A conceptual framework for analysis of primary commodity markets

WorMDevelopment, Vol. 12, No. 1, pp. 25-41, 1984. Prinfed in Great Britain.

0305-750X/84 $03.00 + 0.00 © 1984 Pergamon Press Ltd.

A Conceptual Framework for Analysis of Primary Commodity Markets ALFRED MAIZELS*

Fellow, Institute o f Development Studies at the University of Sussex Summary. - The paper begins with a critique of the traditional neoclassical theory as a tool for analysing commodity markets, and suggests that a viable theory needs to be developed from the concept of 'market power', as expressed in the relative bargaining strengths of transnational corporations active in the commodity production and trade of developing countries, and of the governments and business interests of host countries. A discussion of the principal elements comprising relative bargaining strength, divided into commodity-specific, country-specific and international, is followed by a brief consideration of alternative policy options for developing countries aimed at improving their share of the benefits from their commodity exports to developed countries. countries in this respect, but rather to emphasize the relevance and importance of having a realistic theory of economic development which pays adequate attention to the conditions under which the primary sector could make a significant contribution to industrialization and, in particular, the role that primary exports could play in this process. It is against this background that doubts arise as to the usefulness of the traditional neoclassical approach to the analysis of commodity markets. The usefulness of any theory in this context must be judged in terms of the extent to which it provides a good explanation of real world phenomena and, as a consequence, the extent to which its prescriptions provide a reliable guide to development policy. The principal theme of the present paper is that the traditional theory is not credible on either count, at least as regards the markets for primary commodities, and that we must build a more realistic analytical framework based on the concept of relative bargaining power. Section 2 summarizes some of the main difficulties arising from the neoclassical theory, while Sections 3 and 4 develop an alternative approach. Finally, Section 5 considers some policy implications.

1. INTRODUCTION As is well known, the majority of developing countries are still heavily dependent on primary commodities for their export earnings, and for the external financing of their development programmes. With relatively few exceptions, the domestic economies of these countries also retain much of the lopsided structural character that is a hangover from the period of colonial rule, with agriculture or mining occupying the greater part of the active labour force. In the colonial period, the task of the .dominant primary sector was to supply cheap food and raw materials for the metropolitan power, as well as to provide an assured market for a reverse flow of manufactured goods. This two-way process also engendered a flow of profits from trade which deprived the colony of much of the potential for domestic saving and investment. In the post-colonial period, the primary sector can assume a new role in economic development - either by providing an increasing volume of food and raw materials to support domestic industriahzation, or by providing, by export, the foreign exchange required to finance such industrialization (or, of course, by some combination of both). The key role that the primary sector can play in the industrialization process has often been under-estimated or even ignored. It is not the purpose of the present paper, however, to analyse the policies pursued by different

* The author is greatly indebted to David Evans and Gerald K. HeUeiner for their critical comments on an earlier draft which had been presented to the 1982 Annual Conference of the Development Studies Association. Any remaining errors and omissions are the sole responsibility of the author. 25

26

WORLD DEVELOPMENT

2. INADEQUACIES OF THE TRADITIONAL NEOCLASSICAL APPROACH Standard neoclassical theory proceeds from a set of assumptions which define a 'perfect' idealized market in which buyers and sellers trade in order to maximize their subjective utilities, while each factor is remunerated according to its marginal productivity. The main assumptions made are the existence of perfect competition in both goods and factor markets, perfect foresight, perfect factor mobility and price flexibility, full employment, the absence of 'externalities', and an unchanged original income-distribution. On these assumptions, neoclassical theory demonstrates that the 'free play of market forces' will result in an equilibrium situation characterized by both an optimum allocation of resources and an optimum level of welfare or real income. Moreover, exogenous shocks will cause adjustments which will move the system towards a new equilibrium. Market 'imperfections', such as oligopoly or 'externalities', are then analysed as aberrations from the ideal world of perfect competition. Given its assumptions, the neoclassical theory is a self-contained logical system. However, as has frequently been pointed out, those assumptions are so far removed from reality that the theory has little relation to actual economic change. Moreover, as soon as some of the basic assumptions are modified so as to reflect realworld phenomena, the theory is faced with logical inconsistencies. As one eminent neoclassical theorist has recently explained, neoclassical general equilibrium theory faces logical limitations if there are increasing returns which are large relative to the size of the economy (since this will lead to the emergence of large firms with m o n o p o l y power); if there is market power for individuals to exploit (since this will lead to individuals influencing equilibrium prices); if market information differs as between buyers and sellers (in which case certain contingent m.arkets cannot logically exist); if there are public goods (for which the fundamental welfare theorems cannot hold); and if there are externalities, i.e if one individuars actions affect the welfare of another.1 Thus, neoclassical theory faces a dilemma that as a logical system it is unrealistic, while as it moves towards reality it becomes subject to logical limitations and inconsistencies. Some of these limitations arise in acute form where, as in many of the international c o m m o d i t y markets, large transnational enterprises use their oligopoly or oligopsony positions not only

to increase their profits but to extend their market power on a global scale. What Perroux (1950) has called the 'domination e f f e c t ' where a dominant economy exerts its influence on another economy, either through market operations or by more direct means, to serve its own purposes - is only a particular example of a rejection of the neoclassical notion that economic activity consists of buyers and sellers of equal power exchanging goods and services through a market process. The basic assumptions of the neoclassical approach have tended to result in analyses of primary c o m m o d i t y markets being focussed on price and income fluctuations of buyers and sellers in marketclearing situations with ancillary consideration of the effects of market 'imperfections'. 2 In a 'free market' system, price performs a dual function, a First, changes in relative prices indicate changes in relative scarcities in goods and factor markets, thus providing signals and incentives for new investment and for appropriate changes in relative outputs of different goods and services. Second, changes in prices necessarily involve changes in the distribution of the benefits of trade between buyers and sellers. The neoclassical approach raises some major problems for the analysis of each of these two functions of price, essentially because actual (or 'real world') c o m m o d i t y markets do not conform to the neoclassical paradigm. The central difference between actual and neoclassical international markets for primary commodities is that a wide range of actual markets have tended to become dominated by oligopolistic and/or oligopsonistic elements, reflecting the dramatic postwar expansion of the market power of transnational corporations (TNCs), as well as - for a number of important commodities - the intervention of governments in market operations. While neoclassical price theory can explain price changes when competition prevails on the side either of buyers or of sellers (as well as, of course, when both sides of the market are competitive), it becomes ineffective when a group of oligopsonistic buyers confront a small number of oligopolistic sellers. The o l i g o p s o n y - o l i g o p o l y situation is fairly typical in a number of commodities (particularly so in minerals exported by developing countries), for which prices are determined, to a greater or lesser extent, by the relative bargaining power of the two sides. In such situations, the theory of bargaining related to two-person zero-sum games can provide useful insights. But, as Labys (1980) argues, bargaining theory tends to be ineffective

