Many developing countries have elected to foster the growth of local industry behind the protection of high tariffs and other import restrictions over the past few decades. Recently, however, this strategy has come in for a considerable amount of criticism. An excessive emphasis on industrialization through import substitution (building up local industry behind tariff walls) has been blamed for distorting resource allocation and for widening income inequalities in developing countries, while failing to provide adequate employment. Import substitution is also blamed for generating substantial idle capacity in the industrial sector of many of these countries.
Development theorists are now urging the developing countries to shift from inward-looking to outward-looking strategies, and to place high priority on encouraging the growth and diversification of their export trade. Many are heeding the advice. But, the path they have chosen to achieve this goal has not involved the substantial readjustment of overvalued exchange rates and the dismantling of import barriers that would have been favored by free trade advocates. Instead, developing countries have increasingly adopted fiscal incentive measures designed specifically to offset the handicaps suffered by exports in economies which have a long history of protectionism.
Trade policies for economic development
Proponents of both inward-looking and outward-looking development strategies basically agree on the need for some form of government intervention in favor of industry in developing countries. The quarrel with import substitution is not about the ends but the means. As many theorists have pointed out, the arguments for departures from free trade are not truly arguments for protection but are merely arguments in favor of departures from strict laissez-faire. The correction of distortions in the operation of the market mechanism and the adjustments in the domestic cost structure which are required to stimulate industrialization are best secured, in theory, with an appropriate combination of domestic taxes and subsidies. These can be designed to attack market imperfections at their source and to align internal costs with international prices. Protection, instead, uses tariffs and other trade barriers to raise the prices paid to domestic manufacturers to the same level as domestic costs. This creates a divergence between domestic and foreign prices, which generates distortions of its own and hinders exports.
The potential use of domestic taxes and subsidies to produce essentially the same results as protection, without simultaneously discouraging exports, has been extensively analyzed in theory. The protective effect of trade barriers-the stimulus to domestic production-could be achieved on a nondiscriminatory basis through direct assistance to industry. Domestic tax instruments would then act both as substitute revenue sources (in the place of tariffs), and as a means to provide the additional fiscal resources needed to finance the industrial subsidies. The redistribution effect of import tariffs, which in fact places the burden of subsidizing domestic industry on domestic consumers, would be avoided. Instead, domestic tax policy could be designed to obtain the equivalent consumption effect—that is, the contraction of private demand required to release the amount of resources needed by the public sector-with due consideration for equity and other tax policy objectives. The balance of payments effect of protection, would be sought primarily through exchange rate adjustments. Currency devaluation would raise the domestic prices of both imports and exports, thus drawing resources into both import-substituting and export industries, instead of placing the entire burden of balance of payments adjustment on imports (as in the case of protection).
Gains from outward-looking strategy
Countries adopting this open strategy can theoretically expect all the classical gains from trade. These include benefits from international specialization according to comparative advantage, stimulus to efficiency as a result of exposure to foreign competition and technology, and the prospect of a world-wide market for their products. Their industries would also reap the benefits of internal economies of scale that could not have been achieved by producing only for the limited home market available under protectionist policies. Import substitution and export promotion would no longer appear as mutually exclusive industrial strategies, but would proceed simultaneously and reinforce each other.
In practice, however, there are substantial obstacles to the adoption of the nondiscriminatory domestic tax-subsidy approach to industrialization in developing countries. A major obstacle is related to public finance. There are also difficulties in the use of exchange rate adjustments as instruments of balance of payments policy in developing economies.
The fiscal reason why foreign trade taxes are universally used in preference to domestic subsidies for providing industrial incentives is that import duties are usually a prime source of government revenue in developing countries, and one of the few major taxes whose collection does not impose undue administrative burdens. Broad-based domestic taxes are considerably more difficult to administer and are not nearly as productive in the early stages of development. The simultaneous operation of a direct subsidy system for the industrial sector would presumably also place considerable strain on weak fiscal and administrative structures. In contrast, tariffs provide not only revenues for the treasury but a seemingly costless subsidy to the protected industries. These are cogent, if highly pragmatic, arguments in favor of protection as the only feasible measure at the start of the industrialization effort.
With the growth of domestic industry on the other hand, customs revenue naturally falls off with a decline in import of consumer goods, while the domestic tax base widens. Import substitution, although it is successful in promoting industrial development, eventually costs the government its most traditional and convenient source of revenue; this loss must normally be compensated by increased domestic taxation, a usual trend in tax structure change accompanying economic development. As this evolution takes place, the fiscal argument for protection automatically loses some of its relevance. However, it may be quite some time before a domestic tax system is strong enough to generate the additional resources to finance domestic industrial subsidies on a significant scale.
