The ‘Latte Revolution’? Regulation, Markets and Consumption in the Global Co?ee Chain

STEFANO PONTE * Centre for Development Research, Copenhagen, Denmark
Summary. — Co?ee is a truly global commodity and a major foreign exchange earner in many developing
countries. The global co?ee chain has changed dramatically as a result of deregulation, new consumption
patterns, and evolving corporate strategies. From a balanced contest between producing and consuming
countries within the politics of international co?ee agreements, power relations shifted to the
advantage of transnational corporations. A relatively stable institutional environment where
proportions of generated income were fairly distributed between producing and consuming countries
turned into one that is more informal, unstable, and unequal. Through the lenses of global commodity
chain analysis, this paper examines how these transformations a?ect developing countries and what
policy instruments are available to address the emerging imbalances.
2002 Elsevier Science Ltd. All
rights reserved.
Key words — co?ee, commodity chains, development, globalization, regulation
1. INTRODUCTION
Every day, about 2.25 billion cups of co?ee are consumed in the world (Dicum & Luttinger, 1999, p. IX).
Yet, the act of––and symbols attached to––co?ee drinking are not the same as they were 20 years ago.
New consumption patterns have emerged with the growing importance of specialty, fair trade, and organic
co?ees. Co?ee bar chains have spread dramatically, although the relative co?ee content of the ?nal
consumption ‘‘experience’’ in these outlets is extremely low. 1 Co?ee bar chains sell an ambience and a
social positioning more than just ‘‘good’’ co?ee. In short, the global co?ee chain has gone through a
‘‘latte revolution,’’ 2 where consumers can choose from (and pay dearly for) hundreds of combinations
of co?ee variety, origin, brewing and grinding methods, ?avoring, packaging, social ‘‘content,’’ and
ambience. At the same time, international prices for the raw product (‘‘green’’ co?ee) are the lowest
in decades. Co?ee industries in developing countries are in disarray. Co?ee farmers are losing a source
of livelihood. This paper explores this contradiction through the analysis of the changing features of
the global co?ee-marketing chain. It examines the consequences of the shift that has occurred in the
last two decades in the regulatory framework at the international level––with the end of the quota
system managed by the In
ternational Co?ee Organization (ICO). It also explains how market liberalization and deregulation in
producing countries has decreased their capability of controlling exports and building up stocks,
therefore weakening their market power. Finally, the paper examines how new consumption patterns and
changing strategies by key corporate actors (adoption of supply-managed inventory, consolidation,
branding) a?ect other actors in the chain. These major shifts in international and domestic regulation,
consumption, and corporate behavior are assessed in relation to the organizational features of the
chain, its mode of governance, the ownership characteristics at various ‘‘nodes,’’ and the distribution
of income along the co?ee chain. The next section explains the main features of the global commodity
chain (GCC) analysis (also known as ‘‘value-chain analysis.’’) Section 3 lays out the fundamental
characteristics of the
World Development Vol. 30, No. 7, pp. 1099–1122, 2002
2002 Elsevier Science Ltd. All rights reserved
Printed in Great Britain 0305-750X/02/$ – see front matter PII: S0305-750X(02)00032-3
www.elsevier.com/locate/worlddev
*I would like to thank Henry Bernstein, Niels Fold, Deepa George, Gary Gere?, Peter Gibbon, Michael
Friis Jensen, Poul Ove Pedersen, and two anonymous referees for helpful comments on earlier drafts of
this paper. I am also thankful to the Danish Social Science Research Council and the Centre for
Development Research, Copenhagen for funding the research project under which this paper was generated.
Final revision accepted: 20 February 2002. 1099
global co?ee chain. The following section analyses some of the consequences of the switch in co?ee
trade ‘‘regimes’’ that took place starting in the late 1980s. Section 5 focuses on market power and
corporate strategies in the current con?guration of the global co?ee chain. This is followed by the
examination of how co?ee consumption is evolving in the industrialized economies (the ‘‘latte
revolution’’). Section 7 assesses the insights o?ered by the restructuring of the global co?ee chain to
wider debates in the GCC literature. The ?nal section assesses what the co?ee study has to say about
the role of commodity trade in development and provides several policy options to address the emerging
imbalances in the global co?ee chain.
