Imports, and to a lesser extent exports, have been controlled by almost all developing countries since World War II. Yet the evidence indicates what theory suggests, that the more open the economy the higher is income and the better the growth rate. But reliance on controls persists. Why? In this article I shall examine the legacy of the intellectual case for import controls which was first made in the 1940s, and which still influences both the debate and the policy-making of today. I will then discuss the drawbacks of controls in the light of new analysis which contradicts the still widely accepted thesis that open trade between unequal partners damages the poorer partner.
For some countries in Latin America, controls were introduced in the depression of the 1930s or during World War II, because of worsening terms of trade. But almost invariably the controls instituted by developing countries since World War II have been the result of a balance of payments (BOP) crisis brought on by increased spending for development. Of course, World War II also left most Western industrialized countries with an almost complete array of import and exchange controls. But within a decade the import controls and most exchange controls in the developed countries had been dismantled. Why did the developing world intensify import controls long after the extreme dislocations caused by the war had vanished?
My impression is that this reliance arose mainly from the conviction among leaders of opinion and policymakers in the Third World that a poor country could not develop and manage its foreign trade and payments without controls. Of course, all controls protect, and sometimes protect absolutely. But there are several reasons (among them the timing of the controls, and the fact that the countries instituting these controls had few industries to protect anyway) that lead us to believe the primary motive to have been BOP management—not protection.
Furthermore, an intellectual heritage has probably played an important role in maintaining the control system, as the new intelligentsia, taught by the development economists of the 1940s and 1950s, rose in support of the system. In many semi-industrialized countries this has resulted in a strange meeting of the minds of intellectuals, businessmen, and bureaucrats.
The case for controls
The rationale for poor countries to control imports was developed in the 1950s, mainly by Hans Singer and Raul Prebisch. Briefly, Singers’s thesis was that developing countries had not got much out of international investment and trade—above all, not much industrialization. There were dark hints that these countries might have got less than nothing out of it, as foreign trade diverted activities away from industrialization, which would have played a catalytic role in development. Singer also promoted the thesis of a secular worsening in the terms of trade of the developing countries, basing his argument on a United Nations document, Relating Prices of Exports and Imports of Underdeveloped Countries, published by the Department of Economic Affairs in 1949. Singer presented this thesis as an “indisputable fact,” and explained it essentially by an ever-increasing degree of monopoly (with labor sharing in it) in the production of manufactured goods. However, he drew no inference for BOP management.
Prebisch, and the Economic Commission for Latin America headed by him, emphasized the same worsening in the terms of trade of developing countries from 1949 onward. This thesis became enshrined in the United Nations Conference on Trade and Development (UNCTAD), and in repeated pronouncements by many development economists. The peaks in commodity prices during the Korean War, the subsequent decreases in prices followed by recovery after the early 1960s, the historically high levels of today, and the fact that all the scholarly work I know of denies any long-run trend, has probably still not fully exorcised the myth created by the 1949 UN study (which was based on the United Kingdom’s terms of trade from 1876 to 1946).
To the argument of declining terms of trade was added the view that the demand for exports from developing countries would inevitably be sluggish compared with their own demand for imports—especially, of course, if domestic demand were to be steered toward investment goods for development. To this was added the acceptable argument that the price elasticity for exports was very low (at least for developing countries taken together). Thus, exports of developing countries were determined by external forces not under their control, and this was also true of aggregate capital imports. Therefore, the level of imports was also determined. The conclusion was that their own import controls and other restrictions did not restrict trade but served only to control the pattern of imports in the interest of development. Also, with the declining terms of trade, BOP trouble would be endemic, and frequent devaluation and inflation would result. Moreover, devaluation would further worsen the terms of trade in the face of an inelastic foreign demand.
However, even if we accept the elasticity assumptions, it still does not add up to a case. Export taxes could be used to prevent devaluation from reducing export proceeds. Luxury taxes could prevent low priority imports. If industrialization needed a special push, this could be achieved by tariffs or, better still, by subsidies which would not discriminate against exports. Finally, it is probably an illusion that the reduction of consumption associated with price rises that result from controls is less inflationary than price rises resulting from other policies.
