2. THE EXTRAVERSION OF THE UNDERDEVELOPED ECONOMIES

Extraversion must not be reduced to the quantitative importance of exporting activities in the underdeveloped economies: with “import­substitution” industrialization, extraversion assumes new forms.

The Historical Origin of Extraversion

Colonial trade. The Industrial Revolution was preceded in Europe by an agricultural revolution that released part of the labor force from the countryside, created the proletariat, and at the same time established the conditions for autocentric industrialization with the surplus that provided sustenance for the towns. The new industry ruined the old handicrafts, but at the same time recruited its labor force from them. Although this twofold process was accompanied by poverty and unemployment, it nevertheless repre­sented an advance in the development of the productive forces, and the new socioeconomic equilibrium emerging from this process of transition to central capitalism was a higher equilibrium than that of the precapitalist society that existed previously.

The transition to peripheral capitalism follows a different pattern. Transformation of a natural subsistence economy into a commodity economy is never a spontaneous consequence of the introduction of new manufactured goods, causing the peasants to produce agri­cultural products for export in order to satisfy new wants.

I therefore methods of primitive accumulation. The obligation to pay taxes in money form is the most widespread device employed. In the same connection, however, we should not forget “compulsory crops”: in Tropical Africa, for example, the “commandant’s fields,” with the obligation to grow export crops. In extreme cases the cultivators have been simply expropriated: the creation of inade­quate “native reserves,” so that the African peasants are obliged to go and sell their labor power to the mines, factories, or plantations owned by the Europeans, belongs in this context. It has played a determining role in South Africa, Rhodesia, and Kenya. Rey gives the name of “colonial mode of production” to the totality of economic and political relations at this stage of transition.

Specific distortions make their appearance, altering the original society and depriving it of its traditional character. In general, the

The Development of Underdevelopment 205 "prestige goods” in which the surplus of the traditional mode was embodied can henceforth be bought. This is the case, for example, with the total property accompanying exchanges of women between social groups. This Hiercantilization of precapitalist relations becomes a powerful factor in the penetration of capitalist relations. It compels people to go in search of money, and so either to become commodity producers or to sell their labor power. The land itself tends to become an object of private appropriation, to become a commodity, and ground rent makes its appearance.

The transition to commodity economy in Europe was accompanied by an advance of the productive forces, for it resulted from an improvement in the productivity of labor in agriculture. In the cases we are now considering, however, it is most often found that an increase in production per man is accompanied by an increase in the amount of labor put in. This is so in the agriculture of Tropical Africa, where, almost always, the growing of export crops is added to that of traditional subsistence crops, and does not take their place. Thus, a civilization based on a certain annual contribution of labor is superseded by one based on a larger contribution. This transition is painful and sometimes resisted by those involved: hence the use of extraeconomic methods like compulsory cultivation of certain crops. Large-scale landowners have favored the transition from subsistence to commodity-producing agriculture, and have to a large extent annexed the profits arising, without much improvement in agricultural productivity. This being so, the primary money income acquired gives rise to only a modest demand for local products, and affects mainly the. demand for imported goods. Extension of the domain of commodity exchange where local products are concerned is thus a slow process.

■ This distortion of the traditional mode evicts part of the population from the land, proletarianizing it, but without creating a demand such as to provide employment for this surplus of population caused by the subjecting of the precapitalist structures to the requirements of foreign capital. This lack of a way forward through autocentric industrialization accounts for the increased “pressure on the land” that is so frequently to be observed in the Third World. The increase in the density of population in the countryside leads to a retro­gression in agricultural technique, for progress in agriculture is

usually expressed in the use of more capital and fewer men per hectare. The concentration of landownership and increase in the rate of ground rent reflect this agrarian crisis, perpetuating and reinforcing it.

Where a prosperous body of local craftsmen existed, its destruc­tion through competition from imported manufactures brings about a second retrogression, to be contrasted with the progress that the destruction of the crafts by local industry represented in Europe. The history of the ruin of the crafts in. India and Egypt has been written by Dutt, Clairmonte, and Issawi. Whereas in Europe society found a new equilibrium that ensured employment for its labor force, what we see here is a regressive equilibrium that casts a part of the labor force right out of the productive system.

Foreign investment. The conditions for unequal exchange — that is, for the reproduction of underdevelopment — are thus gradually assembled. The distortion of precapitalist agrarian relations and the ruin of the crafts bring about urbanization without industrialization. The low level of the reward of labor, at one pole, and the concentration of capital, at the other, encourage foreign capital to establish modem sectors in the periphery, producing for export.

To be sure, unlike private U.S. investments, which during the last two decades have been channeled, to the extent of more than 50 percent, into oilfields and mining, only a third of British capital abroad is invested in direct export activities: public services, railways, trade, and finance constitute together a much larger fraction of this capital invested outside Britain. For France the proportion of investments in tertiary activities is even bigger: during the nineteenth century the bulk of capital invested abroad went into loans to govern­ments, public services, trade, railways, and banking. However, it quickly becomes apparent that the (mainly tertiary) sectors that have received, along with the plantations and mines, the major part of the capital originating from the center are to a very large extent grafted upon the exporting economy, to which they form a necessary complement.

The Development of Underdevelopment 207 foreign investment assigned to these sectors is.about 15 percent of the total foreign investment in the underdeveloped world.

In some capitalist countries of the periphery — the oil-producing and mining countries and some of those with a plantation economy — the bulk of foreign investment has gone directly into the export sectors. In others, where the principal exporting activity is indigenous agriculture, this investment is concentrated in the “ancillary” tertiary sector. Consequently there is great unevenness in the degree of penetration by foreign capital. Thus, Cuba before the national­ization measures (a typical plantation economy), and Zaire, Zambia, and Chile (typical mining economies) received, per head, between five and thirty times as much capital as Brazil, Indonesia, Senegal, India, or Egypt; and the oil-producing countries received even more.

While in the second type of countries of the periphery a large proportion of local capital has been invested in export activities, the amount of these investments is often underassessed when they take the form of scattered investments in land improvement. Thus, in Egypt, agriculture, the chief source of exports, absorbed 30 percent of the nation’s gross investment between 1882 and 1914, 12 percent between 1914 and 1937, 14 percent between 1937 and 1947, 4 percent between 1947 and 1960, and a higher percentage since then, with the building of the Aswan High Dam. These investments played a decisive part in growth, at any rate down to the First World War, after which the development of light industries producing goods previously imported took over the function of economic driving force: in 1882 agriculture absorbed 58 percent of national capital, in 1914 it absorbed 48 percent, and still 21 percent in 1960.

Before the First World War a considerable part of the capital exported from the old centers of Europe was invested in the public debt of other countries. On the eve of the Second World War the proportion of the public debt of the colonial and semicolonial countries held in the great finance markets of Europe and North America ranged from 40 to 100 percent of the total amount of this debt, and accounted for between 15 and 70 percent of foreign investment. This investment corresponded Iargelyto public expendi­ture on the infrastructure, occasioned by the integration of the periphery into the world market, an example being the great irrigation works undertaken by the Khedive Ismail in Egypt.

