A number of readers seem to be confused by my previous post on who gets the benefit of a gas tax holiday. If inelastic demand means that the consumers are paying a lot of the gas tax, how come when we lower the gas tax, they don't get that money back?
Because the supply curve is kinked, that's why.
Supply is elastic on the downside--companies can take the stuff off the market and store it if they don't like the price, or throttle back their refineries. But there's no way to expand the supply, because Americans can't import gasoline from abroad; each mix must be specially formulated to the air quality regulations of its regulatory region.
Because supply is unresponsive to price on the upside, and prices are already quite high enough for companies to make a profit, the price of oil is currently basically set by consumer demand: they bid the price up to the point where they want to consume the maximum amount that refineries can supply. Oil companies can't sell more gasoline by lowering prices, and they also will not sell any less gasoline, because the current price is the price at which consumers want to consume all the gasoline they produce. Hence, if you lower the tax, the price stays the same, and 18.4 cents goes to the oil companies for every gallon.
Now, one might say that this is good because it will incent them to find more oil. But this is not, in my opinion, a very good argument. First of all, we're also considering mucking around with windfall taxes, which are a much bigger disincentive to invest than any piddling 18.4 cents per gallon. Second of all, oil companies can discover more oil, but they are hard put to increase refinery capacity, because no one wants any refineries near anyone; virtually all of our refineries are decades old, with improvements coming from throughput enhancement rather than new built capacity. The limiting factor on gasoline right now is refinery capacity, not oil supplies. And third of all, they're already making really quite a lot of money. We don't need to give them even more.
Fourth of all, of course, a gas tax is a user fee for roads, which is good, and also encourages people to invest in more fuel efficient cars, which I think is also good. So it's a rare example of a tax that should be much higher, not lower, IMHO.
Why is this getting so much attention? Does it really matter? Well, I don't think that fiddling with the gas tax is going to bring the US economy to its knees. But it offers some interesting lessons about markets and elementary price theory, which many of my readers seem to be interested in. Also, I really like making charts.
That's an unusual-looking supply curve. Where are you getting it from?
This may not be of interest to anyone but me, but sometimes I get disoriented about what blog I'm reading. Say, if I see a proofreading lapse, I tend to think I'm reading Yglasias. Apparently discussion of the inelastic supply of gasoline are a strong signal I'm reading Kevin Drum. What finally clued me in was the vehemence of the 'step away from the pander' post.
My head. Refinery throughput is fixed, and there are regulatory barriers to importation. No matter how much you increase the price, supply won't go up.
MM,
You might be of further service if you flesh out: "Refinery throughput is fixed, and there are regulatory barriers to importation".
What you say is emminently reasonable and logically unassailable. The problem is, refineries are running at below capacity by choice, gas inventories are very high, and demand is down.
I'm not saying there is some devious conspiracy. I think it shows that there are other effects in play. I don't know what they might be, but I don't think they change the conventional wisdom on the gas tax holiday. If people are willing to pay record high prices, they'll be willing to let the gas tax holiday go to the suppliers.
Megan's post is a good one. I just have one caveat to add.
The government needs some revenues, and a gas tax is about as good a way as any to raise some of those needed revenues, but I would only support a gas tax increase if it were coupled with a tax cut somewhere else.
A gas tax increase in itself would likely only lead to more government spending, and those of us who guard our liberty jealousy believe that an ever growing government carries its own negative externalities.
So raise the gas tax, yes, but use those revenues to reduce taxes on work, savings or investment, not to enlarge an already gargantuan welfare state.
I can only go off what I read, of course, but my understanding is that while consumption has eased, the refineries are about where they're expected to be in advance of the summer driving season, and that inventories are running up in anticipation of demand. I'm happy to be corrected by other information, of course.
I am curious about the argument that lack of refinery capability is a significant contributor to high gas prices in the US. Given that a new refinery has not been built in approx. 20 years (although I suspect existing US refineries have expanded), why don't the vertically integrated US Oil companies build a refinery on a friendly Caribbean island (e.g. Curacao, Puerto Rico, etc..)? There must be an island that would want the construction jobs and additional revenue such a facility would bring. I have never heard a satisfactory response to this idea.
it will incent
Please, NO! You are a graduate of a famous university with a degree I believe in something related to the English language. You have a blog at The Atlantic, not quite The New Yorker in terms of writing but still one of our most distinguished publications. "to incent" is the language of sleezy lobbyists and imbecilic functionaries.
Megan,
Shouldn't you make a distinction here between integrated oil (exploration, production & refining) companies and pure-play refiners (e.g., FTO)? The pure play refiners are making a lot less now, since crude prices have risen so much faster than gasoline prices and consequently their crack spreads have narrowed.
And Gene, "incent" is common in business too. American English tends to truncate words, so what's wrong with "incent"? It's not as if "incentivize" is some mellifluous word.
One other problem with gas taxes is that proponents seem to think that ever higher gas prices must always be good. On wonders, at what point would they say "No, this is too much, the taxes are doing more harm than good"? According so standard economic theory, there must be some optimal level of tax that internalizes all externalities--and beyond that level, the tax is no longer Pigouvian, it starts creating distortions. Gary Becker, for one, has argued that the gas tax is high enough, and might even be too high.
I suspect that Becker would prefer a gas tax (or a consumption tax in general) to other, more distorting, taxes as a means of raising revenue, but his arguments cast further doubt on the idea that, ceteris paribus, a gas tax hike would do more good than harm.