ANALYSIS OF PRIMARY COMMODITYMARKETS where there is mutual dependence between buyer and seller, where more than two decisionmaking entities operate on each side of the market, or where there is a non-zero sum involved. Nonetheless, a credible explanation of price in its division of benefit function would seem to require the incorporation of some form of bargaining theory. A second difficulty with the neoclassical approach in relation to commodity markets arises because, by its treatment of 'externalities' as market 'imperfections', it implies that governments can eliminate such imperfections by appropriate policy measures. On closer analysis, many of these imperfections are seen to be integral parts of the economic system, reflecting essentially the underlying power structure and its institutional manifestations. A related issue is the frequent assumption by many neoclassical analysts that the absence of government intervention in a commodity market can be equated with perfect, or nearperfect, competition. 4 In fact, actual markets perform inefficiently in many important respects, not least in their allocation of resources. Particularly where market forces reflect an oligopolistic market structure, decisions on investment, output and prices can diverge substantially from those which would obtain in a perfectly competitive market, thus distorting both the allocation of resources and the division of benefit. Equally, when the 'free play of market forces' gives rise to large price fluctuations, the price mechanism fails to provide a sure guide to new investment and to rational resource allocation. Such commodity markets move from one disequilibrium situation to another, in a continuous series of shortages and gluts, rendering the neoclassical thesis that market forces will automatically re-establish an equilibrium position unrealistic and unhelpful. A third difficulty is that even where there is effective competition among many sellers and among many buyers, it is not possible to ensure that all decisions are rational in the sense that they correctly take future developments into account. Uncertainty about future prices is a major perennial problem of actual commodity markets and this, together with supply timelags in adjustment to price changes, inevitably leads to a greater or lesser degree of misallocation of new investment. While futures and contingency markets have a useful role to play in shifting the risk of future loss from traders to speculators, s the existence of such markets will reduce the probability of resource misallocation only to the extent that they correctly

27

anticipate future market conditions. Such markets exist only for a limited, though important, range of commodities, while their efficiency in anticipating future price changes varies considerably.6 The degree of price uncertainty may also be reduced as a result of intergovernmental intervention, for example by means of a commodity agreement which effectively maintains prices within an agreed range. To the extent that uncertainty is thereby reduced, such market intervention may in fact result in less misallocation of resources than would otherwise occur. Moreover, because of its ahistorical approach, and its focus on the idealized market in which both buyer and seller gain, neoclassical theory can throw no light on the nature and significance of the dependency of the commodity sector of the economies of developing countries on the economic interests of the main developed countries. Whether or not one accepts all the conclusions of the dependencia school, it seems clear that the principal manifestations of underdevelopment - including excessive dependence on one or a few primary commodities - cannot be divorced from the manner in which the economies of developing countries have historically been integrated into the world economic system.

3. THE INSTITUTIONAL FRAMEWORK OF INTERNATIONAL COMMODITY TRADE The development of a satisfactory theoretical framework for the analysis of international commodity markets would need to take account of the limitations of the neoclassical approach discussed above. More specifically, such a theoretical framework must take into account the influence of market structures and of government intervention, i.e. the institutional context within which market forces operate. Thus, theory must take large enterprises, typically TNCs, acting as oligopsonists in their purchases of commodities from developing countries or, as vertically integrated entities, controlling both the production and trade of developing countries in particular commodities, as the norm, rather than the neoclassical firm which is not large enough to influence market prices by its own actions. Likewise, cognizance must be given to the development of countervailing market power by the governments and enterprises of the developing countries, as well as to the role of governments or parastatal agencies, not only as traders themselves, but also as powerful influences in determining the

28

WORLD DEVELOPMENT

general terms and conditions of international trade. A corollary of a decision to take TNCs as the typical firm is that it can no longer be assumed that an individual country is the sole appropriate unit for international trade theory. Such an assumption is no longer valid with the penetration of TNCs in the commodity production and trade of developing countries, with output and processing becoming dominated by vertically-integrated corporate structures, and with transfer pricing being used as one of a package of instruments designed to maximize the global profits of a TNC. In this context, the concept of 'export value', usually measured f.o.b., itself needs to be modified to take account of remittances abroad in the form of royalties or management fees to a parent TNC, since these are, in effect, leakages from domestic incomes. 7

(a) The main trading channels The main combinations of the three principal 'actors' in international commodity markets (i.e., the TNCs, other enterprises and governments) are summarized in Table 1. The main channels in which commodities are traded internationally are (i) sales between independent entities, for either immediate or forward delivery at a stated future date (the so-called 'arm's length' trade); (i_i) sales under long-term contracts (which contain clauses allowing for prices to be adjusted in accordance with changes in market conditions);

and (iii) shipments between different parts of the same enterprise (intra-TNC trade), or between affiliated enterprises (related-party trade). For many commodities, the bulk of world trade is handled by a small number of multi-commodity trading corporations, which are interposed between the producers and the consumers in importing countries. These powerful trading conglomerate transnationals are generally in a position to influence market prices by the volume of their operations on the spot and futures markets. In some commodities, governments or quasi-government agencies participate directly in international trading (though in a number of cases TNCs manage nationalized productive enterprises, or market their output under contract), while governments are also generally involved in regulating and monitoring the terms and conditions of private long-term contracts between domestic firms and TNCs. The relative importance of the major trading forms varies among the different commodity markets, and there has also been substantial change over the past few decades. Reliable estimates are, however, available for relatively few commodities. A study by Bosson and Varon (1977) found that in the late 1960s, roughly 40% of world trade in iron ore was intra-firm, another 40% was covered by longterm contracts (of up to 20 years), and the remaining 20% was sold on an 'arm's length' basis. The market for bauxite/alumina/aluminium is even more dominated by large TNCs, which are vertically integrated from bauxite mining in developing countries to smelting and

Table 1. The main channels of trade in commodities exported by developing countries Developed country actors Developing country actors

TNCs (productive and trading)

Other enterprises

Governments (incl. state enterprises) (i) Barter deals (ii) Long-term supply contracts (iii) Arm's length trade

TNCs subsidiaries and affiliates

Intra-firm and related party trade

Arm's length trade

Other enterprises

(i) Long-term supply contracts (ii) Arm's length trade

Arm's length trade

Governments (incl. state enterprises)

(i) Long-term supply contracts (ii) Arm's length trade (iii) Sales under management or marketing contracts [Terms and conditions of TNC operations]

[Agreements or arrangements affecting trade flows, e.g. international commodity agreements, market access arrangements]

ANALYSIS OF PRIMARY COMMODITYMARKETS fabrication of aluminium products, the six largest of these TNCs controlling over one-half of world capacity in the industry outside the socialist countries (Labys, 1980). TNC dominance is also found in a number of agricultural products. In sisal, for example, while free market sales account for about two-thirds of world trade, over half the total (or 70% of the free market) is handled by four or five large trading corporations (UNCTAD, 1981). In bananas, almost 70% of total world trade is in the hands of only three transnationals (UNCTAD, 1974), in cocoa one transnational accounts for a quarter of world trade, while in diamonds - perhaps the extreme case - one transnational controls the marketing of 80% of the world's rough diamonds. Data for the United States reveal that for a range of important commodities cotton, rubber, bananas, bauxite - two-thirds or more of total imports from developing countries consist of 'related party' trade (Helleiner, 1980), while for some other commodities - tea, iron ore, manganese and certain tropical timbers - the 'related party' share was over 20%. On the other hand, there are many commodities - particularly those of agricultural origin - where market conditions, at least on the supply side, apparently aproximate to those assumed by neoclassical theory, i.e. there are a large number of sellers, no one of which can influence the price by his own actions. This appears on the surface especially true for commodities sold by an auction system (e.g. tea and tobacco), but more detailed studies have revealed that auction prices may be substantially influenced by the (possibly collusive) actions of a limited number of large-scale buyers (UNCTAD, 1975 and 1978). For commodities with terminal markets, prices are influenced and often manipulated - by transactions in futures contracts (often reflecting speculative activity) as well as by the supply/demand balance for the physical product (UNCTAD, 1980). All commodity markets are sensitive to expectations about future trends in supply and demand, thus allowing traders or speculators with 'inside' knowledge to reap disproportionate gains, a A recent new development has been various attempts, usually shortqived, by producer interests to influence current market prices by operations on the futures market. Such operations have taken place in coffee, rubber and tin futures contracts. An objective analysis of the efficiency of such operations, in terms of the net gain to producing countries, would be