After the Second World War, the direction given to the use of what was thenceforth called “aid,” while varying from country to country, tended to result in more attention to the financing of industry, including industry working for the home market. Soviet policy has played an important role in this connection, leading the West to revise its own attitude. It remains the case, however, that the doctrine laid down by the International Bank of Reconstruction and Development stipulates that investment must facilitate an improvement in the balance of payments such as to ensure repayment of the loan, together with the interest on it. The USSR itself has been moving toward this attitude for some years now. This confers a new dimension on the distortion toward the external market, within the context of an international specialization that concedes to the countries of the periphery certain industrial activities hitherto denied them.

At the outset, when contact was first established between the center and the periphery, if real wages (or real rewards of labor) were equal, the center, whose productivity was higher, could export, whereas the periphery was not competitive ijn any sphere, and could export nothing but exotic agricultural products or crude minerals. It was in that form that international exchange began — first with exotic agricultural products, then, later, when the cost of inter­continental transport had been sufficiently reduced, with crude mineral products, which required investment of foreign capital on a scale unknown until then.

Following in the wake of colonial trade, the creation of enclaves of foreign capital, especially in the mining sector, engendered no greater monetary demand for local products than had arisen from previous colonial exchange, for the primary income distributed by this type of enterprise largely evaporated in external “leaks.” A substantial.proportion of the expenditure of the foreign enterprises went directly into the foreign market — for the purchase of plant and for the payment of exported profits. Furthermore, part of the wages paid locally made its way out of the country, when the manufactured goods sought by the new workers were imported. Only part of this wage fund entered into local demand (mainly for foodstuffs); this made an important contribution to the spreading of commodity relations.

In the case of the exploitation of bauxite in Guinea by the Fria complex, for instance, only 12 percent of the total investment costs and barely 25 percent of the total value of the alumina exported remain within the country. As regards the exploitation of the oilfield in the Algerian Sahara, local expenditure occasioned by investment did not exceed 44 percent of the total, and then half of this local expenditure subsequently transformed itself into imports. The pro­portion represented by local costs in the value of the current exports of oil is even less — hardly 22 percent.

In the case of large-scale mining or oilfield exploitation, the major part of the fraction of “primary monetary” expenditure that remains inside the country is ultimately represented by the income the state takes, either in the form of royalties or as taxes, direct and indirect.

Industrialization by import-substitution. The industrialization of the Third World conforms to a pattern of import-substitution, which follows a process that “climbs up” from the light consumer­goods industries to the industries producing semifinished goods, and then to the industries producing equipment, whereas at the center the process of industrialization embraced all forms of industry at the same time, when it did not “descend” from the heavy equipment­producing industries to the consumer-goods industries downstream.

This industrialization of the periphery occurred late in the day — between the two world wars, in the case of Latin America, and after 1945 in that of Asia and Africa. The delay cannot be attributed to the meagerness of markets, owing to the low reward of labor, for this does not in itself constitute an obstacle to industrial­ization. The market is not composed of consumer goods alone: production goods play a big role there. Low wages mean high profits, and so it is possible for entrepreneurs to save and invest, that is, to create a market. In Europe industrialization took place on the basis of very low wages, to begin with, and the same is true of Japan. When productivity in the enterprises established in the periphery is similar to that in the countries of the center, lower wages make possible a higher rate of profit there. But the divergence between the rewards of labor became sufficient to matter only at a period when the concentration of industry at the center was itself well advanced. Under these conditions, it was the same monopolies that exported commodities to overseas countries that invested capital in those countries. They endeavored to maximize their profit on the whole of their activities, at the center and in the periphery, and this led them to prefer investing in the periphery in production for export. As for local capital, this was insufficiently centralized, and did not exist on a scale enabling it to compete with the foreign monopolies, and so chose, where possible, to go into sectors that were not competitive but complementary, such as comprador trade, or services.

In so far as there existed an industry focused on the local market, this faced a market that was distorted in character, owing to the low level of wages, and developed in response to the demand of the privileged strata, neglecting that of the masses. Moreover, the import-substitution industries employed modern techniques that were too capital-using to absorb the unemployment caused by the aggression of the capitalist mode of production, and so reproduced the conditions of a market in which abundant labor supply kept wages low.

The generalizing of the pattern of import-substitution industrial­ization opened up new opportunities for foreign capital without essentially affecting the extraversion of the economy. Beginning with the production of consumer goods that had previously been imported, the country was content to replace these imports by imports of capital goods and intermediate goods. An autocentric strategy must base itself simultaneously upon the production of consumer goods and capital goods. External trade then involves, as regards exports and imports alike, both consumer goods and equipment goods, so establishing the conditions of equal exchange.

The international division of labor within the multinational firm. The transnational firm, which appeared after the Second World War, is characterized by the scattering of its production activities across the world. It is made up of establishments distributed among the five continents, thus realizing a model of vertical integration that is often total. These establishments supply various items of a category of products the growing demand for which is typical of the “age of consumption.” These are durable goods (domestic appliances, electrical and electronic devices, vehicles, etc.), always sharply distinguished by a particular trademark and the organization needed for after-sales service. The international dis­persion of the.different stages of production of these goods is a sign of the birth of a world production process in the full sense of the term: the old international division of labor, materialized in the exchange of products, is being replaced by a division inside the firm.

The choice of locations for these integrated activities is based on comparisons between wages for the same productivity. In East Asia the hourly wage in the textile industry varies from $.10 to $.30, as against $2.40 (or between eight and twenty-four times as much) in the United States, for the same productivity; and in the electronics industry the ratio is 1:7. It is thus to the firms’ interest to establish those “links” in their production that demand relatively the largest amount of labor in the countries where labor is cheap.

From the standpoint of the international division of labor, this dispersion leads to a new form of inequality between nations. At the center are gathered the strategic activities, those which the jargon calls “software” (technological research and innovation, manage­ment), the “gray matter” in one form or another, and the production of the most complex types of essential equipment, those that require highly skilled labor. To the periphery falls the “hardware” — the production of those elements which, given the help of imported equipment, require Onlyordinarylabor. For, despite the name given to it, the transnational firm remains national in its origins and in its top management. It is usually American, and less frequently Japanese, British, or German. The old division of labor, in which the underdeveloped countries supplied the raw materials and the advanced countries supplied the manufactured goods, is being replaced by a different division, in which the former supply the primary productsαnc? the manufactures, while the second supply the equipment and the “software.” This division reinforces the functions of the centralization of decision-making authority and technological innovation. Thereby it reproduces its own conditions, splitting the world labor market into watertight national markets with big differences in rewards. Itdeepens unequal exchange by internalizing this in the firm.

The effects of this new inequality are several. In the first place, the international division of labor deprives the periphery of any initiative in its own development, and thereby reduces -to nil all chances not merely of “catching up” in terms of consumption but even of aspiration to some sort of autonomy, if only of a cultural and political order. Then, it adds to the transfers of values from the periphery to the center. The visible transfers alone, in the forms of reward of capital and payments for "software,” and arising from the monopoly of specific kinds of equipment, are enormous. UNCTAD, which associates these transfers with technological domination, estimates their annual increase at 20 percent. This division of labor disintegrates the economies and societies of the. periphery. The missing links in them are multiplied through the centralization of the directing links at the center and the dispersion of the dependent links among many different locations, so as to bring about competition between the “little nations” and weaken their negotiating power. The transnational firm intensifies competition between underdeveloped countries, reproducing in them parallel structures that make impossible the development of integrating complementarities within broader structured economic spaces, which is the condition necessary for an independent development of these countries. On the plane of unevenness between regions, sectors, and labor markets, concentration in a few towns, where the “external savings” are greatest, aggravates existing distortions, especially as between town and country. Employing little labor, and preventing the transformation of agriculture and the backward sectors in the underdeveloped economy, these implantations con­tribute no solution to the unemployment problem: on the contrary, they worsen it by accelerating the disarticulation of the society in which they make their appearance.

