Luc De Wulf and Garry Pursell
Developing countries have become increasingly conscious of the limitations of industrialization based on policies of import substitution, and have, therefore, adopted a variety of measures to promote industrial exports. An important reason for this shift in focus is that the package of policies used to stimulate import substitution, and the economic situation that it typically creates and maintains, discriminates significantly against the development of exports.
This discrimination exists for a number of reasons. The first is that the import regime (based on tariffs, quotas, and exchange controls) typically supports the exchange rate at a level significantly above what it would be with a more liberal policy. This reduces the profitability of exporting and in many countries contributes to an “inefficiency illusion”—the feeling that domestic production costs exceed world prices and that local industry would therefore be unable to compete (let alone export) in a less protected economic climate. Moreover, existing export opportunities may not be developed because production for the domestic market is typically more profitable and more secure. Further, import-substitution policies often allow the establishment of activities that are simply not suited to a particular country or its particular stage of development, and may permit technical and managerial inefficiency to prevail even in industries in which the country has a fundamental comparative advantage. Slack management, underutilized capacity, unexhausted economies of scale and specialization, lack of flexibility and adaptability, and resistance to innovation and new investments in already overcrowded markets often prevent firms from exporting their products.
Exports may be further discouraged by relatively high prices for intermediate inputs supplied by protected domestic firms, duties on imported inputs, licensing, or other import controls. The protection of domestic sales often allows industries to pass on cost increases to consumers, and so to support relatively high wages, which further increase the cost of living and the salaries required to attract foreign managerial and technical labor. Finally, import-substitution policies typically include low or zero tariffs on imports of capital equipment. Combined with relatively high industrial wages, this encourages the establishment of capital-intensive industries and the use of capital-intensive techniques, which are often quite inappropriate for the development of export markets.
These problems relating to the export sector could be solved by devaluing the exchange rate while removing import controls and reducing tariffs. However, reducing tariffs and dismantling import controls is usually not politically feasible, and in some cases a devaluation policy is ruled out. Consequently, many developing countries have opted for a more pragmatic approach by implementing various combinations of export promotion measures, sometimes combined with periodic currency devaluations. In this way a number of countries—for example, Brazil, Mexico, the Republic of China, and Korea—have succeeded in obtaining a rapid and sustained growth of exports, which in turn has contributed greatly to relatively high rates of growth of national income.
The export promotion measures typically implemented can be catalogued in five groups:
(1) those that increase the gross receipts from exports (such as export grants and exemption from export taxes);
(2) those that reduce the exporter’s costs (such as input subsidies, drawback or exemption of customs duties paid on inputs, and credit subsidies);
(3) those that specifically reduce the profit tax liability of the exporter (such as extra tax deductions, or favorable tax treatment of export-related profits);
(4) concessions on domestic sales in return for export performance (such as government backing of cartel pricing on domestic sales);
(5) other measures (such as export promotion targets and the establishment of national export promotion organizations).
Because export incentives are so varied, they are often difficult to analyze and compare, yet such a cost-benefit analysis is important and should be part of the framework within which the industrialization policy is cast. It should inform policymakers how the existing, or the planned, production incentives affect production for the export versus the domestic market and how the relative incentives vary among industries. Both the average level of these incentives and their structure deserve full attention.
|(1) Export price (f.o.b.)||10||10||10||10||10|
|(2) World value of intermediate inputs||2||8||7||11||5|
|(3) Value added in world prices (VAw)||(1)–(2)||8||2||3||–1||5|
|(4) Export grant||2||2||2||2||2|
|(5) Duty on intermediate inputs||—||—||2||—||3|
|(6) Net subsidy (S)||(4)–(5)||2||2||0||2||–1|
|(7) Value added in domestic prices (VAd)||(1) + (6)–(2)||10||4||3||1||4|
|(In per cent)|
|Nominal rate of subsidy||(6) X 100||20||20||0||20||–10|
|Effective rate of subsidy||(6) X 100||25||100||0||–200||–20|
Effective rate of subsidy
In principle, the various methods of promoting exports can be expressed as direct subsidy equivalents. For items such as import duties on inputs, export taxes, and direct export grants, the subsidy is the difference between the price paid or received by the firm and the world price. For indirect subsidies, the subsidy equivalent is evaluated in relation to an average or normal situation (such as nonpreferen-tial interest rates). The sum of direct subsidies and direct subsidy equivalents, expressed as a percentage of the f.o.b. (free on board) value of the exported product, is the nominal rate of export subsidy. Since some components of the net subsidy may be negative (such as import duties on inputs), this rate may be negative or positive.