The second basic factor behind the intensive use of import barriers in developing countries is that, in addition to their protective function, they also bear the burden of ensuring balance of payments equilibrium. Exchange rate policy is subject to a particularly rigid set of constraints in developing economies whose exports still consist essentially of primary products. If the price elasticity of demand for these products is low on the world market, there is a terms of trade argument for keeping an overvalued rate of exchange to maximize export receipts and for using trade taxes and direct controls to adjust the balance of payments. Generally, the windfall rise in income of the primary export sector and the increase in the domestic price of imports caused by devaluation tend to generate inflationary demand and cost pressures that spread throughout the economy, preventing the desired reduction of imports and the expansion of exports in the short run.
In such a situation, trade restrictions have often been used as a substitute for exchange rate adjustments. However, the constraints on exchange rate policy in the export economies have actually been used by Kaldor and others to make a case for differential exchange rates instead of protection. Their argument is that, if no single exchange rate can maximize primary export earnings and simultaneously provide enough inducement to import-substitution and export activity in the industrial sector to ensure overall balance of payments equilibrium, the solution would be to adopt two or more exchange rates.
Alternatively, the same results could be achieved with a combination of import duties and export taxes and subsidies designed to have the same effect as differential rates of exchange. Development theorists, such as Bela Balassa, have suggested the adoption of a basic exchange rate that would give adequate inducements to the production of nontraditional primary products, supplemented by taxes on traditional exports of primary products, as well as import tariffs and export subsidies on manufactured goods. The export taxes would be designed to take into account the lack of elasticity of foreign demand for the country’s main primary products, while the import tariffs and export subsidies on manufactured goods would amount to a de facto devaluation of the exchange rate on industrial products only. The tariffs and subsidies would be set at the levels required to offset any overvaluation of the basic exchange rate, as well as any particular market distortions or externalities, and to ensure temporary protection of infant industries, without discriminating in favor of import substitution in place of export production.
Under such a scheme, exchange rate adjustments can again be used in balance of payments policy. While the foreign trade tax-subsidy system preserves the effective exchange rate structure for the appropriate incentive to each sector of the economy, continuous adjustments must be made in the basic exchange rate to keep pace with the rate of domestic inflation. Thus, the real value of foreign exchange will remain constant! and renewed biases in favor of import substitution instead of exports will be avoided.
Within this general framework, exchange rate and foreign trade tax-subsidy measures may be viewed as largely interchangeable instruments of either industrial protection or balance of payments adjustment. The choice of a particular combination of instruments depends essentially on fiscal and other practical considerations. The level of the basic exchange rate itself is relatively unimportant as long as taxes and subsidies on commodity trade flows produce the appropriate effective rates, unless the exchange rate affects invisibles and capital transactions. It would be possible to add a uniform tariff on all imports purely for revenue purposes or to avoid the need for export taxes on primary products, with a correspondingly higher degree of overvaluation in the exchange rate and an equivalent increase in the level of subsidies for manufactured exports. All these steps would leave the effective exchange rates unchanged. At the other extreme, it might be possible to devalue the basic exchange rate to the point where it would provide sufficient incentive for exports of manufactured goods without the help of subsidies. Export taxes on primary products would then have to be higher, and import tariffs and quantitative restrictions would have to be more moderate, to compensate for the depreciated exchange rate.
Compensated devaluation approach
The latter combination, which has been called “compensated devaluation,” constitutes a well-defined alternative to the adoption of export subsidies in developing countries that originally took the protectionist approach to industrialization and must still resort, because of their dual economic structure, to foreign trade taxes as substitutes for exchange rate differentiation. Compensated devaluation consists of “a simultaneous and offsetting adjustment of the financial exchange rate and the trade restrictions such that all the commodity exchange rates for imports and traditional exports stay unchanged. The only net change takes place in the financial rate and in the nontraditional export rate. As a result, the latter obtain the equivalent of a subsidy.” (Daniel M. Schydlowsky, “Latin American Trade Policies in the 1970’s: A Prospective Appraisal,” Quarterly Journal of Economics, Vol. 86, May 1972, p. 283.) The potentially undesirable terms of trade and inflationary effects of the devaluation do not occur, since the accompanying trade measures prevent any change in the domestic price of imports and exports (except for industrial exports, which are being encouraged) and the level of effective protection for import substitution industries is not affected because it is merely extended to the export sector.
