This paper is prepared for staff use and is not for quotation. The views are those of the author and not neces- sarily those of the Bank INTERNATIONAL BANK FOR RECCNSTRUCTION AND DEVELOPMENT INTERNATICTAL DEVELOPMENT ASSOCIATION Economics Department Working Paper No.95 TRADE POLICIES IN DEVELOPING COUNTRIES January 25, 1971 In this paper, presented at the December, 1970 meeting of the American Economic Association, the author briefly describes the trade policies followed during the postwar period in nine developing countries (Argentina, Brazil, Chile, Mexico, Korea, Western Malaysia, Pakistan, the Philippines, and Taiwan) and ex- amines the effects these policies had on their exports and econo- mic growth. Furthermore, the guidelines are suggested for trade policies in the developing countries. In writing this paper, the author has drawn on the findings of research he has carried out at the Bank in recent years. This includes a comparative study of the structure of protection, to be published for the Bank by the Johns Hopkins University Press, as well as studies on development strategies in semi-industrial coun- tries and on the industrial policies of Taiwan and Korea, circula- ted as Working Papers Nos. 34 and 86. Prepared by: Bela Balassa Consultant The purpose of this paper is to examine some of the implications of the author's findings concerning the effects of trade policies followed by developing countries on their export performance and economic growth. In the discussion, use will be made Qf the results of several studies carried out in recent years: an investigation of the growth strategies of six semi-industrial countries two of which, Argentina and Chile, are customarily included in the developing 1/ country category; a comparative study of the structure of protection in Brazil, 2/ Chile, Mexico, VWestern Malaysia, Pakistan, the PhilippiiLis, and Norway; and a 3/ study of industrial policies in Taiwan anld Korea. Following a brief summary of the findings, gu .delines will be suggested for trade policies by developing countries. I Instruments of Grade policy employed vy developing countries include import tariffs and surcharges, export taxes and subsidies, multiple exchange rates, as well as quotas and licenses. By affecting the relative prices of inputs and outputs, these measures influence the allocation of resources, including new investment, and provide incenbives -- or disincentives -- to import-substituting and export acti'vities. Discrimination among economic activities introduced by the use of trade 4/ measures exists to varying degrees in the nine countries studied. Argentina, Brazil, Chile, Pakistan, and -- to a lesser extent -- the Philippines provide considerable incentives to manufacturing industries at the expense of primary activities. The extent of discrimination in favor of manufacturing and against primary production is relatively small in Mexico; it is even less in Korea and Taiwan; and, on the average, virtually nil in Western Malaysia (for short, Malaya). -2- Discrimination in favor of manufacturing entails the protection of domestic manufacturing industries against imports whereas primary exports are penalized by tariffs on their inputs and by the lower exchange rate associated with protection. The protection of manufactured goods and dic rimination against primary exports usually go hand-in-hand, the principal exception being Malaya that levies a tax on its major primary exports, rubber and tin. In Argentina, Brazil, Chile, and the Philippines, there is further a substantial bias in favor of import substitution and against exports in protected manufacturing industries. Firms producing for home markets can get the domestic price rpised by high import protection while, in the absence of export subsidies, they could obtain only the world market price in exporting. By contrast, the bias against exports in manufacturing industries is relatively small in Mexico and MlIalaya where levels of protection are low and it is practically nionexistent in Pakistan, Taiwan and Korea. In Pakistan, manufacturing industries receive high protection irrespective of whether they produce for domestic or for foreign markets; in Taiwan and Korea the extent of import protection is rather low and, on the average, it is matched by subsidies to manufactured exports. II 'The trade policies followed during the postwar period have affected export performance and economic growth in the countries under consideration. The unfavorable treatment of primary exports has contributed to the decline in the shares of Argentina, Brazil, Chile, and Pakistan and, to a lesser extent, the Philippines in the world market for their major export commodities and has retarded the development of new exports. Malaya, too, has experienced a decline in the market shares of its major primary exports -3 - but it has expanded its minor exports which receive more favorable treatment. Finally, Mlexico and to an even greater extent Taiwan and Korea have been successful in raising the world market share of their major primary exports and in introducing new export products. In the latter two countries, these developments have taken place during the sixties following P. shift in economic policies from imrport substitution