These new tendencies in the international division of labor are as yet not very visible in the Third World as a whole. Only in

The Development of Underdevelopment 213 East Asia (South Korea, Taiwan, Hong Kong, and Singapore) and in Mexico is it possible already to study its effects. The installation of “runaway industries,” originating in the United States, Japan, and Britain, in these territories has been sufficiently systematic to ensure, during the 1960s, a growth of manufacturing industry at exceptional rates — between 16 and 35 percent per year — and an overall growth in production, based on this type of industrial­ization, at rates varying from 7 to 10 percent per year. These five countries represent on their own nearly three-quarters of all the exports of manufactured goods from the Third World, which amount to about $4.4 billion. Their industries export to the advanced countries, especially the American market. They are, in the main, light industries (textiles, apparel, and leather goods, $1.6 billion; food and drink industries, $0.8 billion; wooden goods and furniture, $0.4 billion; etc.). But the very fact that they are concentrated in a few underdeveloped countries rules out the possibility of this being a development that could be extended to all the countries of the Third World.

The five countries mentioned above have nevertheless been put forward by the West as a model for the Third World, in reply to Chinese communism and Latin American nationalism. However, these hopes have already been dashed. Despite a big increase in exports, the balances of payments of the countries concerned have remained very vulnerable. First, because the exclusive allocation of investments to industries of this type has been made at the expense of agriculture and the industries aiming at the home market, so that there has been a rapid growth in imports in those sectors. Second, because imports of equipment and semifinished goods have increased at the same pace as industrialization. Finally, and above all, because transfers of profits, visible and invisible, largely wipe out the benefits of exports. The external balance worsens when the rate of inflow of new capital slows down, repro­ducing the familiar pattern of the blocking of independent growth. Although they have achieved high growth rates, none of these countries. has approached the stage of independent and self­maintained growth, dear to the theory of “take-off.” On the contrary, they are even more dependent than they were twenty years ago.

Looked at from another angle, this type of industry engenders

a “semiaristocracy” of labor — few in numbers, badly paid by I Western standards, for the same productivity, yet privileged by virtue of its security of employment in comparison with the pro- Ietarianized masses doomed to unemployment and casual work. This privileged situation ensures the docility of the proletariat, a condition for the reproduction of the system. The pattern of these industries rules out, moreover, any technical advancement, since the center keeps for itself those links that require skilled labor. Finally, the strengthened domination of central capital forbids any formation of a bourgeoisie of national entrepreneurs. These industries do, however, engender a. middle class of salaried pro­fessionals — executives, engineers, office staff — who cleave to the patterns of consumption and ideology of the world system to which they organically belong. The “elitist” ideology that is grafted onto this type of dependence, and the degeneration of the national culture, lead to acceptance of a reduction in the sphere in which decisions are taken by the nation itself.

While these tendencies are also to be seen in the West,'they do not have the same decisive implications as in the Third World, because they are grafted upon a different historical substratum. The flight of American (and also British) industries to Continental Europe, particularly Italy, does not alter the fundamental social structures already established there, and is often carried out in association with the local bourgeoisie. Moreover, this flight has already effectively slowed down the rate of growth both in the United States and Britain and created zones of depression and unemployment. Thereby it includes its own limitation. It ensured the U.S. domination over Europe for a certain period, but at the same time it created the conditions for a challenge to itself to arise. Indeed, the rapid growth of Europe and Japan in recent years is due in part to this redistribution of industry. Because the historical conditions are not those of the Third World, this revival has set in motion a wave of progress and technological innovations, especially in Japan and Germany, which have finally put an end to American domination. It is not possible to equate the asym­metrical processes of domination and-dependence that characterize the relations between the center and the periphery with the uneven process of development inside the center.

Sectoral Unevenness in Productivity and Transmission of the Price Structure from Center to Periphery

Ifwe break down production (value added), on the one hand, and the occupied labor force, on the other, into sectors, and compare the average sectoral product per head in the advanced countries and the underdeveloped ones, we are struck by the relative concentration of product-per-head around their national average in the countries of the center, and their very marked dispersion in the countries of the periphery, as is shown in the table below, compiled by Anibal Pinto:

Gross Product per Occupied Person (1960)

In Latin America the extreme ratios of productivity observed, at the level of aggregation employed, are 1:11, between agriculture and extraction industry, as compared with only 1:1.4 for Britain and 1:3 for the United States.

Actually, this phenomenon is due to what I shall call “sectoral, unevenness in productivity.” It is not, of course, possible to compare productivities in the strict sense of the word except between two enterprises that produce the same product: the productivity of one will be said to be higher than that of the other if the total amount of labor (direct and indirect) necessary to ensure the production of one physical unit of the same product is less in the former case. Between one branch and another one can speak only of different profitabilities, as Emmanuel has reminded us. AΠ the same, if, with a given price structure, conditions are such that labor, or capital, or both, cannot be rewarded in one branch at the same rate as in another, I say that productivity is lower in that branch. In the capitalist mode of production, distinguished as it is by the mobility of the factors, that is, the existence of a market for labor and for capital, the effective tendency is for labor and capital to be rewarded in all branches at the same rates. If, however, this price structure, corresponding at the center to homogeneous rewards for labor, and capital, is transferred to the periphery, the result will be that the factors cannot be rewarded at the same rate in the different branches if the technical conditions (and so the productivity) are distributed otherwise than at the center. Direct comparisons of productivity are sometimes possible if the product is, if not exactly identical, then at least comparable as regards its use-value and the techniques that can be employed to produce it; If, for example, a quintal of wheat, produced at the center, requires a certain total quantity of labor (direct and indirect), and if a quintal of millet — a product of the periphery that is com­parable both in use-value (a grain crop with the same nutritional potential) and in respect of the techniques that can be used to produce it — requires a larger quantity of labor, then this is because the production techniques in the periphery are backward. We are justified in speaking of a difference in productivity. In contrast to this, productivity will be the same at the center and in the periphery in the textile industries, where techniques are similar. For other products, of course, direct comparison of productivities is not possible: for example, for coffee, which is produced only in the periphery and cannot be compared to any product of the center.

Now, the price structure of the center is, in fact, transferred to the periphery. For there is a world market through which trans­ference is inevitably effected to the periphery of the essential structures of relative prices that prevail at the center.

There is no reason why product per head should be the same in the different branches of a central capitalist economy. For this product is made up of two components, the reward of labor and the reward of capital, and, for product per head to be identical, five conditions would have to be fulfilled: (1) that the quantity of labor provided per occupied person (per annum, for instance) be the same; (2) that the organic composition of labor (Emmanuel’s expression), meaning the proportion of kinds of labor with differing levels of skill, be the same; (3) that the rates of reward of labor (with the same skill) be the same; (4) that the amount of capital used, per worker (the organic composition of capital), be the same; and (5) that the rate of reward of capital be the same.