The argument could thus be fully convincing only to those who had acquired faith in planning—and by “planning” I mean here trying to manipulate quantities with some end in view, with prices as a byproduct rather than vice versa. To express a disbelief in planning, at least for developing countries, in the 1950s, was a confession of confusion or worse. In Europe, Thomas Balogh and Gunnar Myrdal were among the conspicuously successful teachers of the need for controls and planning and of the view that trade between unequal partners might damage the poorer partner.
The emphasis on quantities was also connected with “structuralism.” This seemed to involve the belief that a country’s structure of production, and of imports and exports, was not only inappropriate but also unchangeable, except in the long run by investment, which must be controlled in order to produce eventually a more desirable structure. As far as trade and its effects on the proper pattern of production is concerned, structuralism essentially implied pessimism about the capacity of developing countries to expand exports. Export pessimism, together with the fact that most developing countries had only a tiny capital goods industry, led to the view that growth in the typical developing country was limited by foreign exchange and not by savings (which logically required the view that savings could not be transformed into investment).
Structuralism may have been conceived in Latin America, but the principles behind the Second and Third Development Plans in India were also essentially structuralist. Export pessimism, India’s relative lack of minerals, and its desire to become independent of aid led structurally to the conclusion that anything needed for growth must be made at home. There was no choice! Consequently, plans based on this reasoning must be optimal. It also followed that relative prices and most other key concepts of traditional economics were irrelevant. The argument was logical, but hopelessly wrong.
Quite a few of the undesirable extravagances of the resultant import substitution policies had been recognized by Prebisch by 1964. These recognized drawbacks included capital intensity, low value added at international prices (but not value subtracted!), loss of scale, lack of competition, and the growth of “inessential” production behind the barriers to trade. There was, however, no recognition of the effect of protection on exports, nor that the developing countries’ falling share of world trade was due either to this effect or to lagging agricultural production. The idea of the long-term decline in the terms of trade was still being advanced, and policies of import substitution were still favored. No distinction was drawn between protection by controls and by tariffs. The philosophy was still one of controlled trade. Neither then nor since has UNCTAD favored free trade—not even for industrialized countries. For how then could the developed countries grant preferences? Any reciprocity was and has remained anathema and a part of the asymmetry argument—what was bad for the developed was good for the developing world.
In the 1960s the great majority of developing countries were married to control systems and high protection (as indeed they still are, although to a lesser extent). Thus, it is to the challenge to the con-trolled-trade establishment that we must now turn. On a theoretical plane the seeds were sown by 1963 by J. Bhagwati and V.K. Ramaswami, among others. At the more influential applied level it did not acquire real force until 1970, when the extraordinary average heights and variability of effective protection were exposed in a book by Ian Little, Tibor Scitovsky, and Maurice Scott, and in another by Bela Bal-assa, et al (see the related reading list). The former work also discussed the inhibiting effects of general control regimes (which had spread from simple import controls). Both books laid some stress on promoting rather than protecting industry so as to achieve, as far as possible, neutrality between the domestic market and exports. The Little-Scitovsky-Scott volume also discussed how a transition to more liberal trade, which would be more favorable to exports, might be made, and suggested that it would be very difficult to justify effective subsidization of industry, whether directly or by tariffs, of more than 20 percent.
More recently there has been a massive ten-country study of trade regimes guided by Jagdish Bhagwati and Anne Krueger (yet to be published). Brazil, Chile, Colombia, Egypt, Ghana, India, Israel, the Philippines, the Republic of Korea, and Turkey were the countries covered between 1950 and 1972. The study found that the open or relatively open countries grew faster—faster than when they were less open and faster than the chronically “closed.” Brazil, Israel, and Korea performed best of the group. The main reasons given for superior performance were as follows. Exports proved to be highly responsive to the reduction or elimination of the bias against them. The partly consequential increase in imports reduced the chaos in the pattern of import substitution incentives and ensured a freer flow of inputs, with production benefits resulting from greater capacity utilization and a reduction in required stocks. The greater value of exports also made it easier to borrow. In some countries, especially in Korea, more direct foreign investment was attracted to the relatively labor-intensive export sector.
