+ All Categories
Home > Documents > Talbot 1995.pdf

Talbot 1995.pdf

Date post: 06-Nov-2015
Category:
Upload: aivlysman
View: 74 times
Download: 11 times
Share this document with a friend
Popular Tags:
38
Regents of the University of California is collaborating with JSTOR to digitize, preserve and extend access to Berkeley Journal of Sociology. http://www.jstor.org Regents of the University of California Regulating the Coffee Commodity Chain: Internationalization and the Coffee Cartel Author(s): John M. Talbot Source: Berkeley Journal of Sociology, Vol. 40, Globalization (1995-1996), pp. 113-149 Published by: Regents of the University of California Stable URL: http://www.jstor.org/stable/41035513 Accessed: 24-03-2015 18:14 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTC All use subject to JSTOR Terms and Conditions
Transcript
  • Regents of the University of California is collaborating with JSTOR to digitize, preserve and extend access to BerkeleyJournal of Sociology.

    http://www.jstor.org

    Regents of the University of California

    Regulating the Coffee Commodity Chain: Internationalization and the Coffee Cartel Author(s): John M. Talbot Source: Berkeley Journal of Sociology, Vol. 40, Globalization (1995-1996), pp. 113-149Published by: Regents of the University of CaliforniaStable URL: http://www.jstor.org/stable/41035513Accessed: 24-03-2015 18:14 UTC

    Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp

    JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of contentin a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.For more information about JSTOR, please contact [email protected].

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • Regulating the Coffee Commodity Chain: Internationalization and the Coffee Cartel^

    John M. Talbot

    Introduction

    Li almost every city in the US today, there is a specialty coffee shop selling freshly roasted coffee beans from countries around the world: Costa Rica, Kenya, and Indonesia, to name but a few. Is this sudden growth of the specialty coffee market just another aspect of the globalization of economic activity? Not really, because the coffee commodity chain has always been globally organized. In fact, the recent popularity of specialty coffees represents, in a sense, a reaction against globalization of the coffee industry. But the most surprising reaction against globalization was the formation, in August 1993, of a coffee producers' cartel which attempted to raise world market prices of coffee. This paper explores the nature of the coffee commodity chain and the reasons for these recent developments.

    Two promising approaches for studying the globalization of the world economy have recently emerged. The commodity chain approach (Gereffi and Korzeniewicz, 1994) focuses on the structures of the global networks of production processes and transactions which result in production of particular types of finished commodities. The regulation approach (McMichael and Myrhe, 1991; McMichael, 1992) focuses on the role of states and multilateral institutions in regulating globally-organized production systems. This paper proposes a synthesis of these two perspectives, called the regulating the commodity chain approach. This approach uses the commodity chain as the unit of analysis, but focuses on

    lrThe research reported here was partially supported by NSF Dissertation Improvement Grant No. SBR-9300877. Portions of this analysis were presented at the annual meetings of the Association of American Geographers, San Francisco, March 29 - April 2, 1994, and the Pacific Sociological Association, San Diego, April 14-17, 1994. I would like to thank the discussants at those sessions, Terry Marsden and Walter Goldfrank respectively, for their comments on those presentations. I would also like to thank Peter Evans, Brian Wright, and Laura Enriquez for their comments on an earlier draft of this paper. Finally, I would like to thank C. P. R. Dubois, Marianne Bradnock, Rebecca Adams, and Trevor Nash of the ICO Secretariat Staff for their assistance with my research at the International Coffee Organization.Of course, none of the above bear any responsibility for the interpretations and conclusions presented here.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 1 14 BERKELEY JOURNAL OF SOCIOLOGY

    the roles of states and multilateral institutions in regulating the nodes and links of the chain. This approach is illustrated through an analysis of the coffee commodity chain.

    The commodity chain approach highlights the role of transnational corporations (TNCs).2 The TNCs exercise control over commodity chains through their access to capital, their dominance in R&D and technology, and/or their oligopolistic control of distribution networks in the core markets. Through this control over key nodes of the chain, they organize global production and have the power to change the structure of the chain or the location of its nodes. They use this control to appropriate the majority of the surplus generated along the chain. The commodity chain approach acknowledges the importance of state action in shaping commodity chains, but tends to focus on the actions of individual peripheral or semiperipheral states as they attempt to counteract or adapt to the global production strategies of the TNCs.

    For example, Gereffi and Korzeniewicz (1989) and Gereffi (1994) analyze the dynamics of "buyer driven" chains organized by the TNCs that control the core markets for footwear and appareL The apparel TNCs organize sophisticated global networks that produce the variety of apparel items the TNCs market. Production is allocated between countries based on some combination of labor costs, quality of output, and reliability for timely delivery. The TNCs distinguish between a "semi-core" of the "four tigers" (Taiwan, South Korea, Hong Kong and Singapore) which produce high quality on time, but at higher labor costs, and several different levels of periphery, differentiated by their labor costs, consistency of quality, and reliability (Gereffi, 1994: 111). This global sourcing strategy allows the TNCs to obtain the mix of different qualities of apparel items they need to supply their markets, on time and at the lowest possible cost, thereby maximizing their profits. Third World states enact policies intended to integrate local production into this network, and to move up closer to the semi-core. At stake is the larger share of value added and surplus which is located at those levels closest to the semi-core, although it represents only a relatively small share of the total value added and surplus generated along the entire chain. But these actions by Third World states are reactive~they attempt to adapt to the global strategies of the TNCs.

    2This approach is best exemplified by the studies collected in the Gereffi and Korzeniewicz (1994) volume, but also includes the earlier work of Gereffi and Korzeniewicz (1989) and Hopkins and Wallerstein (1986). Arrighi and Drangel's (1986) study of the stratification of the world-economy does not use a commodity chain approach in its analysis, but the concept of the commodity chain is central to their definitions of core and peripheral activities.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 115

    The regulation approach highlights the roles of states and multilateral institutions (particularly the World Bank and IMF).3 The multilateral institutions serve as regulatory agencies for globally-organized capitalist production. They subordinate the national regulatory attempts of peripheral and semiperipheral states to the global logic of capital, using the carrot of access to international credit and the stick of structural adjustment. Individual states lose the power to regulate the fractions of globally-organized production processes which are located within their own borders.

    For example, McMichael (1992) analyzes the foundations of a new global food regime. He argues that they are "the related processes of internationalization of food production and markets, on the one hand, and the substitution of global for local regulation on the other." (1992: 345) Basic food grains are produced in the core and exported to the Third World, while Third World countries increasingly specialize in high-value "non-traditional" fruits and vegetables for export to the core. One of the key regulatory agencies that forced this change was the IMF. During the 1980s debt crisis, the IMF, on the one hand, forced Third World states to open their markets to imported food, and on the other, pushed them to increase their agricultural exports in order to make payments on their debt. According to McMichael, the Uruguay round of GATT (and by extension, the new World Trade Organization) represents a further step in c

  • 1 16 BERKELEY JOURNAL OF SOCIOLOGY

    commodity chain approach focuses on the locations of specific production processes and the nature of the flows between them, which are affected by the actions of specific states and TNCs. The regulation approach focuses on the larger political-economic system within which these processes are situated, particularly on the regulations promulgated and enforced by multilateral institutions. However, neither of these approaches incorporates the possibilities for coordinated action by peripheral states to challenge the regulatory role of transnational capital or multilateral institutions; nor do they incorporate a role for core states, aside from their indirect influence as exerted through the multilateral institutions, hi the analysis of specific commodity chains, the interests and strategies of these diverse institutional actorsr-TNCs, states, and multilateral institutions4~must be taken into account. The political interactions among all of these institutional actors result in the construction and modification of regulatory regimes governing linked sets of production processes.

    In this paper, I combine the commodity chain and regulation approaches with a focus on collective action by groups of states. I use this synthetic "regulating the commodity chain'9 approach to analyze the changing forms of regulation of the world coffee market. In the case of coffee, producing and consuming states constructed a different type of multilateral institution, the International Coffee Organization (ICO), to regulate the market by administering a system of export quotas. Because the quotas were politically negotiated, the ICO regulatory regime caused reactions along the commodity chain which eventually undermined the regime. The collapse of the ICO regime revealed the inherent instabilities in an unregulated market, which reverberated along the chain and created a new commonality of interests among coffee-producing states in renewed international regulation. The coffee-consuming states, by this time firmly committed to a global "free trade" regime, refused to cooperate. The producing states responded by forming the Association of Coffee Producing Countries (ACPC), the coffee cartel.

