Tradle P'ol'cies in Developing C:ountr'e -F eprinted from the Ame carn .ev;ic n n Pi rw-2ew Vokjmd ii LXI, No. 2, May '1971 THE AMERICAN ECONOMIC REVIEW VOLUME LXI May, 1971 NUMBER 2 REPRINT The Journal of THE AMERICAN ECONOMIC ASSOCIATION Trade Policies in Developing Countries By BELA BALASSA* Johns Hopkins Uniiversity The purpose of this paper is to examine summary of the findings, guidelines will some of the implications of the author's be suggested for trade policies by develop- findings concerning the effects of trade ing countries. policies followed by developing countries on their export performance and economic growth. In the discussion, *use will be Instruments of trade policy employed made of the results of several studies car- by developing countries include import ried out in recent years: an investigation tariffs and surcharges, export taxes and of the growth strategies of six semi-indus- subsidies, multiple exchange rates, as well trial countries, two of which, Argentina as quotas and licenses. By affecting the and Chile, are customarily included in the relative prices of inputs and outputs, these developing country category;' a compar- measures influence the allocation of re- ative study of the structure of protection sources, including new investment, and in Brazil, Chile, Mexico, Western Malay- provide incentives-or disincentives-to sia, Pakistan, the Philippines, and Nor- import-substituting and export activities. way;2 and a study of industrial policies in Discrimination among economic activ- Taiwan and Korea.8 Following a brief ities introduced by the use of trade mea- * This paper was written as part of a consultant ar- sures exists to varying degrees in the nine rangement with the World Bank. countries studied.4 Argentina, Brazil, IBela Balassa, "Growth Strategies in Serni-Indus- Chile, Pakistan, and-to a lesser extent- trial Countries" Quarterly Journal of Economncs, the Philippines provide considerable in- February 1970, pp. 24-47. 1 Bela Balassa and Associates, The Strcture of centives to manufacturing industries at Protection in Developing Countries, Baltimore, Md., the expense of primary activities. The ex- The Johns Hopkins TJniversity Press, 1971. For a t discussion of some of the methodological issues and tent of discrimination in favor of manu- a summary of the estimates of effective rates of pro- facturing and against primary production tection, see "Effective Protection in Developing Coun- is relatively small in Mexico; it is even tries" in Trade, Balance of ' ayments and Growth:l Papers in International Economics in Honor of less in Korea and Taiwan; and, on the Charles P. Kindleberger (J. Bhagwati, R, Jones, R. A. average, virtually nil in Western Malaysia Mundell, and V. Vanek, eds.), Amsterdam, North (for short, Malaya). Holland Publishing; Co., 1971-In the present paper, Discrimination in favor of manufactur- results pertaining to Norway will not be considered. Furthermore, in the case of Argentina, Brazil, and ing entails the protection of domestic Chile, the paper will consider the situation existing manufacturing industries against imports prior to the changes in polities introduced in the late sixties. These policy changes will be briefly referted to at a later point in the paper. 'The results cited in the paper have been taken 'Bela Balassa, "Industrial Policies in Taiwan and from the four studies cited above and do not include Korean," Weltwirtschaftliches Archiv, Band 105, Heft recent changes in the policies that w:ll be noted below. 1, 1971. To be reprinted in a volume in honor of Some of the estimates on exports and growth have Raul Prebisch. been reproduced in Tables 1 and 2. 178 LESS DEVELOPED COUNTRIES 179 whereas primary exports are penalized by ico and to an even greater extent Taiwan tari-s on their inputs and by the lower ex- and Korea have been successful in raising change rate associated with protection. the world market share of their major pri- The protection of manufactured goods mary exports and in introducing new ex- and discrimination against primary exports port products. In the latter two countries, usually go hand-in-hand, the principal ex- these developments have taken place dur- ception being Malaya, which levies a tax ing the sixties following a shift in eco- on its major primary exports, rubber and nomic policies from import substitution to tin, but does not protect manufactures. export promotion. In Argentina, Brazil, Chile, and the The expansion of primary exports in in- Philippines, there is also a substantial bias dividual countries has further been af- in favor of import substitution and against fected by world demand conditions in the exports in protected manufacturing in- markets for their major export commodi- dustries. Firms producing for home mar- ties. The main beneficiaries of favorable kets can get the domestic price raised by world market trend& have been Chile high import protection while, in the ab- (copper) and the Philippines (oilseeds) sence of export subsidies, they could ob- while the slow growth of world demand