29

valuable in assessing the necessary conditions under which 'spot' prices can be stabilized, or their trend changed, by trading in futures by producer interests.9 The channels and terms of international commodity trade are also subject to change or modification as a result of government actions or inter-governmental agreements. International commodity agreements, even if directed solely to price stabilization, may also in certain circumstances significantly influence the price trend, and thereby the division of benefit between exporters and importers. 1° But the major influence of government action up to the present, has arisen from measures which change the terms or conditions of market access for the commodity exports of developing countries. The sugar policy of the EEC is perhaps the outstanding example of this, preferential access for the associated ACP countries being combined in this case with severely restricted access for other developing countries, and greatly expanded domestic sugar production. The EEC has in recent years become the largest sugar exporting area, surplus sugar being sold on the world market at subsidized prices. As a result of the protection of domestic sugar production in the developed countries, the world market has become subject to wide price fluctuations and with a significantly lower average return than would have obtained in the absence of the market access restriction. (b) Market structure and price formation The relative importance of different trading channels is associated in a general manner with the market structure for a given traded commodity. The term 'market structure' in this context relates to the institutional organization of supply and demand, with particular reference to the resultant degree of competition or monopoly on each side of the market. Market structure in turn, is a central determinant of the process of price formation and of the division of benefit. The general relationships between different types of market structure and the price formation process is summarized in Table 2. In markets which are fully competitive on both sides (Type 1), none of the individual traders can influence the price by his own" actions, so that the price is that ~vhich equalizes total demand and total supply. Where there is oligopoly or oligopsony on one side of the market, and competitive conditions on the other (Type 2), the price will depend to a substantial extent on the degree of explicit

WORLD DEVELOPMENT

30

Table 2. Market structure and price formation Demand structure

Supply structure

Price formation

1

Competition

Competition*

Price determined by interplay of total supply and demand.

2(a) 2(b)

Oligopsony Competition

Competition* Oligopoly

3(a) 3 (b)

Monopsony Competition

Competition* Monopoly

Market power of oligopsony or oligopoly used (e.g. by 'price leadership' of dominant firm) to decrease effective price paid or increase that received. Monopolist can raise prices to increase revenue (though if price raised too much, new entrants may appear). Monopsonist can reduce prices below competitive level (though if reduced too much, producers may reduce or halt production).

4

Monopsony or oligopsony

Monopoly or oligopoly

5

Vertically-integrated TNCs

Type

Price reflects relative market power of buyers and sellers within limits set by underlying cost and demand conditions. Developing country export price set to maximize global profit and minimize risk.

Source: Adapted from Labys (1980), pp. 64-68. *Assuming no barriers to entry and free availability of price information.

or implicit collusion among the oligopolistic or oligopsonistic enterprises concerned. The question of collusion does not arise where a monopoly supplier confronts competitive buyers (Type 3b), the monopolist being able to raise prices, and so increase his revenue, so long as demand remains inelastic, though in practice there are often limits to the price increase arising from the need to avoid the encouragement of substitutes or the eventual entry of new suppliers in the mediumor longer-term. Similarly, there are likely to be corresponding limits to the extent to which a monopsonist can force down prices below the competitive level (Type 3a). Where there is a high degree of concentration (one or a few entities) on both sides of the market (Type 4), then price will reflect essentially the relative market strengths, or bargaining powers, of buyer and seller within limits set by the underlying cost and demand conditions. Inter-governmental agreements on supplies of particular commodities can be subsumed under this heading. Finally, there is the case of the buyer (in a developed country) and the seller (in a developing country) being the same firm or,

more typically, different firms within the same TNC (Type 5). The recorded export price then becomes an internal transfer price, which is open to manipulation in the interests of maximizing the TNC's global profits. The actual situation in particular cases can, however, be a complex one, with the recorded export price being only one among several instruments used to attain the TNC's profit objectives. If, for example, a TNC subsidiary in a developing country imports equipment, intermediate products or specialist services from its parent company, it can reduce its declared profits by over-valuing such imports, by under-valuing its commodity exports, or by some combination of the two. One major incentive for the manipulation of foreign trade prices in the intra-firm transactions of a TNC arises because of differences in tax rates (import duties, corporation and profit taxes, etc.) in the different countries in which it operates. For any given set of tax rates (and tax allowances) in those countries there is a corresponding pattern of declared profits that will minimize the TNC's global tax payments. 11 Once the desired level of declared profits has been determined for the operations of a TNC

ANALYSIS OF PRIMARY COMMODITY MARKETS subsidiary in a developing country, appropriate adjustments can be made in its transfer prices to arrive at this desired level. A different incentive for the manipulation of transfer prices would arise if a TNC wished to repatriate part, or all, of its capital investment in a subsidiary in a developing country while avoiding official restrictions on capital outflows. This could be done by under-valuing exports from the subsidiary, and over-valuing its imports, thereby accumulating credits abroad. To the extent that tax rates on TNC operations in developing countries tend to be higher than in 'home' (developed) countries, and TNC investment remains susceptible to possible nationalization or other uncertainties, there would be powerful incentives for the undervaluation of exports from developing countries, thus minimizing the relevant tax receipts of these countries. 12

(c) Market fragmentation Many - perhaps most - of the international commodities markets are complex packages of different trading channels and different market structures. Nonetheless, in many cases it is possible to identify a dominant pattern (of trading channel and market structure) which acts as the 'price leader' in the market. For commodities with competitive conditions, even in only a minor trading channel, the price so determined would tend to be used as a reference point for long-term contracts, or for TNC-government bargaining, in other trading channels. This is the case, for example, in the copper market (outside the United States) where the London Metal Exchange (LME) price is normally the current reference price, even though the LME is only a marginal market. But for many other important minerals, there is either no generally recognized price (e.g. iron ore), or virtually all trade moves in intra-firm channels (e.g. bauxite), and for these the bargaining process assumes major importance in price determination and/or in the division of benefit. For a number of agricultural products (e.g. cocoa, coffee, cotton, sugar), markets in futures contracts operate in conjunction with the markets in the commodities themselves. 'Spot' quotations in the latter are inevitably influenced by hedging and speculative activity in futures, as well as by the current supply/demand balance. Such markets are generally competitive, though often subject to 'corners' or 'squeezes'