to export promotion. The expansion of primary exrnorts in individual countries has further been affected by world demand conditions in the markets for their major export commodities. The main beneficiaries of favorable world market trends have been Chile (copper) and the Philippines (oilseeds) while the slow growth of world demand and unfavorable price changes for rubber a:nd tin have depressed export earnings in MIalaya. The trade policies followed, together with changes in world market conditions, largely explain intercountry diflerences in the rate of growth of pnimary exports. Table 1 shows 2Uorea and Taiwan in the lead, followed bar Chile, the P1hilippines, 114exico, Malaya, Argentina, Brazil, and Pakistan. Comparisons of growth rates are less meaningful for manufactured exports because several of the countries under consideration started from a ver-y small base. Instead, we use the share of manufactured exports in malufacturing output and in total exports as indicators of success in exporting manufactures. In countries with a substantial bias against the exports of manu- factured goods, these commodities continue to account for less than 3 percent of manufactured outout and prrovide less than 10 percent of total exports. Among countries with a lesser bias, MIexico's manufactured exports have reached 5oercent of manufacturin5 output and 25 percent of total exports. In .Malaya 10 perc'i t of manufactured output is exported although, given the relatively7 low share of maniufacturing and the high share of exports in GNP, manufactured goods orovide no more than one-tenth of total exoorts. O @ Mra i " 5ar A xc~ ' t 'S r r. . ... . .............................. SItr Shrat- 1rL *azz ............................ r l%ffi ZJ -k.;a>js lux- ss.tfirA.:At~s+\.. ....................................L Lefr tc.z .... 1W.za x o.s.. ...... ...+se.z Table 1 EXPORTS IN SELECTED DEVELOPING COUNTRIES Argentina Brazil Chile. Mexico Korea Malaya Pakistan Philippines Taiwan Average annual rate of growth of exports Primary goods 1950-60 0. L -0.9 5.3 1 5.4 -0.2 1960-69 3.6 4.2 8.4 L xi t - 6.o 8.6 1950-69 1.7 0 6.8 5.7 8.1 Manufactures 1950-60 -7.6 15.6 11. 4.7 29.5 1950-69 17.3 19.1 10.7 AVV\ 4 . 25.0 34.2 1950-69 3.5 16.3 10.9 13.3 32.5 All cornodities N - 1950-60 -0.5 -0.6 5.6 J t 5.4 3.o 1960-69 6.823. 10-94.v6 5. 1 8 05 6 .83 23.e1 1950-69 1.9 1.1 6. 14.7 Manufactured exports as a percentage of outpivt- 1969 2 1 3 5 16 10 8 3 36 total exports - 1969 10 9 6 25 76 10 51 11 67 fiote: For Taiwan and Korea the base year is 1953 instead of 1950. For Brazil, mile, 1ala a, Pakistan, and the Philippines, the terminal year is 1968 instead of 1969. Source: 1%Tational and international trade statistics. Table 1 EXPORTS IN SELECTED DEVELOPING COUNTRIES Argentina Brazil Chile Mexico Korea Malaya Pakistan Philippines Taiwan Average annual rate of growth of exports Primary goods 1950-60 0.1 -0.9 5.3 4.0 -3.5 1.9 -7.5 5.4 -1.2 1960-69 3. 4.2 8.4 5.5 21.0 0.6 2.5 6.0 16.4 1950-69 1.7 0.6 6.8 lj 7 16.7 1.1 -2.9 5.7 6.3 Manufactures 1950-60 -7.6 15.6 11.1 12.2 0.2 29.0 35.0 4-7 30.5 1960-69 17.3 19.1 10.7 19.9 09.0 1247 14.5 25.0 34.0 1950-69 3.5 16.3 10.9 15.8 35.6 19.0 24.0 13.3 32.5 All commodities 1950-60 -0.5 -0.6 5.6 5.0 -2.8 2.0 -2.3 5.4 3.6 1960-69 4.6 5.1 8.5 7.2 38.9 1.2 6.3 6.8 24.0 1950-69 1.9 1.1 7.0 6.0 18.2 1.7 0.4 6.o 14.9 Manufactured exports as a percentage of output -1909 2 1 3 5 18 10 8 3 36 total exports -1969 10 9 6 25 76 10 51 ID0 o7 Note: For Taiwan and Korea the base year is 1953 instead of 1950. For Brazil, Chile, Malaya, Pakistan, and the Philippines, the terminal year is 1968 instead of 1969. Source: iNational and international trade statistics. Mn Taiwan anld Korea, manufactured exports have been stimulated to a considerable extent by the adoption of export-oriented policies around 1960. As a result, the exports of manufactures have increased to a considerable extent in the two countries, both as a proportion of total exports and of manufacturing output. By 1969, these proportions reached 67 anld 37 percent in Taiwan and 76 and 18 percent in Korea. Follolwing the introduction of the Export Bonus scheme in 1959, manufactured exports have also assumed importance in Pakistan, with jute and cotton textiles being the principal items. But, owing to Dolicies penaliz- ing primary exports, foreign sales of raw Jute and cotton have declined in an amount exceeding the rise in textile exports. M-Ioreover, higlh subsidies to manufactured exports together with the high protection of impor- substitutes, have imposed a substantial cost on the national economy. Owing to their favorable performance in both primary and manufactured exports, Korea and Taiwan are far ahead of the other countries studied in terms of the expansion of total exports. (Table 1) Dports have increased more than the average also in Mexico where the growth of tourism and border trade (not included in the export figures) have further contributed to increases in foreign exchange earnings. For reasons noted earlier, exports have risen relatively rapidly also in Chile and the Philippines while increases have been small in Argentina, Brazil, and Pakistan. III The expansion of exports contributes to economic growth directly by raising