Do border adjustment taxes work?

If the Gillard/Garnaut plan for a car­bon tax is imple­ment­ed would bor­der-adjust­ment tax­es (BAT) help to off­set the trade impacts includ­ing loss of export com­pet­i­tive­ness and “car­bon-leak­age”?

It turns out BATs might notl have any impact, good or bad, on trade. In the late 1960s, adis­agree­ment between Europe and the USA over the pro­tec­tion­ist impact of bor­der tax adjust­ments start­ed to shape up as the first (of many) trans-Atlantic strug­gles over the exter­nal impact of Europe’s mar­ket inte­gra­tion. But, to every­one’s sur­prise, the issue fiz­zled out before the launch of the 1973 Tokyo Round of GATT negotiations.

A group of aca­d­e­m­ic ana­lysts includ­ing the leg­endary Har­ry John­son, using sim­ple mod­els of pro­duc­tion, demand and trade showed that remit­ting a uni­form, econ­o­my-wide prod­uct tax on exports (to pre­serve com­pet­i­tive­ness) and impos­ing the tax on imports (to cre­ate a ‘lev­el play­ing field’) had no impact on demand, on the rel­a­tive prices of imports or on local pro­duc­tion in the import­ing coun­try and no export sub­si­dis­ing effect.

The the­o­ret­i­cal analy­sis showed that it did­n’t real­ly mat­ter whether gov­ern­ments taxed pro­duc­tion (an ‘ori­gin-based’ tax) or con­sump­tion (a ‘des­ti­na­tion-based’ tax sys­tem, such as Europe had adopt­ed for VAT); rel­a­tive prices would not be affect­ed. Ulti­mate­ly, the econ­o­mists said, con­sumers’ bud­get con­straint in the tax­ing coun­try would be unchanged. Import demand, too, would be the same after adjust­ment in the rel­a­tive exchange rate (and pre­sum­ably, real wages). Nor would there be any change in the oppor­tu­ni­ty cost of exports, even with a remis­sion of the tax; because there would be no change in rel­a­tive prices in the tax­ing coun­try, there would be no export sub­sidy.

“Don’t wor­ry about BATs”, was the econ­o­mists’ mes­sage. GATT Par­ties let the mat­ter drop; by the mid-1970s a Work­ing Par­ty on the issue had decid­ed that bor­der adjust­ments were com­pat­i­ble with the GATT Sub­sidy rules, the rules on Sched­ules and on Nation­al Treatment

But would this slight­ly counter-intu­itive result hold for sim­i­lar bor­der adjust­ments hold for the pro­posed Gillard/Garnaut car­bon tax? If the pro­posed car­bon tax were off­set for exporters and imposed on imports would it real­ly make no dif­fer­ence to rel­a­tive prices and demand in Aus­tralia? More impor­tant, per­haps, would the prob­lem of “car­bon-leak­age” dis­ap­pear? Would the intend­ed com­pen­sa­tion to producers—to “lev­el the play­ing field” against imports and to pro­tect the com­pet­i­tive­ness of exports— be need­ed at all?

If the mar­ket mod­els used by the econ­o­mists advis­ing GATT applied accu­rate­ly to the pro­posed Gillard/Garnaut car­bon tax, there would prob­a­bly be no fur­ther grounds for con­cern about “car­bon-leak­age” and no con­cerns about tax­ing imports (assum­ing the right rate of tax can be struck) or remit­ting the tax on exports.

But are these sim­ple mod­els accu­rate in the case of a tax that is not uni­form but applies only to ener­gy use and is com­pen­sat­ed, prob­a­bly via social-pay­ments to con­sumers but nev­er ful­ly com­pen­sat­ed (oth­er­wise it would have no car­bon-abate­ment effect)? 

Because the tax increas­es the prices paid by con­sumers rel­a­tive to the prices received by pro­duc­ers, a car­bon tax should affect the sup­ply of real mon­ey bal­ances, in addi­tion to reduc­ing expen­di­tures. This looks like a recipe for eco­nom­ic con­trac­tion, to me.

Of course, the price of imports from all sources (in fact all trad­ables) will go up, thanks to the bor­der adjust­ment tax, by the same amount as the price of non-trad­ed goods and ser­vices. So, although our pref­er­ences won’t change between domes­tic or import­ed goods as the GATT econ­o­mists not­ed, our real effec­tive exchange rate (the price of goods in Aus­tralia com­pat­ed to the aver­age price of those same goods in our trad­ing parters’ mar­kets) will fall. 

The remis­sion of the full tax bur­den on exports of ener­gy resources, man­u­fac­tures etc will help main­tain export com­pet­i­tive­ness and, in prin­ci­ple, the fall in the real exchange rate should even assist exporters. But giv­en that the import tax match­es a tax on con­sump­tion push­ing up input prices for pro­duc­ers, the net impact of the tax remis­sion on exports, as the GATT econ­o­mists not­ed won’t make any change in the propen­si­ty to export.

