W. Max Corden
Professor of Economics
Australian National University
Despite the tariff reductions that resulted from the Tokyo Round, there has been a revival of protectionism in developed countries since the first oil shock and the recession of 1975. Macroeconomic developments help to explain this, even though the immediate motives for protection and the arguments used mostly concern particular sectors. The relationship between protection and the macroeconomic variables, especially the exchange rate, is not always understood, and has been underemphasized in theoretical work. There is a tendency to use protection as a bandage for problems that are essentially macroeconomic in nature, and, moreover, to ignore the macro-economic implications of trade liberalization proposals. In this article I shall try to clarify various relationships between protection and macroeconomic policy, especially the exchange rate.
Effect of monetary contraction
General unemployment usually gives rise to demands for protection of importcompeting industries. The unemployment may have resulted from policies of monetary stringency that were designed to reduce wage and price inflation—unemployment being an unavoidable, if temporary, effect, and possibly the principal channel through which wage restraint is obtained. If the monetary stringency is imposed by a country with a floating exchange rate, and is greater than in other countries, its exchange rate will appreciate, and this will increase unemployment in the tradable-goods producing sectors of the economy, as in the United Kingdom in 1980 and 1981. There are then pressures to mitigate the unemployment effects by protecting industries that are particularly hard hit or that are backed by particularly influential pressure groups. Sometimes the supposed justification will be that competing industries in foreign countries have some “unfair” advantage, but the real causes are domestic.
In such cases it is not always realized that protecting import-competing industries, in order to offset the adverse effects of the monetary contraction on them, harms the export industries. The exchange rate appreciates as aggregate demand falls, and capital is attracted into the country because of higher interest rates. Protection leads to further appreciation, reducing the competitiveness of the export industries even more.
The central message of comparative cost theory, that there should be no discrimination between export and import-competing industries, remains true even when the tradable sector as a whole is being squeezed, and even though there is unemployment. Protection in this case will simply shift more of the unemployment to the export industries and to those import-competing industries that do not obtain higher protection.
Budget deficit and exchange rate
One might also consider the type of situation in which the United States has found itself lately, notably in 1983 and 1984, where the issue is not one of overall employment but specifically the exchange rate. There is a budget deficit and something of an investment boom. Domestic savings are inadequate to finance both the deficit and private investment. If capital could not flow in from abroad, the interest rate would rise until private investment were sufficiently “crowded-out” and savings increased sufficiently for a private sector surplus to finance the public sector borrowing requirement. But with capital mobility and a floating exchange rate, the exchange rate appreciates and a current account deficit emerges, matched by a foreign capital inflow equal to the domestic financial gap. The appreciation of the dollar is part of the mechanism by which the current account deficit is generated: it reduces profitability in the tradable goods industries, and helps to shift resources into nontradables, as required by the extra demands generated by the fiscal expansion and the rise in private investment. The extra demand for tradables is met by imports.
The result of the appreciation is, again, to generate pressures for protection from various parts of the tradable sector. But again, if one part is protected, other parts would suffer more. Successful protection of some import-competing industries would move the dollar even higher.
The adverse effect on tradable goods industries that results from the process which generates the current account deficit may be of concern. It is this effect that gives rise to the protectionist pressures. Yet it is bound to be reversed eventually, since eventually the current account, excluding interest payments, must go into surplus as the interest bill payable abroad comes to exceed new borrowing abroad, and, even more, once new borrowing ceases and some repayment begins. Rational expectations should lead investors and industrial decision makers to take a long view about the prospects of import-competing and export industries that happen to be currently in difficulty. The alternative view is that the current exchange rate is sending out wrong signals and will lead to the decline of industries that may need to expand again later. If a temporary reduction in profitability of these industries is regarded as undesirable (because it sends wrong signals or because the temporarily low profits are unacceptable), then the remedy is to change the fiscal policy stance.
Exchange rates and protection
Distinct from the specific consequences of a budget deficit is the more general issue of whether exchange rate fluctuations increase protection. This has attracted particular attention in the United States, where it has been argued that fluctuations in the value of the dollar give rise to protectionist pressures. This then reinforces the usual arguments for policies to stabilize exchange rates. Many economists have argued explicitly that whenever the dollar appreciates, pressures for protection increase. This has happened in three periods: in the late 1960s and early 1970s (leading to the import surcharge of 1971 and a concerted, though eventually unsuccessful, attempt to introduce protectionist legislation); in 1975-76; and, again, recently.