31

by large trading corporations. For many other agricultural products (e.g. jute, sisal, vegetable oil seeds and oils), a substantial proportion of trade moves under one-year contracts, within an oligopsony-dominated market structure. Where markets for commodities exported by developing countries are dominated by TNC oligopsony structures, there tend to be two separate, though interrelated, sub-markets. There is, first, the market in the developing producing countries, where a few large TNCs often the large trading conglomerates - can exercise their market dominance by reducing the producer price to little more than production cost. Then, second, there is the final market in developed countries, where the same TNCs can often act as oligopolists. Bananas are probably exceptional in as much as the perishable nature of the commodity has resulted in a fiercely competitive final market in the developed countries - this has, in turn, led to intense pressure on the TNCs involved to reduce costs and producer prices in the developing countries. The diamond market is another example of two sub-markets, but in this case a single TNC operates on both, as a virtual monopsony in buying diamonds produced in developing (and other) countries, and as a virtual monopoly in selling these, as well as its own diamond output, on the world market. Market fragmentation also arises as a result of institutional barriers of one kind or another, the most important being national (or regional) protectionist measures. Where trade barriers are widespread, the 'world' market may, in fact, be simply the aggregation of a number of national markets. This seems to be essentially the case for meat and wheat, la the prices of which may bear little relation across national markets. Where, as for sugar, a substantial proportion of trade moves in preferential channels, while major consuming areas impose high barriers on imports (and the EEC subsidizes its exports), the 'world' market assumes a marginal character, as already mentioned, with extremely large price fluctuations, and with price levels appreciably lower than they would be in the absence of protection. An additional constraint on the prices of commodities exported by developing countries arises where such commodities are facing competition from synthetic materials. To the extent that natural and synthetic materials are substitutable in particular end-uses, the synthetic price often becomes the effective upper limit to the price of the natural product. Moreover, for some commodities, an additional constraint arises from the market structure itself, in so

32

WORLD DEVELOPMENT

far as the marketing and distribution channels are under the control of firms with an interest in promoting the sales of the synthetic substitutes. This is, for example, the case with sisal. 14

4. BARGAINING POWER, THE DIVISION OF BENEFIT AND COMMODITY PRICE MOVEMENTS With the increasing importance of TNCs in the world economy, and in the economies of developing countries, the role of 'market power', as expressed in the relative bargaining strengths of the principal actors, has become a central feature of the process of price determination, and of the division of benefit, in a wide range of commodities exported by developing countries. This is particularly the case for minerals, where TNC operations in developing countries are typically conducted under the terms of contracts with governments. In such cases, the 'price' of the mineral export is not simply the f.o.b, or transfer price, but is rather to be interpreted as including royalty and other tax payment.s on production or export. Thus, for the TNC, the relevant 'price', in the sense of net unit return, is the product price less the sum of the extraction cost and royalty/tax payments, whereas for the host country, it is the retained unit value, consisting of royalty/ tax receipts plus payments to domestic factors (Brodsky and Sampson, 1980). The relationship between the initial net return to the TNC and that to the host country can be substantially altered by subsequent renegotiation of the terms and conditions of TNC operations, or by unilateral action by the host government, for example, by increasing the royalty or tax on profits. Alternatively, the host government can buy part of the equity of a TNC subsidiary, or even majority, or full, control. However, where the international marketing of minerals is contracted out (there are several examples of the marketing of minerals produced by newly-nationalized mines being contracted out to the same TNCs that had previously owned the mines), there could be heavy losses for the nationalized enterprise if mineral prices are stagnant or declining, while TNC marketing fees are highJ s When one turns from price (or net return) to the wider concept of the 'gains from trade', there are a considerable variety of indirect effects of TNC operations to be taken into account, including 'backward-linkage' effects

on the domestic economy, 'income spread' effects, etc. (Hirschman, 1958 ; Baldwin, 1963 ; Beckford, 1972; Singer, 1950, 1975; Thoburn, 1977; and many others). Here again there may be scope for bargaining, since the host country government can try to ensure that a new TNC investment will have maximum linkage effects and minimum leakages to the foreign sector. This wider concept of the gains from trade thus leads directly to the relationship between trade and development - a theme which is far wider than the subject of the present paper. Given that bargaining power plays a major role in price formation and the division of benefit (in the narrower sense) in a range of primary commodity markets, it is useful to hypothesize the general form of the relationship involved. The curve in Figure 1 illustrates what seems to be a reasonable first hypothesis: assuming for the moment that the 'proportion of the benefits of trade retained by the developing country' and the 'relative bargaining power of the developing country and TNCs' can each be accurately assessed on a scale from 0 to 100. The intercept a on the vertical axis represents an extreme situation where domestic interests in a developing country have zero bargaining power, so that the terms and conditions of TNC operations in that country are determined solely by the TNC concerned. In this situation, local costs would represent little more than subsistence wage levels plus local materials used. Government taxes borne by the TNC are also likely to be minimal, government expenditures being devoted largely to the provision of the necessary infrastructural facilities. Diamond purchases in African countries by the De Beers 100%

>. o

o

~, .b

i/,~ ~ ~

~

~._~/

.~

~'Petroteum

Diemonds

o

O0 %

ReLative bargaining power of host country vis-a-vis TNCs

Figure 1. Hypothetical relationship between bargaining power and the division of the benefits of trade.

ANALYSIS OF PRIMARY COMMODITY MARKETS organization would seem to approximate to this situation, as would mining operations by TNCs in Namibia. 16 An analogous situation would arise in politically independent developing countries where transnational corporations arrange contract prices through corrupt intermediaries; in such cases, the payment of bribes is, in effect, in exchange for the virtual elimination of the bargaining position of the host countries. 17 At the other extreme, intercept b represents the minimum profit accruing to a TNC operating in a developing country (measured as a proportion of the benefits of trade) whiGh will allow it to continue its operations, is It is difficult to envisage a situation in which TNCs had zero bargaining power, but there could well be situations in which individual TNCs had very little bargaining power vis-d-vis a group of developing countries with closely coordinated production policies. Certainly, the OPEC countries still retain considerable bargaining strength vis-d-vis the petroleum TNCs, provided that they maintain a unified strategy. Action by OPEC has shifted the position of petroleum in Figure 1, from a position on the lower part of the curve before 1973 to a considerably more favourable position for the producing countries in terms of their share of the benefit, though the bargaining power of OPEC has been appreciably eroded since 1980 as a result of the economic recession, the growth of production by non-OPEC countries and efforts at energy conservation. Bauxite is another example of a commodity whose position on the curve in Figure 1 has been shifted quite substantially by action on the part of host country governments.