national income and indirectly by providing foreign exchange for the import needs of the domestic economy. An export-oriented policy also permits specialization according to comparative advantage--both between primary and -,anufactured activities and within the manufacturing sector. In particular, exports of manufactured goods enable finrs to lower costs by employing large-scale prcduction methods, reducing product variety, and participating in the international division of the production process throughl the manufacturing of parts and components for assembly abroad. Last but not. least, familiarity with foreign markets provides incentives for technological change and product improvement. Import substitution, too, can be a source of economic growth in particular cases. A number of developing countries attained rapid rates of growth of manufacturing output and, to a lesser extent, national income, in the early stage of import substitution which entails replacing the imports of nondurable consumer goods and their inputs by domestic production. Industries producing such commodities are the prime candidates for import substitution in developing countries since they employ chiefly unskilled and semiskilled labor, do not require the application of sophisticated technology, and need few in- puts from ancillary industries. Nor does the limited size of national markets constitute an important handicap for the development of these industries since the efficient scale of operations is relatively low and costs are not substantially higher in smaller plants. But, in the absence of exports, the expansion of industries producing nondurable consumer goods and their inputs necessarily slows down after imports have been replaced since domestic production cannot continue to grow faster than home demand. MIoreover, in the small domestic markets of developing countries, increasing difficulties are encountered in import substilution in other intermediate products, capital goods, and durable consumer goods. These cormmodities have higher technological and skill requirements, require the availability of materials, parts, and components from other industries, and need large-scale production for efficient operations with costs being -7- cs'-bslanbially higher at lower outout levels. Last but rnot least, in the everit of continuing, orotection,u there will be few inducements for technological improvement's. These considerations help to explain intercount-y differences in rates of Cco10.tioc - rr0LC ('Table 2). In Taiwai and Kforea, the growth of GDP has accelerated to a considerable extent followin- the ado-ption of e, ort-oriented Dolicies. In exland.in- the e-mmorts of nondurable cons-nier goods, the two co-tltbries have utilized their educated manpower while the capital requirements of -these industries are relatively low. The leading role of e,xorts in the -rowth process is iidicated by the high incemental ratio of exports to GDP; 5/ in 1960-69, this ratio was 39 percennt in Taiwan and 29 percent in Korea. Si-orts have also importantly contributed to '-exicor ' economic growth an-d the rela ively low protection of :nanuufacturing 'industries has limaited the cost of import substituation. In turn, the low deoree of discrimination a'nong econom,lic activities has rnade it possible for M-alaya to attain a rate of growth of national income substa-ntially above that for exports. The remaining countries of the group are characterized by import substitution b'ehind high protective barriers. In these countries, the relationship between import substitution and economic growth has been influenced by -their maarket size and the level of their economic development. Thus, the expansion of manufacturing outout has slowed do-vm to a considerable extent in the Philippines after the mid-fifties by which time the "easy" stage of import substitution had been by-an-d-large completed. Despite rapid increases in e<;norts due to favorable market conditions, there has been a decline in the rate of growth of GDP as well. Table 2 ECO.Y0.-ITC GROWTH IN SELECTED DEVELOPING COUNTRIES Argentina Brazil Chile MJIexico Korea Malaya Pakistan Philippines Taiwan Average annual rate of growth of value added Agriculture .1950-60 2.3 4.7 1.2 5.4 2.3 3.2 1.4 5.1 3.9 19-S0-69 2.0 4.2 2.5 4.0 4.6 3.9 3.7 4-/ 5.0 1950-59 2.1 4.5 1.8 4.5 3.6 3.0 2.5 4.9 4.5 , Manufacturing 1950-50 4.7 8.8 3.3 8.o 13.6 5.1 7.8 10.2 10.1 1930-69 4.6 5.9 5.9 2.0 16 . 0 11.7 8.6 4.5 16.1 i 1950-69 4.6 7.5 4.5 8.4 15.0 8.0 8.2 7.8 13.5 GDP 1950-60 3.4 5.8 3.7 5.8 5.0 4.1 2.5 6.8 6.9 1960-69 3.4 4.3 4.5 7.1 9.2 5.7 5.6 5.1 9.9 1950-69 3.4 5.1 4.0 6.4 7.4 4.8 4.0 6.1 8.6 Per capita GDP 1950-60 1.4 2.8 1.2 2.8 3.0 1.1 0.4 3.5 3.1 1960-69 1.8 1.3 2.3 3.6 6.4 2.5 2.9 1.6 6.6 1950-69 1.6 2.1 1.7 3.2 4.9 1.7 1.6 2.7 5.1 Populationi 1950-6O 2.0 3.0 2.3 2.9 2.0 2.9 2.1 3.2 3.6 1960-69 1.5 3.0 2.2 3.4 2.6 3.1 2.6O 3.4 3.0 -950o-69 1.8 3.0 2.2 3.1 2.3 3.0 2.3 3.3 3.3 ilote: For Taiwan and Korea the base year is 1953 instead of 1950. For Brazil and Malaya, the terninal year is 1968, for the Plhilippi-nes, 1967. Source: National and international statistics. Argentina and Chile had reolaced prac-tially all nondurable consumer goods and their inputs before the period under consideration, and