2. GLOBAL COMMODITY CHAIN ANALYSIS
The main methodological instruments used in this paper are drawn from GCC analysis. The GCC approach
was developed by Gere? and others within a political economy of development perspective. In this body
of work, the international structure of production, trade, and consumption of commodities is
disaggregated into stages that are embedded in a network of activities controlled by ?rms and
enterprises. The systematic study of commodity chains seeks to explain the spatial organization of
production, trade and consumption of the globalized world economy (Gere?, Korzeniewicz, & Korzeniewicz,
1994, p. 2). A commodity chain in this context is seen as ‘‘a network of labor and production processes
whose result is a ?nished commodity’’ (Hopkins & Wallerstein, 1986, p. 159). Speci?c processes within a
commodity chain are represented as ‘‘nodes’’ linked together in networks. Therefore, we can see a
commodity chain as ‘‘a set of interorganizational networks clustered around one commodity or product’’
(Gere? et al., 1994, p. 3), in which networks are situationally speci?c, socially constructed, and
locally integrated. Gere? identi?es four dimensions of GCCs: the input–output structure, the
geographical coverage, the governance structure (Gere?, 1994, p. 97), and the institutional framework
through which national and international conditions and policies shape the globalization process at
each stage in the chain (Gere?, 1995). The input–output structure and the geographical coverage of GCCs
have been used
mainly descriptively to outline the con?guration of speci?c chains. The governance structure has so far
received the most attention, since this is where the key notions of entry barriers and chain co-
ordination appear in the analytical framework, and where the distinction between ‘‘producer-driven’’
and ‘‘buyer-driven’’ GCC governance structures is introduced. Producerdriven chains are usually found
in sectors with high technological and capital requirements, where capital and proprietary know-how
constitute the main entry barriers (automobiles, aircraft, computers). In these chains, producers tend
to keep control of capital-intensive operations and subcontract more labor-intensive functions, often
in the form of vertically integrated networks (Gere?, 1994). Buyer-driven chains are found in generally
more laborintensive sectors, where information costs, product design, advertising, and advanced supply
management systems set the entry barriers (garments, footwear). In these chains, production functions
are usually outsourced and key actors concentrate on branding, design, and marketing functions (Gere?,
1994). The producer-driven versus buyer-driven dichotomy, while useful as a point of departure, should
not be strictly and statically interpreted. First, some commodity chains may exhibit the tendency to
move from one category to the other. In some producer-driven chains such as automobile, computer, and
consumer electronics, producers are increasingly outsourcing portions of component manufacture.
Sometimes, they even outsource supply-chain logistics and ?nal assembly, and keep control of promotion
and marketing of the brand names on which market access is based––a peculiar trait of buyer-driven
chains. Second, this dichotomy does not adequately explain some of the characteristics of service
chains and some of the changing features of chain governance that relate to e-commerce operations. This
has led Gere? (2001a,b) to explore the possibility that another category of governance is emerging, the
‘‘infomediary-driven’’ chain. He also entertains two other possibilities: (a) that ecommerce and the
internet may accelerate the tendency to make all chains more buyer driven; or (b) that e-commerce may
just be captured by established leaders in both producer-driven and buyer-driven chains. The fourth
dimension of GCCs, the institutional framework surrounding the chain, is used to delineate the
conditions under which key (or ‘‘lead’’) agents incorporate subordinate agents
1100 WORLD DEVELOPMENT
through their control of market access and of information––both technological and regarding markets
(Gere?, 1999b). Under the rubric of ‘‘institutional framework’’ Gere? also discusses how subordinate
participation in a GCC can provide indirect access to markets at lower costs than individual small-
scale producers would face, and how technological information and learning-by-doing allow (the more
favored) producers to move up the chain hierarchy. This suggests that participation in a GCC is a
necessary, but not su?cient, condition for subordinate agents to upgrade, and one which involves
acceptance of terms de?ned by key agents as a condition for participating in the chain, especially for
those aiming to progress toward higher (technology, value added) positions in the chain (Gere?, 1999a,
p. 39; see also Gibbon, 2001a; Humphrey & Schmitz, 2000; Tam & Gere?, 1999). The GCC approach has
generated a number of case studies. Although GCC theory originally centered on analyses of the
manufacturing and service sectors, 3 it has recently started to be applied to agro-food systems as
well. 4 Agricultural commodities tend to fall into the category of buyer-driven chains, 5 in which
large retailers in industrialized countries, brandname merchandisers, and international trading
companies are the key actors in setting up decentralized networks of trade in developing countries.