(a) The elements o f bargaining power As indicated in Table 2, the bargaining situation becomes an important element in the process of price formation whenever competitive conditions are not present on both sides of the market. Typically, this represents a situation of bilateral oligopoly, including bilateral contract arrangements between a TNC and a host country government. What, then, determines the relative bargaining strengths of the two sides? Clearly, there are many factors involved, their relative importance no doubt varying substantially from one commodity to another, as well as varying over time, and depending also on which countries and which TNCs are involved. However, it is useful for analytical purposes to group the

33

elements of bargaining power into three broad categories, viz. factors specific to the individual commodity concerned, those specific to the host country - including, in particular, its macro-economic situation and prospects - and those (if any) which related to international action, either by developing country governments (or other entities), or l b y developed country governments (or TNCs). In his analysis of the bargaining situation between host governments in mineral-exporting developing countries and foreign mineralinvesting TNCs, Labys (1980) distinguishes nine relevant factors, all of which relate to the characteristics of the particular commodity market structure in question. These nine factors, which are listed first in the right-hand panel of Figure 2, relate essentially to the bargain struck between host governments and TNCs on the terms and conditions of new investment in mining, and on the bases for revisions of existing mining contracts. 2° The degree of dependence of a country on the export of a particular commodity is relevant, since the higher the dependence, the more vulnerable is the government of that country in bargaining with a foreign enterprise. If a new mine involves a large initial capital investment, in relation to a country's financial resources available for investment, this by itself will tend to strengthen the bargaining position of the foreign investing enterprise. However, once a large investment is made by a TNC, the balance of bargaining power will shift, thus allowing the host country to extract more favourable terms. 21 When the technology involved is complex and changeable, it may be available only from one, or relatively few, TNCs which give the latter an i m p o r t a n t bargaining lever. On the other hand, it may be possible for a host government to play one TNC off against another in order to obtain more favourable terms. When the technology involved is simple, stable and diffused, however, the host country has a better chance of increasing its share of the benefit. Host governments can also attempt to extract a larger benefit by extending their control over mineral reserves and/or production by foreign enterprises, for example, by restricting the latter's exploration rights, by taking up shares in these enterprises, or by establishing State trading or marketing entities. Another avenue for deriving a larger benefit would be for host governments to promote local processing of minerals. In appropriate circumstances, provisions for such local processing can be evolved, either in bargaining with a mining TNC, or by

34

WORLD DEVELOPMENT

I Mocroeconomic position and

I Export dependence

prospects a Financial resources b BaLance of payments c ALternative trade Links 2

i c°'r'-I

r o-o-.t_

specific factors

specific

factors

2 Magnitude of fixed investment 3 Nature of technology 4. Control over reserves and production

Degreeof corruption

5 Opportunities for increased processing 6 Material share in product

prices 7 ObsoLescing bargain B Nature of competition 9 Government Learning process

I0 Transparency of world market I Joint action by developing countries 2 Joint action by TNC%or by

developed countries

InternationaL action

II Control of marketing and distribution 12 ALternative TNC options

Figure 2. The elements of bargaining power between TNCs and governments and enterprises of developing countries. introducing fiscal incentives for such development. Where the commodity in question is a relatively unimportant cost item in the final product price, it is usually possible for an increase in the commodity price to be passed on to the final user. In such cases, it is easier for host governments to extract a higher 'price' in bargaining with a TNC on the terms and conditions of a new project. The concept of the 'obsolescing bargain' (Vernon, 1971) relates to the changing balance of bargaining power over time. The presence of risk, and the need for a large initial investment (in mining exploration and infrastructure), gives a potential foreign investor a predominant bargaining power. But once the investment is made, the TNC will be concerned to come to an accommodation with the host Government should the latter seek to renegotiate the original terms and conditions of the agreement - in order to protect its investment. The bargaining strength of a host Government would also be increased if new supplies appear in the market, thus reducing the oligopolistic or monopolistic power that a TNC may have. The degree of competition in the final product market is also relevant, since a TNC which has a strong market position for its final product can more readily pass on cost increases to end-users. Finally - in the list of factors -

discussed by Labys - the learning process of host governments can also be an important element in shifting the balance of bargaining power away from TNCs. The 'learning process' relates to improved understanding of the coststructure and operating techniques of the industry, as well as increased negotiating skills, and this in turn provides an opportunity for increased host participation in productive operations, in control over marketing and in the regulation of profit remittances. In addition to these nine factors elaborated by Labys in relation to mining developments by TNCs in developing countries, there would seem to be some further elements that can usefully be identified, particularly perhaps in relation to contractual arrangements between TNCs and domestic enterprises (as distinct from governments) in developing countries. In Figure 2, three such additional elements are listed. First, there is the degree of 'transparency' in the operation of the world market for a particular commodity, especially as regards the differential availability of information on world market prices and prospects for a given commodity. Very often, domestic producers in a developing country do not know what the final prices to consumers in developed countries are, let alone being equipped to estimate the probable future trend of

ANALYSIS OF PRIMARY COMMODITY MARKETS prices. In such situations an oligopsonistic TNC has considerable room for manoeuvre in setting a buying price favourable to itself. Asymmetry also often exists in the availability of information to a host government and to an interested foreign corporation in regard to the dimensions and quality of local resources, reflecting the technical, research and marketing expertise of the foreign interests concerned. Equally, when a developing country wishes to invest in commodity-processing plants, there is often a lack of knowledge about the relevant market structures - which often limit access of processed commodities to major markets as well as about the costs and benefit of alternative technological options. Some developing countries have, however, created some countervailing power by creating Technology Information Centres which provide relevant information to domestic enterprises. Second, the nature of competition in the international marketing and distribution system for commodity exports of developing countries can also influence relative bargaining strengths. Where international marketing channels are virtually in the control of a few TNCs, the domestic enterprise in a developing country has little alternative but to sell its output through the established channels. In some cases, however, the established channels can be by-passed, at least to some extent, for example by selling direct to retail stores or other end-users in developed countries, thus reducing the bargaining strength of TNC marketing and distribution entities. Third, TNC bargaining strength in a particular developing country will depend inter alia on the TNC's alternative options for concluding contracts in other countries. These options may be very limited in some cases, for example, where the particular country possesses a mineral deposit of higher-grade ore than can be found elsewhere, or where alternative locations involve more political risks. But where viable alternatives exist, and can be demonstrated, the threat of diverting investment can give the TNC a predominating bargaining strength. Where governments of host countries are active pdrticipants in the bargaining process, the various commodity-specific factors discussed above are by no means the only ones to be considered. The general economic position and prospects of the host country will also be relevant. 22 Three elements of the general economic position seem particularly relevant (see top left-hand panel in Figure 2). These are the financial resources available to the host country, the current and prospective balance

35

of payments positions, and the possibility of establishing alternative trade links. Countries which have little or no monetary reserves, and which find it difficult to raise loans on world capital markets, may thereby find themselves in a generally weak bargaining position, which could well offset any bargaining advantage they might have in a particular commodity setting. In such a situation, a host country government would find it extremely difficult to refuse a TNC offer which would increase the country's net export income, even though the major share of the benefit would accrue to the TNC. On the other hand, those (relatively few) developing countries with ample financial resources would not suffer from this disability in their negotiations with TNCs (which normally also have large financial resources). Equally, countries which are in heavy deficit on their current payments account, due for example to a serious deterioration in the terms of trade and/or to a significant increase in the proportion of their export earnings which has to be devoted to servicing their external debt, may also find their general bargaining power seriously undermined, particularly if the prospects are for continuing external deficits. Countries which have adjusted relatively successfully to the external 'shocks' of recent years, and have continued their economic expansion without incurring major current account deficits, can adopt a generally more active bargaining stance in negotiations with TNCs. Another macro-economic element to be considered is the possibility of a developing country establishing new trading links through, for example, the conclusion of bilateral trade agreements, or adherence to a regional economic grouping, thus opening the way for bypassing traditional dependence on one or a few industrialized countries. Trade agreements between a number of Asian and African developing countries with socialist countries, for example, are likely to have improved the bargaining position of the developing countries concerned in negotiations with TNCs. As already indicated, the bargaining power of a government in its dealings with a TNC can be dramatically weakened, or even reduced to zero, by means of bribery and corruption. Many TNCs regard bribes paid as a necessary marketing cost and budget accordingly. Well-known cases in recent years include bribes paid in both developed and developing countries. Where bribes have become an essential part of the system, recorded prices of international transactions can be seriously at variance with actual or effective prices, while the division of