    The Coffee Commodity Chain

    Coffee is a tropical crop; it is grown in Third World countries, and most of it is consumed in core countries. It is the second most valuable primary commodity exported by "developing" countries, after oiL The coffee commodity chain is thus an excellent case for analysis of the conflicts

    4In this analysis, TNCs, states, and multilateral institutions are generally treated as unitary actors. I use the phrase ''institutional actors" to emphasize that they are all complex bureaucratic organizations, and that different divisions or agencies of these bureaucracies may have conflicting interests. These conflicts are beyond the scope of this analysis.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 1 17

    between core and percherai actors over control of the chain and the profits it generates. Figure 1 presents a simplified picture of the coffee commodity chain; there are four major processing stages, shown on the right side of the Figure. Coffee is produced by growers in the form of a red cherry. When the fruit is stripped of two seeds or '"parchment" coffee beans remain. Next, the white shell of the parchment is milled of leaving a green coffee bean, the form in which most coffee is traded on the world market. Finally, green coffee is processed into roasted and ground (R&G) or instant coffee for final consumption.

    Three major types of institutional actors influence the structure of the chain. The TNCs dominate the chain through their control of key nodes - some are oligopsonistic buyers of green coffee, and others are oligopolistic sellers of processed coffee products in the core consumer markets. The importers buy coffee from a large number of producing countries, ship it to the consuming countries, and sell it to the coffee manufacturers. The manufacturers blend coffees from several different origins and manufacture roasted or instant coffee. This blending allows the manufacturer to substitute coffee from one origin for that from another in response to changes in price or availability, while maintaining a consistent taste in the final product. TNC coffee manufacturers thus require TNC importers who can deal in large volumes of coffee from many different producing countries and supply the different varieties of coffee needed by manufacturers. The coffee commodity chain has been globally organized by TNCs since at least the 1950s, through what Friedmann (1991) has called global sourcing, obtaining generic raw material inputs from a variety of different sources for use in industrial food production.

    Third World producing states influence the structure of the chain both directly, through their state marketing boards, and indirectly, through their regulation of coffee production and exports. Marketing boards, common in Africa and also found in some Asian countries, have legal monopolies on coffee exporting. They set prices and quality standards and buy all coffee destined for export from the growers. They also do some processing, usually from the parchment to green form, and then deal directly with coffee importers to sell the country's coffee. State coffee agencies, found in most Latin American countries, often set prices and quality standards, and they may buy and export some coffee. But they mainly regulate private export firms, controlling the amounts that they are allowed to export and collecting export taxes. Through this regulatory control, they may indirectly set minimum export prices for the private firms. Consuming states are less visible in this picture, but their interactions with both TNCs and producing states, and their regulation of coffee imports, processing, and sales, also influence the structure of the chain. These three types of institutional actors each have contradictory and sometimes conflicting interests in relation to

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 1 18 BERKELEY JOURNAL OF SOCIOLOGY

    Figure 1: The Coffee Commodity Chain

    Coffee Consumers

    /, R&G

    I 1 1 Instant Coffee Manufacturers Roasted/Ground/Instant

    Traders s& and Brokers

    Coffee Importers / and Traders (TNCs) /

    / V x/ X. CORE / X. Green / / x

    PERIPHERY / / '. Coffee

    Marketing Private Boards Exporters

    ' / / Parchment ' Intermediaries / ~ - ' Coops/Traders/Agents /

    ' I / Coffee ' / / Cherries

    Coffee Growers

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 1 19

    the regulation of different stages of the chain, and in particular, to the international regulation of the world market.5

    Coffee-consuming core states, particularly since the early 1980s, are generally champions of a liberal international "free trade" regime. This is also usually in the interests of TNC coffee traders and coffee roasters which are based in these same core countries. However, the inherent instability of a "free" world coffee market, and the serious effects it can have on the economies of peripheral producing countries, can provide a geopolitical rationale for core states to support political regulation of the market, in order to stabilize client percherai states. Similarly, the TNCs which dominate the coffee commodity chain tend to prefer a "free market" regime. But they also require stable supplies and consistent qualities of coffee to be available on the market, and these requirements can be jeopardized by either the economic instability of the market or the political instability of major producing countries. Thus TNCs may at times also share the geopolitical interests of their home states in supporting political regulation of the world market.

    Coffee-producing Third World states have the most complex set of interests regarding regulation of the market. They generally favor a politically regulated market if it will stabilize and raise the general level of world prices, because coffee is a labor-intensive crop which provides a major source of rural income and employment in most producing countries. In addition, export taxes on coffee, or marketing board profits, are often a major source of government revenues. But political regulation of the market has historically been accomplished through export quotas, which limit exports and necessitate government expenditures to purchase and maintain coffee stockpiles in order to meet their quota obligations. Depending on the relative sizes of a country's production and its export quota, political regulation may impose costs on the state which outweigh the benefits obtained through a higher world price. Further, the fact that coffee revenues provide a major source of income to both coffee growers and the state means that the division of these revenues may become a source of conflict within the producing country.

    The result of these contradictory and conflicting interests of the three types of institutional actors has been alternating periods of "free market" and "political" forms of regulation of the coffee commodity chain. In a previous

    ^he ICO and ACPC are also institutional actors in this story. However, because the regulations administered by the ICO were politically negotiated between the producing and consuming states, and the ACPC regulations are politically negotiated among the producing states, neither the ICO nor the ACPC are considered as independent institutional actors in this analysis.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 120 BERKELEY JOURNAL OF SOCIOLOGY

    paper (Talbot, 1995a), I showed how these conflicting interests led from a "free market" period beginning after World War H, to a regulated market under the first two International Coffee Agreements (ICAs) from 1962 to 1972, to a deregulation which lasted until 1980, when the market was reregulated under the third and fourth ICAs. In this paper, I show how developments during the 1980s undermined the most recent ICA regulatory regime, leading to its collapse in 1989 and precipitating a price crash. I then show how the resulting instabilities of a "free" market led first to attempts to renegotiate the ICA, and then to the formation of the carteL To provide the context for this stoiy, the next section describes the new internationalization of the coffee processing industry since 1980.

    The Internationalization of Coffee Production since 1980

    The regulation approach focuses attention on four aspects of internationalization: the increasing internationalization of capital, the increasing power of finance capital, the increasingly important role of multilateral institutions, and the reorganization of percherai states. This section concentrates on the first two aspects, as well as on a segmentation of international trade in coffee which both reflects internationalization and represents a reaction against it. The roles of multilateral institutions and peripheral states are somewhat different for coffee than would be expected on the basis of the regulation literature; these are dealt with in the following sections.

    The internationalization of coffee production has been manifested primarily in the concentration of TNC coffee roasting and coffee importing firms, and their increasing control over markets in all consuming countries. Concentration of roasters in the consuming markets was well underway in the 1960s and 1970s (Talbot, 1995a). It accelerated during the rash of mergers and buyouts of the 1980s, and continued through the free market period of the early 1990s. Meanwhile, the concentration of coffee importers was also increasing, and it was greatly accelerated by the 1989 price crash. Five TNCs now account for well over 60% of total coffee sales across all major consuming markets. The largest is Nestl, the world's largest food processing company. It has always been a leader in instant coffee in the US and EC markets, and in recent years has expanded its share of roasted and ground (R&G) markets. In the US, Nestl has acquired Chase and Sanborn, Hills Bros., MJB, and Sark's, and in Europe it has acquired Zoegas, a Swedish roaster with large shares of Scandinavian markets. Nestl also has a significant share of the Japanese instant and canned coffee markets. Close behind Nestl is Philip Morris. It began to diversify out of tobacco and into food processing in the 1980s, with its acquisitions of Kraft and General Foods, owner of Maxwell House, long the leader in the US R&G market. It

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 121

    also acquired Jacob Suchard, one of the largest roasters in France, with a large share of the EC market, and Gavalia, another major Swedish roaster. General Foods, in a joint venture with food processing giant Ajinomoto, is also the largest coffee company in Japan. The third largest coffee processing TNC is Proctor and Gamble, which owns Folgers, the second-largest coffee company in the US market. Procter and Gamble does not have large coffee sales outside the US. Fourth is Sara Lee, the US food processing giant which owns Superior Coffee, and which acquired Douwe Egberts, a Dutch roaster with large market shares throughout Northern Europe, which itself had previously merged with Van Nelle, another major Dutch roaster and food processing conglomerate. Finally, Coca Cola is a major manufacturer of instant coffee in all core markets (Tea and Coffee Trade Journal January 1989: 16-17; April 1989: 7; September 1991: 58; November 1992: 40, 76; April 1993: 76; World Coffee & Tea January 1990: 28-30; March 1990: 23; January 1991: 26-3 l;Landell Mills, July 1991; Talbot, 1995b).