There is, however, a tendency in the capitalist mode of production toward the fulfilment of precisely these conditions. Capitalism does in fact tend to make labor-time uniform, to reduce it to its simplest, least skilled category, and reward it at a uniform rate, just as it tends toward equalization of the rate of profit. Further­more, there is a tendency toward intensive use of capital in all branches of the economy; this constitutes the way in which pro­ductivity is increased. True, between one branch and another the organic composition of capital is different, and the higher the degree of disaggregation in the analysis, the wider is the range, with the new dynamic industries having the organic composition that increases most rapidly. It is this “scatter” of organic compo­sitions that explains the fact that sectoral productivity is unevenly distributed at the center. But this “scatter” is even more pronounced in the periphery. If we break down the national economy into about ten branches, we find that the organic compositions at the center range from 1 to 4, and so, with an average rate of profit of 15 to 20 percent, productivities vary from 1 to 2, whereas in the periphery, with the same breakdown, the range of organic compo­sitions extends from 1 to 35, and that of productivities from 1 to 10. Such a big divergence for the organic compositions of capital in the periphery is possible only if the capitalist mbde of production has not taken hold of all the branches of production, as it has at the center. It is this circumstance that accounts for the sectoral differences in reward, *and constitutes the principal aspect of the problem of unevenness in the distribution of income in the Third World.

To this main cause of unevenness in the distribution of income must be added some other causes which are also important, and which are connected with the incomplete extent of the development of capitalism: a low level of Uniformization of labor-time (especially between agriculture, where capitalist forms of organization do not prevail, and the urban sector), different rates of profit for foreign monopoly capital and dependent national capital, etc.' There are also some factors of a secondary order, such∙ as: (1) the respective levels of employment in the rural and urban areas, which have a determining influence in the division of income between wages and incomes of enterprise and of ownership; (2) the structures of distribution of ownership of capital and enterprise, which mainly determine the way income of enterprise is distributed in the urban areas; (3) the structures of distribution of landownership and of the way the land is exploited which mainly determine the distribution of nonwkge incomes in the rural areas; and (4) the distribution of the labor supply in accordance with the levels of skill and degree of trade-union and political organization of the different groups, which is what largely determines the structure of the distribution of wages.

The considerable divergences that are sometimes to be seen in the underdeveloped countries, between the average wage and the average income of the most deprived strata, especially the peasantry, are the inevitable consequence of the juxtaposition of two economic systems that belong to different epochs, and whose levels of pro­ductivity are not to be compared. It would be wrong to draw the hasty conclusion that one of the aims of economic policy must be to reduce the level of wages. In fact, a higher level of productivity not only makes possible higher wages but also, to a large extent, demands this. The Marxist conception of the value of labor power brings out this connection. This is why comparisons between standards of living, when types of income are too different, become dubiously valid — not to mention comparisons between standards of satisfaction, welfare, or happiness, which often lure economists beyond the realm of science. Such comparisons overlook price levels, which differ markedly between rural and urban areas in the underdeveloped countries; the foodstuffs provided by a food­gathering economy, which very easily are gathered in some parts of Tropical Africa, but which can be sold at high prices in the towns; the high cost of housing in the urban centers, even in the shanty-towns; the products of the food-gathering economy and of hunting, which do not figure in the national accounts; the different way of life in the towns, which entails new demands.— fares, enter­tainment that has to be paid for, etc. The intensity of labor also has to be taken into consideration. It is often forgotten that the income of the traditional peasant corresponds to a hundred working days per annum, whereas that of the town worker corresponds to three hundred working days. If we allow for all these factors in the problem, the difference between recorded incomes, which is sometimes of the order of 1:10, often loses its dramatic character.

The problem of the “privileged wage earners” lies elsewhere. The hierarchy of wage levels is usually more pronounced in the underdeveloped countries than in the central economies. In the modern sector of the economy, both the plantations and the towns, the mass of unskilled workers, relatively more numerous, make up the most deprived group in the nation. It is in relation to this mass, and especially where unemployment in the towns and the underemployment of the landless peasants attain vast proportions, in relation to this mass of underemployed persons, who are mainly unskilled, that the wages of the skilled workers (both manual workers and office workers) confer a sense of privilege. TTie same applies to the groups employed in public service, especially where there is a widespread feeling that their numbers are too large and that recruitment is governed by the pressures of the “little society of the town,” anxious for jobs. If, in addition, incomes of national capitalist enterprise are not present, then the privileged situation of these groups of workers acquires political significance.

It is generally supposed that there is bound to be an increasing divergence in the underdeveloped countries between the average income of the mass of the workers, the growth of which can only follow the very slow growth of the national product, and that of the more highly skilled categories, which will follow the income of similar workers in the advanced countries. Actually, these imitation effects are confined to the most highly skilled categories, the ones who are in a position to emigrate from their own country: this is what is called the “brain drain.” The few data available for ¹ assessing long-term movements give cause to think that the gap was very large from the outset, perhaps just as large as today, especially wherever the heterogeneity of the two worlds, the tradi­tional one and the modern one installed by colonization, caused the supply of labor to be inadequate in the new sector. Gradually the gap becomes narrower for the mass of unskilled workers in the modern sector, as emigration from the country to the towns develops, whereas it widens for the more highly skilled categories.

In the advanced countries, wage earners constitute the great bulk of the working people, between 60 and 90 percent of the occupied population. Consequently, in the long run the average wage cannot evolve very differently from the national product per head. Again, in the industrialized countries, the working class is, broadly speaking, fairly solid as regards unity in struggle, through its organization in trade unions — except where, as a result of racial differences (as with the black and white workers in the United States, for instance) or national ones (as with native and immigrant workers in some European countries), this solidarity is broken or at least impaired. The growth of wages therefore tends to be fixed at a uniform rate for workers in all branches of the economy, around the average growth rate of productivity, rather than around the varying growth rates of productivity in each separate branch. This being so, wages policy is a fundamental element in national policy on income distribution.

The situation is different in the underdeveloped countries, where wage earners make up only a small part of the occupied population — between 1 percent and a maximum of 30 percent — and where solidarity is not so strong, owing to the backward state of trade-union organization and the gap that exists between the worlds of the country people and those of the towns. There is therefore no obvious relation between the long-term evolution of wages and that of the national product. We find that in some countries in some recent periods a very low, or medium, growth of the national product (between 0.2 and 3 percent) was accompanied

The Development of Underdevelopment 221 by a marked growth of real wages (over 6 percent per year in Jamaica and in Colombia; 4.5 percent in Ceylon; over 8 percent in Zambia_j Rhodesia, Nigeria, and Tanzania), or else, contrariwise, very low rates of growth of real wages, even negative rates, in cases where the growth of production per head was relatively better (Taiwan, Burma, South Korea, India, Philippines, etc.). Phenomena like this are not open to simple explanations, for there is not the slightest correlation between the movement of wages and the pace of industrialization, or even the movement of profits. Cases are known (Belgian Congo, Puerto Rico) where a steady increase in wages has stimulated enterprises to choose more efficient methods of production. As regards response to chronic inflation, we find every possible case: belated adjustment of wages, steady advance of real wages, or, on the contrary, steady reduction of real wages. Elastic behavior by wages, upward or downward in real terms, is only possible because the problem of wages does not constitute the main axis of income distribution in the given country.

The large divergences, both absolute and relative, in the levels of remuneration of the different categories of working people in the underdeveloped countries — between those in the countryside and those in the towns, between the skilled, and the unskilled, between those employed by certain large enterprises and the rest — even if explicable on strictly economic grounds, constitute an obstacle to the building of a coherent nation. It is conceivable that an economic policy of development might aim to work sys­tematically against the “natural laws” of the economy, seeking to reduce these divergences in order to ensure national cohesion. But this policy can be justified only if the reduction in the rewards of the privileged categories of workers that it undertakes to achieve is not effected for the benefit of other categories of income, such as incomes of private enterprise, whether national or foreign, but for that of the community as a whole, and provided that the categories affected by this policy are clearly aware that this is so.