their small domestic markets have made the expansion of industries producing other intermediate products, capital goods, and durable consumer goods both difficult and costly. These countries have built up an industrial structure which entails the use of small-scale and often out dated production methiods, inadequa-te specialization, and the manufacturing of 6/ products of low quality. Discrimination against agriculture, associated with the high protection of manufacturinlg activities, has further hindered their economic growth and helps to explain that they have experienced the lowest growth rates among the nine countries studied. Brazil, too, had completed the first stage of inport substitution prior to the period under consideration. Its large domestic market, how- ever, provided possibilities for the continued expansion of manufacturing during the fifties, mostly in intermediate products, capital goods, and durable consumer goods. But, as the possibilities for irnport substitution have keen increasingly exhausted, industrial expansion has slowed down in this countr-y also. Pakistan had practically no industry prior to independence and it was able to achieve rapid rates of economic growth by substituting domestic production for the inrports of nondurable consumer goods and their inputs. Subsequently, the adop-tion of tlhe iuport Bonus scheme has contributed to its relatively rapid industrial expansion. Continuing discrimination against agriculture has however adversely affected the growth of the Pakistani economy. If national income is measured at world market prices rather than at the do- mestic prices distorted by protection, increases in per capita termls appear to have been small. - 10 - IV The experience of the countries under consideration suggests the conclusion that while the protection of the manufacturing sector may permit rapid growth at an early stage of import substitution, it will eventually have adverse consequences for economic growth. Discrimination among industries does not permit specialization according to comparative advantage;the high protection of domestic industry induces the establishment of high-cost import- substituting activities; and the bias against exports retards the development of manufactured exports. Finally, in the absence of foreign competition, there will be little incentive for technical progress in small protected domestic markets. The increasing difficulties experienced by countries at higher stages of import substitution have recently led some governments to reconsider their economic policies. In Argentina, the extent of discrimination against pri- mary production and exports has been reduced ald the protection of mainLfac- turing industries has been moderated through a simultaneous devaluation and a lowering of tariffs. Manufactured goods also receive export subsidies in Argentina and such subsidies have assumed an important role in Brazil. Furthermore, in Chile, an effort has been made to lessen the degree of over- valuation of the currency while subsidies have been used to promote the exports of manufactured goods. Efforts made to reform the structure of protection, however, have Epne only part of the way and further progress is made difficult by resistance on the part of vested interests. Businessmen are opposed to changes in the status quo which ensures corfortable profits, and -they demand continuing protection from foreign competition, whlether this comes from the industrial countries or from developing nations as in the case of LAFTA. Additional problems are that transition to a more open economy would entail dislocation in particular industries and regions. It appears, then, that once an industrial structure geared to import substitution has been established, change becomes increasingly difficult. This observation points to the need for making appropriate policy choices at the time when a country embarks on an industrialization program. In the following, guidelines will be suggested for trade policies by developing 8/- countries. Apart from their application to countries at an early stage of industrialization, the guidelines can provide a basis for improvements in the policies of countries presently engaged in import substitution. V International trade theory tells us that sm4ll countries whth do not affect the prices of their exports and imports will maximize welfare by specializing in accordance with price relations on the world market. Deve- loping countries can generally take their import prices as given and they will not affect the prices of most of their manufactured exports either. This will not be the case, however, for traditional primary exports whenever increa- ses in the country's exports lead to a fall in prices. For these exports, then, the relevant decision rule will involve equating marginal costs to marginal .revenue from exports rather than to price. This call be accomplished by converting foreign exchange eami:gs from such exports at a less favorable exchange rate or -- what amounts to the same - - imposing an export tax on them. Export tax rates on individual commodities should be set by allowing for the elasticity of world demand, the country's share 9/ in world exports,, and the possible reactions on