    All five of these TNCs are multi-product conglomerates, and despite the fact that they are the largest coffee manufacturers in the world, coffee is not their main product. These statistics underestimate the degree of TNC control, since, in some of the major markets, coffee and food-processing TNCs of only slightly smaller scale also have significant market shares, for instance Tchibo and Eduscho in Germany, and Ueshima and Key Coffee in Japan. Italy is the major exception, with many small regional roasters, but even that is beginning to change (World Coffee & Tea, August 1992: 68; Tea and Coffee Trade Journal, August 1990: 52; January 1991: 52-6; June 1991: 16; February 1992: 47; October 1992: 56; F.O. licht, 23 December 1993; ANECAFE, 15 May 1993).

    Through the 1980s, the increasing concentration of coffee roasters and instant coffee producers in consuming markets began to lead to concentration of coffee importing firms. The few large manufacturing TNCs prefer to buy their own coffee directly, or to deal with a few large importers who can provide the full range of coffees and the full range of services they require. Most of the time, they deal with the TNC trading firms, which can assume the risks involved in international trading, such as currency fluctuations, late or small harvests in producing countries, shipping delays, or poor quality coffee sold as high quality. This tends to drive out of business or marginalize smaller importers who are unable to compete for the high volume sales to TNCs.

    Additionally, the 1989 price crash drove many importers, large and small, out of business; some of them were already in precarious positions after the October 1987 stock market crash (F.O. licht, 3 November 1987). Some were holding large stockpiles purchased at high quota prices, which declined precipitously in value after the end of the quotas, and had to be

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 122 BERKELEY JOURNAL OF SOCIOLOGY

    sold at a loss. Others were holding large speculative positions in the futures markets, and also took large losses. But most importers work on percentage commissions, and when the price fell by 50%, so did their commissions. Thus by 1992, the five largest coffee importers (Neumann, Volkart, ED&F Man, Cargill, and J. Aron) controlled over 40% of total world imports (ANECAFE, 15 May 1993). Neumann, already a large multi-commodity trader, became the largest after taking over another of the largest coffee importers, Rothfos. ED&F Man is the world's largest cocoa trader, and also a major sugar importer. Cargill, the grain trader, instantly became one of the world's largest coffee importers when it purchased ACLI Coffee; J. Aron is now owned by Goldman, Sachs, In addition, the largest sogo shosha, C. Itoh, Marubeni, and Mitsubishi, control most coffee imports into Japan. All of these importers are large muM-commodity TNC traders, and again, although they are the world's largest coffee importers, coffee is generally not their most important commodity (Tea and Coffee Trade Journal, July 1990; December 1990; January 1992; World Coffee & Tea, January 1991; F.O. licht 15 December 1987; 11 June 1993).

    Consistent with the arguments of the regulation theorists, this concentration of coffee importing and processing TNCs has gone hand-in-hand with an increasing role for finance capital Access to large amounts of capital is crucial for the TNCs, both for pursuit of their merger and acquisition strategies, and for financing the purchases of the huge volumes of coffee with which they deal This requirement has been particularly important for the coffee importers, and is best exemplified by Goldman, Sachs' acquisition of J. Aron. At the same time, the large banks and financial houses prefer to deal with the large trading TNCs with which they have ongoing relations; this further disadvantages smaller import firms and reinforces the trend toward concentration.

    The other important role played by finance capital is in the coffee fixtures markets. The futures markets, in New York and London, have been a part of the world trade in green coffee since the early 1900s. But since the early 1980s, the importing TNCs have increasingly pegged the prices they pay for green coffee to the New York or London price on the day the purchase contract is finalized. This has increased the centrality of the futures markets to the coffee trade. Not only must importers and exporters follow trends in the market, but they must buy and sell futures to protect themselves against any changes in the price which occur between the time the contract is finalized and the time the coffee is delivered. Once again, following and participating in the fixtures markets on a daily, or even hourly basis, requires financial expertise and access to capital, giving a further advantage to the largest TNCs.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 123

    This development has also given finance capital a larger role in setting coffee prices. The volume of coffee futures transactions soared during the 1980s, due in large part to the activities of commodity funds, huge conglomerations of financial capital seeking the highest and most rapid profits available, by trading in financial, oil, metals, and agricultural fixtures markets. These funds shift large amounts of money in and out of the coffee futures markets, for reasons which are usually only marginally related to the actual global situation of supplies o and demand for, coffee. This combination of pegging the price of coffee to the fixtures markets and the increased weight of the commodity funds in these markets, has probably increased the instability of world market prices (Marazzi, 1984; Kuchiki, 1990).

    The other aspects of internationalization identified by the regulation approach, the role of multilateral institutions and the reorganization of percherai states, have been quite different for coffee than for many other commodities. As we will see in the following sections, structural adjustment programs pushed by the World Bank and IMF did not have much effect on producing states' regulation of the early stages of the commodity chain until the coffee crisis of the early 1990s. State coffee agencies and marketing boards continued to exercise control over coffee growing and exporting through the 1980s. And the primary multilateral institution regulating the world coffee market, the ICO, continued to function during this time.

    The other important internationalization tendency in coffee production, which is not considered by the regulation theorists, involves a segmentation of the world market. A new trade in industrial coffee products has grown rapidly alongside the still-important 'traditional" trade in green coffee; and the green coffee trade itself has become segmented with the growth of the specialty coffee market in the consuming countries. The former is the result of internationalization, but consists of finished products rather than the industrial inputs which are the focus of the regulation approach. The latter is a reaction against it, which provides hand-crafted coffees to the upper-middle classes of the consuming countries. In the face of a creeping standardization of class diets (Friedmann, 1991), coffee consumption becomes a means for elites in certain core countries to reassert class differences.

    The new trade in industrial coffee products is linked, in part, to the increasing concentration of coffee processing TNCs in the consuming markets. As they have acquired significant shares in a number of markets, they have rationalized their operations, closing roasting or instant coffee plants with small capacity or outdated equipment, and expanding their more modern plants or bunding new state-of-the-art facilities. These new plants are strategically located near major ports and are often designed to produce

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 124 BERKELEY JOURNAL OF SOCIOLOGY

    for major markets in several different consuming countries. In international trade statistics, this rationalization shows up as an increased trade in processed coffee products among core consuming countries. Thus, exports of processed coffee by importing members of the ICO increased from a little over two million bags (green bean equivalent) in the early 1970s to just under five million in the late 1980s. In the early 1970s, about three-fourths of these exports were instant coffee; by the late 1980s, they were evenly divided between instant and roasted.

    The concentration and internationalization of coffee manufacturers has increased the demand for green coffee as an industrial input, produced in large quantities at the lowest possible price. Some coffee growing states have responded to this demand by promoting the adoption of new high-yielding varieties of coffee. These new varieties produce large volumes of coffee, but their taste is generally inferior to the traditional varieties which they replace. This standardization of coffee for a global market has sacrificed quality for quantity. But this trend has also generated a countertrend in the growth of the specialty market. Specialty coffees are carefully-tended traditional varieties produced in regions which are particularly well-suited to coffee growing, such as the Antigua region of Guatemala or the Sulawesi highlands of Indonesia. Combinations of soil and climate characteristics in these regions produce fine coffees with distinctive flavor characteristics, analogous to the distinctive flavors of wines produced in famous European wine regions. And particular coffee "estates" in these regions are beginning to gain reputations as the producers of the world's finest coffees; probably the most famous of these is La Minila in Costa Rica's Terrazu region.

    Specialty coffees are traded in much smaller quantities than industrial coffees, generally by relatively small export and import firms, which are more likely to trade only in coffee. And as small roasting companies have been driven out of the mass market by the TNCs, the specialty market has created a niche for small roasters who regard coffee roasting as an art. Peet's Coffee, which was started in Berkeley in the 1960s by coffee expert Alfred Peet, was a pioneer of this market segment. These roasters are willing to pay higher prices to obtain the highest quality green coffees available, and are also able to charge their customers much higher prices for their hand-crafted single origin coffees than the TNCs charge for their standard industrial blends.