36

WORLD DEVELOPMENT

benefit can be significantly shifted in favour of the bribe-giver (i.e. the TNC). The third category of factors distinguished earlier as influencing the bargaining situation, viz. international action, relates to joint measures by governments of developing countries, or joint action by interested TNCs or by their home country governments. Intergovernmental action by governments of developing countries would include measures to regulate supplies coming on the world market by means of a cartel-type arrangement; common fiscal policies, such as agreements among host governments to impose a uniform tax or royalty on TNC production or export of a given commodity; the pooling of import purchases (e.g. of fertilizers or tractors) from TNCs for commodity production; or the establishment of regional technology transfer centres to advise small developing countries in their negotiations with TNCs. Joint action by TNCs to influence a bargaining situation is likely to arise only in extreme cases where competing TNCs all feel themselves threatened, for example, by nationalization without their claims for 'fair and adequate' compensation being met. In some cases, their home country governments may also intervene to put political and/or economic pressure on host governments not to nationalize the TNC interests in question, or to pay compensation in full or, more generally, to modify existing or proposed policies so as to favour their homeTNCs or to protect their interests.

on the world market. An analysis which focusses solely on shifts in supply and demand, and thus on changes in market prices, will not, however, reveal the underlying relationships between the TNCs and producers in developing countries. For that, it is necessary to place the supply/ demand analysis in the context of the structures of control and decision-making which govern the production, trade and marketing of a given commodity, and to show how these structures influence the price outcome, and the division of benefit between developed and developing countries.

5. POLICY IMPLICATIONS An analysis of the structure of control and of decision-making in different commodity markets, and of the process of price formation and the consequent division of benefit, could form the basis for a consideration of alternative policy options for developing countries. These options are best discussed in the context of a choice of longer-term strategy alternatives. The two major strategies that could be adopted by developing countries are: (i) to improve their share of the benefit of the present integration of their commodity sector with the economies of the developed countries; or (ii) to 'delink' from the latter, to a greater or lesser extent, by redeploying their commodity production increasingly towards their own needs and those of other developing countries a strategy of 'Collective Self-Reliance' (Fortin, 1980). -

(b) Commodity price movements The balance of bargaining power is thus a complex phenomenon, the elements of which are likely to change in relative influence over time. In commodity markets substantially dominated by oligopoly/oligopsony, the division of benefit will be heavily influenced by the relative bargaining strengths of the main actors. The mechanism by which the benefit is distributed is the negotiated 'price' which, in many situations, must be interpreted in terms of net unit return. Moreover, such a 'price' is essentially a longrun concept. It represents the intended division of benefit over the period of a contract, or over a period long enough to amortize the capital investment of a TNC in a specific project in a developing country. In practice, market prices - and thus the TNC's net unit return - will fluctuate in the short-term as,a result of shifts in the supply/demand balance

(a) Measures to improve the share o f the benefit accruing to developing countries In principle, one can distinguish two categories of measures, viz. those which operate within the existing institutional framework governing economic relations between developing and developed countries, and those which are designed to change the institutional framework so that the system operates differently, providing greater benefits for developing countries. Measures in the first category - operating within the existing institutional framework - relate essentially to joint action by exporters and importers of particular commodities., Of the 12 commodity-specific factors listed in Figure 2, such joint action could be envisaged for only three, viz. on processing, information and marketing and distribution.

ANALYSIS OF PRIMARY COMMODITY MARKETS As regards processing, there are three major reasons why, for many commodities, processing is carried out mainly in developed importing countries. First, where the trade is largely dominated by TNCs, these are often verticallyintegrated with links to user industries. In such cases, it is generally more efficient to locate the processing stage in the TNCs' home markets, near to the main users. Second, many processing activities are highly capital- or energy-intensive and require specialist skills, and the scarcity of these resources in many developing countries may effectively rule out their establishment in these countries. Third, there is a well-documented tendency for trade barriers of developed countries to escalate with the degree of processing (Yeats, 1979). 23 There may be scope for negotiations between developing and developed countries designed to reduce the degree of protection of processing industries in developed countries. However, in spite of much discussion of this issue for many years in international fora, the developed countries have so far been unwilling to make any substantial changes, and probably are unlikely to do so in the present economic climate. Moreover, where user industries are highly integrated, and especially where they have connections with synthetic materials producers, processing plants established in developing countries may have major difficulties in securing market outlets. In recent years, some developing countries have established copper product plants in Western Europe as joint ventures with European refiners (Fortin, 1980) as one way around these trading impediments, and this might be an option increasingly followed in the future. Collaboration between importing and exporting countries in the provision of adequate and timely information on the world market for individual commodities is clearly a 'positivesum' game, and one of the traditional functions of International Commodity Agreements (ICAs), and of international c o m m o d i t y Study Groups, has been the dissemination of regular information on market prices, production, consumption and trade. Both the ICAs and the Study Groups would also be concerned with estimating the world supply/demand balance for the coming year, and with basing their policy recommendations on these estimates. For commodities with terminal markets, virtually continuous information on market prices and on market conditions is essential for effective producer participation, but information service is costly and needs special organization. An interesting question here is

37

whether producer countries will be able to utilize the new satellite and communication technologies to reduce the present asymmetry in the availability of market information, or whether these technologies will be used by consumers to increase the existing disparities in the availability of such information. As regards marketing and distribution channels, there would appear to be scope for negotiations on a Code of Marketing which would set out general principles to govern the international marketing by TNCs of the commodity exports of developing countries. Such a Code could also specify the normal terms and conditions of sale to be included in longterm contracts between producing enterprises in developing countries and consuming enterprises in developed countries, including procedures for adjustment in the event of unforeseen changes in market prices, exchange rates, etc. Joint action on the above lines by developed and developing countries could improve the share of the benefit of commodity trade accruing to developing countries in certain respects, but their overall impact is hardly likely to be more than fairly marginal. Turning now to the second category of measures - those which involve changes in the existing institutional framework these relate essentially to measures which could be taken by developing countries themselves. 24 Here, the possibilities for significant action are wider than those subject to negotiation with developed countries. Of the 12 commodityspecific factors listed in Figure 2, there would seem to be good potential for action as regards five (viz. control over reserves and production, processing, the obsolescing bargain, the government learning process, and marketing and distribution), particularly if allowance is made for the possibility of joint action by groups of developing countries as well as by individual governments. There are a number of instances in recent years of governments of mineral-exporting countries improving their bargaining position by exercising greater control over reserves and production. For example, exploration rights held by TNCs have on occasion been withdrawn, while there have been important instances, particularly in copper, of governments taking partial or full control of TNC operations. The recent purchase of UK rubber estates by the Malaysian government is another example of this type of action. Action by developing countries to create or expand processing facilities could be envisaged as part of

