly. As I pointed out above there’s no reason to expect that future policy will be hampered by present taxes. Just as likely, it will be enhanced. The goal is to reduce gas consumptions. Taxes work. Fuel efficiency less so (and yes there’s empirical work on it – higher fuel efficiency lowers cost per mile driven and so increases the demand for total miles driven which crowds out some or even all of the gasoline saving). It just seems like Deval is getting the most accomodating reading and Mankiw is getting the least generous reading possible. People are disagreeing with the man, not the statement.

And come on, there’s like 4 or 5 posts above which are just asking for the ALL CAPS and the “HARD SCIENTIFIC FACT”.


Henry 12.23.06 at 1:59 am

There’s a good bit of reason to trust Econ101 on this as deviations from it in this case get quite convoluted quite quickly.

Because the world is complicated, nicht wahr? Simplicity in a theory is a virtue – but it surely doesn’t make that theory more _trustworthy_. Far from it.

And I dunno exactly how Mankiw is getting the least generous reading possible here. I haven’t said anything rude about him or his arguments for Pigouvian taxes; but instead have pointed out that he has presented someone else’s arguments in a way that suggests that they are economically illiterate when they seem in fact to be making a different (and perfectly reasonable) argument. If I had wanted to be ungenerous, I could surely have phrased things quite differently than I did. What exactly is the ungenerous part of the post???


radek 12.23.06 at 2:35 am

Like I said above, the argument given by Deval can be used to argue against taxing or fining any socially harmful activity. Which means the burden of proof is on the other side. Furthermore, as I pointed out above, it’s not clear that implementing taxes today will make politicians reluctant to institute other policies later on, it could go either way. So the burden of proof is on the other side. Then, even if it does make politicians less likely to subsidize hybrid vehicles later on, so what? The point is to have less gasoline consumption and an optimal tax is an, um, optimal tax. Finally there is evidence that a gas tax is pretty much the best way – compared to alternatives – to reduce gas consumption. So if someone’s gonna argue for some alternative the burden of proof…
I mean seriously, Deval is basically trying to argue that higher gas taxes will somehow lead to higher gas consumption in the future, given alternatives. Which does violate “Econ101 Principles” (demand curves slope downward) and that’s all Mankiw is apparently saying.
I guess it could be true. But hey, I’m nearly certain that there’s some model of the climate out there which says that rising Co2 emissions *could* lower temperatures. But that would violate Climate Science 101.


Henry 12.23.06 at 7:02 am

ummm no radek. I’m sorry to break it to you but there isn’t any particular reason why the burden of proof is on Deval’s side rather than yours. The arguments made in Econ 101 aren’t fundamental physical laws of the universe or anything like it. Never have been. Never will be. But it’s clear that you disagree, so I will bow out – no particular point debating this further. And in any event I have to get on a plane.


Thomas 12.23.06 at 11:41 am

I take Patrick’s argument seriously: to move to energy independence, we should cut gasoline taxes to zero. I’m sure I’ll find takers for that in this string. Well, I might, if I tell you some right-winger has endorsed it.

I can’t help but wonder how to properly deploy Henry’s convenient skeptical argument. If we can’t foresee with any acceptable degree of confidence the effects of any change in policy, then we’ll have to keep what we have, or at least agree that we’re not going to have a meaningful argument about the change. Certainly we can end any discussion of a new carbon tax, since those opposed (Exxon, say) can apparently argue in good faith (and with support here at CT) that they are worried that the feedback effects might swamp whatever good would come of the new tax, and there’s no good way to dispute that claim. (It is interesting to note that policy shifts create new constituencies, but the related bit is that current policies have their own constituencies–and that not subtle insight may have a bit more to do with Patrick’s thinking on the subject. Patrick is no fool, but he can make some very smart people look to be in their defense of him.)

(Finally, as radek notes, the proposed feedback effect just doesn’t work in this case. To the extent that we’d worry about feedback effects, it seems to me that we’d worry that MA would set the gasoline tax at a rate that would maximize revenue, but not match the negative externalities of the use of gas. Choosing the revenue maximizing level might reduce the mitigating effects, but it wouldn’t eliminate them. And again note that we might already be at the “wrong” level–we have policy and politics already, and they are being defended.)


bi 12.23.06 at 12:02 pm

To show what exactly is wrong with radek’s argument:

“The Earth is flat! Saying that the Earth is round will mean that the geometry of the Earth’s surface is non-Euclidean, and that violates Geometry 101!”