I would put rather more stress than Bhagwati does in his summary volume on the supposition that exports are simply good business for the country. Social (that is, shadow priced) profits and savings are higher on exports than on import substitutes at the margin. That restrictive regimes result in more investment in capital-intensive sectors and plants is empirically clear for a number of countries both within and outside the Bhagwati-Krueger ten. This also has implications for spreading the benefits of growth, as well as for growth itself.
Krueger, in her summary volume, deals primarily with the conditions under which liberalization attempts have been made. Twenty-two liberalization efforts are reported for the ten countries between 1950 and 1972. All involved packages. Devaluation was combined with import liberalization, deflation, reduction of tariffs, and export subsidies, in varying degrees. Most efforts were made in periods of economic crisis, usually in a situation in which the government had committed itself to an overvalued exchange rate, and in which there was a loss of reserves or debt rescheduling. Many efforts included attempts at stabilization.
Of course, if a simple devaluation is to be effective, inflation has to be stopped. It was the consequent deflationary measures, and the foreign (International Monetary Fund or World Bank) involvement, which often made these attempts unpopular, and resulted in a reversal. It must be remembered that the Fund norm was then a fixed exchange rate. But a floating rate, or a sliding peg, was used in several countries on and off.
The reasons for failure to liberalize are divergent. Either the effective devaluation was inadequate, or the bias against exports was not removed or was not much reduced (as in India and the Philippines), so that exports did not respond sufficiently. Inflation and a fixed rate continued or was reimposed (Chile under Allende). There was insufficient political and intellectual commitment (India, Chile), or an actual reversal for political reasons (Ghana), or bad monsoon luck (India again). Needless to say, manufacturers require some expectation of continuing profitability for exports if they are to invest for export production, make products designed for export, and spend money on a marketing organization. Only in a few countries was the government commitment to the change of policy sufficient for this expectation. But although failure was frequent, there was progress among the ten. The Bhagwati-Krueger phase analysis suggests an increasing degree of liberalization after the mid-1960s and comes down in favor of an export promotion strategy. (Oddly enough, Bhagwati means by “export promotion” a strategy that is neutral between the domestic and export markets.)
Further relevant evidence is added by Balassa in a preliminary report on his World Bank studies of exports in developing countries, where he discusses Argentina, Brazil, Chile, Colombia, India, Israel, Korea, Mexico, Singapore, Taiwan, and Yugoslavia. More material is provided by two chapters (by Maurice Scott and myself) in a new book on Taiwan. A few major points which seem to me to emerge from including a few more countries, and even more recent information, in the review are as follows:
Of the above, only Korea, Singapore, and Taiwan have created virtually free trade regimes for exports. In these countries exporters can buy not only imports but also domestic inputs at world prices. Singapore and Taiwan are now as free trading as most developed countries. Korea and Taiwan also have free labor markets, barely affected by trade unions or labor legislation. These three countries sustained gross national product growth rates of about 10 per cent for as much as a decade prior to 1973—a performance shared only by Hong Kong, Israel, and Japan.
Although Israel is fully liberalized, tariffs still result in a significant bias against exports. None of the other countries mentioned are fully liberalized, and none have created the free trade regimes for exports of Korea, Singapore, and Taiwan (and, of course, Hong Kong). All, however, have made some effort at export promotion. The staggering export performance of Hong Kong, Korea, Singapore, and Taiwan is well known. However, the manufactured exports of Argentina, Brazil, and Colombia have also grown very fast, at about 30 per cent per annum in 1967–73, but from very low levels. As a consequence, the proportion of manufactured output exported remains very low: in 1973 the ratios were 3.6 per cent, 4.4 per cent, and 7.5 per cent, respectively, compared with 49.9 per cent for Taiwan and 40.5 per cent for Korea.