Moreover, according to these arguments, the situation is asymmetrical: appreciations lead to increased protectionist pressures, and presumably actual increases in protection, while later depreciations do not fully reverse the process. There is an exchange rate protection “ratchet effect.” If this effect operated in all countries, then presumably protection would always increase somewhere: when the dollar appreciated, protection would rise in the United States, and when the other currencies later appreciated, protection would rise in those countries. In fact there are some signs that the effect is not wholly one way, so that protection in some countries (e.g. Japan) has tended to be reduced when their currency has been in its depreciation phase. But it seems plausible that there is some asymmetrical response, it being easier to increase than to reduce protection.
As stressed earlier, any increase in protection at a time of appreciation will only increase the appreciation and so exacerbate its effects on export industries and import-competing industries that do not obtain extra protection. But, insofar as there is such an effect, it must be regarded as a cost of exchange rate fluctuations, and strengthens the more common arguments for stability and greater synchronization in macroeconomic policies.
Protection via the exchange rate
In a floating exchange rate system without significant intervention in the foreign exchange market, the level of the exchange rate is a by-product of fiscal and monetary policies, and also of protection policies, in different countries. The exchange rate can itself be a policy instrument if the float is managed through intervention or, even more so, if it is pegged by the authorities. When the rate is pegged, sustaining a desired exchange rate has well-known domestic monetary implications. Here I am concerned with sterilized intervention, i.e. intervention that does not change the money supply because it is associated with open market operations or changes in reserve requirements that offset the direct monetary effects of the intervention. Hence, the domestic money supply is insulated from changes in the exchange rate, and intervention can affect the current account.
One motive for altering an exchange rate when it is used as a policy instrument is to protect particular industries. This can be described as exchange rate protection. The exchange rate may be depreciated—or an appreciation that might otherwise take place may be averted through intervention—in order to raise the profitability of the export and import-competing industries. In this case the protection is at the expense of the nontradable industries, whose relative profitability will decline. At the same time a current account surplus (or a smaller deficit) will be generated as a by-product, so that excessive lending, or lower borrowing, abroad is a cost of this form of protection. Sometimes this is described as “export-led growth” induced by exchange rate policy. Of course a depreciation can only succeed in protecting the tradable goods sector if nominal wages are slow to respond. If the rate is floating, governments rarely intervene to reverse the direction in which the market is moving it (which would be “aggressive intervention”); they aim instead to moderate its movements, i.e., to “lean against the wind.” When the intervention is to moderate an appreciation, the intention is usually to provide some exchange rate protection.
Exchange rate protection is to be contrasted with ordinary protection, which involves favoring one part of the tradable sector over others, and does not necessarily alter the current account. The cost of protection in that case is caused by the distortion within the tradable sector.
The macroeconomic implications of trade liberalization need also to be clearly understood. In many countries advocates of reduced tariffs and quotas must usually confront the question: “When jobs are lost in the protected industries, where will the new jobs come from?” Partial equilibrium analysis suggests that reducing protection would lower employment, not just in particular industries but in the economy as a whole, and would generate a current account deficit. This, broadly, would be true if the exchange rate were fixed and the general level of nominal wages were given. The unemployment could be eliminated by sufficient expansion of aggregate demand, or the deficit by sufficient reduction of it, but it is true that, with a fixed exchange rate and nominal wage level, and if the initial situation were one of balance, macroeconomic imbalance would result from large-scale trade liberalization.
The missing variable is the exchange rate. This must depreciate in real terms, whether through a reduction in nominal wages or through nominal depreciation. The depreciation will partially restore profitability of the previously protected industries and, more important, increase profits and hence employment in export industries and in import-competing industries that were previously not protected or protected below average. It is in the latter group of industries that most of the new jobs will be created.
There are also more complex, general equilibrium, repercussions to consider in answer to the “where will the jobs come from” question. The export industries may be relatively capital intensive and in the first instance may not open up as many new jobs as the decline in import-competing industries has closed down. There will be a shift to profits. A part of the profits will go in corporate tax payments, and if these are spent by the government, demand both for nontradables (and hence employment) and for tradables will go up. This will have the effect of depreciating the exchange rate further, and so will raise employment. If the net effect of trade liberalization is to raise the efficiency of resource use and hence real national income, demand for nontradables must rise, so that eventually some new jobs will be created in that sector. In the extreme case where export industries are highly capital intensive and foreign owned, greater dividend remittances will help depreciate the exchange rate and “re-protect” import-competing industries; higher tax payments will generate employment through extra government spending; and greater corporate savings will lead to more investment, either directly or through the capital market. Thus the new jobs can be produced at many different points, but the exchange rate adjustment is crucial.