“Intelligent Design is true! If ID is false, then life was created by spontaneous generation, and Redi’s experiment clearly shows that spontaneous generation cannot occur! Don’t you know your Biology 101?” (Someone actually made this argument, though not in snark mode. The answer to this is that the conditions of primitive Earth are vastly different from the conditions of modern-day Earth, and Redi’s results only apply in the latter.)


Inbal 12.23.06 at 12:14 pm

Another element to consider in gasoline tax: how is the revenue used? If we’re worried about the long terms constituencies that feed off revenue sources, it is useful to evaluate the interests that they serve. Perhaps for a gasoline tax to be optimally effective, revenue would be invested in mass transit alternatives or in funding for alternative energy sources- vested interests created that would feedback into the direction of reducing gasoline consumption.


radek 12.23.06 at 1:42 pm

Um, bi, I don’t know if you’re incapable of picking up on sarcasm or what but you sort of make my point pretty well. Parallel lines do cross in some universe but that’s sort of irrelevant when you’re building a bridge.


radek 12.23.06 at 2:04 pm

re: 29 Just want to make it clear that I gave like 4 “particular reasons” why the burden of proof would be on Deval.


bi 12.23.06 at 2:23 pm


Is the Earth’s spherical shape “relevant” or “irrelevant” when deciding what’s the shortest route from Japan to America? And why? Do you have any principled way to decide what’s “relevant” and what’s “irrelevant”? No?

Now just to use your own argument: Your above line of reasoning means you can accept or reject Economics 101 principles as you wish. This is of course bogus. Which means the burden is on you to prove that Economics 101 principles are “relevant”, and other models are “irrelevant”, in this specific case of MA’s gas tax rates.


SF 12.23.06 at 4:31 pm

I think Mankiw is partially correct here. If you want to artificially reduce demand for a commodity like oil, taxation is an appropriate measure. Nevertheless, increased taxation on producers is also a consideration, since that will increase consumer cost indirectly. Also, it’s difficult to factor in the risk/benefit of the possibility that the legislature may be corrupt or inept.


Johan 12.23.06 at 7:59 pm

A similar case from Sweden: Some decades ago, a major Swedish trade union (TCO) suggested taxing pornographic films, and channel the revenues to less morally depraved film production. Their very Swedish and prudish proposal was intended to fight porn and support culture at the same time.

However, the obvious objection to their model was, of course, that serious film makers would then have had to rely on increasing audiences not for their own films, but for porn. Luckily, the TCO proposal was never implemented.


Jasper 12.24.06 at 12:06 pm

The simplist way to go about accomplishing what both Mankiw and Patrick want is for the government to institute a cap and trade style system for gasoline purchasing. Require purchasers of gasoline to buy permits, in other words, and gradually (or not so gradually, if you prefer) reduce the total quantity of permitted purchases.


James 12.24.06 at 2:01 pm

After reading all this, I’m a bit curious concerning the distrust for the use of “econ 101” thinking. can someone tell me specifically which non-econ-101 concepts are appropriate for analyzing the positive economics of a gas tax?


wkw 12.24.06 at 5:12 pm

Mankiw’s chief ideal for instituting a Pigouvian tax is not to raise revenue, tho he admits that it is a nice side-effect that may allow a reduction in more-distortionary taxes elsewhere (e.g. income taxes). the chief reason is to combat a series of externalities. this may be done with a gas tax. sweetening the pot is the fact that the incidence of a gas tax will be shared by oil-exporting countries, so the entire burden will not fall on American people.

Patrick’s main focus does not seem to be battling externalities. his focus is on how to improve a crumbling transportation infrastructure, and how to fund its repair. therefore, he is right to say that it is unwise to come to depend on revenue from a gas tax in the long-term.

Patrick is not saying that a gas tax would not be good, in general; he is saying that it won’t fix the transportation problem in Mass.

Mankiw is not arguing that it will fix that problem; he is saying that it will help to fix a bunch of other problems.

and so they are arguing about two different things, and aren’t really contradicting each other at all.

of course, no one has brought this question: how efficacious would state-by-state gas taxes be? shouldn’t these sorts of things be done by the Fed government? otherwise, all sorts of other adverse effects may occur.


derrida derider 12.24.06 at 10:20 pm

What a load of nonsense – the empiric experiment has been done. The US attempts to restrict gas use by regulation. The Europeans do it by taxation. Just ask yourself who has the more fuel-efficient transport.

Again, just ask yourself – since governments started heavily taxing tobacco has its use increased or decreased?