The extent to which compensated devaluation constitutes a realistic alternative to export subsidies depends on the circumstances in the country concerned. When the industrial cost disadvantage to be overcome is very large, the size of the required devaluation may pose problems in terms of the invisibles and financial transactions. A substantial increase in export taxes on the agricultural sector may not be feasible, for political or other reasons. And holding the line on domestic prices of imported goods may in some cases require import subsidies, which could pose budgetary problems. From the fiscal point of view, however, the overall impact of compensated devaluation versus export subsidies would depend on the interaction of a number of factors, including the level and structure of imports and exports, the predevaluation level of foreign trade taxes and restrictions, and the relevant demand and supply elasticities. It is by no means obvious which of the two alternatives would have the higher fiscal cost.
Fiscal, administrative, and political factors are usually the decisive elements in the choice between export subsidies and the alternative of a straightforward multiple exchange rate policy. Although in principle they have the same effect on trade, multiple exchange rates differ from foreign trade taxes and subsidies in their practical aspects, in particular in their fiscal implications, since, under the usual institutional arrangements, only the foreign trade taxes have a direct impact on treasury revenues and expenditures. Tariffs and export subsidies might also be preferred on administrative grounds, in view of the practical difficulties of operating a multiple exchange rate system.
The extension of protection from the import substitution to the export segment of the industrial sector through export subsidies avoids anti-export bias at a smaller fiscal cost than the domestic tax subsidy. It offers an intermediate approach to redressing the imbalance between import substitution and exports fostered by protection. While it does not directly affect the causes of the domestic distortions that hinder industrialization, it makes more moderate demands on the domestic tax system.
At a subsequent stage, as the fiscal and administrative operations permit it, part of the burden of industrial protection could gradually be shifted from tariffs and export subsidies to modest domestic subsidy schemes closely related to the objectives of industrial policy. The first step in this direction has recently been taken in Brazil, which had been giving heavy emphasis to industrial export incentives and has now started to grant assistance to the relatively unprotected machinery and equipment industry on a nondiscriminatory basis by extending to domestic sales the fiscal advantages previously accorded only to export sales. Thus, the overall level of protection for the sector is increased through government subsidies rather than through tariff barriers.
Domestic industrial subsidy schemes could be limited, like the Brazilian one, to industrial activities that generate the greatest indirect benefits to the economy. A more ambitious plan could involve the creation of a labor subsidy designed to offset distortions in the industrial labor market and to promote more labor-intensive techniques of production. Tibor Scitovsky has proposed a system of industrial protection based on labor subsidies for medium-scale and large-scale enterprises, possibly ranging between 10 per cent and 50 per cent of the unskilled wage bill. The extent of the subsidy required would depend, in each country, on the difference between the shadow price of labor (its true social cost) and the prevailing level of industrial labor costs. This subsidy could still be combined with a small, industrial tariff designed to provide some protection to firms too small to be included efficiently in the labor subsidy scheme and, if necessary special infant industry protection for other well-defined sectors.
Protection is also used sometimes to compensate industry for the poor quality or for the absence of public infrastructure and services in developing countries. This is still another area where potential trade-offs exist between import duties and export subsidies for industry, on the one hand, and domestic government expenditure, on the other hand. In the Philippines, for instance, export manufacturers establishing their plants in the less developed sections of the country are allowed to deduct from their taxes the full amount of their expenditure for construction and for maintenance of the necessary public infrastructure. Although undertaken on private initiative, such infrastructure is in fact a public investment, fully financed with tax money and transferred to the state upon completion.
Export subsidies are regarded by many theorists as having a definite role in the industrial and trade policies of developing countries. Since they are negative export taxes, export subsidies have essentially the same incentive effect on exports as protective tariffs have on imports. The same arguments that can be advanced to justify government intervention to protect industry in the domestic market may be used in defense of the production of goods for the export market.
Some form of export subsidization is also likely to be required to overcome the illusion of inefficiency that may affect the industrial sector in a protectionist setting when it confronts the world market. This illusion of inefficiency, as pointed out by Daniel Schydlowsky, stems from making international price comparisons at conversion rates that overstate the domestic costs and understate the domestic currency value of exported production. The growth of manufactured exports may also help domestic producers to become accustomed to external competition and may permit a more economical scale of operation, thus also contributing to the relief of the efficiency handicap associated with import protection. In Brazil, for instance, within a few years of the adoption of an export incentive system, manufacturers testified to the efficiency effect of export promotion policies by stating that, if government incentives were abolished, exports would not decline to preincentive levels, owing to the economies of scale realized. The growth of production for both domestic and export markets had reduced unit costs to such an extent that some producers had become competitive, even without the fiscal incentives.