An egalitarian policy of this sort is politically rational, since the aim of national cohesion is essential for development. It means, however, that a price system must be adopted that differs from that of market prices. The actual price system in the under­developed countries, which is largely determined by that prevailing in

the advanced countries, as a result of international competition and the substitution of some products for others, corresponds to a relatively uniform distribution of productivities. In view of the much wider “scatter” of productivities, in the underdeveloped economies, the rewarding of labor and of capital, respectively, on a uniform basis would result in a system of prices that would not be rational from the standpoint of economic calculation when it came to choosing the sectors of the economy to be developed. Two price systems would therefore be adopted, with rationalities situated on different planes: a system of actual prices intended to “iron out” inequalities in rewards and ensure national cohesion, and a system of reference prices serving the needs of economic calculation. Only in the course of development would the unevenness between productivities be reduced and the two systems draw close together.

The nature of the political relations between foreign capital, the local business bourgeoisie, the privileged strata of wage earners, and the administrative bureaucracy is what ultimately determines important aspects of the evolution of this social distribu­tion of income. When there is no business bourgeoisie, as is often the case in Black Africa, the privileged wage-earning strata may become, along with the administrative bureaucracy, the principal transmission belt of domination from without. But this does not always happen. In Zaire, for instance, between 1960 and 1968 it was the bureaucracy that grabbed- the lion’s share, while the condition of the working class, along with that of the peasantry, was worsened.

In the capitalist mode of production, the equilibrium prices that ensure supply is adapted to demand are prices of production in the Marxist sense. These prices presuppose equal rewards of labor in all branches (a single labor market) and an equal rate of profit on capital (equalization of the rate of profit). Consequently, if the same fraction of profit has to be saved in order to ensure expanded reproduction in all branches (let us say, for simplicity’s sake, if all the profit is reinvested, eliminating consumption by the capitalists), the structure of growth — the sharing of invest­ment between the different branches — is determined by the struc­ture of prices. If there were no capital market to ensure the circulation of capital from one branch to another, there would be no guarantee of coherence between the structure of growth and that of demand, modified in its proportions by this very growth. The circulation of capital is therefore a necessary law of the way the capitalist mode of production functions. But this circulation comes up against a permanent obstacle: the ownership of capital. The enterprises and branches that are called upon to grow faster, as a result of the evolution of demand, are afraid, should they need, in order to finance their investments, to draw upon external capital to too large an extent, that they may lose control of their affairs. They therefore try to include in their prices a margin large enough to make possible an adequate volume of self-financing. The circumstances of competition make such an operation possible, more or less. A price system that would be rational from the standpoint of growth would imply (leaving out of account any consumption by the capitalists) a structure of prices by which each branch could finance its own growth, in accordance with demand, without calling in external capital, and so different rates of profit, or else, on the contrary, an equal rate of profit and a perfect circulation of capital. The actual price system in the capitalist countries is neither the one nor the other, but something in between; the margins for self-financing vary widely, depending on many factors — including, for instance, the degree to which the given branch is monopolized. To this must be added the distortions that uneven indirect taxation causes in the price system.

It will be said that an enterprise or a branch has a higher productivity than others if it ensures, with equal rewards of labor, a higher rate of profit; and this is indeed the tendency if the branch has to have a bigger growth rate in order to meet a change in the size of demand.

Now, the price structure at the center is largely transmitted to the periphery for the reasons that also explain the mechanisms of transmission of the value of the dominant currency: psychological mechanisms connected with consumption models, competition by imported goods with local products that are more or less subject to substitution, etc.

This transmission of the price structure of the center determines in the periphery unevennesses of productivity between branches that express the uneven degree of modernization — of penetration by the capitalist mode of production. These unevennesses of productivity are often reflected in unequal rates of profit, but also in unequal rewards of labor, especially where sectors that do not belong to the capitalist mode are concerned, as is often the case with rural production. This price structure has, therefore, nothing rational about it from the standpoint of the needs of a growth organized in order to put an end to the historical Iagging-behind — uneven between one sector and another — which is characteristic of the periphery.

The transmitted price system is seen to be even more irrational if we consider that, with the generalization of monopoly in the central capitalist economies, the tendency to equalization of the rate of profit is continuously under challenge. The Hiarginalist theory of general equilibrium was constructed on the basis of an assumption of perfect competition. Taking the hypothesis of sudden cartelization of an economy in a state of competitive equilibrium, Joan Robinson drew the twofold conclusion that the national income would be redistributed in favor of the entrepreneurs and that the orientation of production would be altered. If, indeed, we assume that elasticity of total demand for products varies from one branch to another, just as elasticity of the supply of the factors varies, then it follows that more products for which the demand is less elastic will be manufactured than products for which the demand is more elastic; while, similarly, those sectors where the supply of labor is highly elastic will develop, whereas those where it is less so will decline.

It must be added that increasing the degree of monopoly in the economy does not increase the volume of saving in the apparent proportion resulting from Joan Robinson’s analysis, according to which what is lost by the factors of production is gained by the entrepreneurs. Indeed, when distribution is altered to the advantage of profit, the technique of production utilized tends to become more primitive, as Sraffa has shown. The level of the national product thus declines, and the entrepreneurs do not recover every­thing that the factors lose. Full development. of the productive forces is fettered. Furthermore, this more uneven distribution of income aggravates the contradiction between the capacity' >to produce and the capacity to consume — a contradiction that becomes a supplementary reason for an equilibrium of underemployment. Baran and Sweezy show that in monopoly capitalism the actual surplus is less than the potential surplus.

Analysis of the phenomenon of monopolization in terms of elasticity of the demand for products is, of course, far from satisfactory. The overall conception of the degree of monopoly in the economy takes into account the fact that every system contains, potentially, a certain degree of monopoly. There is always a curve of total demand for every commodity, whether this commodity be produced by a single enterprise or by an infinity of enterprises. The hypothesis of complete cartelization merely reveals the degree of internal, monopoly in the economy, and renders this effective. The method would theoretically enable us to measure the degree of monopoly in an economy in which production was entirely in the hands of the monopolies. It does not enable us to follow the actual evolution of the process of concentration. Now, analysis of this evolution, and of the distribution of monopoly superprofit — uneven as between different branches — is of decisive significance for the theory of prices. While analysis of the elasticity of demand enables us to understand the extent to which transfer of the profits of one branch to another can take place, it is in terms of the overall strategies of the firms concerned that we need to analyze the struggle — by means of prices and by other means (investments, purchase of blocks of shares, monopoly of trade­marks, etc.) — which goes on between monopolies in one and the same branch.

Looked at in another way, the world system of relative prices is, in part, the result of unequal exchange. This exchange — and the unequal specialization on which it is based ∙— serves a function, namely, to increase the rate of profit on the scale of the system as a whole. It is in these terms that we need to interpret the results of Kaleckfs analysis of the effects of monopoly. He believed he could define the reasons why the share of the national income obtained by labor had remained stable in the advanced countries in. the course of history: the progressive increase in the degree of monopoly had been balanced by an evolution in the terms of trade to the disadvantage of raw materials.