    The specialty coffee segment within the consuming markets, particularly the US, is a growing segment, while supermarket sales of the TNCs' industrial coffees are declining. This segment, which also includes flavored coffees and the sales of espresso bars and carts, now accounts for over 15% of the US market. But as this market segment has become a larger and more

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 125

    profitable part of consuming markets, the degree of concentration is increasing, and the TNCs are also entering it. Starbucks, by far the largest specialty roaster, has grown rapidly over the last few years, through an aggressive strategy of expansion and acquisitions. It has plans to open new stores in the US at the rate of nearly one per day through 1996. Nesti has moved into the specialty market with its purchase of Sark's Gourmet Coffees, and Proctor and Gamble has acquired Millstone Coffee. Both of these brands are sold in special sections of supermarkets. Supermarket chains such as Safeway, and discount chains such as Price Club and Costco, have also developed their own house brands of specialty coffees. As these developments create a demand for larger volumes of specialty green coffees, quality may become compromised in the same way it has been in the supermarket brands.

    The internationalization of coffee production since 1980 has displayed some of the aspects identified by the regulation approach with the new internationalization of agriculture. But it has also proceeded in some ways not anticipated by these authors, because they do not take into account the possibility that different commodities may be affected quite differently by globalization. In the case of coffee, because of the tropical nature of the commodity and the structure of its chain, the impact of globalization has been limited to the upper half of the chain, beginning with the importing of green coffee (see Fig. 1). Here, the increasing concentration of the TNCs, the globalization of the scope of their operations, and the increasing importance of finance capital, are consistent with predictions of the regulation approach. The growth of trade in processed coffee products has been facilitated by the liberalization of trade promoted by GATT, but this development has not supplanted the ''traditional" trade in green coffee. The growth of the specialty trade is a reaction against globalization, but it has occurred in a very different way from that forecast by Friedmann (1993), because of the inherently global nature of the chain.

    All of this internationalization has left the lower half of the chain relatively untouched. Coffee is still a labor-intensive crop produced mainly by peasants and small farmers in most countries, using methods that have changed relatively little since the 1950s. Most important, coffee producing states have maintained regulatory control over the production and exporting of green coffee. This has led to a complex interaction between the regulation of the early stages of the chain and the changing forms of regulation of the world market. The story of how this interaction developed, leading to the formation of the coffee cartel, is told in the following sections.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 126 BERKELEY JOURNAL OF SOCIOLOGY

    Internationalization Undermines Political Regulation, 1980-89

    Coffee producing states began to cooperate in the 1950s, in an attempt to halt falling world market prices. These efforts led to the signing of the first ICA in 1962. The ICA was an agreement between coffee exporting and coffee importing states to regulate world market prices by controlling the flow of green coffee onto the market. This was accomplished by means of an export quota system Under the ICAs, producing and consuming states first agreed on a total global quota for coffee exports, which implied a target price range to be maintained by the agreement. If the world market price moved outside this range, the quota would be cut or increased in order to maintain the price within this range. The global quota was then divided into export shares allocated to the individual exporting countries; quota cuts or increases were shared proportionally. This quota system was reinstated in 1980 under the third ICA and was continued under the fourth ICA signed in 1983.6

    The division of quota shares was always a source of conflict among producing states, and created the potential for a "free rider" problem The collective interests of coffee exporters were best served by a reduction in the total quota, which would limit supply and drive up the price. But within the total quota, each individual producing state wanted to obtain the largest possible share. These individual shares were politically determined, but had to be based on some "objective" criteria, in order to secure the agreement of all producers. These criteria had traditionally been some measures of a country's "historic" world market share, and its "capacity" to produce and export coffee. Thus, the ICAs provided economic incentives for individual producing states to intervene in their coffee-growing sectors to maintain or increase their production, in order to demonstrate their export capacity and defend or increase their quotas. The regulatory regime attempted to limit total coffee exports while simultaneously creating conditions under which producers would tend to produce coffee in excess of the amount needed to fill the global quota (F.O. licht, 22 March 1988).

    ^he ICAs are international treaties which, strictly speaking, have been in force continuously since 1962. They established the International Coffee Organization (ICO) as an international organization funded by its member states, to collect statistics, promote coffee consumption, and serve as a forum for discussions among coffee producing and consuming countries. During the periods 1972-1980 and 1989 to the present, the so-called "economic clauses" of the ICAs, which establish the quota system and the rules for its operation, have been suspended. This allows for the continued operation of the ICO, but removes all regulation on the world coffee market. I date the periods of ICA regulation and of the '"free market" on the basis of whether or not the quotas were in effect. The quotas as defined under the third and fourth ICAs were in effect through the 1980s except for a brief period in 1986-87 when they were suspended following a severe drought in Brazil which caused a temporary shortage of coffee on the world market.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 127

    During the 1970s and 1980s, there was an internationalization of coffee consumption, as Western patterns of coffee drinking spread to Eastern Europe and the more affluent Third World countries.7 The ICA quotas were policed by the "traditional" importers (Le., the original importing members who had signed the 1962 agreement), who monitored their imports and reported to the ICO. The ICO issued export stamps to exporting members based on their quotas; importing members agreed to refuse to accept coffee coming from exporting members that was not accompanied by an ICO stamp. They also accepted a limit on the amount of coffee they could import from non-member exporting countries. The ICO could then keep track of total exports of each exporting member and sanction any exporting member which attempted to exceed its quota. But these new coffee-consuming countries were not ICO members; there was no limit on exports to these countries and no requirement for the use of ICO export stamps, and they were not obligated to monitor or report their imports to the ICO. The ICAs provided incentives for new producers to join (access to major consuming markets), but no incentives for new consumers. The first ICA had been signed by 25 consuming countries accounting for 94% of world coffee imports; the fourth ICA of 1983 was signed by 27 consuming members, but by 1989 they accounted for only about 80% of world imports. Meanwhile, the producing membership of the ICO grew from 37 in 1962 to 48 in 1983, and in 1989 producing members still accounted for about 98% of world exports of green coffee8 (Fisher, 1972; Akiyama and Varangis, 1990; Bohman and Jarvis, 1990; Mwandha, et al., 1985; ICO documents). The increasing importance of non-member importers in the world market, combined with the free rider problem of the quota system, undermined the

    7This was part afa larger "standardization of class diets" identified by Friedmann (1991) as one of the major trends which undermined the second food regime.

    8Members of the 1962 ICA were, as exporting members: Brazil, Burundi, Cameroon, Central African Republic, Colombia, Congo-Brazzaville, Congo-Leopoldville (Zaire), Costa Rica, Cote d'Ivoire, Cuba, Dahomey (Benin), Dominican Republic, Ecuador, El Salvador, Ethiopia, Gabon, Guatemala, Haiti, Honduras, India, Indonesia, Kenya, Malagasy Republic (Madagascar), Mexico, Nicaragua, Nigeria, Panama, Peru, Portugal (Angola), Rwanda, Sierra Leone, Tanganyika (Tanzania), Togo, Trinidad and Tobago, Uganda, Venezuela, and Yemen; and as importing members: Argentina, Australia, Austria, Belgium, Canada, Cyprus, Czechoslovakia, Denmark, Finland, France, West Germany, Israel, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland^ Tunisia, United Kingdom, United States, and USSR By 1983, Bolivia, Ghana, Guinea, Jamaica, Liberia, Malawi, Papua New Guinea, Paraguay, Philippines, Sri Lanka, Thailand, and Zimbabwe had joined as exporting members, while Yemen dropped out. In the late 1980s, Equatorial Guinea, Vietnam and Zambia also joined as exporting members. There was more turnover among importing members, as Argentina, Czechoslovakia, Tunisia and USSR had dropped out by 1983, while Greece, Hungary, Ireland, Portugal, Singapore, and Yugoslavia were added.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 128 BERKELEY JOURNAL OF SOCIOLOGY

    ICA World exportable production exceeded total export quotas through most of the 1980s, and producing member countries began to compete for sales to the new non-quota markets by offering discounts from the prices that were being paid by member importers in the quota markets. For the producing countries, it was preferable to sell the coffee that they were producing in excess of their quotas, even at a discount, rather than paying the costs of stockpiling it, or simply letting it deteriorate in storage and getting nothing in return for it. As production continued to exceed demand in the quota markets, the stocks in exporting countries grew, and so did the discounts, reaching 50% by the mid-1980s according to some estimates. This two-tier market, in which ICO-member consuming countries in effect subsidized exports to non-member importing countries, turned consuming states and TNCs against the ICA (Akiyama and Varangis, 1990).