the part of foreign competitors. - 12 - The application of these measures would take accowit of market limi tations for traditional primary exports, without unduly discouraging their production as has often been the case in the past. A further question is whether mIau- facturing industries should be favored over nontraditional primary production, and if so, to what extent au-d by the use of ihat measures. In this connection, note should first be taken of arguments for infaint industry protection, designed to compen.sate for assumed differences between social and private profitability. On the firm level, such differences may arise if the lack of credit facilities, the overestirmation of risks, or simply, the desire to exclude the possibility of bankruptcy provide disincentives for investment, although eventual cost reduc-tions through the learning process or tlhrough increases in the scale of operations would make the investment socially desirable. Other instances are when some of the benefits of the pioneering firm's activities are enjoyed by others who utilize the knowhow generated by the firm or hire away skilled labor and technicians it has trained. It has often been said that infant industry arguments justify using production subsidies rather thani tariffs since the latter limit the size of the domestic market by raising the price of the commodity in question. '3ut while tariffs contribute to government revenue, subsidies represent a claim on this revenue. Budgetary reasons, then, may explain why developing countries use tariffs in preference to production subsidies. In fact, tariffs often account for a large part of government revenue in these countries, aind their replacement by other forms of taxation may encounter practical difficulties. The arguments for subsidies in preference to tariffs gain in force in cases when a particular distortion or cost disability needs to be corrected. - 13 - This will be so if the cost of industrial labor to manufacturing enterprises exceeds its social cost in the form of the output foregone in primary activities from which labor is drawn. In some overpopulated countries, such a situation may exist on family-type farms where the contribution of the marginal worker is said to be less than his consumption. There is further the possibility that unemployment will persist at the existing wage rate which cannot be reduced lest it decline below a socially acceptable minimun. Mining industries, for example, uhtilize relatively little labor and colmtries relying on minerals exports may not be able to fully emply their labor force without providing special incentives for labor use. In the cases described, the appropriate measure would be subsidizing the use of labor ra-ther than imposing tariffs. Tariffs encourage th.e use labor in such of labor as well as that of capital in protected industries and they favor using/ industries in preference to other sectors of the national economy. Moreover, tariffs may provide incentives for the development of industries that would not be profitable under free trade even if wages were nil. In such instances, there is a trade-off between employment and growth since resources are channelled into industries with relatively high costs. Some of these industries may also have limited possibilities for improving productivity as is said to be the situation in the Indian cottage industry that receives considerable inducements. Finally, while employment-creating measures tend to improve the distribution of income, they may adversely affect savings and hence the prospects for future growth. The choice between employment and growth, then, becomes a choice between present and future employment. - 14 - Subsidizing labor use may take the forn of taxing output and rebating the tax on the basis of the number of employees. This method would encourage the expansion of labor-intensive i:ndustries which use a developinig country's abimdarnt resource, labor, and would also provide incentives for employing labor-intensive production methods. However, there is no reason to restrict the application of this method to manufacturing industries but it should be extended to all sectors other than family farming. Particular cost disabilities or handicaps of manufacturing industries owing to inadequate overhead facilities can also best be corrected by speci- fic action rather than by protection. But again, the provision of such facilities should not be restricted to manufacturing. Thus, roads and electricity are needed for agricultural activities, just as an increase in the educational level of the labor force would contribute to the development of industry as well as to the modernization of agriculture. VI The question remains if, apart from temporary protection on infant industry grounds and the correction of particular cost disabilities, manu- facturing should receive preferential treatment. In support of this pro- position, it has been adduced that productivity tends to rise more rapidly in manufacturing than in primary production and that the expansion of manu- facturing industries provides indirect benefits by inducing investments in other branches of industry and improving the quality of the labor force. The first claim holds true if we compare