PS: I never use "incent". I try to use older words like "induce" when I can, otherwise I'll use "incentivize," but that's just a matter of taste. Languages do change over time and incent is becoming more common. At any rate, Gene should chill out.
"Incent" is a perfectly cromulent word.
I don't think I disagree with your basic premise -- that the supply curve is kinked -- but one statement you make is clearly incorrect. The US does import many refined products, including gasoline, and will import even more in the future as refining capacity is built outside of the US but not here.
We do have formulations in this country that are unique, but that generally is handled through blending, and that can take place domestically even if the refining occurs overseas. It is a bit like Coke making the syrup in Georgia but my store actually blending the final drink in the fountain locally.
Megan,
Unless you are some kind of peak oil nut, I don't see what you care one way or another whether people would buy fuel efficient cars. More fuel efficient cars do not reduce pollution and in some cases may increase pollution by making it cheaper to drive and increasing the marginal value of a gallon of gas. More fuel efficient cars have the significant downside of killing a lot of people because they are generally lighter and less safe. Your shout out to fuel efficient cars is a glaring piece of PC foolishness in what is an otherwise very well thought out post. It is the kind of analysis that newspapers ought to be writing but seem not to have the ability or desire to write.
Obviously languages change over time. I just think that "incent" and other forms of MBA speak belong on the Neanderthal branch of the English language evolutionary tree.
Thank you for tackling the idea of pricing and demand. Here are some facts that might help. The U.S. imports about 1.4 million barrels a day gasoline. Each day the U.S. uses 9.2 million barrels a day when using data from the Energy Department. That means they are importing about 15 percent of what the U.S. uses.
As Fred points out, refiners aren't producing much gasoline because they aren't making much if any money doing it. Oil prices are rising much faster than gasoline prices are. Oil companies that explore for it are doing quite well but the ones that only turn it into fuels like gasoline are losing money. These are two different groups of companies.
Also, supplies aren't unresponsive to prices on the upside. They are just responding more slowly or at a smaller rate. Gasoline prices went up by about 20 percent since last year and demand has dropped less than a percent.
There is some confusion in the comments on short-run versus long-run supply elasticities. Economists who study energy issues agree that the summertime short run supply elasticity for gasoline is extremely low. That is why about a jillion economists with widely different political leanings all signed on to Thoma's letter criticizing the McCain/Hillary proposal for a tax holiday. All of these economists can talk to the energy specialists in their departments, and they all get the same answer that Megan gets on the relevant short run supply elasticity. But the long run supply elasticity of oil to the U.S. is pretty high -- if the Saudis or traders net back a lower price from selling in the U.S. than elsewhere, they will shift sales away from the U.S. That is why consumers bear the incidence of permanent gasoline taxes -- gas prices are a lot higher in Europe than the US because taxes are a lot higher there.
Ken,
The thing with energy demand is that it takes a while to change your behavior. You don't just go out and buy a new car or move closer to your job or start taking public transportation immediately because gas prices went up. Because energy prices effect people's lifestyles in such fundamental ways, like where they work, how they get to work, how they travel, it takes a couple of years of higher prices before people's behavior changes enough to significantly effect demand. Of course the reverse is true as well. Once you have bought the new fuel efficient car or moved closer to your job, you are unlikely to run out and buy a new SUV or move back across town just because gas prices are falling again. This is why I think oil prices are going to crater as some point in the next two to four years. Oil has always been a boom and bust commodity because it takes so long for demand to react to price in either direction.
I agree with Gene. "Incent"? For shame!
These curves have the benefit of (a) being theoretically possible and (b) offering intellectually coherent support for McArdle's position, which the prior charts didn't. But I would require empirical support from somewhere other than McArdle's head for the proposition that right this minute, May 2008, we just happen to be at the price/quantity point where the supply curve kinks. Where were in May 2007? September 2007? Etc. It seems very convenient that just when a difference opens up between Obama and his opponents, we are in the precise market conditions that make Obama correct and his opponents wrong.
Yes, I quite agree that this is only short run supply inelasticity.
John,
It is reasonable to suppose that the short run demand elasticity is also pretty low -- not zero, because even within the constraints you mention, people can adjust. Shorter versus longer planned summer driving vacations, metro versus cars, etc. But take Megan's kinked supply curve, with the nearly vertical portion representing the very low summertime supply elasticiy for gasoline. Now draw a demand curve as inelastic as you want but intersecting the vertical portion of the supply curve. An 18 cents (or whatever) reduction in the gas tax shifts the supply curve down 18 cents, but doesn't affect the position of the demand curve. Where supply is vertical, the downward shift doesn't move anything, so the equilibrium price is unchanged, and the oil companies keep the money from the tax holiday.
Is a gas tax paid by refiners or by the drivers? Your comment about a road use tax belies your claims about the tax incidence.
Your earlier post argued that it is paid by refiners. But, if it is paid entirely by refiners, then it is not a road use tax, because drivers pay the same for gas whether there is a tax or not.
If it is paid in part by drivers, then it is a road use tax, but then a gas tax holiday would reduce the price that drivers pay. But your last post refuted this very position.
What y81 said.
The idea that, at this moment in time, we just so happen to be exactly at the kink seems to be an awful big coincidence. Why were we not at the kink 12 months ago, when gas was priced $.60 less, or 18 months ago when gas was $1.60 less?