Chris 12.25.06 at 12:19 am

Hmm… I see. We don’t want to tax things we view as BAD because it increases the government’s dependance on bad things. Conversely, we should obviously tax activities we view as GOOD as to increase government dependance on GOOD things.

Therefore, for the betterment of society, the government should heavily tax charity and scientific research. Once the majority of the government’s revenues come from these sources, they will be promoted as to maximize taxable income! Of course! Why hasn’t anyone thought of this before?

The whole argument is positively asinine. A substantial gas tax (i.e. a gas tax large enough that the government would derive a significant portion of its revenues from it) would go further to weaning dependance on fossils than any CAFE regulation or ethanol subsidy ever could. Yet the problem is that the true costs of the latter programs are disguised, while a tax on gasoline is direct and simple, and thus politically impossible. Deval Patrick was able to brilliantly sidestep the fact that he is both an enviromentalist AND against a gas tax – but don’t confuse his bullshit with any kind of logical or economically sound argument.


radek 12.25.06 at 2:16 am

the other d^2; But that’s Econ101!!!!!!! Or should that be scare quotes????


bi 12.25.06 at 11:32 am

Chris: Excuse me? Why should anyone want to increase their dependence on anything, whether good or bad? Does the fact that something’s good mean one should purposely make themselves rely more on it? Can you at least make your strawmen look more realistic?

radek: That’ll go under Scientific Method 101, but I guess you can’t tell the difference.


radek 12.25.06 at 1:44 pm

bi: First get things straight – round objects are perfectly compatible with Euclidean geometry. They are not non-Euclidean. Maybe after that we’ll talk.


Seth Edenbaum 12.25.06 at 4:47 pm

Gasoline is a necessity for many. Smoking is a necessity for no one. Sales taxes are regressive. Europe has a massive and well funded public transportation system. The US does not.
Mankiw is a Republican hack who pretends to be a scientist.
Patrick is a politician and a Democrat.
Other than that, they’re talking about two differrent things: long term revenues and policy inertia on the one hand and the Pigovian tax (as such) on the other.

Words are not numbers: there is no ‘science’ of language in use. Read the words. Ask who wrote them, and in what context.
I just read yet another review of Dawkins’ idiot-book, written again as a defense of faith. When will that piece of shit get the secular historically and anthropolically minded trashing it deserves, the sort of trashing David Cole gave Richard Posner? [both in the NYRB]

The relation to the above should be obvious. That it may not seem so to many is another reason to be depressed.


radek 12.25.06 at 7:12 pm

Gasoline is a necessity for many. Smoking is a necessity for no one.

Who appointed you in charge of deciding what is and what is not a necessity and for whom? Come off it. Short to mid term (until all the smokers die off) demand for cigarettes is a lot more inelastic then that for gas and anyway I’m not sure how this is relevant.

Sales taxes are regressive.

So a sales tax on diamonds and yachts is regressive? Dude, go back to principles of Econ101.


Seth Edenbaum 12.25.06 at 9:14 pm

radek, is this a joke?


derrida derider 12.25.06 at 11:33 pm

Ya gotta remember that things can be true even if they’re taught in Econ 101. In Econ 201 you learn about public choice and the incentives facing governments in taxation policy; Mankiw would be very familiar with the Virginia school. The usual arguments here, though, are that these incentives lead to the tax rates being set too high, not too low.

A Pigovian tax on gas really is a no-brainer. It would markedly increase economic efficency and improve US national security. The Arabs would pay much of the tax anyway, as the reduced global demand from the US being sensible would lower the pre-tax price of oil. And if it’s regressive then it’s not hard to use the revenue raised to implement large progressive tax cuts elsewhere to offset it.

I really can’t understand the US population’s opposition to it.


wkw 12.26.06 at 12:40 am

dd –

i can understand the citizenry’s opposition to gas taxes: they don’t understand that the revenue may be used to either rebate portions of the income taxes they pay, or else to provide additional services which benefit them. they don’t understand that the burden of a gas tax would be shared by oil-exporting nations, and so we are (sort of) getting something for nothing. they don’t understand that this is one simple, fairly easy way to improve the environment, national security, income inequality, and economic efficiency in one stroke.

it’s easy to understand their opposition: they don’t understand the argument. i tried to explain this to my (working-class) family today. some of them couldn’t wrap their heads around the concept that they’d be paying more in one tax, but that as a result their net welfare would increase in both measurable and immeasurable ways.

i mean, it doesn’t stop there. decreased oil consumption would free up money currently applied to highway maintenance, state-run inspection services, public parking, and other public funds. also, if the world price of oil falls, then a production cost for our trade partners falls, so our imports are cheaper. but they don’t understand those nuances. all they know is that “gas is already too expensive; why make it higher?”