38

WORLD DEVELOPMENT

a 'package' deal with TNCs. One example, quoted by Labys (1980), was the refusal of Jamaica in the mid-1960s to grant new bauxite concessions unless the TNCs involved agreed to build local alumina plants. 2s As regards the obsolescing bargain, there has been a marked trend for governments to insist on renegotiation of mineral contracts some time after the initial investment by TNCs, the governments using such devices as increased corporate taxes, and increased participation in equity, to negotiate better terms. Governments of developing countries can accelerate the learning process in a number of ways, e.g. by providing technical and managerial training in required skills, as well as incorporating provisions for in-house training of local personnel in contracts with TNCs and their subsidiaries. Governments can also establish specialized institutes to study the technical, economic and marketing problems of particular commodities, so that they are fully aware of these problems when negotiating with TNCs. For some commodities, where international marketing and distribution channels are dominated by TNCs, there may be possibilities for developing producing countries, acting in concert, to establish alternative sales outlets, for example, by promoting their own brand names (which would be possible in tea, and probably in coffee) and selling through their own marketing companies. In other suitable cases, contracts could be arranged directly with mail order firms or retail stores in developed countries. Increased control of marketing channels can also cover the important case of joint action by developing countries, e.g. through Producer Associations, to regulate total supplies of a commodity coming on the world market. Such regulation, if successful - and certain preconditions must be satisfied for this to be the case - can substantially shift the division of benefit in favour of developing producing countries. Alternatively, in suitable circumstances, governments can successfully impose a uniform export tax (as for bananas), or royalty (as for bauxite) to increase their net return. Finally, among the macro-economic factors influencing the bargaining situation, there is one which has been deployed effectively by some developing countries, viz. the opening of quite new trade channels (with China or with Eastern Europe, in particular) which effectively bypass the TNCs or, at least, significantly reduce the latter's bargaining power in negotiations with the developing countries concerned. Measures such as the above, which intro-

duce new institutional forces in the market, can undoubtedly result in significant shifts in the balance of bargaining power in a number of commodity markets in favour of developing countries and, to that extent, they would tend to result in a reduction in the disproportionate share of the benefits of trade which is now captured by oligopolistic TNCs. However, this general strategy implies that the developing countries will become more closely integrated into the existing international economic system, and may result in their becoming even more dependent on the developed countries than they are at present.

(b) The 'delinking' strategy The main alternative strategy for developing countries is to direct their efforts to making the best use of their own resources in order to satisfy the fundamental needs of their own population (i.e. a self-reliant development strategy). This will evidently require dramatic changes in existing patterns of production, consumption and trade, and will almost inevitably involve major changes in the nature of the economic relationships with developed countries and, in particular, with the TNCs. One school of thought has advocated that, to achieve this end, developing countries should 'delink' their economies on a selective basis from the international economic system, thus allowing them to reorientate their production towards meeting pressing domestic needs. 26 Broadly speaking, a policy of economic self-reliance, whether or not based on delinking, or selective delinking, should be directed towards a threefold objective. First, it should greatly reduce the existing degree of dependence of developing countries on the economies of developed countries, in terms of the latter as sources of capital for investment or as markets for primary commodities. Second, it should result generally in an improved bargaining strength of the developing countries in negotiations with developed country interests. And, third, it should lead to significant improvements in living standards of the majority of the population of developing countries. How might such a strategy work in relation to the commodity sector? Here, it is useful to consider broad commodity groups separately, distinguishing in each case those commodities for which demand in developing countries can be substituted for exports to developed countries from those commodities for which such substitution is not practicable, or is unlikely.

ANALYSIS OF PRIMARY COMMODITY MARKETS In the food, beverages and tobacco group, most exports from developing countries are unlikely to find equivalent domestic sources of demand. Tea is, perhaps, an exception in view of the relatively high income - and price elasticities of demand in developing countries. But if exports of coffee and cocoa to developed countries were reduced substantially the shortfall would probably not be made good by exports to other developing countries. On the other hand, a regulated reduction of production in developing countries should be accompanied by a rise in price (demand being price-inelastic), so that 'partial delinking' in tropical beverages should result in higher export earnings. Sugar is, perhaps, a special case in so far as there are important industrial uses which could be exploited in the producing countries. A reduction in the output of export crops should, ideally, be undertaken in conjunction with an increase in the output of food for domestic consumption in the large number of developing countries which are now food importers; the possibilities of such a switch in output pattern will, of course, vary greatly according to local socio-economic circumstances and technical factors. For the relatively few developing countries which are net exporters of basic foods (Argentina being the main example), the possibilities of switching exports of cereals and meat from developed country markets to other developing countries would depend heavily on the degree to which a general policy of 'delinking' raises real incomes and strengthens the payments balances of Third World countries. In the agricultural materials group, producers of the various textile fibres might well find it possible to substitute domestic demand for exports to developed countries. There is clearly an enormous unsatisfied demand in the Third World for clothing and household textiles. The question is, rather, whether government policies can make even part of the potential demand effective in terms of purchasing power. A diversion of only a proportion of the cotton exports of developing countries, to take one important example, could well change the -

39

character of the world textile economy, with cotton then being used as a high quality, highprice, fibre in the developed countries. Most of the other agricultural materials, e.g. rubber, vegetable oilseeds and oils, are not so likely to find substitute markets in developing countries for some time to come. Much will depend on the rate of expansion of domestic user-industries. For such commodities, it might be best for developing countries to attempt to maximize their real foreign exchange return, by strengthening their competitive position in developed country markets and, to the extent that it is possible, by regulating supply so as to improve the trend of prices. As regards mineral exports, an additional option for producing countries would be to limit output to domestic requirements or, rather, to the domestic requirements of developing countries as a group. But such a policy would cause serious payments difficulties for many specialized mineral-exporting countries. 27 Most of these would seem to have little option but to continue to export to developed countries, while attempting to diversify their economies and reduce their external dependence. While a final verdict would have to await the results of detailed studies of particular commodities, it would appear that 'selective delinking' would probably be a feasible approach for certain export commodities (tea, textile fibres and some others) where a reorientation of current development strategies would allow a substantial expansion of effective demand to take place. For other commodities, the demand for which in developing countries would be unlikely to grow sufficiently to offset a loss of markets elsewhere, any delinldng would need to be gradual and spread over a considerable time period. However, the main difficulty in adopting a self-reliant strategy does not lie so much in the characteristics of the various commodities, but rather in the need for fundamental socio-economic change as a precondition for the adoption of such a strategy, since such fundamental change will inevitably be strongly opposed by the dominant social groups.

NOTES 1. F. Hahn (1982), pp. 5-9. 2. More recently, a useful distinction has been made between markets where price is the adjusting element ('flex-price' markets) and those where adjustments are effected by quantity changes ('fix-price' markets).