Oh, right, because Obama picked a policy position. And Megan and her entire circle of friends, as well as the rest of the media, are infatuated by Obama and so will make out a justification to make Obama's policy position look right.
It doesn't matter who pays the tax, drivers or refiners; the point is, the tax goes up as road use goes up.
Y81, Al--this isn't some kind of conspiracy. It's pretty generally known that refiners operate flat out during the summer months, and that in the short term--i.e., the term of the tax--supply isn't very elastic.
Your earlier post argued that it is paid by refiners. But, if it is paid entirely by refiners, then it is not a road use tax, because drivers pay the same for gas whether there is a tax or not.-Rob
Rob, this has to do with the difference between long-run and short-run elasticities. The supply of gas is much less elastic in the short run than in the long run.
A perfectly inelastic supplier will bear the full price of any tax increase and get the full benefit of any tax cut. Thus, a temporary tax cut will benefit suppliers but not consumers (so Megan is arguing).
But since supply is more elastic in the long run, producers and consumers share the burden of the tax more evenly over time. This means that a permanent gas tax hike of say 20 cents per gallon, would raise the price of gasoline for consumers, but by less than 20 cents.
So there is no contradiction (look here for further clarification). Nevertheless, a gas tax hike (unaccompanied by a cut in some other tax) is a bad idea, for reasons I have explained above.
So here's my question: if the retail level tax were eliminated but consumers still paid the same price for gasoline (i.e., the gas companies absorbed 100% of the monies that had formerly gone to to the retail gas tax paid by consumers), then wouldn't a windfall profits tax on the gas companies equal to exactly the amount that had formerly been paid in retail level gas taxes, leave every exactly where they were before?
If that is correct, then what are the implications for the notion that a windfall profits tax discourages production and exploration?
Alternatively, if the risk adjusted cost of capital is x , and a windfall profits tax only applies [marginally] to returns >> x , then wouldn't all projects with a positive NPV without the windfall profts tax still have a positive NPV with the windfall profits tax? If this is correct, then how would that discourage production?
Now I assume that in practice it would discourage production, but the arguments above appeal solely to theory, so how would it in theory?
How does the tax go up? The tax is static per unit. The revenue might increase, but the tax is not progressive. Most taxes are paid at the wholesale level and passed along.
John -
Sort of an aside, but, even though I agree that most of the supposed benefits of fuel-efficient cars are, currently at least, illusory, I would still consider buying one fairly seriously (if I drove - viva the NYC subway system). The reason would be that I might like to use my money to incent (sorry, had to do it) carmakers to develop more and better fuel-efficient cars - someday actually producing cars that truly are fuel efficient and not made out of paper mache - by demonstrating that there is money in it for them.
After all, with all the yuppies buying Priuses because they think are saving the earth, what might the market be for a car that gets better gas mileage than a mid-90s vintage Civic but doesn't have a rubber band for an engine and won't be blown away in a strong breeze, which would appeal not just to PC green-nuts but to drivers who just want a nice car?
In you graph isn't P actually P(net of gas tax) for S bit not for D? Paying more to the government in road taxes doesn't give the oil companies any more money for their gas. Since D is apporximately vertical, the oil companies will make basically the same money at a lower D (or consumer) price?
It seems very implausible to me that the same gasoline refineries that operated flat out last summer are operating flat out this summer, when the economy is troubled and prices are 20% higher.
What has happened in the past year is that there has been an increase in the price of the raw material for gasoline (oil). The effect of this increase is to shift the supply curve straight upward. Meanwhile, there has been some change on the demand side (people buying fewer SUVs), which has the effect of shifting the short-term demand curve downward slightly (but not changing its slope, which is of course much steeper than the long-term demand curve slope). So, if the supply and demand curves were intersecting at the kinked point in the supply curve last summer, they should now be intersecting at a point to the left of the kink, meaning that the incidence of the excise tax falls primarily on the consumer.
But hey, that's just theory. Empirical data about refinery operation could cause me to change my analysis, if someone--maybe someone who gets paid for this discussion--wanted to dig it up.
"Y81, Al--this isn't some kind of conspiracy."-MM
Maybe not, but Megan should realize how her past cheerleading for Obama has damaged her credibility. She has been more critical of Obama lately, but a long period in which she was praising Obama to the skies while taking cheap shots at his opponents has made her conservative and libertarian readers mistrustful. And understandably so.
Gene,
One problem is that (somewhat like pharmaceuticals) a lot of oil projects are high initial exploration costs with a reasonable chance for not finding anything. The problem is that even if the tax is only on successful areas of production with return in excess of x, the company views all the added expense in exploring elsewhere as part of the costs or, in another way to look at it, determines the expected return by multiplying by a probability of success.
The government has no good way to determine the probability of success of an exploration venture, and the oil company would lie like crazy if the government just asked. The tax then tends to tax based on actual return after the project is discovered, but not expected return based on chance of success. (There's also the question of how to amortize exploration costs.) So the windfall profits tax discourages risky ventures for which the return if they succeed is high but the chance of success low.
This makes the tax revenue correlate with miles driven, but you can't call it "a user fee for roads" unless the user is paying the fee.
The idea of "a user fee for roads" is that the driver pays the hidden costs of driving, and thus marginal drivers choose not to drive, but if the tax incidence falls upon a third party, the marginal driver continues to drive.