It needs to be added that the price of a raw material becomes purely conventional if its processing is carried out by firms that are integrated with those providing this raw material. This is the case, for instance, with the bauxite produced in Jamaica, Guinea, arid elsewhere by the same groups that control its transformation into alumina in Cameroon and into aluminum in Canada or Ghana. Depending on whether the group’s interest lies in localizing its profits in the periphery or in the center, it will fix high or low prices for the bauxite or the alumina.

The view that the system of declared (book) prices constitutes an objective criterion by means of which the rationality of economic choices can be judged is without any scientific validity. The tech­niques for evaluating projects, based on calculations of profitability that rely on such a price system, which are advocated by the ₁World Bank belong to the realm of pure ideology.

The Choice of Production Techniques in the Periphery: The Irrationality of the System

Marginalist theory claims that choice of technique is dictated by endowment in factors and that the economic system is rational, that is, that it ought to lead to light techniques being chosen in the underdeveloped countries. It is a matter of observation, however, that this is not so. Why not?

Does investment, under the particular conditions of the inter­national integration of the underdeveloped economies, take the direction that is most favorable to maximizing the pace of accu­mulation? The problem has three aspects. (I) The question of the total rate of investment: What is the mechanism that determines the division of the national income between consumption and investment? Does this mechanism, under conditions of under­development, determine a division that is particularly favorable to investment? Is it possible to determine a priori the proportion of the national income that it would be rational to devote to investment? In other words, up to what point is the restriction of consumption advantageous to a society which wants to speed up the rate of capital formation? (2) The question of the choice of investments: What are the mechanisms that guide investments toward one industry rather than another, differing in capital-intensity, and toward the use of one. technique rather than another? What are the effects on the rate of development of these mechanisms as they function in the setting of underdeveloped economies? Is it possible to establish a priori an order of priority among useful investments? (3) The question of international specialization from the standpoint of the differing capital-intensity of industries: What are the mechanisms that guide a country’s production mainly toward light industry or mainly toward heavy industry, when this country is integrated in the world market? Are the results of these mechanisms of international specialization favorable, in the case of the underdeveloped countries, to development at the most rapid pace? To what extent should organized investment-effort be based on the internal economy, and to what extent should it depend upon international exchange?

Marginalism considers that it is the rate of interest, and that alone, which determines the direction taken by investments — that is its theoretical position. It considers, moreover, that only a rate of interest freely arrived at on the money market is capable of guiding investment in a rational way and determining the rate of growth that conforms to individual preferences —∙ that is its doctrinal position. As the marginalists see the matter, the rate of interest is what adjusts the supply of capital to the demand for it. Now, production methods that are more capital-intensive cause the process of production to be prolonged and require a sacrifice from the consumer, who regularly prefers present consumption to consumption of an equal quantity in the future. The money market thus makes possible, through rates of interest, adjustment of the division of income between consumption and investment in accordance with the rate of “subjective underestimation of the future.” It determines the general rate of development that conforms to individual preferences. Furthermore, it is held, the rate of interest determines, besides the general rate of formation of saving, optimum distribution of investments between branches of production and optimum choice of production techniques. It is the rate of interest that ensures that no capital is invested in any branch beyond the point at which the increase in productivity resulting from the additional investment becomes less than it would be in other branches. Interest is, in fact (say the marginalists), not merely the yardstick of preference for the present but also that of the marginal productivity in value terms of the capital factor.

Does the rate of interest really play the decisive part in deter­mining the total amount of investment and the direction taken by capital? No, it does not. In the context of the capitalist mode of production, the division between consumption and investment is determined by the level of real wages (the rate of surplus value) and not by individuals’ preference for the present.

Indeed, how does Bohm-Bawerk set about proving that this division conforms to “individual time-preferences”? He starts from the principle that more intensive use of capitaΓ goods always makes possible increased production, but also requires a lengthening of the production period. This “length of the production process” is only a clumsy way of measuring the “capital-intensity” of production, something that the Marxist concept of the organic composition of capital expresses more clearly. This being so, Bδhm-Bawerk's claim is no different from Marx’s, namely, that the techniques that are most capital-intensive are also the most productive. The consequence of Bohm-Bawerk’s argument, however, seems less pertinent. Since the longer the production process, the more productive it is, the production of intermediate goods ought to develop ad infinitum. Yet this is clearly not the case. Why not? Because, Bohm-Bawerk tells us, owing to the subjective under­estimation of the future, although the physical volume of production can be increased indefinitely if we lengthen the duration of pro­duction, the value of this production, increasingly large in volume, but also increasingly distant in time, first grows and then shrinks, so that there is an optimum duration of production. For this to be so, however, we must presuppose, a priori, that the rate of subjective underestimation of the future is higher than the rate of growth of physical productivity when the production process is lengthened. To get out of this difficulty, Bohm-Bawerk then puts forward another proposition: the period of production cannot be lengthened indefinitely because the means of subsistence needed by the workers who make the instruments of production have to be produced. What does this new proposition mean? That the labor force can be divided into two categories: one engaged in producing consumer

The Development of Underdevelopment 229 goods, the other in producing equipment for production. Bohm- Bawerk’s new proposition means that it is not possible to reduce the fraction of the labor force engaged in ultimate production below the number needed to produce consumer goods equivalent to the wages distributed. The pace of development then appears as being fundamentally dictated not by the rate of subjective underestimation of the future, but by the rate of surplus value.

And so here we come once more upon Marx’s fundamental proposition. Sraffa proves how the social relation that dictates the level of real wages (the value of labor power) determines at the same time the average rate of profit and the system of relative prices. This “rediscovery” finally exposes the ideological character of the marginalist analysis and strips the practice of economic choice under capitalism of all rationality — or, more precicely, reduces this rationality to what it really amounts to: the means of reproducing capitalism’s own social conditions of reproduction.

The division of available income between immediate consumption and investment, or, in other words, the growth of future consump­tion, is thus a social choice. It is governed, in the capitalist mode of production, by the social relation between bourgeoisie and pro­letariat. In a rational society it can be governed only by a collective choice, effected on the basis of long-term considerations related to society’s aims (which go far beyond the time-prospect of capitalist economic calculation). This is the answer to the first question.

Where choice of production techniques is concerned, current theory still resorts to a type of marginalist analysis. A certain kind of production can be carried on equally well with various combi­nations of factors. If the rewards of the factors are given, it will be possible to choose from among the different possible techniques the one which, with a given stock of factors of production, weighted in accordance with their relative rewards, makes it possible to maximize immediate production.

The employment of a more advanced technique, characterized by a higher capital-intensity, is accompanied by an increase in the productivity of labor. Two cases can then arise. In the first, the improvement in the productivity of labor is less than proportional to the increase in capital-intensity. In this case, the productivity of capital diminishes. Here we have the classical hypothesis: if, in

order to make one unit of the product, one can-employ less labor, then one must necessarily employ more capital. In the second case, the improvement in the productivity of labor is more than proportional to the increase in capital-intensity. In this case, obviously, the productivity of capital is also improved. The first case enables one to choose between different “efficient” tech­niques, while the second enables one to eliminate the “inefficient” ones, that is, those that are inferior, whatever the relative rewards of the factors may be.