manufacturing with agri- cultural activities that employ traditional techniques, although modern advances in agriculture offer possibilities for improvements in productivity. In turn, linkages among industries often favor the establishment of related branches of manufacturing, but one should not condone on this basis the establishment of inefficient industries whidh supply inputs to other industries at a high cost. There is finally some merit to the argument that manufacturing contributes to improvements in the quality of the labor force to a greater extent that does even modern agriculture. From the point of view of long-term policy makinmg, further consideration should be given to possible future changes in the supply and demand of primary products. In some developing countries either the supply of primary commodities or foreign demand for them would eventually prove to be a limiting factor for the country's economic growth. In such a situation, the preferential treatment of manufacturing industry,where supply and demand limitations are negligible, would be warranted not only vis-a=vis traditional primary commodities but also in comparison to the primary sector as a whole. These considerations indicate the difficulties encountered in appraising the claims made for the superiority of manufacturing over primary production; the difficulties are compounded if we attempt to quantify these alleged advantages. Nevertheless, one may argue that manufacturing offers some advantages over primary production in the form of labor training and in encouraging the expansion of related industrY that do not enter into the profit calculations of the firm but benefit the national economry. Moreover, manufacturing will improve the growth potential of the economy whenever supply or demand limitations would eventually impinge on primary activities. - 16 - There i3 some presuaption, then, in favor of promoting manufacturing industry in developing countries. The word "ipromote"? is used advisedly as it includes Drotection of production for domestic markets (import substitution) as well as assistance to firms exporting maniufactured goods. Since, for reasons mentioned earlier, bias against manufactured exports entails an economic loss, equal incentives need to be provided to production for domestic and for foreign markets. This can be accomplished by granting a subaidy to the exports of manufactured goods at a rate equal to the tariff applied to the same comnodity or by using differential exchange rates for the manufacturilng sector. Given the cost and uncertainties of entering foreign markets, it right even be desirable to provide additional incentives to exports of manufactured goods on a termporary basis. A further question is what are "reasonable" rates of tariffs and export subsidies and whether all manufactured goods should receive equal treatment. Assuming that particular measures are used to correct special cost disabilities and that the erployment objectives are served by a direct or indirect subsidy to the use of labor, as a first approximation one may suggest providing effective protection at equal rates to all manufacturing 10/ activities that have passed the infant industry stage. In this way, one would apply the "market principle" in the sense that firms will be established that are profitable under such conditions and existing firms would have to improve their operations, change their product composition, or disappear altogether. At the same time, nonessential imports could be restricted by levying excise taxes that bear also on domestic production. The choice of a "'reasonable"f rate of tariffs and subsidies for mature industries in the developing countries will deDend on the particular circum- stances of the situation and oii the range of other policy measures available to a particular country. It may be sug,ested, however, that since most - 17 - developing countries have small domestic markets, they should aim at eventually reducing the net effective protection of manufacturing to levels observed in countries such as Denmark and Norway, i.e. to approximately 10 percent. Exceptions to the proposed equality of effective rates may be made if there is evidence that profitability on the firm level greatly under- states (or overstates) the contribution of a particular industry to the national economy. But such exceptions should apply to entire industries rather than to individual firms and only in cases that are well-documented so as to avoid a "slippage" in protection. In other words, the burden of proof should be on those who request favorable treatment. Standard rates of protection should be applied also in the case of infant industries and one should avoid "tailor-made" tariffs. Mhile it is difficult to judge how much protection would be justified on infant industry grounds, it does not appear likely that, exceptional cases aside, a rate of effective protection more than double that for mature industries would 11/ be warranted. This additional protection of infant industries should be set on a declining scale so that its eventual disappearance provides incentives for improvements. VII The described scheme may be implemented by using a basic exchange rate for nontraditional primary products, export