However, this distinction has yet to be fully incorporated into mainstream neoclassical theory. 3. In socialist countries of Eastern Europe, pricing is used in distinctly different functions (e.g. as a control mechanism by the planning authorities).

40

WORLD DEVELOPMENT

4. Similarly, the spurious argument that actual commodity markets result in an optimal allocation of resources is one often used by governments (and others) opposed to international commodity agreements. 5. For a useful concise discussion of the role of futures and contingency markets, see Harris, Salmon and Smith (1978). 6.

involving foreign responsibility for management and marketing was terminated in 1974. 16. If a wider concept of the benefits of trade were used, including indirect effects, a could decline to zero, or even become negative in the case of export by a TNC of a depletable natural resource. 17. It is not suggested that corruption in foreign trade is confined to developing countries.

See the analysis in Kofi (1973).

7. The alternative concept of 'retained value' would thus seem to be more relevant than the usual f.o.b. export value, especially for analysis of the division of benefit. The retained value concept was first used by Reynolds (1965). A recent discussion of the concept in relation to depletable natural resources is contained in Brodsky and Sampson (1980). 8. Cocoa seems to be a special case where the largest trader is also the source of the most auth6ritative forecast of the current crop (see UNCTAD, 1975). 9. Such an analysis is not possible from published data, but it may be hoped that the producer interests concerned will eventually make, and publish, this kind of analysis.

18. The two intercepts, a and b, need not, of course, be equal. 19. A broadly similar, though more comprehensive, conceptual framework to that presented here has been developed by the United Nations Centre on Transnational Corporations jointly with the United Nations regional Economic Comissions. The UN schema covers political factors in host countries (such as the political philosophy of the government), the world political order (including decolonization) and the world economic and information orders, in addition to the more specifically commodity-related aspects considered here (United Nations, 1978). 20. The following discussion of these nine factors is based essentially on Labys (1980).

10. For a discussion of this possibility, see Maizels (1982).

21. In extreme cases, this could precipitate a repatriation by a TNC of its original investment.

11. This problem is amenable to solution by linear programming or other appropriate optimizing techniques.

22. The political stability of the host country government could also be an important influence on that government's bargaining position, but political factors are not considered in the present context.

12. One method of avoiding this result, and of expanding tax revenues from TNC operations, would be to relate tax, not to the value of primary production or export, but to the value of the final product on consumer markets. This method has been used for several years by Jamaica in taxing the production of bauxite by TNCs in that country. 13. See Harris, Salmon and Smith (1978), pp. 23. 14. 'Where non-integrated polypropylene twine manufacturers are atso manufacturers and/or importers/ distributors of sisal twine - as is the case in the EEC this aggressiveness [of pricing policies] is further enhanced by multi-tier pricing policies which, while maximizing the overall returns from twines . . . give them ~tremendous power to influence the prices at which sisal twine exporters can sell. This is reinforced when access to markets and marketing is largely under the control of the same manufacturers/ importers/ distributors' (UNCTAD, 1981).

23. It could be argued that the protection of processing industries in developed countries is a consequence of the economic power of the TNCs, but this hypothesis has never been tested empirically. 24. The only exception to this generalization has been the creation of the Common Fund in the context of the UNCTAD Integrated Programme for Commodities. But the Common Fund, as it has emerged from the negotiations, will have greatly reduced powers as compared with the original proposal of the developing countries.

-

15. This seems to have been the case following the acquisition by the Government of 51% of the assets of the copper mines in Zambia in 1969. The arrangement

25. The outstanding case of 'downstream' diversification was, of course, the establishment of petroleum refineries in OPEC countries. 26. For an analysis and critique of the delinking hypothesis, see Diaz-Alejandro ( 1 9 7 8 ) a n d Evans (1981). 27. Unless prices rise more than in proportion to the decline in output (which is unlikely for the majority of minerals exported by developing countries).

ANALYSIS OF PRIMARY COMMODITY MARKETS

41

REFERENCES Baldwin, R. E., 'Export technology and development from a subsistence level', Economic Journal (March 1963). Beckford, G. L., Persistent Poverty: Underdevelopment in Plantation Economies of the Third World (Oxford: 1972). Bosson, R., and B. Varon, The Mining Industry and the Developing Countries (Oxford: 1977). Brodsky, D. A., and G. P. Sampson, 'Retained value and the export performance of developing countries', Journal of Development Studies, Vol. 17, No. 1 (October 1980). Diaz-Alejandro, C. F., 'Delinking North and South: unshackled or unhinged?', in A. Fishlow, et aL Rich and Poor Nations in the World Economy (McGraw ttill: Council on Foreign Relations, 1978). Evans, H. D., 'Trade, production and self-reliance', in D. Seers (ed.), Dependency Theory: A Critical Re-Assessment (London: Frances Pinter, 1981). Fortin, C., 'Third World commodity policy at the crossroads: some fundamental issues', IFDA Dossier 15 (Jan./Feb. 1980). Hahn, F., 'Reflections on the Invisible Hand', Lloyds Bank Review (April 1982). Harris, S., M. Salmon, and B. Smith, Analysis of Commodity Markets for Policy Purposes (London: Trade Policy Research Centre, 1978). Helleiner, G. K., lntra-firon Trade and the Developing Countries (London: Macmillan, 1980). Hirschman, A. O., The Strategy of Economic Development (Yale, 1978). Kofi, T. A. 'A framework for comparing the efficiency of futures markets', American Journal of Agricultural Economics, Vol. 55, No. 4, Part 1 (Nov. 1973). Labys, W. C., Market Structure, Bargaining Power and Resource Price Formation (Lexington Books, 1980).

Maizels, A., Selected lssues in the Negotiation of International Commodity Agreements: An Economic Analysis (Geneva: UNCTAD, 1982). Perroux, F., 'The domination effect and modern economic theory', Social Research, Vol. 17 (1950) [reprinted in K. W. Rothschild, Power in Economics (Penguin Books, 1971)1. Reynolds, C. W., (with Mamalakis, M.), Essays on the Chilean Economy (Homewood, Illinois: Irwin, 1965). Singer, H. W., 'The distribution of gains between investing and borrowing countries', A.E.R. Papers and Proceedings (May 1950). Singer, H. W., 'The distribution of gains from trade and investment - revisited', Journal of Development Studies (July 1975). Thoburn, J. T., Primary Commodity Exports and Economic Development (Wiley, 1977). United Nations, Transnational Corporations in ExportOriented Primary Commodities: A General Conceptual Framework /or Case Studies, Working Paper I of Meeting of Participating Regional Commissions and the Commission on Transnational Corporations (New York: Sept. 1978). UNCTAD, Marketing and L)istribution System /br Bananas (December 1974). UNCTAD, Marketing and Distribution System for Cocoa (January 1975). UNCTAD, Marketing and Distribution of Tobacco (June 1978). UNCTAD, Fibres and textiles." dimensions of corporate marketing structures (November 1980). UNCTAD, The Marketing of Hard Fibres (Sisal and Henequen): Areas /br International Co-operation (October 1981). Vernon, R., Sovereignty at Bay: The Multinational Spread of U.S. Enterprises (New York: Basic Books, 1971). Yeats, A. J., Trade Barriers Facing Developing Coun tries (London: Macmillan, 1979).