tain only the world market price in ex- and unfavorable price changes for rubber porting. By contrast, the bias against ex- and tin have depressed export earnings in ports in manufacturing industries is rela- Malaya. tively small in Mexico and Malaya where The trade policies followed, together levels of protection are low and it is prac- with changes in world mark-et conditions, tically nonexistent in Pakistan, Taiwan, largely explain intercountry differences in and Korea. In Pakistan, manufacturing the rate of growth of primary exports. Ta- industries receive high protection irre- ble 1 shows Korea and Taiwan in the lead, spective of whether they produce for do- followed by Chile, the Philippines, Mex- mestic or foreign markets; in Taiwan and ico, Argentina, Malaya, Brazil, and Paki- Korea the extent of import protection is stan. Comparisons of growth rates are less rather low and, on the average, it is meaningful for manufactured exports be- matched by subsidies to manufactured ex- cause several of the countries under con- ports. sideration started from a very small base. Instead, we use the F'I-re of manufactured II exports in manufacturing output and in The trade policies followed during the total exports as indicators of suiccess in ex- postwar period have affected export per- porting manufactures. formance and economic growth in the In countries with a substantial bias countries under consideration. The unfa- against the exports of manufactured goods, vorable treatmenit of primary exports has these commodities continue to accotunt for contributed to the decline in the shares of less than 3 percent of manufactured out- Argentina, Brazil, Chile, and Pakistan put and account for at most 10 percent of and, to a lesser extent, the Philippines in total exports. Among countries with a lesser the world market for their major export bias, Mexico's manufactured exports have commodities and has retarded the devel- reached 5 percent of manufacturing output opment of new exports. Malaya, too, has and 25 percent of total exports. In Ma- experienced a decline in the market shares laya 10 percent of manufactured output is of its major primary exports, but it has exported although, given the relatively expanded its minor -xports which receive low share of manufacturing and the high more favorable trea,L. ent. Finally, Mex- share of exports in GNP, manufactured 180 AMERICAN ECONOMIC ASSOCIATION TABLE 1-EXPORTS IN SELECTED DEVELOPING COUNTRES Argen- tina Brazil Chile Mexico Korea Malaya Pakistai Philip- Taiwan A.verage 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 3.6 4.2 8.4 5.5 21.0 0.6 2L5 6.0 16.4 1950-69 1.7 0.6 6.8 4.7 16.7 1.1 -2.9 5.7 8.3 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 69.0 12.7 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 1,0 1.1 7.0 6.0 18.2 1.7 0.4 6.0 14.9 Manufactured exports as a percentage of output-1969 2 1 3 5 18 10 8 3 36 total exports-1969 10 9 6 25 76 10 51 10 67 Note: For Taiwan and Korea the base year is 1953 instead of 1950. For Brazil, Chile, Malaya, Paki-tan, and the Philippines, the terminal year is 1968 instead of 1969. Source: National and international trade statistics. goods provide no more than one-tenth of have imposed a substantial cost on the na- total exports. tional economy. In Taiwan and Korea, manufactured Owing to their favorable performance exports have been stimulated to a consid- in both primary and manufactured ex- erable extent by the adoption of export- ports, Korea and Taiwan are far ahead of oriented policies around 1960. As a result, the other countries studied in terms of the exports of manufactures have increased to expansion of total exports (Table 1). Ex- a considerable extent in the two countries, ports have increased more than the aver- both as a proportion of total exports and age also in Mexico where the growth of of manufacturing output. By 1969, these tourism and border trade (not included in proportions reached 67 and 37 percent in the export figures) have further contrib- Taiwan and 76 and 18 percent in Korea. uted to increases in foreign exchange Following the introduction of the Ex- earnings. For reasons noted earlier, ex- port Bonus scheme in 1959, manufactured ports have risen relatively rapidly also in exports have also assumed importance in Chile and the Philippines wvhile increases Pakistan, with jute and cotton textiles be- have been small in Argentina, Brazil, and ing the principal items. But, owing to Pakistan. policies penalizing primary exports, for- eign sales of raw jute and cotton have de- III clined in an amount exceeding the rise in The expansion of exports contributes to textile exports. Moreover, high subsidies economic growth directly by raising na- to manufactured exports, together with the tional income and indirectly by providing high protection of import substitutes, foreign exchange for the import needs of LESS DWEVLOPED COUNTRIES 181 the domestic economy. An export-oriented substitution in other intermediate prod- policy also permits specialization accord- ucts, capital goods, and durable