Chris 12.26.06 at 12:57 am

“Gasoline is a necessity for many. Smoking is a necessity for no one.”

It’s true that gasoline is a necessity, but in the advent of higher prices, people’ll find a way to conserve. Oil’s a necessity the same way food is – sure, people need it, but not all consumption is created equal. People can drive more fuel efficient cars. They could use public transport (which, in the Northeast corridor, IS quite good). They can carpool. They can (gasp!) live closer to work.

Oil is going to $100 anyway. Supply cannot possibly keep up with its skyrocketing demand – therefore, demand must be priced out of the market. With a gas tax, the higher prices can be reinvested back into the country (better mass transit would be a logical place to start), rather than ending up in the pocket of some Saudi Oil Prince.

But perhaps Americans are by nature myopic – allergic to anything that dampens our short term standard of living. Perhaps the reason people are so afraid of a gas tax is the same reason why our personal savings rate is negative.

We can’t go down this path forever. Soon or late, we’re gunna have do the difficult thing and curb our usage of fossils. I hate to break it to you, but cars ain’t ever gunna run on water.


bi 12.26.06 at 2:40 am

I asked, “Is the Earth’s spherical shape ‘relevant’ or ‘irrelevant’ when deciding what’s the shortest route from Japan to America? And why? Do you have any principled way to decide what’s ‘relevant’ and what’s ‘irrelevant’? No?”

radek replied, “First get things straight – round objects are perfectly compatible with Euclidean geometry. They are not non-Euclidean.”

Conclusion: radek is an idiot.

– – –

Seth Edenbaum said, “Sales taxes are regressive.

radek replied, “So a sales tax on diamonds and yachts is regressive? Dude, go back to principles of Econ101.”

Conclusion: radek is still an idiot.

– – –

I said, “That’ll go under Scientific Method 101, but I guess you can’t tell the difference.”

radek replied, “…”

Conclusion: I won’t engage in discussions with idiots, I’ll point and laugh at them until they cry.

– – –

Seth Edenbaum: So can we say that the US’s low taxes aren’t exactly encouraging them to develop a well-funded public transportation system? I guess perhaps the whole idea of developing a good public transportation system is itself a disincentive — it’s going to cost money, no matter what the tax rates are. So that’ll mean that the MA government must _want_ to develop a public transportation system in the first place, and then decide how to get money for it. Hmm…


Slocum 12.26.06 at 8:04 am

Oh, c’mon. Patrick opposes significant new gas taxes because they’d be deeply unpopular among his constituents. His argument is sophistry intended to square the circle so can seem to be for a responsible energy policy and opposed to gas tax increases.

But I think he’s missing an opportunity. As others have hinted at, what he should do is get Eliot Spitzer on the phone and, instead of new taxes, what they can do instead is file a lawsuit intended to ‘punish’ big oil the way the tobacco lawsuit ‘punished’ big tobacco. As with the tobacco settlement, the end result would be a large tax increase (part of it paying for an enormous windfall for a group of well-connected lawyers), but the people will be fooled into thinking that big oil was taken to the woodshed, they won’t really connect the price increases with the lawsuit, AND the politicians will be able to spend the settlement monies like manna from heaven (unlike regular tax revenue).


radek 12.26.06 at 12:39 pm

bi; You haven’t made an ounce of sense this whole discussion. Maybe if you tried to actually, you know, articulate, rather then resort to cutsy irrelevant examples and name calling then maybe, just maybe (in some alternative non-Euclidean universe) having a discussion with you would be…well, not worth the effort, but I think that in the spirit of Holidays I could chalk it up to meaningful charity work.

seth; At this point I’m still willing to be civil, despite your arrogant displays of fake incredulity. If you think gas is a necessity I think these folks would disagree;
But seriously, hippies aside, for most people some gas is probably a necessity. But hell, I’ve known people who would drive two blocks to the convenience store, then drive another block for a parking space. That my friend is not use of gas out of necessity. In my family we walk our asses down the block. And like I said before I don’t see what brining cigarettes into this has to do with anything.
And your statement that “all sales taxes are regressive” is just plain wrong. I’m not sure where the joke is supposed to be so let me spell it out to you. A sales tax on diamonds or yachts is NOT regressive as most of those goods are purchased by high income people. A tax on bread and potatoes is regressive as low income people spend a significant share of their budgets on food. A gas tax is somewhere in between. Gasoline consumption as a share of total budget starts of really low for low income people, rises and peaks somewhere around 80K of income or so, then falls again. So it’s progressive in parts and regressive in other parts. Gee, a complicated world out there isn’t it?