taxes on traditional pri- mary exports, and a combination of tariffs and subsidies on manufactured goods. The same result could be achieved by applying differential exchange rates to the three groups of commodities, with further adjustments made for differences in the elasticity of demand amog traditional primary exports. The choice - 18 - between the two alternatives, or a combination thereof, would have to be made on the basis of considerations of political and administrative feasibility, with further account taken of the implications of. the choice of exchange rates for invisibles and for capital movements. Compared to the policies of industrial protection followed by developing countries engaged in import substitution, the application of these guidelines would entail providing more favorable treatment to nontraditional primary comimodities, reducing, the protection of manufactured produc-ts, and equalizing the incentives for manufactured goods sold in domestic and in export markets. Also, as a general rule, equal incentives would be provided to all branches of manufacturing other than infant industries, and additional protection to infant industries on a temporary basis. For countries that have already embarked on industrialization behind high protective barriers, the application of the guidelines would entail a revaioing of the structure of protection. N,Teedless to say, this could not be undertaken instantaneously but would require a transitional period, the length of which would depend on the particular circumstances of the coLutry in question. There would also be differences in the mode of application of these guidelinles, again depeniding on political and institutional factors. Finally, the relative emphasis on direct measures and on the tariff-sybsidy scheme may differ among countries at different levels of industrialization. - 19 - FOOTNOTES 1/ Bela Balassa, "Growth Strategies in Semi-Industrial Countries?, Quarterly Journal of Economics, February 1970, pp. 24-47. 2/ Bela Balassa and Associates, The Structure of Protection in Developing Countries, Baltimore, NHd., The Johns Hopkins Univer- sity Press, 1971. Ibr a discussion of some of the methodological issues and a summary of the estimates of effective rates of protection, see "Effective Protection in Developing Countries', in Trade, Balance of Payments and Growth: Papers in International Economics in Honor of Charles P. Kindleberger (J. Bhagwati, R. Jones, R. A. Mundell, and V. Vanek, eds.), Amsterdam, North Holland Publishing Co., 1971 - - In the present paper, results pertaining to Norway will not be considered. 3/ Bela Balassa, "Industrial Policies in Taiwan and Korea", Weltwirt- schaftliches Archiv, Band 105, Heft 1, 1971. To be reprinted in a volume in honor of Raul Prebisch. 4/ The results cited in the paper have been taken from the four studies cited above and do not include recent changes in the policies that will be noted below. Some of the estimates on exports and growth have been reproduced in Tables 1 and 2. - 20 - 5/ The incremental exports-GDP ratio (the ratio of the absolute increase in the value of exports to that of GDP) was calculated in constant prices from data given in national and international sources. 6/ For a detailed discussion, see Bela Balassa, "Growth Strategies in Semi-Industrial Countries", op. cit., pp. 45-46. 7/ Far such an adjustment, see Bela Balassa, The Structure of Protection in Developing Countries, Ch. 2. 8/ In formulating the guidelines, the author has drawmn on the results of the studies referred to above, his experience in advising developing countries on trade policies, and the pertinent economic literature. Limitations of space have not peiniitted, however, the detailed considera-tion of particular issues. 9/ A special case is that of coffee where producing countries would be advised to set the export tax (differential exchange rate) at a level calculated to ensure that domestic supply be equal to quota allocations under the International Coffee Agreement. In this way, profits due to the price-raising effects of quotas in international markets accrue to the government and, rather than providing incentives to surplus produc- tion, the proceeds of an export tax can be transferred: to other activities where higher returns are obtained. 10/ In "Decision Rules for Effective Protection in Developing Countries" (mimeo, November 1970), Trent Bertrand provides an elegant proof of the proposition that maximizing welfare subject to the constraint -21 that a certain amount of value added is generated in the manufacturing sector involves equalizing effective rates of protection within this sector. In a more general model, the desired amount of value added in the manufacturing sector and the rate of effective protection of this sector would be jointly determined. 11/ According to an OECD study, economies of scale and external economies can hardly justify effective protection of infant industries exceeding 20 percent even if direct subsidies to labor use are not provided. This figure declines to 10 percent if labor use is subsidized. Cf. Ian Little, Tibor Scitovsky, and Maurice Scott- Industry and Trade in Some Developing Countries, London, Oxford University Press, 1970, pp. 159-59.