consumer ing to comparative advantage-both be- goods. These commodities have higher tween primary and manufactured activi- technological and skill requirements, re- ties and within the manufacturing sector. quire the availability of materials, parts, In particular, exports of manufactured and components from other industries, goods enable firms to lower costs by em- and need large-scale production for effi- ploying large-scale production rnethods, cient operations with costs being substan- reducing product variety, and participat- tially higher at lower output levels. Last ing in the international division of the but not least, in the event of continuing production process through the manufac- protection, there will be few inducements turing of parts and components for assem- for technological improvements. bly abroad. Moreover, familiarity with These considerations help to explain in- foreign markets provides incentives for tercountry differences in rates of economic technological change and product improve- growth (Table 2). In Taiwan and Korea, ment. the growth of GDP has accelerated to a Import substitution, too, can be a considerable extent following the adoption source of economic growth in particular of export-oriented policies. In expanding cases. A number of developing countries the exports of nondurable consumer attained rapid rates of growth of manu- goods, the two cotintries have utilized their facturing output and, to a lesser extent, educated manpower while the capital re- national incorne, in the early stage of im- quirements of these industries are rela- port substitut.'on, which entails replacing tively low. The leading role of exports in the imports of nondurable consumer goods the growth process is indicated by the and their inputs by domestic production. high incremental ratio of exports to GDP; Industries producing such commodities in 1960-69, this ratio was 39 percent in are the prime candidates for import sub- Taiwan and 29 percent in Korea.' stitution in developing countries since Exports lhave also importantly contrib- they employ chiefly unskilled and semi- uted to Mexico's economic growth and the skilled labor, do not require the applica- relatively low protection of manufacturing tion of sophisticated technology, and need industries has limited the cost of import few inputs from ancillary industries. Nor substitution. In turn, the low degree of does the limited size of national markets discrimination among economic activities constitute an important handicap for the has made it possible for Malaya to attain development of these induistries since the a rate of growth of national income sub- efficient scale of operations is relitively stantially above that for exports. low and costs are not substantially higher The remaining countries of the group in smaller plants. are characterized by import substitution But, in the absence of exports, the ex- behind high protective barriers. In these pansion of industries producing nondura- countries, the relationship between import ble consumer goods and their inputs nec- substitution and economic growth has essarily slows down after imports have been influenced by their market size and been replaced since domestic production the level of their economic development. cannot continue to grow faster than home demand. Moreover, in the small domestic 5The incremental exports-GDP ratio (the ratio of markets of developing countries, increas- of GDP) was calculated in constant prices from data ing difficulties are encountered in import given in national nnd international sources. 182 AMERICAN ECONOMIC ASSOCIATION TABLE 2-E CONoxC GROWTH IN SELECTED DEVELOPZNG COUNTRIES Argen-Phlp tina Brazil Chile Mexic,. Korea MIalaya Pakistan pines Taiwan Average annuaI 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 196G-69 2.0 4.2 2.5 4.0 4.6 3.9 3.7 4.6 5.0 1950-69 2.1 4.5 1.8 4.5 3.6 3.6 2.5 4.9 4.5 Manufacturing 1950-0 4.7 8.8 3.3 8.0 13.6 5.1 7.8 10.2 10.1 1960-69 4.6 5.9 5.9 2.0 16.0 11.7 8.6 4.5 16.1 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.S 3.1 196069 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 Population 1950-60 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 2.6 3.4 3.0 1950-69 1.8 3.0 2.2 3.1 2.3 3.0 2.3 3.3 3.3 Note: For Taiwan and Korea the base year is 1953 instead of 1950. For Brazil and Malaya, the terminal year is 1968, for the Philippines, 1967. Source: National and international statistics. Thus, the expansion of manufacturing scale and often outdated production meth- output has slowed down to a considerable ods, inadequate specialization, and the extent in the Philippines after the mid-fif- manufacturing of products of low quality.' ties by which time the "easy'7 stage of im- Discrimination against agriculture, associ- port substitution had been largely com- ated with the high protection of manu- pleted. Despite rapid increases in exports facturing activities, has further hindered due to favorable market conditions, there their economic growth and helps to ex- has been a decline in the rate of growth of