Because of the changes in distribution and retailing in industrialized countries since the 1980s,
agricultural production and trade have involved an increasingly heterogeneous combination of ?rms,
types of ownership, size, and relative access to markets. Therefore, a commodity-based analysis can
provide better insights on the emerging con?gurations of agricultural trade than a sectoral approach
(Raynolds, 1994, pp. 143–144; see also Raikes, Jensen, & Ponte, 2000). The GCC approach emphasizes the
power of di?erent constellations of lead ?rms and how interactions between these ?rms determine some of
the speci?c organizational features of trade. Analyses of commodity markets (including co?ee) based on
neoclassical economics consider trade in isolation from investment, ?nance and other relations between
parties. They also assume that both participants and transactions are separate and independent from
each other. These constraining assumptions generate trade patterns that are determined by each
country’s endowments of production factors (see Raikes et al., 2000). Other contributions,
while acknowledging imperfect competition and information asymmetries, still tend to focus on the
reasons for the existence of institutions (notably transaction costs and barriers to entry) and
perceive institutions primarily as regrettable departures from free trade (Raikes et al., 2000).
Finally, political science approaches employing a rules-based version of neoinstitutional economics
(for co?ee, see Bates, 1997) examine institutions, but have relatively little to say about the internal
organization of commodity trade. The analysis of the co?ee-marketing chain is particularly important in
understanding the political economy of development for a variety of reasons. First, over 90% of co?ee
production takes place in developing countries, while consumption happens mainly in industrialized
economies. 6 Therefore, the production–consumption pattern provides insights on North– South relations.
Second, for most of the post-WWII period co?ee has been the second most valuable traded commodity after
oil. 7 Third, attempts to control the international co?ee trade have been taking place since the
beginning of the 20th century, making coffee one of the ?rst ‘‘regulated’’ commodities. Fourth, a
number of developing countries, even those with a low share of the global export market, rely on co?ee
for a high proportion of their export earnings. Co?ee is a source of livelihoods for millions of
smallholders and farm workers worldwide. 8 Fifth, producing country governments have historically
treated co?ee as a ‘‘strategic’’ commodity; they have either directly controlled domestic marketing and
quality control operations or have strictly regulated them––at least until market liberalization took
place in the 1980s and 1990s. Not all aspects and ‘‘nodes’’ of the co?ee commodity chain are covered in
this paper, for obvious space limitations. The paper aims at mapping the general development of the
chain from the producer to the retail levels and focuses on selected global issues. Detailed analyses
of domestic and local experiences can be found elsewhere (Losch, 1999; Pelupessy, 1999; Ponte, 2002a).
3. THE GLOBAL COFFEE CHAIN
Co?ee goes a long way and changes many hands from bean to cup (see Figure 1). Historically, Brazil and
Colombia have been top world co?ee producers. In the 1990s, however,
‘‘LATTE REVOLUTION’’ 1101
the situation changed with the fast growth of co?ee production in Vietnam (see Table 1),
which has contributed to the dramatic drop in international co?ee prices of the late 1990s
Figure 1. General structure of the global co?ee-marketing chain. Note: With market liberalization,
dotted links are disappearing.