Furthermore, depending on the size of the domestic market and the nature of the resource base, the best infant industry opportunities for some countries may well occur in export sectors rather than in import substitution sectors. It has also been argued that the economic costs of policies biased in favor of export promotion may be smaller and that the pay-offs may be higher than those of one-sided import substitution strategies.
Except in a few special cases, such as Hong Kong and Singapore, the pattern of industrialization in developing countries has always involved an initial phase of more or less intensive import substitution. The neglected export market becomes the focus of attention only at a later stage. Export growth comes about, in part, as the result of a prior buildup of industry behind protective barriers. A “spill-over” of import substitution into exports eventually takes place, after the exhaustion of the more immediate import replacement opportunities in the domestic market. At that point, export promotion often appears more promising in the short run, than the second phase of import substitution (in the intermediate and capital goods sectors), both as a source of continued growth, and as a means of increasing foreign exchange availability. Steps are then taken to remove the roadblocks that stand in the way of exports and supply manufacturers and to give them inducements to explore foreign markets. A number of semi-industrialized countries have now embarked on this phase, and export incentives accompanied by attempts to rationalize exchange rate and import policies have become an important element in the transition to more outward-looking strategies.
“the export incentive policies of the developing countries have thus far been characterized by a fair degree of chaos”
Practical problems seen
On the whole, however, the export incentive policies of the developing countries have thus far been characterized by a fair degree of chaos. This is due, in part, to the haphazard and piecemeal fashion in which these policies have evolved over the years; that is, in a somewhat experimental manner in the attempt to offset the increasingly serious distortions spawned by protectionism. Another factor has been the international disapproval of export subsidies, accompanied by the fear of retaliation by importing countries. Export incentive measures bear a stigma that is not attached to other government trade interventions. This explains why export incentives rarely take the form of outright subsidies, assuming instead a number of indirect forms which may sometimes be dictated by reasons of domestic practical convenience or political expediency and which also fulfill considerations of international acceptability and foreign market reaction.
The array of techniques that have been used by developing countries to encourage industrial exports is considerable. Some of the oldest devices employed to offset the anti-export bias of prevailing policies operated directly through the trade and exchange control system—such as special exchange premiums or bonuses for exporters, or straightforward multiple exchange rates. Import entitlements have also been used, ostensibly to provide manufacturers with scarce materials needed for export production. This use was equivalent to granting exporters a subsidy, except that the government did not need to finance the subsidy directly, merely transferring to exporters part of the gains normally accruing to importers as a result of quantitative restrictions and exchange controls.
At present, the most widely used export incentives are fiscal incentives. The degree to which they are accepted by the international community is somewhat blurred, and the basis for the approval or disapproval of certain fiscal devices is controversial. There is also a difference between the international code of conduct and its application, since some measures, although condemned, are tolerated in practice. Fiscal concessions have the advantage of using the existing tax administration system as a vehicle for their implementation, and they also offer a wide range of possibilities for pursuing particular policy objectives. A variety of tax concessions, rebates, and credits can be used to subsidize exports in a more or less covert fashion. They are often supplemented by a battery of credit subsidies, in the form of preferential facilities by official financial institutions, export credit insurance at subsidized rates, and government exchange guarantee programs. All of these measures are borrowed from the developed countries which resorted to them earlier and continue to use credit incentives much more intensively than the developing world. Finally, governments usually grant varying degrees of assistance to exporters in the form of free market information and sales promotion services.
The overall impact of the various export incentive programs in each of the developing countries is not easy to assess. Many incentives are difficult to quantify, since their costs and benefits are difficult to measure. However, in the same way as indiscriminate import protection can be expensive and wasteful of the economy’s resources, export subsidies that are excessive or ill-designed can be responsible for costly distortions in the allocation of resources and undesirable distributional biases, while their fiscal and administrative cost may be out of proportion to their benefits. The costs, benefits, and potential biases introduced by export incentives vary widely from one technique to another. Only a detailed analysis of the different devices could throw some light on a subject that has so far benefited from relatively little systematic thinking.