The problem of the trade-exposed

In 1991, Larry Summers (then at the World Bank) initiated a storm of controversy when he wrote the following in a memo: “Just between you and me, shouldn’t the World Bank be encouraging MORE migration of the dirty industries to the LDCs [Least Developed Countries]?” Now back then it was an argument about pollution per se and not climate change. But it did give rise to a very interesting paper by Graciela Chichilnisky (American Economic Review, 1994) who demonstrated that differences in the definition and security of property rights between countries could give rise to apparent comparative advantage. In other words, it may be precisely because some countries have ill-defined pollution rights that they appear to have a comparative advantage in some ‘dirty’ production. In particular, asymmetries in property rights may exacerbate global pollution as the North (with well-defined property rights) overconsumes underpriced resource-intensive products imported from the South (without such property rights). Now one thing this analysis did is put ‘consumption’ in a par with ‘production’ over dirty goods. 

All this came back to me as I read Chapter 15 of the Garnaut draft report. As I mentioned, Chapter 15 is where it is all at. And it is there and in a strange appendix that the scheme to deal with the “truly dreadful problem” of the trade-exposed industry is exposited. Garnaut frames the problem of one where, if we price emissions properly in Australia, trade-exposed industries will face falls in production maybe well beyond what would occur if international competitors faced similar emissions pricing. He worries that by the time this occurs, Australian producers may never be able to bounce back. This leads to an suggestion that, absent an international agreement, we should identify and then literally subsidise the production of those industries.

The problem with this is that it frames the issue incorrectly. The problem with an emissions cap including emissions from trade-exposed industries is that emissions reductions may be an illusion. The emissions problem is a global one. We want a cap on Australian emissions below what would otherwise occur because it takes that volume of emissions out of the world economy. But if this leads to the asymmetries Chichilnisky considered, then Australia’s cap is not performing that function. Indeed, emissions may not be reduced by nearly as much across the world. And it is this that is the problem for trade-exposed industries — treat them the same as others and our cap is not as effective in mitigation.

So what do you do? The Garnaut proposal (absent international agreement which we all agree would be the best here) is to literally subsidise the production of these industries. In effect, a firm in that industry would buy a permit at a price, p, and then receive a subsidy, s, for each unit of emissions. In effect, their emissions price would be p – s. Of course, this leads to a real problem as to how to determine whether an industry is trade-exposed and what s should be.

But here is where I started to get really confused with the report. Garnaut worries that trade-exposed industries might not invest if under the cap but if they were protected and did invest then this would push up the permit prices in Australia and harm other businesses. And so protection that encouraged such investment would be bad.

The subsidy to production, however, seems to do the same thing. The higher the subsidy, the greater the demand for permits from trade-exposed industries and the higher the permit price (impacting on others). But the good news, is that the cap is still doing its job in that Australia hasn’t exported emissions production overseas. The point of the subsidy is to ensure that there is a stronger rationing of emissions elsewhere. So Garnaut is wrong about this: we ‘protect’ or ‘subsidise’ trade-exposed industries to ensure our cap is doing its global job. We don’t do this to protect other sectors.

That said, the subsidy arrangement seems complex and contentious. For starters, there is an easier way to deal with industries trade-exposed to imports — put a tariff on goods from countries that do not have an emissions trading regime. And we can work out what tariff should be and base it on Australia’s permit price. In effect, we require a permit for consumption of imported goods. That solves the issue for imports and also provides a natural way of removing the tariff as countries come on board. Indeed, it promotes incentives for them to do so.

For exports, it is trickier. But it seems to me that it would just be easier to identify these industries and leave them out from the initial emissions trading regime. They will already face some constraints because of emissions prices built into domestically supplied inputs like energy. For the rest, leave them out until the rest of the world comes on board rather than administer and negotiate subsidy arrangements. To be sure, this means that our reduction targets will have to be met elsewhere. But my point here is that they will be met whereas a subsidy seems to me to offer a situation whereby the cap could be hidden in a mess of accounting. We need real change and that requires transparency.

9 thoughts on “The problem of the trade-exposed”

  1. Production releasing carbon emissions should be taxed on a destination basis – so exports of alumina should be untaxed and only that portion of production consumed here taxed.

    Ditto for imports – if they are taxed in their home country leave them alone. If they are untaxed then wack a tariff on them at the boundary as Stiglitz and others have suggested. Again you are, in principle, taxing consumption not production. That is the principle.

    This gets rid of most of the carbon leakage phenomena you mention.


  2. Hi Joshua,

    Your instinct is correct. The right thing to do with exporting trade exposed industries is to leave their exports out of our system and allow the importing countries to impose a tax equivalent in impact to their own carbon mitigation tax (just as we tax imports from them). This is a tax subsidy to exports, of course, but has been the sanctioned procedure for ‘border tax adjustments’ in GATT/WTO since the 1960s because it turns out to be more transparent than the procedure of subsidizing exports.

    The overall impact of a system of border-tax adjustments for carbon taxes is, however, difficult to assess a priori and depends a lot on initial conditions in different industries. See here for some comments and a pointer to a recent Brookings conference on the general topic where W. McKibben made an interesting contribution.




  3. BTW Joshua, I remember reading the Chichilnisky paper when it first came out and thinking about the issue of how to achieve fame in economics.

    Her observation – and paper – seem obvious – if you don’t price an input (or underprice it) you will be observed to have a comparative advantage in producing goods that use this input intensively.

    The appearance may be deceptive because you are not internalising all costs. Living in Bangkok for a decade the truth of her claims ran into my lungs every day.

    OK it is probably sour grapes but her paper, is not ‘very’ interesting.


  4. Joshua, sounds sensible and I agree with you on everything but the tariff idea. Going down this protectionist route (and I’m not talking about protection from emissions) doesn’t sound that great for the international trading system.

    I don’t think subsidies would be so much more complex than taxing imports as you suggest. With subsidies you can reduce them over a pre-announced period at pre-determined rate but to give preferential treatment to traders depending on whether they have an ETS would also be messy. Who decides which ETSs in different countries are effective? or even a carbon tax? I can imagine this being a very tricky issue as ETSs and the collection of carbon taxes in developing countries (enforcement, collection and transparency) would not be as effective.

    Also, how does this stack up with current trade agreements? will we raise tariffs against US goods until they implement an ETS? do we reduce the tariff even if their emissions goals are not low enough?

    I’d think we’d have to be very careful about introducing tariffs…think of it as having the the opposite effect of unilateral trade liberalisation: it hurts us more than the country we’re sourcing the goods from! And the first obvious move for someone wanting to export to Australia would be to shift exports to another country. There are other ways to encourage importers to lift their game.


Comments are closed.