As an indicator of the export incentive, the nominal rate of export subsidy is deficient, because it does not express the net subsidy in relation to the value of the export activity. Compare industries A and B, each exporting their product for an f.o.b. price of 10 pesos and receiving an identical export grant of 2 pesos per unit. Assume that there are no other taxes or subsidies, and that no duties are paid on imported inputs. Then the nominal rate of export subsidy is 20 per cent for each industry. However, A and B import different proportions of intermediate inputs on which no import duties are paid (for simplicity, no local intermediate inputs are assumed—see the table). Industry A, for instance, imports 2 pesos’ worth of intermediate inputs per unit of output, while industry B imports 8 pesos’ worth of intermediate inputs. The effective rate of export subsidy allows for this difference by relating the subsidy to the value added of each industry, rather than to the total value of its exports. In this way the concept is a direct extension of the effective protection concept, which is now frequently used to quantify the incentives to import-replacement activities given by tariffs and import restrictions.
More precisely, the effective rate of export subsidy (ERS) is defined as the percentage excess of value added measured in domestic prices (VAd) over value added in world prices (VAw), which is equivalent to the net subsidy (S) as a percentage of value added in world prices. That is,
Comparing industries A and B, it can be seen that while the nominal export subsidy is 20 per cent for both, the effective export subsidy is quite different because of the difference in the proportions of imported intermediate inputs. For industry A, the net subsidy is 2 and value added in world prices is 8, giving an effective rate of subsidy of 25 per cent. Industry B receives the same net subsidy, but value added in world prices is only 2, so the effective rate of subsidy is 100 per cent. Expressed slightly differently, the subsidy enables firm B to have a domestic value added of 4, which is 100 per cent in excess of the maximum possible value added of 2 if the net subsidy were zero.
In industry C, the export grant of 2 exactly offsets an import duty of 2 on intermediate inputs, giving a zero net subsidy. Hence, both the nominal and effective rates of subsidy are zero.
The effective rates of subsidy for industries D and E are both negative, but for different reasons. Whereas for E this result indicates the existence of a net disincentive (import duties on inputs exceed export grant), for D the incentive is positive but the activity is losing foreign exchange. Case E is likely to be the rule rather than the exception in countries in which export taxes or duties or intermediate inputs are not compensated by measures such as export grants or drawback provisions. Case D illustrates the dangers of excessive subsidization where there is a narrow margin between the world price of the finished product and the world value of its principal inputs. World Bank studies have shown that this has occurred in subcontracting assembly operations for export, and also in the local processing of a number of major export crops in West Africa. In the latter cases, the excessive subsidization occurred by providing the exportable inputs to the processing firms at prices well below world prices.
This method of measuring export incentives has the advantage of enabling policymakers to evaluate the overall incentive structure as it applies to production for both domestic and foreign markets. By comparing, for instance, effective subsidies given to import-replacing industries with effective subsidies obtained by the exporting industries, it becomes possible to discuss any discriminatory incentive structure and to assess what policy tools should be drawn upon to obtain an incentive structure that would lead to a more efficient utilization of scarce resources.
Estimating effective subsidies
To calculate the effective subsidy, estimates need to be made, first, of the world value added and, second, of the net amount of the subsidy, taking into account direct and indirect subsidies as well as any other cost-increasing provisions, such as import duties on intermediate inputs.