The highly open economies of Hong Kong, Korea, and Taiwan all weathered the world recession of 1974–75 very well, despite their extreme dependence on imported energy. Korea’s growth rate never dipped below 8.3 per cent. Hong Kong’s and Taiwan’s were brought below 3 per cent, but they recovered to rates of 16.2 per cent and 11.9 per cent in 1976. Despite continued sluggish world demand, and increasing protection in many industrialized countries—some of it specifically directed against these three economies—the U.S. dollar value of the exports of the three rose by 39.4 per cent, 56.2 per cent, and 52.2 per cent in 1976.
One thing at least is certain. The more labor-intensive manufactures of the now semi-industrialized countries need no protection or subsidization. Of course, it will still be argued that the least industrialized countries need considerable promotion of manufactures, if not protection, to get going. This may be true. But need it be very heavy? And need it be protective rather than promotional?
Of course, liberalization and reduced protection cannot achieve all the miracles it has produced for the four Asian countries discussed. It is obvious that the speed of the consequential changes in industrialized countries has some limit. But the fact that there is this limit to the degree of market penetration that will be permitted at any one time by the industrialized countries (and a limit to the size even of their markets) should not be regarded as a reason to continue with a policy bias in favor of import substitution. There are still gains to be made by all countries, even if only a few more miracles can be expected.
The developed countries still produce a very high proportion of the labor-intensive manufactures they consume. Clothing imports from the developing countries account for little more than 5 per cent of consumption in the developed world. The range of labor-intensive goods for which there has been any significant market penetration is still quite limited. Yet for 15 years manufactured exports from developing countries have increased at the rate of 15 per cent per annum. With reasonable goodwill on the part of the developed countries, and appropriate policies on the part of the developing countries, this rate of growth could be maintained for a very long time. The growth of exports from Hong Kong, Korea, Singapore, and Taiwan is certain to slow down greatly because at present they export such a high proportion of their output; hence, the growth rate of exports must soon approach the countries’ overall growth rate, which is unlikely to exceed 10 per cent. In this situation there should be room for sales of comparable goods by other developing countries. It must be remembered that these countries now play a role that is far from small in markets for manufactured exports (the growth of exports from them in 1976 was worth about $7 billion).
Furthermore, developing countries trade little with each other. Such trade is greatly inhibited by their own import substitution policies. This has been of secondary importance because the market in developing countries is so much smaller than the market in the developed world, but its importance is increasing as markets in developing countries for tradeable items are growing faster than those in developed countries. Preferential regional trading arrangements are very much a second-best option, and they are also very difficult to negotiate. Policies of general liberalization and reduced protection can be of far greater benefit and cost much less in terms of scarce administrative talent. Yet organized and controlled regional trading remains the conventional wisdom of leaders of the developing world when trade among themselves is under discussion.
I have discussed mainly manufactured goods, but it seems also likely that considerable gains could result from more intra-trade in agricultural products and even minerals, although this seems to have been studied relatively little. Furthermore, there is little doubt that liberal trade regimes would help even if exports did not increase greatly. All the benefits arising from greater use of the price mechanism—the reduction in administration, in delays, in stocks, in corruption, and the increase in competition—occur anyway.
Transfer of technology
In the 1970s there have been renewed attacks on trade and foreign investment, much of it stemming from England. The focus is now more on the evils of inappropriate technology and its inappropriate products, on multinationals, and, indeed, on the transfer of capital in any form—including aid. It is related to the increased emphasis on income distribution and poverty. But trade is involved. It is claimed that trade has harmed the poor in developing countries. If there were no trade, none of the other alleged evils could result. If the conclusion that no trade should occur is too extreme, nevertheless trade and the transfer of technology should be carefully controlled, just as it has been for 25 years, only, I suppose, on a broader scale and more effectively (see, for example, Singer, 1971, Keith Griffin, 1974, and Paul Streeten, Finance & Development, September 1977).