Car­bon-leak­age would still be like­ly, if only because (absent a sud­den, mirac­u­lous appear­ance of a nuclear pow­er indus­try) Aus­tralia will be a less reward­ing place to pro­duce ener­gy or ener­gy-inten­sive goods. It will be a slow­er-grow­ing econ­o­my.

As the price of all goods pro­duced with tax­able ener­gy ris­es, real wages will fall and there will be pres­sure for an increase in real wages that the Labor gov­ern­ment will find hard to com­bat, despite the com­pen­sa­tion pack­age. I assume that infla­tion pres­sures will rise, too.

I’m not much of a macro-econ­o­mist, but if this analy­sis is right I expect the impact of even the com­pen­sat­ed, export-remit­ted tax to be a blow to our cur­rent prosperity.

Note: Econ­o­mist John Whal­ley has writ­ten a short paper on the his­to­ry of the BAT issue.


  1. Hi Peter,

    As an retired pri­ma­ry pro­duc­er now envi­ron­men­tal­ist and social ecol­o­gist I have no doubt that ear­ly tran­si­tion to car­bon sta­bil­i­ty is vital if we are to mit­i­gate the unmea­sur­able suf­fer­ing our gen­er­a­tion will inflict on future gen­er­a­tions by our greed.

    Pri­ma facie, it seem cen­tral that import BATs be applied in any suc­cess­ful car­bon tax system .

    -So domes­tic pro­duc­ers stay com­pet­i­tive. All domes­tic pro­duc­ers pay­ing the car­bon car­bon tax on their ener­gy inputs.
    ‑Except­ing those export trade exposed for whom adjust­ment grants be made over a twen­ty year peri­od for struc­tur­al adjust­ment (Whilst in business).These indus­tries must have incen­tive to change.
    ‑Equal­ly low­er income mem­ber of soci­ety be giv­en a pro­por­tion­al com­pen­sato­ry adjust­ment lift to pen­sions, social pay­ments and income tax thresh­olds to cov­er increased ener­gy costs.
    ‑House­holds above the thresh­old be giv­en a 20 year cap­i­tal struc­tur­al adjust­ment grant to reduce ener­gy con­sump­tion.
    ‑Numer­ous stud­ies have shown increase in totoal demand for labour with tran­si­tion loss to GDP have been due to leak­age of unsup­port­ed ener­gy inten­sive industries.(Spain)

    Whilst this may mean a con­trac­tion in the for­mal GDP ; Real GDP,(GDP minus “bads”) should even­tu­al­ly grow again after tran­si­tion costs are absorbed; espe­cial­ly in an expect­ed world of sky­rock­et­ing ener­gy costs.

    Levy­ing a car­bon tax cred­it­ed tax on imports would send pow­er­ful sig­nals to the inter­na­tion­al com­mu­ni­ty to get their house in order; ie to bring their tax to parity.

    Not being an econ­o­mist or hav­ing found rel­e­vant mod­els or hav­ing the time to study the Gar­naut report, expect I have missed salient point.

    Your com­ment would be welcome.

    cheers Dun­can

  2. You make no com­ment on mas­sive reduc­tion in exter­nal­i­ties (pol­lu­tion, road trans­port) and per­son­al costs (com­mut­ing) inher­ent in a no car­bon tax econ­o­my.
    Sure GDP per capi­ta may decline but what is the like­ly sit­u­a­tion in nett per­son­al income?

  3. Hi Dun­can,

    I apol­o­gise for not respond­ing to your com­ments. We may have to agree to dis­agree on the need for any action to reduce car­bon emis­sions: I can­not see any cer­tain val­ue in it (or, at most, if you’re high­ly risk-averse, a tiny val­ue that does not war­rant any expen­sive action).

    As to the impact of the Labor/Greens tax on car­bon diox­ide: Treasury’s mod­el­ling of the eco­nom­ic impact on aggre­gate demand (i.e. GDP) of the for­mer ETS plan showed a 1- 2% cut to GDP which SOUNDS small but is, in fact, a huge amount of mon­ey. Rough­ly ten bil­lion dol­lars every year. 

    If the Car­bon Tax has any­thing like that impact (it’s designed to) it will cer­tain­ly cut net per­son­al incomes! We can­not, as a nation, “com­pen­sate” our­selves for a loss in our nation­al income: so some­one has to pay and that “some­one” is, inevitably, house­holds like yours and mine.

    As for the esti­mate from Trea­sury: it is very like­ly an under­es­ti­mate of the costs because it seems that their mod­el assumes (it’s a massve, unre­al­is­tic, assump­tion) that labour mar­kets in Aus­tralia would instant­ly adjust to the new prices of car­bon-inten­sive pro­duc­tion. In oth­er words; they assumed away a big chunk of the real-world costs of this wretched experiment.

    Thanks for your interest.


Leave a Comment

Your email address will not be published. Required fields are marked *