    But at the same time that the coffee TNCs were complaining vociferously to their home states about this unfair subsidy, they were also attempting to take advantage of the existence of the two-tier market to buy some of their coffee at a discount. Some of the exports shipped to the non-quota markets were re-exported to the quota markets in ways which were difficult for the ICO to trace. This so-called "tourist coffee" (because it visited several countries on its journey) had always been a problem in the ICAs, but the increased importance of the non-member market and the large price differential which opened in the 1980s made it an increasingly significant problem One of the most common methods for accomplishing this diversion was for an exporting country to trans-ship coffee ostensibly destined for a non-member market through one of the world's major coffee ports, such as Hamburg or Singapore. While the coffee was in port, high-quality coffee destined for a non-member market could easily be switched with low-quality coffee destined for a member market, thus providing an effective discount. The coffee could also be supplied with forged ICO export stamps, or authentic stamps could be purchased from another exporting member which did not have sufficient coffee to meet its quota for that year. Another method was to sWp for 'Inward processmg"~the coffee would be imported into a member market, supposedly to be processed into roasted or instant coffee to be sold in a non-member market. But once it had entered the member market, its final destination was impossible to trace. Obviously, the importing firms in the quota markets that were in the best position to take advantage of these "tourist" routes were the large TNCs with offices and warehouses in a number of the major coffee ports. But such transactions generally required the complicity of both exporters and importers, and there was a high degree of creativity on both sides of the market.

    Because this was a clandestine trade, it is very difficult to get any data on the volume of the trade or the size of the discounts. The evidence is largely

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 129

    anecdotal and mainly supplied by major participants in the market, whose interests were sometimes served by dissemination of misleading information on the extent of this trade. That discounts were at times as high as 50% seems to be generally accepted, and estimates of the volume range from about 1.5-2 million bags per year, or about 2.5-3.5% of the global quota, by the mid-1980s. (F.O. Licht, 3 November 1987 and 25 May 1988; Landell Mills, December 1986 and Januaiy 1987; ICO documents EB3 159/89 and EB3 196/90.)

    Table 1 presents some imperfect data on this trade, derived from ICO statistics. It shows the volume and value of exports to ICO-member and non-member markets, before and after the imposition of quotas in 1980. The data are imperfect because they are based on reports by exporting states. Some states had an interest in disguising the extent of the 'tourist" trade; others may have been deceived by their export firms. The data show that the volume of exports to non-member markets increased sharply after 1980, and their unit value also fell dramatically. Consumption in non-member countries was increasing during this period, but not as fast as imports; if we assume an average annual consumption of 7 million bags in these countries after 1980, then these data indicate a 'tourist" trade of over 2.5 million bags per year. The quality and price of coffee exported to non-member markets is generally lower than at sent to the core member markets, but this can not account for the huge differential of 20-40% between member and non-member prices during 1981-85. Events of 1985- 86 confirmed that these changes in volumes and value of exports were due to the quotas. In the middle of the 1985-86 coffee year, quotas were again suspended. Exports to the non-member markets quickly fell from over 1 1 million bags to just under 7 million, and the price difference fell to less than 10%.9 Relative to total coffee exports of about 70 million bags, these quantities seem rather small. But combined with the large price differentials, these quantities were large enough to provide a significant competitive edge to the TNCs willing or able to take advantage of the 'tourist" trade.

    The second development that undermined the ICA during the 1980s was a change in "selectivity"~a shift in world demand away from the robusta coffees produced mainly in Africa and Asia, and toward arabicas produced mainly in Central America and Colombia (also Kenya, Indonesia and several other countries). The ICA also indirectly set, as a result

    9In addition, for almost a year after quotas were reinstated in 1987, the US had no customs regulations in place to allow it to refuse entry of any coffee; it is estimated that about 2 million "illegal" bags of coffee entered the US during that year (F.O. Licht, 8 February 1989).

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 130 BERKELEY JOURNAL OF SOCIOLOGY

    Os

    5 A es

    I E

    | I c a u

    E I S

    a

    Vi

    I s

    I 55 8 8 8 g I

    5 i^ co co vo ^ fe 2 &

    kS = 5S Is? ^ s ^ 1 ^2 S3 g:^ jj o CO t***_J QN-^ ON!!!? *rn I

    'E ,8 S^ oon moo m ts 2o*ri (N ^h' Q VnD

    ^ G S

    5 mt^ oom oo^; oQ

    g SI s = .S^g II-;

    m '-HCN OOtJ" 00)^ GoOeS m 25. mvo ooin oo xi .2 95c g 2g ^3 ^M is t*M ^ .t O 4-i G

    g g =

    t> voos ^htj- g^ " g^ 2 t> i? voos 55 ^htj- Sf

    g^ " I g^ Si

    . 1 1 5j 13 c . 1 13 c ^ '3 8

    Je l JJ 's 'S

    o | o ?! lof I I ! 8 2 I II M* J3:3w p^ ^^o ^S'S S'oc -g-oc :o2 ^^hJ ^^ ts eCS

    S3 vi S

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 13 1

    of its individual country quotas, global limits on the availability of the different types of coffee.10 The quota system thus limited the ability of importers to meet the shifting demand in consuming member countries. This was a further reason for TNCs and consuming states to turn against the ICA, but it also deepened the divisions among exporters. In particular, the Central American "other milds" producers and Indonesia felt that they were in a position to significantly increase their world market shares because of the shift in demand, and that they were being limited by the quota system But the evidence suggests that once again, as the TNCs and "other milds" producers complained publicly about this situation, some of them were privately engaged in 'tourist" transactions through which additional "other mild" coffee found its way into quota markets.

    As a result of these developments, when producing and consuming ICO member states met in late 1988 and early 1989 to negotiate a new ICA (the 1983 ICA was set to expire on Sept. 30, 1989), it was impossible to reach an agreement. The consumers, led by the US, demanded that any new ICA put an end to the two-tier market and include a mechanism to automatically adjust for selectivity changes without requiring renegotiation of the quotas. And by 1989, with the Soviet Bloc crumbling, the US and other consumers had fewer geopolitical concerns which would necessitate using the ICA as a means to provide aid to the producing countries, which had been the rationale behind the original 1962 ICA Meanwhile, the issue of the division of quota shares continued to divide the producers. Brazil refused to consider any reduction of its quota of about 30% of total world exports, making it difficult to increase any other exporter's quota. At the same time, Central American producers and Indonesia were demanding increased quotas because of the increasing demand for their coffees, and the expansion of their production capacity during the 1980s. And no exporting country was enthusiastic about a selectivity mechanism which could automatically cut its quota without renegotiation.

    10There are four basic types of coffee recognized in the trade: three are arabicas, which are further subdivided into Brazilian milds, Colombian milds, and other milds; and the fourth is robusta Colombian and other milds are considered best for roasting and grinding; robustas are considered to be of lower quality, but produce a higher yield of instant coffee than the arabicas; and Brazils are intermediate, and are used in both R&G and instant coffees. Colombian milds are produced by Colombia, Kenya and Tanzania; Brazilian milds are produced by Brazil, Paraguay, and Ethiopia. The other milds group includes all other producers of arabicas, but the major players in this group are Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, India, Mexico, Nicaragua, Papua New Guinea and Peru. Robustas are grown in African and Asian countries, but the major players in this group are Indonesia, Cote d'Ivoire, Uganda, Zaire, and Cameroun; by the late 1980s, the increased production of Thailand and Vietnam gave them greater importance in this group. Countries are classified into groups based on the main type they produce, although, for example, Brazil also produces robusta and Indonesia also produces arabica

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 132 BERKELEY JOURNAL OF SOCIOLOGY

    The major division in the ICO had always been between producers and consumers, but the division in 1989 pitted the US and the "other milds" producers' group (see note 10) against all other producers and consumers. The US and the other milds group felt that they had the least to gain from a new ICA and were unwilling to compromise. At the same time, the majority of producers and consumers were not willing to capitulate completely to these minority demands.11 Each side presented its proposal at the final negotiating meeting in June 1989, but because the producer and consumer groups were internally divided, neither proposal could muster the distributed 70% majority necessary for adoption.12 Having failed to renew the quota system, producers and consumers then unanimously agreed to indefinitely suspend the quotas as of July 4, 1989. The 1983 ICA was extended for two years, until Sept. 30, 1991, but without economic clauses. This allowed for the continued existence of the ICO as a data collection agency and forum for continuing discussions between producers and consumers, but removed the economic teeth of the regulatory regime.