How is it, then, that the developing countries have grown so fast as a group, that is by 5.6 per cent in 1960–65 and by 6.0 per cent in 1965–73? Within these averages, half a dozen countries have grown, without special benefit from ample mineral resources, at rates which would have been deemed inconceivable 20 years ago. The industrialized countries do not grow as fast, nor the centrally planned economies either. None but Japan has ever achieved even half the growth rates of the half dozen developing countries with the best performance. Even Japan’s high growth was for nearly a hundred years based primarily on foreign technology, designed for countries with higher wage levels than Japan.
I suggest that the basic reason is plain. Such rapid growth can only be the result of catching up by importing relatively inappropriate technology. The result is technical change. Yet Keith Griffin has recently advanced the theory that growth in developing countries has come predominantly from increases in the amount of labor and capital and not from technical change. Not only are the supporting figures and analysis inadequate but the conclusion flies in the face of common sense. No one can be against making technology more appropriate—that is, more labor intensive—but this should not blind one to the fact that foreign technology can and has produced miracles. These benefits are found not only in the agricultural and industrial spheres but also in advances in public health and life expectancy. We can all agree that the poor in developing countries have not benefited from growth as much as desirable. But the claim that the mass of the poor have not benefited at all is rhetoric. Even in Brazil, where inequality has increased with fast growth, it is clear that the poorest 40 per cent of the population has benefited. Most observers will also agree that they would have benefited more if development had been more labor intensive. I am convinced that with appropriate and open policies on the part of the developing countries, more labor-intensive development would have resulted.
In the open and fast growing Asian economies, the standard of living of the relatively poor has been revolutionized in a period of 15 years. If one wants to look for millions who have scarcely benefited, or even lost out in the postwar period, there can be little doubt that one would find them most easily in the slow-growing, low-trading countries, principally in South Asia and the Sahel. To attribute this primarily to foreign technology is the greatest absurdity. The attack, mainly Western-inspired, on the transfer of technology, and on one of its modes of transfer—the multinational firm—has gone much too far, and threatens the poor.
However, I do not think that these recent arguments for import and technology controls have much effect on leaders in the developing world, as they violate common sense. Even the hatred and fear of the multinationals, and the feeling of being “dominated,” seem to be dying down, as developing countries come to realize their strength. Many countries in Latin America are trying to struggle out of the trap of having established an excessively inward-looking industrial structure. The problem is that change takes many years if it is to be relatively painless for all. At the same time, potential exporters need confidence that exports will remain profitable, and this, in turn, means confidence that a government committed to change will remain in power. The tragedy is that elsewhere, especially in Africa, a number of countries seem to be falling into the old trap of anything goes provided it is capital-intensive import substitution.
Related reading list
H.W. Singer, “The Distribution of Gains Between Investing and Borrowing Countries,” American Economic Review, Papers and Proceedings, May 1950.
J. Bhagwati and V.K. Ramaswami, “Domestic Distortions, Tariffs and the Theory of Optimum Subsidy,” Journal of Political Economy, February 1963.
Toward a New Trade Policy for Development, United Nations, 1964.
D.B. Keesing, “Outward Looking Policies and Economic Development,” Economic Journal, June 1967.
Ian Little, T. Scitovsky, and M.FG. Scott, Industrialization and Trade in Some Developing Countries, Oxford University Press, 1970.
B. Balassa, et al, The Structure of Protection in Developing Countries, Johns Hopkins Press, 1971.
K. Griffin, “The International Transmission of Inequality,” World Development, Vol. 2, No. 3, March 1974.
“The Distribution of Gains Revisited,” IDS, May 1971, published in H. Singer, The Strategy of International Development, International Arts and Sciences Press, New York, 1975.
“Export Incentives and Export Performance in Developing Countries,” World Bank Staff Working Paper No. 248, January 1977.
J. Bhagwati, Anatomy and Consequences of Trade Control Regimes (National Bureau of Economic Research, New York, forthcoming).
Anne O. Krueger, Synthesis, (National Bureau of Economic Research, New York, forthcoming).
Walter Galenson (editor), The Economic Development of Taiwan (Cornell University Press, forthcoming).