What policy should be recommended in an underdeveloped country suffering from a substantial degree of structural unem­ployment — in other words, where shortage of capital is the factor limiting growth, whereas labor is available in unlimited quantity? The techniques that, though the lightest, are inefficient in the sense defined above must, of course, be ruled out. Among the efficient techniques, it is often recommended that the one be chosen that is most economical of the scarce factor, and that therefore maxi­mizes the productivity of capital. This amounts to saying: the lightest technique among all the possible efficient techniques; Choice of a zero reference for wages leads regularly to preferences such as this.

This line of reasoning is highly questionable, even if we accept the assumption that the labor factor is indeed available in unlimited quantity. For, among a variety of efficient techniques, a less light technique may, at the prevailing rates of reward of the factors, enable a surplus to be produced which, allotted to investment, will provide the condition for growth later on. Calculation based on a zero reference price for wages rules out this choice, since it amounts to ignoring the fact that, in reality, wages are distributed that, being allotted to consumption, reduce the nation’s capacity to obtain a surplus to be,allotted to investment. The rule will thus be that the heaviest technique will remain preferable, because the improvement in the productivity of labor that goes with it provides a surplus that, when invested, makes possible growth at the rate desired by the community. Competition drives entrepreneurs to choose the technique that maximizes surplus. This is doubtless why, in economic life, in the present-day world of business, the choices made are not very different in the underdeveloped countries from what they.are in the industrialized ones. Often, when different choices are made, this happens more for reasons connected with the size of the market than for reasons connected with the level of wages. In any case, these choices are almost always — and for­tunately — remote from those that would be dictated by a calculation based on a zero reference price for wages. This reveals that the problem of the choice of techniques is a pseudo-problem, as often happens with marginalism. The real problem is that of the choice ofbranches.

The surplus can be allotted in its entirety to investment, or it can be consumed, wholly or in part. If one regards the growth of wages as the ultimate objective of development, one will endeavor to ensure a parallel growth of the surplus and of wages. Given that the surplus available for investment will grow more slowly in proportion as the rate of wages is authorized to grow faster, and that growth of employment depends on growth of the surplus reinvested, it will be possible to define a social optimum-function that will enable that combination of rates of growth of the surplus and of wages to be chosen that will maximize the wage bill not at the end of a period but during the whole period, of ten or fifteen years, for example.

These arguments are not of merely theoretical interest. The countries that began to industrialize later than others have indeed experienced rapid rates of growth both of productivity and of employment whenever they have granted priority in their develop­ment to the most up-to-date industries, using the most advanced techniques. As a general rule there are no grounds, in an under­developed economy, for making choices different from those that would be made in a country that was already well industrialized: it is necessary to choose the most efficient technique, the one that maximizes surplus, at the rate of rewards of the factors that actually prevails. In fact, accelerated accumulation in the modern sector will be accompanied by an improvement in wages, whereas in the traditional sector, where productivity is relatively stagnant, rewards will increase more slowly, if they increase at all. There is therefore no reason to be surprised if average incomes in the two sectors are very much out of line with each other, and if this unevenness becomes still more marked as development progresses.

Although the spontaneous movement proceeds along this path of increasing differences in the rewards of labor, it is to be considered that, during the long transitional period, a genuine policy of development will not be able to tolerate this increasing inequality, for this breaks up national unity, the very precondition of development. The state must, then, plan prices and wages so as to ensure national cohesion. In order to do this, the local price system will have to be isolated from the world system. But it must be appreciated, at the same time, that planning — the choice of the sectors to be developed — can then not be based upon the system of prices chosen, the rationality of which (the political necessity for solidarity between the workers in sectors with differing productivities) lies elsewhere. It will be necessary to have a system of reference prices, for economic calculation purposes, such that the choices made lead to the development of modern branches. In proportion as the traditional sector is encroached upon and reduced, the system of prices that is rational from the standpoint of political cohesion will approach closer to the system that is rational from the standpoint of economic choices.

We still have to answer the third question, regarding the way in which unequal international specialization takes shape. In a closed economy, a certain level of national income, accompanied by a certain distribution of this income, entails a particular orientation of demand and in consequence requires a particular orientation of production in conformity with this demand. The first industries established in Europe depended on techniques that were relatively light, because these were more profitable. But the development of an industry (e.g., textiles) necessitated increased production in other branches (e.g., the making of machines). The most profitable technique in these branches might be the heaviest. Marx examined this problem when studying the mechanism of the equalization of profit. Equilibrium is obtained when the orientation of production conforms to social demand, on the one hand, and on the other ensures equal reward of all capitals. The tendency for capital to prefer to go into light industry is thus.limited by the necessary development of complementary industries.

It should be noted that this definition is quite different from that which identifies light industry with the making of consumer

The Development of Underdevelopment 233 goods and heavy industry with the production of capital goods. Production of coal, for example, uses more labor per unit of capital than production of plastic objects, or of beer. Nevertheless, there is a link between the two phenomena: if, in any sector or industry, a more modern technique is put into effect, then the national production becomes “heavier” on the average. But then the production of capital goods has increased more than that of consumer goods. The increasing heaviness of techniques runs parallel with the shifting of the productive forces from ultimate production toward intermediate production. Under conditions of international integration, however, when capitalism is developing in a framework dominated by external exchange, the comple­mentary goods may be imported.

It is the search for profit, and that alone, that leads central capital to establish light rather than heavy industries in the periph­ery. With the same productivity, wages are lower in the periphery than at the center. In a given branch of production, using the same techniques, the increase in profit resulting from emigration of capital from the center to the periphery will be the greater in proportion to the “lightness" of this branch. It is this force that accounts for unequal specialization.

Unequal International Specialization, Domination by Foreign Capital, and Transference of the Multiplier Mechanisms,

Disarticulation

In distinguishing between the primary effects produced by an initial independent change in an economic magnitude from the successive waves of secondary and tertiary effects that it induces, theory has brought out the cumulative character of the majority of economic processes. Between the values of the different magnitudes considered in the initial situation, on the one hand, and, on the other, at the end of the infinite series of decreasing waves induced by the initial change, simple mathematical procedures enable us to reveal the multipliers that epitomize the volume of the change effected, and so the inducing power of the initial change.

We have seen the role played by two of these multiplier mech-

anisms in the origin of the economic cycle: the multiplier that measures the relation between the inducing investment and the growth of income induced, and the accelerator that measures the relation between the growth of the inducing income and the induced investment.

Keynes’s analysis of the multiplier assumes the setting of an advanced capitalist economy paralyzed by the inadequacy of demand, and possessing an installed production capacity that has to be set to work. In this setting Keynes assumes that supply responds at once to the solicitations of demand — that is, that production can be increased without any fresh investment. Any initial independent demand (not only an independent investment, blit also the creation ex nihilo of an initial demand by the state, or a surplus in the trade balance, etc.) makes it possible to reanimate the productive system.

In cases in which expansion of production in order to meet the pressure of demand requires an investment, we leave the strict framework of Keynes’s analysis. If, in order to bring about this expansion, all the income saved during the first period has to be invested, we are back in the “classical” case — in other words, the multiplier no longer makes sense,'its “value” becoming infinite. But if only part of this saving has to be invested in order to bring about this expansion, the multiplier recovers a finite value, Keynes’s “propensity to save” being replaced by a propensity to hoard, or the propensity to consume by a propensity to consume and to invest. If we accept that wages are destined to feed the demand for consumer goods, and profits to supply saving (with a view to investment), we shall have an inherent disequilibrium in the very mechanism of accumulation when the ratio of profits to wages increases faster than the ratio between the volume of equipment needed in order to ensure a given increase in consumption and

create its own supply, that only part of the saving be invested. In this case, investment of the entire saving is not profitable. The bridge between the theory of hoarding and analysis of the require­ments of production is established without the need to bring in this psychological factor, the marginal efficiency of capital, which constitutes the weakest point of the Keynesian theory, and without bringing in the rate of interest and of “liquidity preference,” which compel Keynes to accept the quantity theory of money.