RWE: Megan argues (and maybe correctly) that refiners are "perfectly inelastic supplier[s]" because they are at capacity, and the cost of expanding output is cost prohibitive, even at today's gas prices.
And, perhaps more importantly, because margins for refiners right now are unusually low, unlike with oil production. The DOE's Energy Information Agency stats agree with me, but Megan's blog is holding all comments with links for moderation right now, so I can't link to it.
Gasoline prices, while up, have not risen as fast as crude oil prices. So refiner margins are actually down, discouraging production. Of course, that means that they might be able to expand output if their margins increase.
John Thacker
Thank you. That makes sense.
I'm with you, and I have said so in Megan's company before.
Lots of other example at the link above, including rampant misuse of "adverse" and "unique".
I have a colleague who uses the word "contemporaneous" in ways that make me cringe.
Actually, the U.S. imports significant quantities of gasoline from foreign countries. See figures 2.2 to 2.5 and the accompanying text from the report I co-authored for the AGs of some Northeastern states:
http://www.ersgroup.com/media/PDFs/Complete_Petroleum_Report_090707.pdf
I'm not disagreeing with your analysis of the tax; the supply has reached a point of relative inelasticity regardless of the source of gasoline. But most people are indeed surprised to see how much of the gasoline used on the East Coast comes from foreign sources.
Megan, did you actually just respond to Al?
Okay, then, I guess I will too:
Oh, right, because Obama picked a policy position.
Ah, welcome to the general election. We are now to expect several months in which the standard texts of Econ 101 and the opinions of every single economist in America, should they happen to jibe with Obama's policies, will abruptly become the far-left ravings of Muslim Communist black Christian terrorists.
I can hardly wait!
Let me put that a different way: America cannot import gasoline that has been mixed for another country. It doesn't matter where it is processed, but once it's formulated, we can't buy elsewhere.
"we just happen to be at the price/quantity point where the supply curve kinks. Where were in May 2007? September 2007? Etc. It seems very convenient that just when a difference opens up between Obama and his opponents, we are in the precise market conditions that make Obama correct and his opponents wrong."
It is a natural situation for the curves to sit that way.
If the demand curve were moved a bit to the left, people would buy bigger cars, move further out to suburbia to bigger houses or do other things that allow them to get more utility from purchasing more gas. Refiners would suspend less efficient capacity. In the long term, it pushes the demand curve to the right or the supply curve to the left.
If the demand curve were further right, refiners would invest in enhancing capacity, people would eventually replace larger cars with smaller ones, they would try to live closer to their jobs etc. This would push the supply curve to the right, and the demand curve to the left.
In both situations, the long term effects take a while to kick in, but they eventually put the demand curve right across that kink in the supply curve.
Right Njorl,
Many energy market supply curves look exactly that way (I am most familiar with Power).
In energy markets there are very high fixed cost so people build very little spare capacity because that capacity will just sit idle unless there is an unexpected demand surge.
> Let me put that a different way: America cannot
> import gasoline that has been mixed for another
> country. It doesn't matter where it is processed,
> but once it's formulated, we can't buy elsewhere.
Which DOES suggest a certain elasticity - if we are willing to pay much more for the stuff, foreign refineries are going to want to formulate for OUR market.
>Let me put that a different way: America cannot
> import gasoline that has been mixed for another
> country. It doesn't matter where it is processed,
> but once it's formulated, we can't buy elsewhere.
Well, we got along fine without these botique blends once. Change the requirements. Most of these blends are practically indistinguishable anyway.
In addition to Njorl's points, the demand for gasoline is highest in the summer. In the winter gasoline demand is lower (the demand curve shifts to the left). Gasoline capacity (refining, pipelines, terminals, etc.) is sized around this seasonal peak. Also, for each of these, NIMBY has constrained capacity expansion. Refineries have some flexibility to produce more gasoline and less heating oil, but that flexibility is reflected in the more flattish part of the supply curve. Foreign-refined gasoline is used in the U.S., but these refineries face adjustment costs to produce more American-blend gasolines, and the stuff has to be distributed through a system which itself, for gasoline, is near its designed capacity. Believe it or not, Al, gasoline capacity is not sized where it is in anticipation that Obama could come along years later and be right about something. Your conspiracy theory is truly bizarre. How come Hillary, Obama, and McCain can each find economists who support their positions on a whole variety of issues, but there are no economists who support Hillary and McCain on the tax holiday? Maybe its because the tax holiday proposal is so blindingly stupid that even an MBA could figure it out.
Unfortunately, botique blends ARE quite distinguishable, even when chemically similar, because every major city and many states have their own little gaggle of experts with an opinion on what blend of gasohol and additives will minimize specific pollution problems. Since air quality violations fall under EPA jurisdiction, the cities (in particular) have a strong incentive to stick with their guns in regards to what kind of fuel blends can be sold, and make violations legally punishable.
As such, the oil companies have to make a statistical prediction for how much gasoline will be demanded for each region of the country and each botique sub-market for a given period of time at a particular time of year, and then refill and distribute the stocks accordingly. Some blends can be shipped around to other markets if there is a surfeit in one area and a surplus in another, and some cannot. This is why gasoline prices can take seemingly bizarre, unconnected paths in different parts of the country when demand fluctuates sharply.
The only way I could see for this situation to be mitigated is for the EPA to define perhaps four blends, each designed to address a certain scope of air quality problems, and then get the Feds to hand every state an offer they can't refuse: You and every location within your borders can specify any of these blends without changes and no other blends whatsoever, or you can forfeit all federal highway monies for the next year.