plain why they have experienced the low- GDP as well. est growth rates amnong the nine countries Argentina and Chile had replaced prac- studied. tically all nondurable consumer goods and Brazil, too, had completed the first their inputs before the period under con- stage of import substitution prior to the sideration, and their small domestic mar- period under consideration. Its large do- kets have made the expansion of indus- me3tic market, however, provided possi- tries producing other intermediate prod- bilities for the continued expansion of ucts, capital goods, and durable consumer manufacturing during the fifties, mostly goods both difficult and costly. These 'For a detailed discussion, see Bela Balassa, countries have built up an industrial "Growth Strategies in Semi-Industrial Countries," op. structure which entails the use of small- cit., pp. 45-46. LESS DEVELOPED COUNTRIES 183 in intermediate products, capital goods, ernments to reconsider their economic pol- and durable consumer goods. But, as the icies. In Argentina, the extent of discrimi- possibilities for import substitution have nation against primary production and ex- been increasingly exhausted, industrial ex- ports has been reduced and the protection pansion has slowed down in this country of manufacturing industries has been also. moderated through a simultaneous devalb Pakistan had practically no industry tuation and a lowering of tariffs. Manufac- prior to independence and it was able to tured goods also receive export subsidies achieve rapid rates of economic growti- by in Argentina and such subsidies have as- substituting domestic production for the sumed an important role in Brazil. imports of nondurable consumer goods Furthermore, in Chile, an effort has been and their inputs. Subsequently, the adop- made to lessen the degree of over-valua- tion of the Export Bonus scheme has con- tion of the currency while subsidies have tributed to its relatively rapid industrial been used to promote the exports of manu- expansion. Continuing discrimination factured goods. against agriculture has however adversely Efforts made to reform the structure of affected the growth of the Pakistani econ- protection, however, have gone only part omy. If national income is measured at of the way and further progress is made world market prices rather than at the do- difficult by resistance on the part of vested mestic prices distorted by protection, in- interests. Businessmen are opposed to creases in per capita terms appear to have changes in the status quo which ensures been small.7 comfortable profits, and they demand con- IV tinuing protection from foreign competi- tion, whether this comes from the indus- The experience of the countries under trial countries or from developing nations consideration suggests the conclusion that as in the case of LAFTA. Additional prob- while the protection of the manufacturing lems are that transition to a more open sector may permit rapid growth at an economy would entail dislocation in par- early stage of import substitution, it will ticular industries and regions. eventually have adverse consequences for It appears, then, that once an industrial economic growth. Discrimination among structure geared to import substitution industries does not permit specialization has been established, change bcomes in- according to comparative advantage; the creasingly difficult. This observation high protection of domestic industry in- points to the need for making appropriate duces the establishment of high-cost im- policy choices at the time when a country port-substituting activities; and the bias embarks on an industrialization program. against exports retards the development In the following, guidelines will be sug- of manufactured exports. Finally, in the gested for trade policies by developing absence of foreign competition, there will countries.8 Apart from their application to be little incentive for technical progress in countries at an early stage of industrial- small protected domestic markets. ization, the guidelines can provide a basis The increasing difficulties experienced by countries at higher stages of import In formulating the guidelines, the author has substitution have recently led some gov- drawn on the results of the studies referred to above, his experience in advising developing countries on For such an adjustment, see Bela Balassa, The trade policies, and the pertinent economic literature. Structure of Protection in Developing Countries, Limitations of space have not permitted, however, the Ch. 2. detailed consideration of particular issues. 