This may involve complications when the incentives are indirect. For example, if an export firm obtains preferential credit and tax concessions, its subsidy is the difference between the interest and the tax rates actually paid and those that would have been paid in the absence of the incentive legislation. If exportable raw materials are supplied to processors for less than world market prices, the subsidy to the processor is equal to the difference between the prices paid by the firm and the f.o.b. prices at which the materials are or could be exported. In practice, it may be difficult to make the appropriate price comparisons. When an incentive takes the form of government support of monopoly or cartel pricing in the domestic market, the problem is particularly acute, since it is seldom clear to what extent profits on domestic sales would have been lower in the absence of exporting.
Because of these difficulties, it may be necessary to estimate a probable range for the value of effective subsidies rather than single values. While this is unsatisfactory, it is an unavoidable problem that is inherent in any systematic approach to the formulation of trade and industrial policies.
Rules for uniformity
In practice, countries that have implemented export promotion measures have done so in a somewhat haphazard fashion, with the result that there is typically considerable variation in effective rates of subsidy as between exported products. For the reform of existing systems and for countries contemplating the introduction of export incentives, it is consequently of interest to consider the possible objectives of such policies and how these objectives may be achieved.
A primary objective of export-incentive schemes is to compensate the export sector for the bias in the existing incentive system favoring import substitution. This may involve two sets of instruments:
Specific compensatory export-incentive schemes that remove existing export disincentives to the point where the effective rate of subsidy (measured with the official exchange rate) would be zero rather than negative (industry C in the table). Such schemes would ensure that inputs are supplied to exporting industries at world prices (such as through refunds of import duties paid on inputs), and that credit is supplied and taxes levied at average rates. If some inputs must be purchased from domestic producers at more than world market prices, an offsetting subsidy would be required.
General compensatory export incentives that give a positive effective subsidy to exports so as to compensate for the overvaluation of the exchange rate associated with the import substitution policies. One general objective of these incentives is to equalize the positive incentives provided to the export sector with those given by tariff protection and other measures to import-replacement activities.
The level of these general incentives depends not only on the degree of the currency overvaluation but also on the probable response of the export sector to the incentive measures and to the level of protection that is deemed desirable for the import-substitution sector. If exports are highly responsive, the average level of positive incentives required may be quite low, since the expansion of net foreign exchange earnings may reduce the degree of currency overvaluation, thereby allowing a gradual reduction in the average level of protection to the import-substitution sector with only a minor adjustment in the exchange rate. On the other hand, if exports are not very responsive, a larger positive stimulus may be required and the possible rate of reduction of import-replacement protection may be correspondingly restricted. Although it is not an easy matter to estimate the average supply elasticity of exports, these estimates must be attempted in order to assist policymakers in evaluating the desired level of the export incentives.
General compensatory export subsidies should aim at equalizing incentives not only within the export sector but also between the export and the import-replacing sectors. This would tend to equalize the domestic resource costs of earning or saving foreign exchange. As pointed out earlier, a given effective rate of subsidy for all activities will correspond to different nominal rates of subsidy.
Deviations from uniformity
Uniformity of effective export incentives should be the starting point for an export subsidy policy, but there are a number of cases when deviations from this general principle may be appropriate.
Some countries or groups of countries may have a sufficiently large share of the world market of a given product (such as an agricultural or mineral product) that by reducing their supply they are able to increase the world price. In these circumstances a lower than average export subsidy, or even an export tax, may maximize net foreign exchange earnings. Note, however, that an overvalued exchange rate may itself constitute a sufficient disincentive to increased output in such sectors without an export tax.
It may also be desirable to impose an export tax on certain industries (especially agriculture, forestry, cattle, and mining) in situations where alternative taxes are administratively or politically impracticable or too costly.
If the structure of the exporting industry is not competitive, a somewhat lower than average subsidy may be paid in order to limit monopoly profits, especially if these would accrue to foreigners. However, it is important to take a reasonably long-term view of what constitutes a monopoly profit, as above-average profits for a number of years may attract new investment and lead to the expansion of a valuable export industry.
