The Microeconomics of Export Restrictions

The typical protectionist measure is a limit on imports such as a tariff or a quota. Restrictions on exports are less common and less discussed but proposals to restrict the exports of natural gas have been in the news recently so perhaps a quick refresher on the protectionism of export restriction is in order. Restrictions on imports harm domestic demanders and benefit domestic suppliers but the harm is greater than the benefit so restrictions on imports create a net social loss. The basic result on export restrictions is similar, export restrictions benefit domestic demanders and harm domestic suppliers but the harm is greater than the benefit so restrictions on exports create a net social loss. The figure gives the analysis. As with import restrictions, the arguments for export restrictions soon turn to spillovers, networks effects, and other second round arguments. Without dismissing these in any particular case, the basic analysis suggest we should be wary of such arguments–the transfer always creates political opposition and any second round gains would have to be larger than the first round net benefits.


Addendum: Matt Yglesias comments. In brief he argues for the tax on Georgist grounds. Just because natural gas comes from land, however, doesn’t make a tax on natural gas equivalent to a tax on land. It’s the value of unimproved land that should be taxed not the value of the improvements, namely the extraction of the gas.


Nice chart, but what it ignores is any subsidy that may have been given to producers.

If you don't net out subsidies to producers when you permit them to export now at a higher world market price, you subsidize exports! That may be what you intend, but not if you gave the subsidy based on the premise of energy security.

So, let them export...but be sure to remove subsidies of all types, including tax subsidies, that were used to promote production.

Good call.

I also agree to remove all subsidies (direct and indirect) on all energy production, in fact I am very strongly in favor of additional taxes on primary energy production in the form of a carbon tax (which would further favor natural gas over coal I suspect). That said, do you have any evidence that there are any significant subsidies on natural gas production? I would be amazed if natural gas production isn't a significant tax payer already when the value chain from producer to consumer is taken into account.

Not that I really disagree about including subsidy in the calculation, but it's not obvious that having significant export business hurts the cause of energy security. In fact, I'd think it was the other way around. Exports let us maintain a production business larger than it would otherwise be, so that if serious trouble arises we have that capacity to produce available domestically. You can always turn off the exporting later if you need to. So subsidizing exports helps energy security. Whether or not that's something you really want to do is another question, but I think your reasoning doesn't quite work as stated.

But if the domestic supply of natural gas is limited, consuming more of it now (and exporting part of the additional consumption) hurts energy security by diminishing the "total" supply of domestic potential energy.

I suppose so, but the supply of natural gas is very large, so this is a very long-term issue. I think the idea behind energy security is that we can handle a decade of Mideast turmoil, not centuries of autarchy. The need is for more capacity to pull the stuff out of the ground and process it for use, not for more stuff in the ground.

Not to support export restrictions --because I don't favor them-- but this argument requires that caveat that the country has no market power over its export. If the U.S. becomes the Saudi Arabia of natural gas, then an export restriction can be used to keep the world price from being competed down by American supply. Depending on the relevant elasticities, the gain from controlling the reduction in price can more than offset the domestic harm outline in this post. For political and public choice reasons that manipulation is best avoided, but it sure is interesting to contemplate.

I think that opponents of gas exports are well aware that there is a net social cost but they have more important (to them) considerations:

Chemicals producers use natural gas as a feed stock and would love to have a price advantage.

Greens would just like for there to be no extraction of gas by fracking but at least there will be less of it if exportation is banned.

Considering the cost of export--the gas has to be compressed, cooled and liquefied it would seem that if we were to export carbon based energy, coal makes more sense from a physics standpoint. I suspect that in the absence of all restrictions, including problems in building coal export terminals, We would be exporting most of our coal and using most of our gas domestically.

No, all the analysis still shows that LNG is by far more favorable from a net CO2 emissions basis than coal.

I did not mention CO2 in my post but if you give it some thought you will see that the CO2 footprint difference is hardly clear.

My assumption would be that US coal will be burned here or someplace else and the same for CH4 as well. The CO2 impact would then be simply a matter of processing (lots for gas, little for coal) and transportation (about twice as much for coal since it has about half the energy per pound). These two things pretty much cancel each other out.

If you look at pg 9 here:,%208:49.pdf you will see that processing of LNG costs a bit more than transport by sea. Processing for coal is very low but should be about twice the cost/btu to transport due to the btu/lb difference.