Hoarding in the underdeveloped economies is a totally different phenomenon from this forced hoarding, which reflects the capitalist mode of production’s inherent contradiction between capacity to produce and capacity to consume. In the precapitalist economies, such as ground rent. Hoarding, which used to take the form of an accumulation of “real values” (gold and land), now takes the form of accumulating the local currency. The hoarding of precious metals must be identified with luxury consumption, since gold has to be produced, or paid for with real exports. If the hoarders buy land, the sums that are spent in this way pass into the hands of other individuals: demand is shifted but not sterilized. Nevertheless, this attraction to land does increase the inequality of distribution of wealth and income. If, finally, hoarding takes the form of an accumulation of currency, the quantity of money adjusts itself automatically to economic need, so that this hoarding is sterilized as regards its effect on the level of economic activity while retaining its function for the hoarder himself, namely, accumulation of potential purchasing power, reinforcement of his social effectiveness.

Since expansion of production under the conditions of the underdeveloped world obviously necessitates investment, and the specific forms of precapitalist hoarding do not constitute a leak or drain, in the Keynesian sense, the multiplier effects of an indepen­dent investment ought to be maximal. But they are not, and this for two reasons.

The chief reason is that the profits on invested capital are to a

large extent exported, since this capital is foreign-owned. Now, it is profit that constitutes the income destined essentially to finance induced investments. Export of profits thus transfers to the center the driving power of the primary investment.

The second reason for this poor capacity to generate a cumula­tive process reflects the specific contradiction of peripheral capitalism. If, in the periphery, wages are low but the techniques employed are advanced (similar to those employed in the advanced countries), then overall equilibrium between society’s capacity to produce and capacity to consume will not be achieved: profits, which will be high in these countries, will not be reinvested, for lack of outlets.

Turning now to the accelerator, let us recall that its effect is to increase induced investment — the increase in demand for capital goods being more than proportional to that for consumer goods — because modern production techniques require the installation of durable plant that takes many years to depreciate. We have seen that, at the beginning of the cycle, the accelerator helps to slow down the effects of the multiplier, but also to intensify the fluctuations in overall demand.

Once the accelerator has been transferred to the place where the equipment goods are produced, then if, as is the case, the unequal international division of labor assigns this branch of production to the center of the system, it will be at the center that the reinforcement effects that speed up the pace of accumulation will be felt.

This transference of the cumulative effects of investment results in making the underdeveloped economy that disjointed type of economy that has been depicted in inter-industrial tables for the last twenty years. Here, too, structural comparison between advanced and underdeveloped economies has meaning only if these tables, when compared, are compiled at the same levels of aggregation. A qualitative difference in structure is then perceived, which can be summarized by saying that the inter-industrial tables of the under­developed countries are “empty,” or that the “technical coefficients” are negligible. For a level of aggregation that retains fifteen sectors, the total of inputs (those of diagonal being excluded) represents over twice the value added in the advanced economies of the West, and less than half that for the average underdeveloped countries (those where product per head is between $100 and $200). This means, if imports (or exports) represent in both cases about 20 percent of the gross domestic product, that at this level of aggregation, external exchanges make up, in the advanced countries, about 6 percent of all exchanges, internal and external, as against 12 percent in the underdeveloped countries. If we exclude ultimate exchanges, both internal and external — that is, the spending of income on ultimate goods (both for consumption and for investment), local and foreign alike — and if we accept that ultimate goods represent about half of imports, then external intermediate exchanges amount to 5 percent of the total intermediate exchanges (internal and external) of the advanced countries, as against 16 percent in the case of the underdeveloped ones. The higher the level of disaggregation, the bigger the range becomes. At the level of sixty branches, the range is between 3 percent and 15 percent. Although moderate at the overall level, the percentages are, of course, much higher for the main branches of processing industry (here the range is between 10 percent and 60 percent), and are probably even higher still in the case of certain especially important firms.

This means that the advanced economy is an integrated whole, a feature of which is a very dense flow of internal exchanges, the flow of external exchanges of the atoms that make up this whole being, by and large, marginal as compared with that of internal exchanges. In contrast to this, the underdeveloped economy is made up of atoms that are relatively juxtaposed and not integrated, the density of the flow of external exchanges of these atoms being much greater, and that of the flow of internal exchanges very much less.

The consequences that follow from this disarticulation are crucial. In a structured autocentric economy, any progress that begins at any point is spread throughout the entire organism by many convergent mechanisms. Contemporary analysis has stressed the “leading” effects of an increase in primary demand. Formerly, analysis emphasized other channels of diffusion: the reduction in prices resulting from progress, and so, along with this, the change in the structure of relative prices, of demand, and of real income, the possible increase in profits and change in the distribution of investments. If the economy is extraverted, all these effects are limited, being largely transferred abroad. Any progress realized in the oil industry will, for instance, be without the slightest effect on the economy of Kuwait, since nomad stockbreeding sells nothing to and buys nothing from the oil sector. This progress will be diffused in the West, in all the countries that consume oil.

In this sense, one ought not to speak Ofunderdeveloped national economies, but to reserve the adjective “national” to the auto­centric advanced economies, each of which alone constitutes a true, structured economic space, within which progress is diffused from industries that can be regarded as poles of development.

₁ an integrated internal market. Depending on its geographical size and the variety of its exports, the underdeveloped economy may appear as being made up of several “atoms” of this type, all independent of each other (as with Brazil, or India) or of a single “atom” (e.g., Senegal, which is entirely organized around the groundnut economy).

The consequence is that the false, nonstructured economic spaces of the underdeveloped world can be broken up and divided into microspaces without serious danger, something that cannot be done without intolerable retrogression in the case of the integrated spaces of the advanced countries. The weakness of national cohesion in the Third World is often a reflection of this fact, which is also the source of micronationalism: the area interested in the export economy has no need for the rest of the country, which rather constitutes a burden upon it.

The effects of this disarticulation are plainly to be seen in the historical geography of the Third World. The areas interested in an export product that is comparatively important for the develop­ment of capitalism at the center experience brilliant periods of very rapid growth. But because no autocentric integrated entity is tormed, as soon as the product in question ceases to be of interest to the center, the region falls into decline: its economy stagnates, and even retrogresses. Thus, Northeastern Brazil was, in the

The Development of Underdevelopment 239 seventeenth century, the scene of an “economic miracle’’ that led nowhere— the moment that the sugar-growing economy lost its importance, the region fell into a state of lethargy, to become later on the famine area that it is today. Even in little Senegal, the river region was a prosperous one in the days of the gum trade. WhCn gum was replaced by synthetic products, the region became an exporter of cheap labor, this being the only livelihood available to its population. When the iron ore of Lorraine is eventually worked out, this may create a difficult reconversion problem for the region, but it will be able to overcome these difficulties, for an infrastructure of integrated industries has been formed on the basis of the mineral, which could be imported from elsewhere. But when the iron ore of Mauritania is worked out, that country will go back to the desert.