Finally, there may be a case for above-average effective subsidies for “infant” export industries for reasons similar to the often discussed case for protecting “infant” import-replacement industries. But extreme care should be exercised to avoid creating export equivalents of the many import-replacement infants that never grew up.
So far we have discussed effective export subsidies and various policy instruments used to stimulate overall export performance. However, some instruments are more adequate than others. Allowing all exporting firms to purchase their trad-able inputs at world prices and exempting manufactured exports from export taxes are specific compensatory measures and give no effective export subsidy. They should be considered first. While they are relatively simple to administer, they have the disadvantage of removing protection from local suppliers of intermediate products. These local suppliers may have to be compensated accordingly.
As a general principle, subsidies should be provided in such a way as to deal as directly as possible with the distortion they are intended to correct. Since the basic problem is usually the overvaluation of the exchange rate, theoretically the best procedure is to pay an export grant based on net foreign exchange earnings. This has the additional administrative—if not political—advantage that its amount is apparent, both to the government and to the subsidized firms, and that it can be varied according to the firm’s net foreign exchange earnings.
However, if for some reason export grants are not feasible—they may attract countervailing duties in some importing countries, or their budgetary cost may be all too visible compared with the other export-incentive schemes—it may be advisable to resort to other forms of export subsidy, such as input and credit subsidies, tax holidays and concessions, and extra profits on sales to the domestic market. Whatever the form of these subsidies, they should attempt to equalize the percentage of the total value of the subsidies to net foreign exchange earnings between exporting industries. However, the problem with an array of individually small subsidies is that their combined advantages may not be fully appreciated by investors, nor easily calculable by the government. In addition, they may be difficult or even impossible to adjust so as to maintain a fixed relation to net foreign exchange earnings between industries. They may also introduce some undesirable distortions, and their effects on exports may be uncertain. Credit subsidies and tax concessions that reduce the cost of capital, for example, will tend to bias input decisions in favor of capital-intensive techniques. Input subsidies may lead to the overutilization of the subsidized input, while monopoly advantages for sales in the domestic market have an uncertain relation to export performance.
The implementation of an export-incentive program will encounter some administrative and budgetary problems. These are, however, not necessarily more difficult to cope with than those associated with import-replacement industrialization. In particular, the net budgetary cost of a given economic expansion through further import replacement may well exceed the net budgetary cost of the same expansion achieved by an economically desirable level of export promotion.
From an administrative point of view, it will be difficult at times to allocate inputs and profits between domestic and export sales when firms supply both markets, and the firms themselves will be tempted to adjust their accounts to obtain the greatest benefit from the export-incentive program. Appropriate controls should limit such abuses, but care should be taken that these controls do not stifle the activities that are meant to be stimulated.
The budgetary costs of direct export grants, as noted earlier, have the advantage of being explicit. The potential exporter can assess how they alter his export possibilities, and policymakers, aware of the cost of their policy, will be less likely to continue them beyond the point where they cease to fulfill their objective. Indirect export subsidies lack these important advantages. However, although involving a loss of government revenue, and having similar budgetary implications, they may be more attractive politically, precisely because the budgetary cost is less apparent, and because they may be less likely to attract countervailing duties by importing countries.
The export promotion policies of a number of developing countries are currently being studied in a World Bank research project directed by Bela Balassa. The results of this study should throw some light on the problems mentioned in this article.
At the international level, as part of a long-run objective of reducing trade barriers, there is some opposition to the use of export grants. In view of this, and because of the potential advantages to them of a properly administered system of export grants, developing countries may wish to press the issue that export subsidies should be viewed in a development context during the next multinational trade negotiations. One possible way of doing so might be to distinguish between developed and less developed countries and to permit the latter group of countries a limited recourse to some incentive schemes which would continue to be prohibited when used by developed countries.