Seth Edenbaum 12.26.06 at 1:09 pm

radek, at this point I’m not trying to be civil. You’re an ass.
Patrick was referring to, as I said: “long term revenues and policy inertia,” he was speaking as a democratic politician making political statements about economic policy, to which you’ve responded as if he were saying “1+1=3” “the world is flat” or “god made the universe in 7 days.” But you like Mankiw, were not responding to what Patrick said but to what you think Patrick must have said. If we were talking in person, I’d be about to walk away or maybe I’d sucker-punch you just for fun.
Context little boy, context. The people who run this place don’t understand it very much or very well, but you make Henry Farrell sound like a hippie.
Nothing Patrick said was objectively ‘wrong.’ Nothing.

And I haven’t owned a car in 20 years.


bi 12.26.06 at 1:16 pm

What, radek is “still willing to be civil”? Well, who first bullied whom with the “Dude, go back to principles of Econ101” line? Wow, that was so civil!

One thing I hate more than a self-righteous idiot, is a self-righteous idiot who resorts to bullying tactics and yet pretends to be civil.

So, to radek: go away, idiot.


Shelby 12.26.06 at 1:28 pm


You assume that the gas taxes will in fact be used to either reduce other taxes, or provide additional services to those paying the gas taxes. Knowing what I do of US state and federal government, I see no reason to make this assumption. Isn’t it just as likely that the money will be siphoned off via lobbyists and the relevant legislature, to make (say) Archer Daniels Midland richer?


radek 12.26.06 at 1:33 pm

at this point I’m not trying to be civil…maybe I’d sucker-punch you just for fun.

Thanks for clearing that up. See comment 369 in this thread:

bi: I wasn’t even talking to you. And speaking of bullying, you’re not even a bully. You’re like that little scrawny stupid little runt that follows the bully around and cheers him on so that he can live vicariously through him. You’re a supporting actor. Like that guy in sunglasses that follows Biff around in BTTF.


Seth Edenbaum 12.26.06 at 2:06 pm

r dear, I’m sorry but you’re not my type.


InTransit 12.26.06 at 4:52 pm

I hate to inject some facts into a theoretical argument among economists, but here goes.

Right now, gasoline taxes exist in all US states and are used for transportation. (Important details, especially whether revenues can be used only for highways or also for mass transit, vary.) As car mileage has improved, there is less tax revenue per mile driven. This has already put state transportation departments into a financial bind. Remember that it is expensive to maintain highways, even if you aren’t building new ones.

If you expect crude oil prices to keep going up (certainly not an unreasonable outlook even if you disagree with it), then you can expect gasoline usage per mile driven to go down. To the extent you want to use transportation revenues to build mass transit, the problem is exacerbated because if your policy is successful it will further reduce transportation funds (by reducing gas tax revenues) while increasing the need for those revenues (because almost all transit services require operating subsidies). The implication is that the gas tax is a politically unappealing means for solving the transportation funding problem, because the revenues it produces are expected to decline in future years (or grow more slowly than the costs they are supposed to cover).

This is in fact conventional wisdom among state highway departments, and Patrick was not saying anything that you can’t read from others who are found in all parts of the political spectrum.


abb1 12.26.06 at 6:18 pm

Isn’t it just as likely that the money will be siphoned off via lobbyists and the relevant legislature, to make (say) Archer Daniels Midland richer?

But if it’s that bad, if Archer Daniels Midland owns the government, then the issue is essentially moot, then any tax is waste, robbery. To discuss this you have to assume that the government represents interests of the general population to a reasonable extent. And then the question becomes: is this particular government more likely to get addicted to the revenue or to truly represent interests of the population? If it’s the former, then you probably need a better government.


wkw 12.27.06 at 4:38 pm

shelby –

well, in my first post i was talking about Mankiw’s and Patrick’s respective ideals. in the end, that’s what this whole stupid argument is about (are we really arguing about Euclidean geometry?) i guess i should have re-stated that premise.

but, as abb1 pointed out, if you operate under the assumption that all taxes are wasted, then you can’t really participate in this conversation. neither Mankiw nor Patrick are talking about anything other than the best possible way to tax.

in any case, is Archer-Daniels really that much of a presence in Mass.? i really don’t know.

Comments on this entry are closed.