    The ICA regulatory regime had been undermined during the 1980s by a peculiar form of internationalization. Coffee production spread to new Third World countries during this period, but contrary to the commodity chain approach, this internationalization was not driven by TNCs seeking new sources of supply. Since the world market price was fixed by the ICA, TNCs could not attempt to move production to new locations in search of lower costs. Rather, the new exporters entered the market as a result of Third World state initiatives to expand exports. In part this was a response to pressures from the IMF to expand exports, but contrary to a regulation analysis, these pressures led not only to an expansion of "non-traditional" exports from the Third World, but also to a spread of 'traditional" exports to countries which had not previously exported them. Thailand, Vietnam, and Honduras exemplify the results of this internationalization- none had been major coffee exporters before the 1980s, but all became significant exporters during that decade. And all of them had to prove their ability to

    nBy this time, the US position was being decided by the office of the US Trade Representative (USTR), ideologically committed to removing all barriers to 'Iree trade." Up until the mid-1980s, the coffee negotiations had been handled by the State Department. In the EC and Japan, negotiations were still controlled by foreign ministry officials who were still somewhat committed to using the ICA quota system as a means to provide foreign aid to Third World coffee producing countries.

    12In the ICO Coffee Council, producers and consumers each have a total of 1000 votes, divided among countries on the basis of their share of total world exports or imports. All decisions must be approved by at least a majority of both producers and consumers, and on important matters such as the terms of a new agreement, there must be a distributed super-majority of at least 70% of each side. See ICO document ICC 53-7.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 133

    deliver consistent quality on a timely basis before the TNCs would commit to large, regular purchases of their coffees.

    On the consuming side of the market, coffee consumption spread to countries which had not traditionally been coffee importers; and within the traditional consuming countries, the composition of imports was upgraded toward the higher quality arabicas. These changes, as would be expected based on the commodity chain approach, were largely driven by TNCs1 searches for new markets, and for increased profits in established markets. Because of the provisions of the ICAs, the new producing states joined the ICO, but the new consuming states did not. This situation created an economic incentive for both TNCs and producing states to promote the growth of a "tourist" trade which subverted the regulatory regime.

    The tendency toward overproduction of coffee emerged from the contradictory nature of the politically-determined regulatory regime, winch provided incentives for producing states to maintain production above their quotas. But this tendency was equally due to producing state responses to the organized pressure of coffee growers to maintain high internal prices for coffee. The flip side of heavy economic dependence on a primary commodity export is political dependence on the social classes involved in its production. For a labor-intensive commodity like coffee, this meant not only that local capitalists involved in the growing, processing, and exporting of coffee had a great deal of influence on state coffee policies; but also that a cutback of coffee production or a decline in the world market price would have significant negative effects on agricultural income and employment. The response of producing states under these conditions was to support coffee production internally and turn to collective action writh other producers to maintain the world price. The new "free market" period which began on July 4, 1989, revealed die high costs of a regulatory regime which tended to reinforce coffee export dependence.

    The Economic Impacts of Deregulation Undermine the Free Market: 1989-93

    As the world coffee market entered the new free market period in July 1989, prices immediately crashed. Producing countries were anxious to unload the stocks that had been accumulating under the quota system, and TNCs demanded prices closer to those which had prevailed in the non-quota markets. At these low prices, the TNCs were eager to stock up on coffee; and within the first few months of the free market, there was a massive transfer of coffee stockpiles from producing to consuming countries, at bargain prices. Table 2 shows the extent of the price crash. For most market participants, the failure of producers and consumers to agree on a new ICA by March 1989 signaled that the end of the ICA regime was

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 134 BERKELEY JOURNAL OF SOCIOLOGY

    near, and most coffee buyers held back on new purchases, expecting the price to fall Consequently the price began to slide after ApriL The June- July price decline shows the magnitude of the crash which immediately followed suspension of the quotas, and the April-October decline shows the total magnitude of the loss which can be attributed to the end of the quota system- prices were cut in hah0 within six months.

    The internationalization of coffee production had solidified TNC control over the core markets, which they already dominated. As a result of this increased control, they were able to maintain the level of the retail prices which they charged for their final processed coffee products in the consuming markets, despite the crash in world market prices for their major raw material input, green coffee. The overall average level of retail prices in ICO consuming member markets declined by only 1% in the two years following the end of quotas, despite a decline of over 50% in green coffee prices (ICO EB 3338/92).

    Producing states were able to partially oflset the impact of the price crash in the short run. By exporting their accumulated stockpiles of coffee, they cut the costs of maintaining these stocks, and earned some additional income through the increased volume of exports. Still, the total coffee earnings of all producing countries fell from $9.2 billion in coffee year 1988-89 to $6.7 billion in coffee year 1989-90, a 27% decline in earnings despite a 13% increase in the volume of exports. Meanwhile, the TNCs used this increased volume to build up their coffee stockpiles, so that they were in no hurry to buy additional coffee unless the price was right. These huge stockpiles held by the TNCs exerted further downward pressure on world market prices (ICO; F.O. Licht, 4 April 1990: 22; ED&F Man 3 August 1989; World Coffee & Tea, September 1989: 8-13).

    Table 3 shows the effects that the coffee price crash had on the ten countries which depended on coffee for more than 30% of their total export earnings. El Salvador, Colombia, Guatemala, Rwanda and Tanzania were the arabica producers who were able to increase the volume of their coffee exports to partially oflset the decline in prices; but all of them, even El Salvador, with a more than 50% increase in volume, lost money. Robusta producers Uganda and Madagascar were hurt the worst. Those countries which were somewhat less dependent on coffee were able to increase exports of other goods to partially oflset the effects of the M in coffee prices, but the most dependent countries suffered a significant decline in total export earnings.

    Prices recovered somewhat in 1990, after the selling frenzy by producers had subsided, with the ICO Indicator price averaging just over 71 cents for the year. But beginning in April 1991, and continuing through 1992, prices

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 135

    Table 2: The 1989 Coffee Price Crash

    Type Average Monthly Price Percent Change of Coffee April June July October June-July Apr.-Oct.

    Colombian 151 134 94 74 -30 -51 Other mild 144 125 88 69 -30 -52 Brazilian 131 115 79 60 -31 -54 Robusta 91 83 65 54 -22 -41 ICO Indicator 118 105 77 61 -27 -48

    Source: International Coffee Organization, monthly price series. Prices in US cents per pound. Colombian and Brazilian prices are for New York delivery; other milds and robusta prices are averages of delivery prices for several major importing ports. ICO Indicator price is a weighted average of the other milds and robusta prices. See note 10 for definition of coffee types.

    Table 3: Effects of the Price Crash on the Ten Most Coffee-Dependent Countries

    Coffee % Change Coffee Exports %Change Country and Dependence 1988/89-1989/90 Total Export Coffee Typea 1988-89b Volume Value Value, 1988-89

    Uganda(R) 96.1 -24 -54 -31 Burundi(O) 82.9 -8 -41 -50 Rwanda(O) 80.1 31 -14 -12 Ethiopia(B) 61.5 -1 -34 15 El Salvador(O) 59.6 53 -8 -8 Tanzania(C) 35.7 14 -36 -7 Guatemala(O) 32.4 22 -18 8 Madagascar(R) 31.9 0 -48 14 Colombia(C) 30.5 34 -18 14 Nicaragua(O) 30.4 5 -30 5

    aCoffee Types: R = robusta, O = other milds, B = Brazilian, C = Colombian milds. See note 10 for definition of coffee types. Value of coffee exports as a percentage of value of all exports, for calendar years 1988 and 1989. cCoffee years, Oct. 1 - Sept. 30. Sources: Coffee dependence from UNCTAD, Handbook of International Trade and Development Statistics 1991; change in volume and value of coffee exports from ICO EB 3338/92; change in total export value from IMF, Direction of Trade Statistics Yearbook 1993.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 136 BERKELEY JOURNAL OF SOCIOLOGY

    resumed their slide, reaching historically low levels by the summer of 1992, when the ICO Indicator price fell below 50 cents per pound for the first time since 1972. In real terms, the price of coffee was lower than it had been for over 100 years. Meanwhile, the TNCs continued to maintain their retail prices at pre-1989 levels, reaping windfall profits of $2-3 billion per year.