Given the fact that the US is a quite large economy and a quite large natural gas source, shouldn't the world demand curve on this chart be downward sloping, albeit at a shallower slope than domestic demand? If world demand is downward sloping, does that change the analysis at all?

Peter H, The world demand curve IS downward sloping. It's not drawn here. It just happens to be downward sloping at a higher (greater) intercept. You are right though, it would be flatter.

Man that was long-winded. Could you not just have used the term "deadweight loss"?

"In brief [Matt] argues for the tax on Georgist grounds. Just because natural gas comes from land, however, doesn’t make a tax on natural gas equivalent to a tax on land. It’s the value of unimproved land that should be taxed not the value of the improvements, namely the extraction of the gas."

A possibly more stirring argument would be there is nothing Georgist about American nationals – for the sake of borders – benefitting from randomly (by some measures illegitimately) acquired unimproved gas. We should be doing everything we can to make those things available internationally. I stop short of asking for total unimproved land income distribution only because of huge public choice and rentier problems in emerging markets.

Excellent post, Alex. Another point - although one that goes beyond the standard supply-and-demand analysis that is the heart of your post - is that export restrictions might cause additional economic losses over time by making uneconomical domestic producers' expansion of their operations to take advantage of any greater scale economies that might be available with a larger market - a market beyond that of the U.S.

Lerner Symmetry Theorem.

But the value of unimproved land changes over time as technology develops. Much of the deep south was uninhabitable in 1900; but the invention of air-conditioning made it livable and the land value rose. New York city's land was already fairly valuable when we could build 10-storey blocks, but now we have the technology to build over 100 floors, and the land value is thus much greater. And so it is with our fracking fields: ten years ago they were dull farmland, today each one is a little goldmine.

The standard way of measuring the value of unimproved land is to ask competing users to bid for the right to use it. We already see this happening: drillers are bidding against each other to secure drilling rights from landowners. Some landowners have gained significant cash sums from fracking. Under a Georgist system that money would go to the state. Thus Yglesias is correct.

As Ritwik says, gas is land. Ignoring the surface, the mineral/extraction rights include a component of economic rents when the market price is over what would incentivise extraction. That is what drillers are prepared to pay land/mineral rights holders in rents, as you say.

However Yglesias is proposing *reducing* rents through export restrictions, and distributing the benefits at random, to whoever consumes the most gas. I don't believe that's in line with Georgist economic policy, even if it has effects on rents.

"Extraction of the gas" isn't an improvement, the facilities to extract the gas are an improvement. Matt's not talking about taxing those. Taxing the resource being extracted from the ground seems pretty similar to taxing a piece of land with a good view - the view is the resource.

Import restrictions favour largish businesses and their stakeholders (e.g. employees) at the expense of the broad public. Export restrictions do the opposite. So at least export restrictions are a more rational kind of populism.

Extracting the gas should not be taxed, of course. But the gas itself should be taxed. Or lets turn it topsy turvy: nationalize it and let the companies pay the government for the right to extract and sell it. Works well:

Can you imagine a restriction on the export of agricultural products? Agricultural products are subject to the impacts of various forms of subsidies.

Would cornflakes be cheaper?

What ethanol be cheaper?

Would more nations like Egypt the in uproar over the increasing costs of foodstuffs?

The argument for export controls on gas is that the US as a nation gains a greater competitive advantage - industrial and otherwise - from access to gas at half the price paid by others than it loses by foregoing exports of this single commodity. We lose that advantage if gas becomes priced globally like oil. What am I missing here?

A tax on mineral-land values would accelerate the extraction of natural resources, and an extraction royalty, at some point, would slow extraction. (Rich deposits would still be extracted as always, but marginal deposits would not, and exploration would be reduced.)

The western provinces of Canada do both. Corporations bid on the lease rights to mineral lands (mostly oil and gas lands). They also pay a royalty that is indexed to rise and fall with world prices. The lease rights are for ten years each.

The economics of this is well laid out in *Wealth of Nations*, book 1, chapter 11, "The Rent of Land." There is a section called "The Rent of Mines" which is a good starting point for analysis. Ricardo improved on Smith, but Smith had the basics down.

There's another author who similarly confused the value of fossil fuels in the ground with the value of the land.

Comments for this post are closed