1102 WORLD DEVELOPMENT
(see below). In 1999/2000 Vietnam replaced Colombia as the world second largest producer. The ICO
categorizes exports by type of co?ee. As we can see in Table 2, Mild Arabica coffees are divided into
‘‘Colombian Milds’’ and ‘‘Other Milds.’’ Colombian Milds comprise co?ees produced in Colombia, Kenya
and Tanzania. The main players in the Other Milds category are Guatemala, Mexico and India. ‘‘Brazilian
Naturals’’ basically consist of Hard Arabicas from Brazil and Ethiopia. The last
category includes Robusta co?ees from all origins. Here, Vietnam is by far the main producer, but
C^ooted’Ivoire,Indonesiaand Uganda are also major players. 9 In normal supply conditions, market prices
are highest for the Colombian Milds category, followed by Other Milds, Brazilian Naturals, and ?nally
the wide spectrum of Robustas (McClumpha, 1988, p. 14). Among consuming countries, Scandinavian
countries (which have the highest level of consumption per capita in the world) and Germany prefer Mild
Arabica co?ees in their blends. Robusta co?ee is a key component in espresso blends and darker roasts,
therefore important in Southern Europe. The US and UK markets prefer lighter roasts in general, but
require a wide spectrum of qualities. Historic trading links are still important in shaping the
international co?ee trade. A sizeable proportion of East African co?ee ?nds its way to Germany and the
UK. France maintains close links with C^ooted’IvoireandotherFrancophonecountries. Dutch trading links
with Indonesia remain important as well (McClumpha, 1988, p. 12). Most international co?ee trade
consists of ‘‘green’’ co?ee packed in 60-kg bags. 10 Green co?ee is available to buyers either directly
from its origin or via the spot markets in the United States and Europe. In theory, physical co?ee can
also be accessed to via the futures market, but this happens only rarely. The purpose of these markets
is to provide hedging against risk rather than being a supply source (McClumpha, 1988, p. 8). Two sets
of international
Table 2. Exports by major ICO-exporting member to all destinations (60-kg bags)
March 2000–February 2001
Colombian Milds 11,539,133 Colombia 9,499,242 Kenya 1,214,199 Tanzania 825,692
Other Milds 28,059,771 Guatemala 4,771,031 Mexico 4,659,096 India 4,460,021 Honduras 2,915,806
Brazilian Naturals 19,999,823 Brazil 18,154,618 Ethiopia 1,834,205
Robustas 29,008,946 Vietnam 11,958,220 C^ooted’Ivoire5,793,381 Indonesia 5,248,067 Uganda 2,641,651
Total 88,607,673
Source: ICO (2001a).
Table 1. Total production of top 10 ICO-exporting members (ranked by 1999/2000 production); crop years
1995/96–2000/01 (thousands of 60-kg bags)a
Crop year commencing
Type of co?ee
1995 1996 1997 1998 1999 2000 (estimates)
Share of world production (1999)
Total 85,647 102,495 95,969 106,508 114,218 112,901
Brazil (A/R) 15,784 27,664 22,756 34,547 32,353 31,100 28.3 Vietnam (R) 3,938 5,705 6,915 6,947 11,264
11,350 9.9 Colombia (A) 12,878 10,876 12,211 11,088 9,336 12,000 8.2 Mexico (A) 5,527 5,324 5,045 5,051
6,442 6,338 5.6 Indonesia (R/A) 5,865 8,299 7,759 8,463 6,014 7,300 5.3 C^ooted’Ivoire
(R)2,5324,5283,6822,0425,4634,1674.8 India (A/R) 3,727 3,469 4,735 4,372 5,407 4,917 4.7 Guatemala
(A/R) 4,002 4,524 4,218 4,892 5,201 4,500 4.6 Ethiopia (A) 2,860 3,270 2,916 2,745 3,505 3,683 3.1
Uganda (R/A) 3,244 4,297 2,552 3,298 3,097 3,200 2.7
Source: ICO (2001b). a A¼Arabica; R¼Robusta.