    This prolonged period of low prices, combined with protests from coffee growers, did more to deregulate the early stages of the chain than the World Bank had done through the 1980s. World Bank structural adjustment programs during the 1980s pressured many Third World states to devalue their currencies, open their economies, and decrease their intervention in the economy. These changes did not have a major impact on the coffee sectors of most Third World countries, and the Bank did not directly target the coffee sectors for reform, because the world market was being regulated by the ICAs. But after the 1989 price crash, coffee growers began to protest the low prices they were receiving for their coffee, and states became increasingly concerned about the effects that prolonged low prices would have in the coffee growing regions. Most coffee-exporting states significantly reduced their export taxes after 1989, although this led to a decrease of government revenues. In African and Asian countries where coffee was purchased and exported by state marketing boards, the boards were either abolished or their coffee monopolies were ended, and private exporters were allowed to compete with them These changes allowed growers to receive a larger share of the export price than they had before 1989, and somewhat cushioned the blow of the price crash. But these changes also meant that the state could no longer afford to provide the range of services that many growers had come to expect, such as agricultural extension and credits for imported fertilizers or for renovation of coffee trees. These actions by producing states mitigated or delayed the full impacts of the price crash on growers, but as the period of low prices dragged on, their impacts began to 'trickle down" to the coffee growers, who responded by cutting back on their variable costs: labor, fertilizers, and pesticides used to maintain their coffee trees.

    Because of the biology and economics of coffee production, these decisions by growers had little immediate effect on output. The "coffee year" begins on October 1 of each year, near the beginning of the first harvest, and runs through September 30 of the following year, with a second harvest in the spring. Thus, as the first effects of the end of quotas began to be felt among growers, the 1989-90 crop was already on the trees, and the flowering of the 1990-91 crop was beginning. And because of producing states1 responses to the price crash, growers in many countries did not feel the full effects of the low world prices until 1991, when the 1991-92 crop was already on the trees. Reduced maintenance could have had a minor effect on the output for 1990-91 and a somewhat larger effect for 1991-92, but

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 137

    would not produce a serious effect on world supply until at least 1993. As the reports of reduced maintenance reached the major participants in the world market, producing states, large growers, exporters, and some traders and roasters began to warn that continued low prices threatened the quality, and eventually the quantity, of coffee which would be available {Tea and Coffee Trade Journal, June 1992: 26). But in the short run, these effects were minimal, and TNCs and consuming states generally saw no reason to be dissatisfied with the functioning of the free market, or to consider renewed political regulation.

    But on the ground in the producing countries, the effects of the prolonged low world market prices were devastating. Consider just one example from one of the less coffee-dependent countries shown in Table 3, the country with the largest economy of these ten, and the one which was arguably the least affected by the price crash-Colombia.

    The municipality of Lbano in the department of Tourna lies in the Central Cordillera of the Andes, in the largest coffee-producing region of Colombia. Lbano was once the fifth-largest producing municipality, in terms of total volume, in the country, with most of this coffee grown on small peasant holdings and family farms. As the prices peasants were paid for their coffee began to decline, especially after 1990, many were forced to sell their mules to pay old debts and supplement their subsistence production.

    Then, in April 1992, the coffee berry borer worm invaded. This is a parasite that burrows into the coffee cherry to lay its eggs on the coffee beans, thereby destroying them. By this time most of the growers had no money for the pesticides needed to fight the coffee borer, and it rapidly spread to all firms in the area. Already in debt, with no way to repay it, and with their coffee trees dying, many began to abandon their firms. Some went to the towns to try to scratch out a living in the informal sector; others left for the jungle to go to work in the drug trade. The repercussions from this migration drove many other small businesses in the municipality out of business. The community organized to demand assistance from the government, but the government officials had few resources with which to provide assistance, and were faced with similar demands from all sides. To the leaders of the community movement, the officials seemed as sensitive as marble statues to their plight.

    The guerrillas of the ELN took control of much of the rural zones surrounding Lbano. None of Colombia's various guerrilla groups, some of which date back to the early 1960s, has ever had a very significant presence in the coffee-growing zones of Colombia, because they have always been among the more affluent rural areas of the country. But the guerrillas were the only force capable of stopping the banks from confiscating peasants'

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 138 BERKELEY JOURNAL OF SOCIOLOGY

    farms when their debts were declared in default. Some of the youth from the area went to join the guerrillas, because they at least offered a steady salary (El Tiempo, 17 April 1994). This story, with variations, could be repeated thousands of times over in the coffee-growing zones of countries around the world.

    Colombia is the best example of a country with state coffee policies controlled by large growers and exporters which has always had the capacity to respond effectively to such situations. Its Federacin Nacional de Cafeteros (FNC) has consistently followed both a countercyclical internal policy goal, collecting heavy taxes while world prices were high, which enabled it to maintain high internal prices for coffee, and in effect provide subsidies to growers, when world prices were low; and an external policy goal of political regulation of the world market to increase and stabilize prices (Cardenas, 1991). The FNC has used its income to invest in infrastructure in coffee-growing zones and to fund major research efforts and extension services. In the 1970s, it developed high-yielding varieties of coffee and provided credit to growers so that they could replace their older varieties. In the 1980s, when coffee leaf rust struck in Colombia, it again developed new rust-resistant strains, and again provided loans to support conversion to the new variety. It has maintained an impressive extension effort to educate growers about proper cultivation techniques. Through 1990 and 1991, the FNC cut export taxes and used its accumulated reserves to subsidize growers. But as the period of low prices dragged on and its reserves dwindled, it was forced to lower prices to growers and cut back on its services. By 1993, it was in such desperate straits that it was forced to begin selling its interests in the Banco Cafetero and the Colombian shipping line, Flota Mercantil Grancolombiano, to obtain money to pay its debts (F.O. licht, 15 March 1993; Latin American Economy and Business, May 1993: 27; Osrio, 1994).

    But through 1992, world production of coffee remained higher than the stagnant levels of world consumption. TNCs maintained their large stockpiles of coffee, and these large stocks in the consuming countries depressed world market prices. By 1991, coffee growers worldwide were complaining that the prices they were receiving were below their costs of production.13 Producing states had liberalized their coffee sectors and exhausted their reserves, and were less able to cushion their growers from

    13Landell Mills Commodity Studies estimated, based on an exhaustive study of production costs in most major coffee producing countries, that most growers were receiving about 50-75% of their annual costs of production (i.e., not including the initial investment in planting the trees). This percentage was significantly lower for some African countries with poor transportation systems and thus high transport costs (Landell Mills, April 1992).

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 139

    the low prices, or from any future fluctuations in the market. Few new investments were being made in coffee, and in some countries, growers were beginning to uproot their coffee trees to plant other crops. As the domestic economic impacts of low world prices began to spread and become more severe, producing states increasingly turned back to attempts to construct a new international regulatory regime.

    Formation of the Coffee Cartel

    Early efforts to organize a coffee cartel were sporadic and uncoordinated; all producers had been caught off guard by the severity of the price crash. Uganda, the most coffee-dependent country and one of the most severely affected, began discussions with other producers, particularly in Africa, about the possibility of forming a cartel, almost as soon as the quotas had ended ( World Coffee & Tea, December 1989: 27; Januaiy 1990: 18). The five Central American producers, who had played a key role in the demise of the ICA and had expected to benefit from a free market, signed an agreement in December 1989 to withhold 15% of their exportable production (F.O. Licht, 24 Januaiy 1990: 133-4). But without the participation of Brazil and Colombia, this initiative had little chance of affecting the level of world prices. And Brazil and Colombia were in no mood to cooperate to ease the suffering of the dissident producers who, in concert with the US, had been responsible for the end of the quota system (F.O. licht, 2 June 1989: 308; 20 October 1989: 30). None of the consuming states felt that there was much to be gained from a new round of negotiations so soon after the 1989 impasse. This situation continued through 1990 and 1991-producers made feeble attempts to reduce their exports, and consumers displayed little interest in new negotiations. The most that they could agree on was to continue to extend the 1983 ICA (without economic clauses) to keep the ICO alive as a forum for future discussions.

    By 1992, the consuming countries, even the US, were beginning to be concerned about the economic impacts of continued low prices on the producing countries. The TNC coffee importers and roasters were beginning to report that the quality of coffees available worldwide was declining, as the reduced maintenance began to have an effect {Tea and Coffee Trade Journal, January 1992: 54). The US was being pressured by Colombia, as it played an increasingly central role in the Bush Administration's "war on drugs" in the Andean region; the Colombians pointed out that they were losing more income through low coffee prices than they were gaining in aid from the US anti-drug efforts.14 They also

    14"Drug Policy in the Andean Nations," Joint hearings of the Senate Judiciary Committee and the Caucus on International Narcotics Control. This was the main topic at the second

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 140 BERKELEY JOURNAL OF SOCIOLOGY

    pointed out that the EC had exempted the Andean countries from import tariffe on coffee in 1990, to help support their anti-drug efforts (World Coffee & Tea, February 1991: 22).