That might restore some sanity to the gasoline distribution markets in this country, but if it were that easy in practice, I expect it would have been done already.
Not only does the United States import significant amounts of gasoline -- but we import some of it from France. And here's the fun part: Some surplus French wine is turned into ethanol, which they mix with their gasoline. So you may have just put a little Beaujolais in your tank.
[Insert your own wine joke here.]
As for Megan's general argument, I don't think I am alone when I say that it would be more persuasive if she sullied the purity of her theory with a few facts. The diagrams are pretty, but insufficient, by themselves. Especially considering how extreme her assumptions are.
The role of EPA in the use of boutique fuels is not very well understood. The choice of a boutique fuel as an anti pollutant strategy is in lieu of other strategies. There does not appear to be a very good analysis of the broad costs of boutique fuel blend use in comparison to the costs of other implementation strategies. It may well be at the current price levels there really isn't very much net benefit to be gained from the continued use of boutique fuel blends as an implementation strategy.
It should also be noted that some of the imported gasoline is available because European refineries end up with more than is needed when they produce diesel fuel, which is in great demand in Europe. Availability of surplus gasoline is thus determined in part by the European refinery run cycles.
It should also be noted that gasoline usage in the United States does seem to be dropping ,or at least not growing,as the economy drops. This leaves the Highway Trust Fund, the beneficiary of the federal gas tax, even more at risk and unable to support the levels of federal transportation spending now scheduled. The gas tax moratorium, unless replaced, will leave the Trust Fund in greater jeopardy and some impact on job creation in transportation capital investment down the road.
According to the EIA's most recent figures (05/02), refineries are operating at 85% capacity. A superficial scan of the historical data suggests this is somewhat low.
So supply may be elastic enough to accomodate increased demand at a marginally lower price point during a "tax holiday." Granted, it wouldn't be a full 18.4 cent saving to the consumer at equilibrium, but it would likely be something, no?
I can understand why economists would oppose it because of negative externalities. But if the price "won't drop" because demand is almost completely inelastic, why would they simultaneously argue that it will discourage conservation? Wouldn't it only discourage conservation if the price fell to begin with?
What am I missing?
Megan,
Contra to: "It's pretty generally known that refiners operate flat out during the summer months, and that in the short term.."
You & your Carbon-Tax friends, should begin to appreciate that consensus doesn't equate to Fact.
http://oilspot2.dtnenergy.com/e_article001031240.cfm?x=b11,0,w
If you learned anything during 'Enron', you may have picked-up on their, related-party, supression of generation capacity--driving nominal spot pricing higher.
Megan, did you actually just respond to Al?
Okay, then, I guess I will too:
Oh, right, because Obama picked a policy position.
Ah, welcome to the general election. We are now to expect several months in which the standard texts of Econ 101 and the opinions of every single economist in America, should they happen to jibe with Obama's policies, will abruptly become the far-left ravings of Muslim Communist black Christian terrorists.
I can hardly wait!
Posted by brooksfoe | May 7, 2008 1:43 PM
brooksfoe, just out of curiosity, are you really a Communist?
Any gas-tax revenue proposals must include language limiting those revenues to use by the highway department. California's roads are disintegrating even though we have the highest gas prices and gas taxes in the nation. This is because our gas taxes are being used to pay the salaries of schoolteachers and bus drivers.
Here's a blog by an oil/natgas analyst that posts a weekly update of refinery production, crack spreads, etc.
http://zmansenergybrain.com/2008/05/06/tuesday-still-more-earnings/
Have fun picking through real data, instead of "what is generally known."
Cheers,
Mark Hoffer objects to the empirical statement that oil companies operate "flat out" during the summer. "If you learned anything during 'Enron', you may have picked-up on their, related-party, supression of generation capacity--driving nominal spot pricing higher."
This really doesn't undermine the prediction that a temporary oil tax holiday in the summer of 2008 will not be passed through (to any meaningful degree) to consumers. If one believes the oil industry is reasonably competitive, and that refiners are at capacity for gasoline during the summer, then the competitive model predicts that the producers will keep most or all of the tax reduction as higher profits. If one believes that summertime capacity is tight because the oil companies exercise market power by strategically withholding the capacity from service, why would that change during a gas tax holiday? It is not plausible that the conspiracy to withhold capacity would be undermined by a temporary tax holiday (which will temporarily increase industry margins). Since we are now in the arena of cartel theory, just about anything is theoretically possible, including the possibility that the increased profits from the tax holiday will tilt each refiner's calculation toward cheating on the cartel by returning its withheld capacity to production. But does one really believe that oil companies clever enough to cartelize gasoline markets every summer by withholding capacity won't also be clever enough to hold the cartel together when it becomes even more profitable during a gas tax holiday? This seems an awfully slender reed for public policy. Especially when, to test it, you first have to give the oil companies a windfall increase in profits.
Clay says "I can understand why economists would oppose it because of negative externalities. But if the price "won't drop" because demand is almost completely inelastic, why would they simultaneously argue that it will discourage conservation? Wouldn't it only discourage conservation if the price fell to begin with?
What am I missing?"