This analysis has two implications. First, proposed policies for trade liberalization should normally provide for exchange rate adjustment, since in few countries can one rely on nominal wage flexibility. Second, the common argument against reducing protection, namely, that this would increase unemployment, is essentially based on partial equilibrium analysis and ignores the general equilibrium consequences which would follow if the exchange rate were flexible and appropriate macroeconomic policies were being followed. There will, of course, be transitional effects if protection were suddenly removed, especially in a context of low overall growth. Employment lost at one end would not be instantaneously offset by employment gained at the other end. Thus, there can be short-term localized unemployment, but this is the normal effect of any kind of change in the economy.
One qualification must be noted. Where real wages tend to be rigid, it is possible to construct examples where the net effect of trade liberalization, combined with exchange rate adjustment sufficient to maintain external balance, would be either to raise or to lower total employment. But even if the net effect is for employment to fall, there must be some new jobs created while old ones are lost.
Restrictions or devaluation?
For many developing countries the problem at present is not so much whether to liberalize, or how to do so without increasing a current account deficit, but rather how to reduce a deficit, or produce a noninterest surplus, at minimum social and economic cost. It is well understood that real spending, whether public or private, has to fall, and associated with this there must be a relative price change that makes production of tradables more profitable relative to nontradables.
The outward-looking approach is to allow some real depreciation of the exchange rate; this will foster both import substitution (and economize on imported materials) and export promotion. But extra exports may, at least in the short run, worsen the terms-of-trade, and the more barriers developing countries encounter to their export expansion in other developing countries, the more they have to lower their prices in order to sell extra exports in the markets that remain open. Thus a major world supplier may be justified in restraining exports in cases where their large expansion would have severe terms-of-trade effects; but by how much is a matter of judgment. This is a version of the classic terms-of-trade argument for protection applied to a situation where free trade, operating through a devaluation, would involve a substantial export expansion. The evidence suggests that, in spite of protection in the developed countries, manufactured exports from developing countries to developed countries have increased. Furthermore, countries that have taxed or otherwise discouraged their exports on the grounds of “export pessimism” have forgone market opportunities and reduced their growth rates. Hence it does not seem to me that one should place heavy weight on this terms-of-trade argument.
It must also be stressed that protection of import-competing industries is a very inefficient way to achieve the terms-of-trade objective—even if the objective is desirable. The terms-of-trade effect will vary with different exports, and may be important only for one or two exports—probably of primary products rather than of manufactures. Protecting import-competing production would hold back all exports relative to what would have happened if the exchange rate had been devalued more.
Protection not a remedy
The unemployment problem in Europe and the difficulties of import-competing and export industries in the United States have revived protectionist arguments. The European unemployment problem is probably caused primarily by particular real wages being too high in relation to the productivity of labor, the value of labor’s product having fallen because of low investment, technological changes, increased competition from newly industrializing countries, and so on. Tight macroeconomic policies may also have played a role. In the United States the problem is clearly attributable to the appreciation of the dollar. In both cases the remedies are to be found not in sectoral protection but in changed macroeconomic policies, in real wage moderation, and in a willingness to adjust to changes in comparative advantage. The revival of protectionism is understandable, but it must always be remembered that protection means helping one industry at the expense of another. This is ignored by the partial equilibrium view which is so common. Severe and unexpectedly adverse “shock” effects on particular industries may justify temporary subsidization, even temporary “safeguard” tariffs (as allowed by GATT rules), and, more reasonably, adjustment assistance, help in retraining labor, and so on. The danger about the provision of short-term protection is that it may be difficult to eliminate it once the need has passed.
It is possible on the basis of careful analysis to show that in certain cases there can be an overall national gain from protection, whether in terms of employment or aggregate output valued at world prices. This requires particular forms of protection, carefully devised. Much international trade theory has been devoted to the analysis of such cases, usually concluding that subsidies are to be preferred to trade restrictions. Lately, a new and sophisticated argument connected with strategic behavior among oligopolists competing in an international market has been explored in the academic literature. But there is little evidence that the pattern of protection that usually results from the interaction of pressure groups combined with popular partial equilibrium theories produces clear-cut national gains, even when the potential for some gains may exist.