184 AMERICAN ECONOMIC ASSOCIATION for improvements in the policies of coun- itional primary production, aild if so, to tries presently engaged in import substitu- what extent and by the use of what meax tion. sures. In this connection, note should first r be taken of arguments for infant industry protection, designed to compensate for as- International trade theory tells Us that sumed differences between social and pri- small countries which do not affect the vate profitability. On the firm level, such prices of their exports and imports will differences may arise if the lack of credit maximize welfare by specializing in accor- facilities, the overestimation of risks, or dance with price relations on the world simply the desire to exclude the possibility rmarket. Developing countries can gen- of bankruptcy provide disincentives for erally take their import prices as given investment, although eventual cost reduc- and they will not affect the prices of most tions through the learning process or of their .manufactured exports either. through increases in the scale of operations This will not be the case, however, for would make the investment socially desir- traditional primary exports whenever in- able. Other instances are when some of creases in the country's exports lead to the benefits of the pioneering firm's activi- a fall in prices. For these exports, then, ties are enjoyed by others who utilize the the relevant decision rule will involve know-how generated by the firm or hire equating marginal costs to marginal rev- away skilled labor and technicians it lhas enue from exports rather than to price. trained. This can be accomplished by converting It has often been said that infant indus- foreign exchange earnings from such ex- try arguments justify using production ports at a less favorable exchange rate or subsidies rather than tariffs since the lat- -what amounts to the same-imposing ter limit the size of the domestic market an export tax on them. Export tax rates on by raising the price of the commodity in individual commodities should be set by question. But while tariffs contribute to allowing for the elasticity of world de- goxverinmenit revenue, subsidies represent a mand, the country's share in world ex- claim on this revenue. Budgetary reasons, ports, and the possible reactions on the then, may explain why developing coun- part of foreign competitors.9 tries use tariffs in preference to produc- The application of these measures tion subsidies. In fact, tariffs often ac- would take account of market limitations count for a large part of government reve- for traditional primary exports, without nue in these countries, and their replace- uinduly discouraging their production as ment by other forms of taxation may en- has often been the case in the past. A fur- counter practical difficulties. ther question is whether manxufacturing The arguments for subsidies in prefer- industries should be favored over nontrad- ence to tariffs gain in force in cases when 'An extreme case is that of coffee where producing a particular distortion or cost disability countries would be advised to set the export tax (dif- needs to be corrected. This will be so if ferential exchange rate) at a level calculated to ensure the cost of industrial labor to inanufactur- that domestic supply be equal to quota allocations in enterprises exceeds its social cost in under the International Coffee Agreement. In this ing way, profits due to the price-raising effects of quotas the form of the output foregone in pri- in internatiornal markets accrue to the government mary activities from which labor is drawn. and, rather than providing incentives to surplus pro- In some overpopulated countries, such a duction, the proceeds of an export tax can be trans- ferred to other activities where higher returns are situation may exist on family-type farms obtained where the contribution of the marginal LESS DEVELOPED COUNTRIES 185 worker is said to be less than his con- it should be extended to all sectors other sumption. There is further the possibility than family farming. that unemployment will persist at the ex- Particular cost disabilities, or handi- isting wage rate which cannot be reduced caps of manufacturing industries owing to lest it decline below a socially acceptable inadequate overhead facilities, can also minimum. Mining industries, for example, best be corrected by specific action rather utilize relatively little labor and countries than by protection. But again, the provi- relying on mineral exports may not be sion of such facilities should not be re- able to fully employ their labor force with- stricted to manufacturing. Thus, roads out providing special incentives for labor and electricity are needed for agricultural use. activities, just as an increase in the educa- In the cases described, the appropriate tional level of the labor force would con- measure would be subsidizing the use of tribute to the development of industry as labor rather than imposing tariffs. Tariffs well as to the modernization of agricul- encourage the use of labor as well as capi- ture. tal in protected industries and they favor using labor in such industries in preference VI to other sectors of the national economy. The question remains whether, apart Moreover, tariffs may provide incentives from temporary protection on infant in- for the development of industries that dustry grounds and the correction of par- would lnot be profitable under free trade ticular cost disabilities, manufacturing even if