First, the prediction that price won't drop is based on inelastic supply, not demand. Demand is also pretty price inelastic, but that is not where the price prediction comes from. As for the externality argument, perhaps the best way to interpret it is that any small price reduction from the gas holiday (say 2 cents out of the 18.4 goes to consumers because supply is very inelastic, but not perfectly inelastic) is a step in the wrong direction for the carbon tax guys. Not that it makes a huge difference in terms of pollution or externalities because the effects of a 2 cent price reduction would be trivial. So the concern is more on the level of principle -- we should be moving toward much higher carbon taxes, and any small step in the other direction is bad because it is going the wrong way. Put differently, under a tax holiday, the trust fund loses 18.4 cents per gallon. Consumers get at most a small percentage in lower prices (say 2 cents). This is stupid distribution policy -- surely we can find a way to give the consumers the equivalent of 2 cents per gallon in lower prices without having to give the oil companies 16.4 cents! And its bad environmental policy because it is a (tiny) step in the wrong direction. This tiny step is seen by some economists as horrible precedent -- they would argue that we should be educating the public as to the necessity of far higher carbon taxes, so any reduction in those taxes is badly off message.
Please explain the following:
A) Florida had a month-long gas tax holiday in August of 2004. $.10 a gallon dropped of the price of gas in Florida as a result, then promptly reappeared when the tax resumed in September. Why didn't the oil companies raise the price of gas to cover the droppped tax if your theory is true?
B) Why do prices vary by state (due to state taxes) anyway? Presumably supply and demand are not much different in Lakeland FL than in Valdosta GA, just up I-75, but check out the difference in gas prices should you happen that way.
John F -- A) the incidence of a tax depends (among other things) on the supply elasticity to customers in the geographic area of the tax. When supply is highly elastic, consumers bear the tax. Florida is a much smaller taxing unit than the whole US. Gasoline supply to the state of Florida should be pretty elastic, even when the elasticity of supply to the US as a whole is low. Refiners serving Florida can presumably serve other states in the southeast. Importers serving Florida can also presumably import to other ports on the Gulf Coast or Southeast. To maximize profits these suppliers have to earn the same per gallon net back profit in Florida and elsewhere. If the netback in Florida is Florida is suddenly higher because of a tax holiday, other suppliers will quickly begin selling in florida until the rough equality in netbacks is restored. Your facts suggest this was true. Was the tax also 10 cents per gallon, so it was passed through 100%? The economic studies of the Illinois tax holiday apparently find that about 25-50% of the tax cut was passed through to consumers. Given the difference in supply elasticities, the Florida and Illinois experiments do not imply that a tax holiday for the whole US would have similar effects.
John F -- B) There are two interrelated answers. First, the incidence of permanent taxes depends on long run demand and supply elasticities, while the incidence of temporary tax holidays depends on short run elasticities. Since long run supply elasticities are higher than short run elasticities, differences in permanent taxes will be born more by consumers than temporary changes in taxes. Long run supply elasticities are higher because in the long run suppliers can build new refineries, build new import terminals, build or expand product pipelines, modify refineries to produce more gasoline, etc. None of this is possible in the summer of 2008 in response to a tax holiday announced in May. Second, as discussed in my previous post, even the short-run supply elasticity to small geographic areas will tend to be high, so consumers will bear most or all of the costs of differential local or state taxes (and be the beneficiaries of local or state tax holidays).
Interesting update, but perhaps proof of how much people might go through hoops to avoid saying, "gee, maybe I shouldn't have held those views so strongly."
My state is well known for very low state taxes. If you go to any of the bordering states, the price of gasoline is higher by just about the same amount as the state tax differential.
The prices do not "equalize" in the summertime, much, whether or not refineries are operating flat out.
The reasons for this are multiple, but it's strong evidence that tax-increases and tax-decreases are pretty-much passed along, penny-for-penny. The rest is noise.
JonF. This would sem to some support for my analysis in my comment of 12:05; thanks.
Ken, your explanation makes some sense as to why a temporary gas tax holiday does indeed lower the price of gas. However it does not explain why people in adjacent cities in different states (e.g., Mobile and Pensacola or Lake City and Valdosta) pay different prices for gas due to permanent tax differences between states.
By background is in physics and I believe that all theories must be tested by real-world results. Too much economic theorizing flunks such tests so I am fairly skeptical about high-flown economic theorizing. (see also: the old chesnut about raising the minimum wage causing an increase in unemployment. Sounds good in theory, but reality doesn't obey that rule.)
Since this discussion is broadly about gasoline price elasticity, I wonder how many folks have read Robert Zubrin's book "Energy Victory" where he proposes that a government mandate that all new cars sold have Flex Fuel capable engines so that ethanol or methanol could be substituted for a substantial fraction of the gasoline when the price difference gets high enough.
Tight now, gasoline is kind of a Giffen good. When the price goes up, we pay more for it and buy less of other things. Flex Fuel engines allow substitution based on price.
To some extent, gasoline terminals are doing that right now with ethanol. They can mix up to 10% ethanol with gasoline at the terminal without legal consequences as long as the resulting mix continues to meet their local boutique requirements. As long as the wholesale ethanol costs less than the wholesale gasoline, they have an economic incentive to do this regardless of the arguments of energy content, mileage, emissions, volatility or any other considerations other than price per gallon. The market will tend to push the wholesale price of ethanol up to very close to the wholesale gasoline price unless there is a surplus of ethanol.
