Gottfried Haberler, “Import Border Taxes and Export-Tax Refunds versus Exchange-Rate Changes”

This file is a single-authored scholarly economics chapter, with an appended conference-prompted essay. Its scope is theoretical, historical, and policy-critical: Haberler asks whether import border taxes paired with export-tax refunds can replace exchange-rate changes, and whether border adjustment of general internal taxes is justified. His answer is that the tax-subsidy device is a useful equivalence in a model but a poor institution in practice.

It is well known that a uniform ad valorem tax of X per cent on all imports plus a uniform ad valorem subsidy of X per cent on all exports is equivalent, as far as commodity trade is concerned, to an X per cent devaluation of the currency.

That opening equivalence supplies the benchmark and the limitation. It concerns commodity trade, not services, tourism, or capital movements, and it assumes a uniformity that actual policy rarely preserves. Haberler traces the idea from Keynes’s 1931 proposal for import tariffs and export bounties through Hicks, Triffin, European VAT practice, and the 1968 Franco-German Ersatz adjustments. The genealogy matters because it shows how a seemingly neutral balance-of-payments instrument can slide into managed and discriminatory trade.

The chapter’s main conceptual move is to separate two questions often conflated: microeconomic tax neutrality at the border and macroeconomic balance-of-payments adjustment. On the micro side, Haberler distinguishes specific from general taxes. A tax confined to whisky may alter comparative costs and call for border adjustment; a general income, turnover, sales, or value-added tax does not, merely by being domestic, disadvantage national producers.

A perfectly general tax does not distort the comparative cost situation and, therefore, does not require adjustment at the border.

Ricardo is invoked as the classical authority for this incidence argument. Haberler rejects the GATT-style distinction between indirect taxes as border-adjustable and direct taxes as not border-adjustable. The relevant issue is whether a tax changes relative costs across goods or industries, not whether it is nominally shifted.

As can be seen, Ricardo also knew that it makes no difference whether the tax is direct or indirect so long as it is a general tax.

On the macro side, Haberler asks whether border taxes and export refunds should be used openly as external-adjustment tools. His answer is emphatic. The required rate depends on the size of the payments disequilibrium and trade elasticities, not on a country’s VAT or income-tax rate. More importantly, real schemes omit services, invite exemptions, and permit implicit discrimination among commodities and countries.

But now I come to the basic point: I maintain that the case against using the tax-subsidy system for balance-of-payments adjustment is overwhelming.

This is an institutional objection as much as a theoretical one. For countries without general export-subsidy machinery, especially the United States, the scheme would create a new administrative apparatus and a permanent target for industry lobbying. A mere import surcharge is worse: by omitting exports it is protectionist, politically sticky, and may provoke speculation if markets expect later devaluation.

The fundamental objection to the tax-subsidy system is that, in actual practice, the rate of tax and subsidy will never be uniform.

The appendix broadens the critique. Written after a conference and stimulated by Robert Z. Aliber’s proposal, it reviews Keynes, Hicks, Triffin, European tax-refund practices, and Soviet advice to Egypt. The same test governs each case: if the scheme is comprehensive and uniform, it is simply a thinly disguised exchange-rate change; if it is politically usable as a disguise, it becomes differentiated, discriminatory, and a path toward exchange control.

The work remains relevant for debates over VAT border adjustments, import surcharges, adjustable pegs, and speculative capital flows. Haberler does not deny the need for external adjustment; he denies that disguising exchange-rate policy as tax policy makes adjustment cleaner. His liberal economic conclusion favors transparent price adjustment, especially exchange-rate flexibility, over controls that pretend to be neutral while remaking commercial policy.

I conclude that the border tax on imports and tax refund on exports is an inferior, messy, wasteful, and inefficient substitute for exchange-rate adjustments.

The final methodological warning is that a valid analytical equivalence is not automatically a sound policy recommendation. Currency depreciation may be modeled as a uniform import tax plus export subsidy, but Haberler’s point is that abstraction becomes distortion once administered through actual tax and trade institutions.