wages were nil. In such instances. should receive preferential treatment. In there is a trade-off between employment support of this proposition, it has been ad- and growth since resources are channelled duced that productivity tends to rise more into industries with relatively high costs. rapidly in manufacturing than in primary Some of these industries may also have lim- production and that the expansion of ited possibilities for improving productivity manufacturing imdustries provides indi- as is said to be the situation in the Indian rect benefits by inducing investments in cottage industry that receives considerable other branches of industry and improving inducements. Finally, while employment- the quality of the labor force. creating measures tend to improve the dis- The first claim holds true if we compare tribution of income, they may adversely af- manufacturing with agricultural activities fect savings and hence the prospects for fu- that employ traditional techniques, al- ture growth, The choice between employ- though modern advances in agriculture ment and growth, then, becomes a choice offer possibilities for improvements in pro- between present and future employment. ductivity. Furthermore, linkages, among Subsidizing labor use may take the industries often favor the establishment of form of taxing output and rebating the tax related branches of manufacturing, but one on the basis of the number of employees. should not condone on this basis the es- This method would encourage the expan- tablishment of inefficient industries which sion of labor-intensive industries which supply inputs to other industries at a high use a developing country's abundant re- cost. There is finally some merit to the ar- source, labor, and would also provide in- gument that manufacturing contributes to centives for employing labor-intensive improvements in the quality of the labor production methods. However, there is no force to a greater extent than does even reason to restrict the application of this modern agriculture. method to manufacturing industries, but From the point of view of long-term 186 AMERICAN ECONOMIC ASSOCIATION policy making, further consideration change rates for the manufacturing sector. should be given to possible future changes Given the cost and uncertainties of enter- in the supply and demand of primary ing foreign markets, it nmight even be de- products. In some developing countries ei- sirable to provide additional i.ncentives to ther the supply of primary commodities or exports of manufactured goods on a tem- foreign demand for them would eventu- porary basis. ally prove to be a limiting factor for the Further questions are what are "reason- country's economic growth. In such a situ- able" rates of tariffs and export subsidies ation, the preferential treatment of manu- and whether all manufactured goods facturing industry, where supply and de- should receive equal treatment. Assuming mand limitations are negligible, would be that particular measures are used to cor- warranted not only vis-A-vis traditional rect special cost disabilities and that the primary commodities, but also in compari- employment objectives are served by a di- son to the primary sector as a whole. rect or indirect subsidy to the use of labor, These considerations indicate the diffi- as a first approximation one may suggest culties encountered in appraising the providing effective protection at equal claims made for the superiority of manu- rates to all manufacturing activities that facturing over primary production. The have passed the infant industry stage.'" In difficulties are compounded if we attempt this way, one would apply the "market to quantify these alleged advantages. principle" in the sense that firms will be Nevertheless, one may argue that manu- established that are profitable under such facturing offers some advantages over pri- conditions and existing firms would have mary production in the form of labor to improve their operations, change their training and in encouraging the expansion product compositioin, or disappear alto- of related industry that do not enter into gether. At the same time, nonessential im- the profit calculations of the firm but ben- ports could be restricted by levying excise efit the national economy. Moreover, taxes that bear also on domestic produc- manufacturing will improve the growth tion. potential of the economy whenever supply The choice of a "reasonable" rate of or demand limitations would eventually tariffs and subsidies for mature industries impinge on primary activities. in the developing countries will depend on There is some presumption, then, in fa- the particular circumstances of the situa- vor of promoting manufacturing industry tion and on the range of other policy mea- in developing countries. The word "pro- sures available to a particular country. It mote" is used advisedly as it includes pro- may be suggested, however, that since tection of production for domestic mar- most developing countries have small do- kets (import substitution) as well as