John F -- I tried to explain why a temporary nationwide gas holiday will NOT lower price of gas. Since you thought the explanation makes "some sense", hopefully you just left the NOT out of your sentence. And thanks for the compliment, but this is not high flown theorizing. It's applying the proper basic tool out of undergraduate micro. As usual, the trick is knowing which tool to apply. As for Mobile vs. Pensacola, let me try again. [I thought it was pretty clear, but if it's not clear to a reader who trying, it is not clear.] In the short run, producers have ample supplies that can easily be used to serve each city (send his gas trucks one way or the other, or tell the guy on the pipeline to take off more in Mobile and less in Pensacola). This is generally true in the gulf coast and southeast, because they are close to huge supplies headed to the rest of the country. The supplier will not make as much money as possible unless she/he earns the same profit margin from sales in both cities. If Pensacola cuts taxes by 10 cents per gallon, consumers in Pensacola should quickly see the full 10 cents in lower prices. If not, then suppliers have the ability (huge available supplies relative to Pensacola demand) and incentive to divert production there. If only half the tax cut is passed through, then suppliers earn 5 cents more in Pensacola than elsewhere, and additional supplies are diverted there. Put differently, the elasticity of supply to Pensacola is very high, and the economists' basic toolkit for tax incidence says when supply is highly (perfectly) elastic, consumers bears most (all) of the burden of excise taxes. But Pensacola is not a good analogy to the entire US. Where is the huge availability of more gasoline formulated to US regulatory requirements supposed to come from that would generate a high short-run (summer 2008) supply elasticity for gasoline in the US?
I asked: "I can understand why economists would oppose it because of negative externalities. But if the price "won't drop" because demand is almost completely inelastic, why would they simultaneously argue that it will discourage conservation? Wouldn't it only discourage conservation if the price fell to begin with?
What am I missing?"
Ken responded: "First, the prediction that price won't drop is based on inelastic supply, not demand. Demand is also pretty price inelastic, but that is not where the price prediction comes from."
In Ms. McArdle's analysis it seemed to come from both, though I may be mistaken. Isn't the basic logic that consumers will continue to demand all the gas at the same price point, so why would producers pass along the tax savings? And even if profitability increases for the producers, they can't put more gas on the market because they're running at capacity?
First, it seems they're not running at capacity. In fact, for the first week in May, their production is unusually low. I'm not sure why this is the case, but isn't it possible that an increase in profit (which the "tax holiday" would admittedly cause) would induce them to refine and sell more gas, thereby ultimately lowering the price to the consumer? I guess you think it might as well, just not by very much. I'm not sure I understand why you pick 2 cents though.
Ken wrote: "Not that it makes a huge difference in terms of pollution or externalities because the effects of a 2 cent price reduction would be trivial. So the concern is more on the level of principle -- we should be moving toward much higher carbon taxes, and any small step in the other direction is bad because it is going the wrong way."
Fair enough, and I understand the externalities/carbon tax issue. But it still seems intellectually dishonest to argue (in the same open letter) that consumers won't be helped price-wise, and that it will discourage conservation. It's essentially saying that while the effects will be tiny in both cases, when it's a tiny help to the consumer it's pointless, yet when it's a tiny disincentive to conserve it's disasterous. Can't have it both ways.
Thanks for your response.
Megan-this statement in your story is not true-"But there's no way to expand the supply,"-read www.fuelguru.com or fuelguru.blogspot.com to educate yourself on why that is.
fuelguru
Clay asks "In Ms. McArdle's analysis it [prediction that consumers will not benefit from the gas tax holiday] seemed to come from both [low demand and supply elasticities], though I may be mistaken. Isn't the basic logic that consumers will continue to demand all the gas at the same price point, so why would producers pass along the tax savings? And even if profitability increases for the producers, they can't put more gas on the market because they're running at capacity?"
I can't speak for Ms McArdle. But the correct basic micro answer is that, holding everything else equal, the pass thru to consumers will be greater with a low demand elasticity than with a high one. Can't draw a picture in this comment. But draw an upward-sloping supply curve, and draw two demand curves through a point P on that curve. One demand is more elastic (flatter) than the other. Now shift down the supply curve by an equal amount, to depict an X cents per gallon reduction in suppliers costs due to a tax holiday. Price falls more for the inelastic demand than for the elastic one.
Clay notes "First, it seems they're not running at capacity. In fact, for the first week in May, their production is unusually low. I'm not sure why this is the case, but isn't it possible that an increase in profit (which the "tax holiday" would admittedly cause) would induce them to refine and sell more gas, thereby ultimately lowering the price to the consumer? I guess you think it might as well, just not by very much. I'm not sure I understand why you pick 2 cents though."
I picked 2 cents as a plausible number when the supply curve is very inelastic (very steep), but not perfectly inelastic (supply curve is perfectly upright). Various commentors pointed out that they think this summer's capacity will not be that tight, and therefore supply might be more elastic that in the past. I no longer follow energy markets closely enough to have personal knowledge of capacity utilization. (Although, as you can probably infer from my comments, I have studied energy issues in a prior life.) But energy/oil is a huge industry, followed by many quite capable economists. If it were plausible to suppose that energy supply will be pretty elastic over the relevant range this summer (i.e. that unlike in most years, refineries will not be running at their full gasoline capacity this summer), Hillary and McCain could easily have found economists to confirm their stated belief that the tax holiday would benefit motorists. They can't find any economists to come forward! Economists like to disagree. And on many policy issues, one can find economists generally supportive of the candidate's position. Hillary and McCain each are advised by well-trained and capable economists, who have every incentive to find experts willing to support their candidate's tax holiday. Hillary, Obama and McCain can each find well-trained economists who support their differing approaches to health care. But no credible economists (or even an incredible economist) can be found in favor of the gas tax holiday. As an economist, I find that very compelling.