    Through the spring and summer of 1992 and into early 1993, the momentum toward a new ICA seemed to be picking up. Producers grudgingly accepted the consumers' concept of a ''universal quota," which would cover all exports regardless of their destination. Then the consumers, led by the US, refused to play any role in the policing of the quotas or to restrict their imports from non-member exporters. Producers argued that this would remove incentives for coffee producing countries to join the ICO, and would create the conditions for a new two-tier market. A compromise was reached whereby the consumers would be free to accept all coffee imports, but would report sufficient information to the ICO so that it could determine whether producers were exceeding their quotas. Next, the producers accepted consumers' demands for a selectivity system, which would adjust the quotas for different types of coffees depending on world demand for them But the negotiations ultimately broke down in March 1993 over the issue of "continuity," or the power of the ICO Coffee Council to review and revise the selectivity mechanism Consumers wanted to minimize this power to make the agreement more "market-oriented;" producers feared that without this power, a new ICA would institutionalize the prevailing low level of world market prices for the duration of the agreement (F.O. licht, 7 July 1992; 18 December 1993; 19 April 1993).

    The negotiators for the producing states felt betrayed by the consuming states' representatives. They felt that they had given in to most of the consumers' demands, but had gotten little in return. The US, as the clear leader of the consumers in this round, came in for particularly harsh criticism, Myles Frechette, the head of the US delegation, admitted that the producers had unilaterally conceded to more than 90% of consumers' demands, but defended consumers' intransigence by stating that, "according to the exporters' scorecard, it was the importers' turn to make concessions, regardless of the magnitude of the concessions, and whether or not the proposals were consistent with a new market-oriented agreement." (F.O. licht, 26 February 1993: 186) Colombian negotiators in particular felt that they had had a firm commitment from the US to negotiate in good faith toward a new agreement, but that the US government had put the interests

    session, January 18, 1990, at which the Ambassadors of Colombia, Bolivia, and Peru testified It is also mentioned in an exchange of letters between President Bush and Colombian President Barco in September 1989, reprinted as ICC 54-1, and in a letter from Bush to Colombian President Gaviria in August 1992, stating the US commitment to negotiating an new ICA (F.O. Licht, 7 October 1992; World Coffee & Tea, August 1992: 27).

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 141

    of its own coffee TNCs ahead of this commitment at crucial stages of the negotiations (F.O. licht, 19 April 1993). 15

    This sense of betrayal unified the producers, and after March 1993 momentum toward unilateral action by the producers began to build rapidly. Brazil and Colombia agreed to freeze their current stocks, as well as continuing to retain 10% of their exports. The Central Americans, including even recalcitrant Guatemala, agreed to retain 15% of their exports. And the Inter-African Coffee Organization called a meeting for August in Kampala to discuss formation of the cartel. Agreement was reached August 17, 1993 in Kampala, by 24 exporting countries, soon joined by Indonesia, bringing their share of world green coffee exports to 85%. They formed the Association of Coffee Producing Countries (ACPC), and agreed to withhold 20% of their exportable production from the market. This agreement went into effect with the new coffee year on October 1, 1993 (Financial Times, 18 August 1993; Wall Street Journal 18 August 1993; F.O. licht, 27 September 1993; 11 January 1994).

    The retention plan was based on the level of the ICO Indicator price, and set a price floor of 750/tt>. If the price stayed above 750 for 20 consecutive days, the retention percentage would be cut from 20% to 10%. If the price stayed above 800, the retention would be cut to zero; and if the price stayed above 850, countries would be allowed to sell from their retained stocks (Journal of Commerce 18 August, 1993; F.O. Licht, 12 October, 1993). Given that prices were generally over $1.00/Ib. through the 1980s under the quotas, this seemed like a very moderate goaL

    Even though the ACPC controlled about 85% of coffee exports, some major producers remained outside the agreement, and this created problems for its potential success. Mexico is the fourth largest producer and by far the largest one which did not join the ACPC, because it did not want to jeopardize the passage of NAFTA (Journal of Commerce, 8 July 1993). Ironically, the peasants of Chiapas, where about half of Mexico's coffee is grown, did not think that NAFTA was such a great deal, and the Zapatista rebellion which began on January 1, 1994, severely cut into Mexico's production, minimizing the effects of Mexico's non-membership on the amount of coffee available outside the ACPC (Journal of Commerce, 20 January 1994; 2 May 1994). The Zapatista rebellion was in part a response

    15There was also a change of US administration at a crucial stage just before the negotiations broke down. But US negotiators insisted that they had full instructions from their government and that the change would not hamper their participation. This was probably accurate; Clinton's early focus was almost exclusively on domestic matters, and his trade policies, coordinated by the USTR, were, if anything, more neoliberal than his predecessor's.

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • 142 BERKELEY JOURNAL OF SOCIOLOGY

    to the structure of the coffee commodity chain in Chiapas. Peasant growers who produce most of the coffee there were unhappy with the low prices they had always received for their coffee, from agents of the large coffee processors and exporters who came in from outside the region. This sense of exploitation was obviously heightened by the low world market prices which prevailed through the early 1990s. But the rebellion had much broader and deeper roots than just this: racial discrimination, widespread poverty and economic exploitation, and a fear that the opening of Mexico's markets under NAFTA would devastate peasant producers of food crops.

    The other large bloc of producers which did not join the ACPC were the Asian countries, particularly Thailand and Vietnam {Journal of Commerce, 9 August 1993, 24 August 1993). These countries were '"non-traditional" exporters of coffee-they had only begun to extensively plant coffee in the 1980s. Their coffee exports were growing rapidly in the early 1990s, and agreeing to the retention would have hampered their efforts to expand and diversify their exports. Further, as new exporters, they had no history or experience of cooperation with other 'faaditionaT exporters in the ICAs, and no real commitment to the principle of producers' collective action. The largest Asian producer, and the third largest producer overall, Indonesia, was reluctant to participate, because of concerns that its ability to expand coffee exports had been stifled under the quotas. This was why it had joined with the "other milds" group in 1989 to oppose a new ICA. But Indonesia experienced some solidarity with other producers, and finally did join the ACPC, and began to comply with the retention on April 1, 1994.

    World market prices began to rise in the fell of 1993, even before the retention went into effect, and they continued to rise through the spring of 1994. While the retention undoubtedly decreased the supply of coffee on the market, its effects were multiplied by declining production worldwide, the result of three years of reduced maintenance and uprooting of trees. During the second week of May 1994, futures prices in both New York and London hit their highest levels since the end of the quotas five years before {New York Times, 14 May 1994: 28). In early April, the ICO Indicator price passed 850/Ib., and the ACPC announced that it would begin a controlled release of 50% of the retained stocks. As prices continued to rise in May, it announced plans to release the remaining stocks {New York Times, 14 May 1994; Financial Times, 25 May 1994). Then, in late June and early July, two severe frosts struck the coffee-growing regions of BraziL The true extent of the damage would not be known until the 1995-96 harvest was underway, but early estimates were that as much as one-third of the 1994- 95 crop had been destroyed. By mid- July, the New York futures price for arabicas had soared to over $2.70/B).

    This content downloaded from 169.234.103.50 on Tue, 24 Mar 2015 18:14:48 UTCAll use subject to JSTOR Terms and Conditions

  • TALBOT: COFFEE COMMODITY CHAIN 143

    Producing states and coffee growers were able to begin to recoup some of the losses they had suffered during the preceding four years. Of course, the TNC coffee importers who held large stocks of coffee in late 1993 were not unhappy with the price trend either, as they reaped huge windfall profits. Part of this enormous price increase was driven by speculation, and some of this speculation was also being done by these same TNCs; thus prices soon began to fall back from this peak, stabilizing around $2.00/Ib. in the Fall before beginning to slide further in November and December of 1994. In January 1995, the ACPC decided to once again begin retaining coffee to halt this slide. Through 1995, producers were not as unified as they had been in late 1993, and the results of the retention were more mixed. Prices declined slowly through 1995, but generally remained above pre-1989 levels; at these prices, producers had less incentive to comply fully. For 1995-96, worldwide exportable production will be below demand for the third straight year, and prices are likely to remain relatively high. The next real test of the solidarity of the ACPC will be whether its members can colle


Recommended