as- mestic markets, they should aim at even- sistance to firms exporting manufactured tually reducing the net effective protection goods. Since, for reasons mentioned ear- "'In "Decision Rules for Effective Protection in lier, bias against manufactured exports Developin- Counties" (niimeo, MNvember 170)i entails an economic loss, equal incentives Trent Bertrand provides an elegant proof of the need to be provided to production for do- proposition that maximizing welfare subject to the constraint that a certain amount of value added is mestic and for foreign markets. This can gencrated in the manufacturing sector involves equal- be accomplished by granting a subsidy to izing ceffecLiv%e rates of protection within this sector. the exports of manufactured goods at a In a more general model, the desired amount of value added in the manufacturing sector and the rate of rate equal to the tariff applied to the same effective protection of this sector would be jointly commodity, or by using differential ex- determined. LESS DEVELOPED COUNTRIES 187 of manufacturing to levels observed in for differences in the elasticity of demand countries such as Denmark and Norway, among traditlonal primary exports. The i.e., to approximately 10 percent. choice between the two alternatives, or a Exceptions to the proposed equality of combination thereof, would have to be effective rates may be made if there is evi- made on the basis of considerations of po- dence that profitability on the firm level litical and administrative feasibility, with greatly understates (or overstates) the further account taken of the implications contribution of a particular industry to of the choice of exchange rates for invisi- the national economy. But such excep- bles and for capital movements."2 tions should apply to entire industries Compared to the policies of industrial rather than to individual firms and only in protection followed by developing coun- cases that are well-documented so as to tries engaged in import substitution, the avoid a "slippage" in protection. In other application of these guidelines would en- words, the burden of proof should be on tail providing more favorable treatment to those who request favorable treatment. nontraditional primary c -rnmodities, re- Standard rates of protection should be ducing the protection of manufactured applied also in the case of infant indus- products, and equtalizing the incentives for tries and one should avoid "ctailor-made" manufactured goods sold in domestic and tariffs. While it is difficult to judge how in export marlkefs. Also, as a general rule, much protection would be justified on in- equal incentives would be provided to all fant industry grounds, it does not appear branches of manuifactuirinig other than in- likely that, exceptional cases aside, a rate fant industries, and additional protection of effective protection more than double to infant industrie.s on a temporary basis. that for mature industries would be war- For countries that have already em- ranted.1" This additional protection of in- barked on industrialization behind high fant industries should be set on a declin- protective barriers, the application of the ing scale so that its eventual disappear- gui(lelines would entail a revamping of the ance provides incentives for improve- structure of protection. Needless to say, ments. this could not be undertaken instantane- ously but would require a transitional pe- VII riod, the length of which would depend on The described scheme may be imple- the particular circumstances of the coun- mented by using a basic exchange rate for try in question. There would also be dif- nontraditional primary products, export ferences in the mode of application of taxes on traditional primary exports, and these guidelines, again depending on polit- a combination of tariffs and subsidies on ical and institutional facters. Finally, the manufactured goods. The same result relative emphasis on direct measures and could be achieved by applying differential on the tariff-subsidy scheme may differ exchange rates to the three groups of com- among countries at different levels of in- modities, with further adjustments made dustrialization. 'According to an OECD study, economics of sc:le "In some cases, however, neither of these alter- and external economies can hardly justify effecct ie natives might be feasible because of constraints in protection of infant industries exceeding 20 percent policy making. In Chile, for example, the government even if direct subsidies to labor use are not provided. is said to have obligated itself not to levy special This figure declines to 10 percent if labor use is sub- taxes on copper, the major export commodity. Ac- K lized. Cf. Ian Little, Tibor Scitovsky, and Maurice cordingly, the basic exchange rate would have to be Scott, Industry and TrcIe in Some Developing Coun- applied to copper while tariff-subsidy schemes would tries, London, Oxford University Press, 1970, pp. need to be used both for nontraditional primary prod- 158-$9. ucts and for manufactured goods.