Regarding the economists open letter criticizing the tax holiday, Clay observes "But it still seems intellectually dishonest to argue (in the same open letter) that consumers won't be helped price-wise, and that it will discourage conservation. It's essentially saying that while the effects will be tiny in both cases, when it's a tiny help to the consumer it's pointless, yet when it's a tiny disincentive to conserve it's disasterous. Can't have it both ways."
Two things. First, intellectually honest people sign these letters when they think the position is right, and doesn't mistate anything. The letter lists X reasons why the tax holiday is bad policy. Some signers may think A is the really important reason, and B is also correct, but isn't all that important. Others may think B is the important reason. I don't see that as intellectually dishonest. I really doubt that anyone signing that letter thinks the tax cut will flow mostly to consumers, but it is bad policy nonetheless. That is a defensible position, but would be clearly inconsistent with the analysis presented in the letter. Second, when the tax holiday is only a tiny help to consumers, the right diagnosis is not that the policy is pointless. The holiday is far worse than pointless -- the intended benficiaries gain only a tiny amount, but the government loses a lot of revenue, and the oil companies, who presumably we are not trying to help out here, get to keep almost all of the revenue that the government loses. Certainly Hillary and McCain do not defend their proposal on the grounds that the oil companies will get most of the benefit, but that is the intent.
Ken,
I'm with you now on the elasticity piece.
But on the letter, let's look at what it says:
"First, research shows that waiving the gas tax would generate major profits for oil companies rather than significantly lowering prices for consumers.
Second, it would encourage people to keep buying costly imported oil and do nothing to encourage conservation."
These are still at odds, verging on being direct contradictions, no? If it won't "significantly" lower prices for consumers, then it also won't encourage people to keep buying oil. Maybe it's true that it will "do nothing to encourage conservation," but it won't do anything to discourage it either if Reason #1 is valid. Basically, if the price drop is so insignificant that it won't help consumers at all, why would they change their behavior to buy more fuel? Or in reverse, if they do buy more fuel, the price drop was obviously "significant" to them.
Either reason would certainly be enough to oppose the policy, I just can't agree that they're co-possible, at least as worded.
A lot of the opposition seems to hinge on the idea that it will increase profits for oil companies. I'm not sure why this, in and of itself, is negative. A lot of people work for, or are shareholders in, said companies. Would the decrease in government revenue outweigh the increase to the companies? Now the debate just goes back to competing philosophical perspectives on the proper role of government. It depends on how effectively you expect the government to use the revenue. I personally suspect they'd still find money for roads somewhere without it, but it's also true that would require either cutting spending elsewhere or running up higher deficits.
Clay,
I see your point. If written by one person, the letter would seem at least schizophrenic and unfocused. Even then, I wouldn't call it internally contradictory. But I can see how one can read it that way. Maybe I'm being too generous because I'm sympathetic to both criticisms. The tax holiday will accrue completely or almost completely to the benefit of the oil companies, and any tiny reduction in effective carbon taxes is a step in the wrong direction because of global warming is a serious problem. But even if the tax holiday went entirely to consumers, I can't see how an 18 cents cut in gas prices for two or three months by itself would be anything more than a scar on a pimple on the elephant's butt of global warming. So my belief on externalities and carbon taxes, which maybe I am too sympathetic in projecting into the group letter, is that America's carbon taxes are already far too low, a temporary cut this summer could easily get transformed into a permanent cut (the incidence of which would be largely on consumers, but that's a whole other story), and that would become a permanent step in the wrong direction.
As for the opposition to the tax holiday hinging on the fact that it increase profits for the oil companies, well, the opposition is heated because the McCain and Hillary justifiy the proposal on the grounds it will put more money in the pockets of suffering motorists. Since that empirical claim is false, Hillary and McCain take justifiably heated criticism. Hillary is especially exposed here -- how can a co-called populist get anyone to take seriously a proposal to help the average motorist that starts with step one -- oil companies can stop writing checks for about $9 billion in taxes in the summer of 2008, and let's setback in awe as this doesn't merely trickle down to consumers, it floods down immediately in the form of 100% pass through in lower prices. Especially when every economist in the world will tell her that the economics of this are pretty straightforward, and the oil companies will keep the money (not because they are evil or unpatriotic, or some such nonsense, but because short run supply is extremely price inelastic.)
Your last paragraph raises bigger issues on taxes,role of government,etc. However, a good place to start with sensible tax reform is another area where economists mostly agree, but no one listens to them. Eliminate the corporate income tax, keep sensible user taxes on companies and consumers (like gas taxes) in order to send sensible price signals on services (roads) the government provides, and hash out the efficiency/equity trade off issues in income or consumption taxes (to me, preferably progressive taxes on consumption). But I see no good argument in any of this for reducing gasoline taxes.
Having a professional online presence is key in the success of your business. Our Basic Package includes up to 5 pages with search engine optimization and more! Have your site up in 2 weeks or less! $300 Flat
Having a professional online presence is key in the success of your business. Our Basic Package includes up to 5 pages with search engine optimization and more! Have your site up in 2 weeks or less! $300 Flat