‘‘LATTE REVOLUTION’’ 1103
co?ee prices are available: (a) ICO-published prices: these are indicators of the physical trade, where
each contract refers to a speci?c quality, origin, shipment, currency and destination; and (b) prices
determined by futures markets: these are short-term syntheses of market fundamentals (production,
consumption and stocks) and technical factors (hedging, trend following, reactions to trigger signals).
Prices in the physical trade of Arabica co?ees from various origins are set as di?erentials in relation
to the futures price quoted at the New York Co?ee, Sugar and Cocoa Exchange. The reference price for
Robusta co?ees is set at the London International Financial Futures and Options Exchange. The
international co?ee market is characterized by relatively low price elasticities of supply and demand
(McClumpha, 1988). Supply elasticities are low in the short run and higher in the long run because it
takes at least two years for new trees to be productive and several others before they reach full
production levels. Therefore, the supply response in the short term is possible only by changing the
quantity of resources used for inputs and labor application, not by increasing the productive area––a
feasible option for annual crops. Demand elasticities are also low, with co?ee demand dropping signi?
cantly only at times of large increases of co?ee prices. The peculiar characteristics of the price
elasticities of supply and demand lead to highly variable prices in the world co?ee market. A situation
of supply shortage results in high co?ee prices without a signi?cant reduction of consumption.
Likewise, supply reacts slowly in the short run while new plantings take place. In the long run, this
leads to a higher than necessary response as new co?ee trees mature. A situation of supply shortage may
then be followed by one characterized by oversupply and low prices. An opposite bust period then
begins––usually lasting longer than the boom period (Daviron, 1993; McClumpha, 1988). Another important
feature of the co?ee market is that consumption tends to increase as income rises, but levels o? at the
highest income levels. For this reason, the co?ee market is considered ‘‘mature’’ due to the relatively
stable and low level of growth of consumption––about 1% per year in 1987–97 (van Dijk, van Doesburg,
Heijbroek, Wazir, & de Wol?, 1998). Low levels of growth of consumption have led roasters and retailers
to invest in product innovation and segmentation in order
to increase value added and also in e?orts to ‘‘cultivate’’ markets where the potential for growth of
consumption is most promising–– especially Eastern Europe and the traditionally tea-drinking countries
of Asia (van Dijk et al., 1998).
4. CHANGING TRADE REGIMES
(a) The international co?ee agreements
Co?ee was one of the ?rst commodities for which control of world trade was attempted, starting in 1902
with the ‘‘valorization’’ process carried out by the Brazilian state of S~aaoPaulo. This process
involved state action to raise the price of co?ee, which was made possible at that time by the large
share of production (between 75% and 90%) of S~aaoPaulointermsofworld co?ee production (Lucier, 1988,
p. 117). PreWWII attempts at manipulating the world co?ee market were all centered around Brazil. In
the post-war period, control schemes involved other Latin American countries as well. The ?rst
international co?ee agreement (ICA) was ?nally signed in 1962 and included most producing and consuming
countries as signatories. Under the ICA regulatory system (1962– 89), a target price (or a price band)
for co?ee was set, and export quotas were allocated to each producer. When the indicator price
calculated by the ICO rose over the set price, quotas were relaxed; when it fell below the set price,
quotas were tightened. If an extremely high rise of co?ee prices took place (as in 1975– 77), quotas
were abandoned until prices fell down within the band. Although there were problems with this system,
most analysts agree that it was successful in raising and stabilizing co?ee prices (Akiyama & Varangis,
1990; Bates, 1997; Daviron, 1996; Gilbert, 1996; Palm & Vogelvang, 1991). The relative success of the
regime has been attributed to various factors: (i) the participation of consuming countries in the
working of the quota system; (ii) the existence of producing countries as ‘‘market units,’’ where
governments were in control of decisions concerning exports; (iii) Brazil’s acceptance of a shrinking
market share that resulted from successive ICAs; and (iv) a common strategy of import substitution in
producing countries, which required maximum mobilization of export earnings––therefore high commodity
prices (Daviron, 1996, pp. 86–89).
1104 WORLD DEVELOPMENT
At the same time, the ICA system was undermined by free-riding and squabbling over quotas. Other
problems were the increasing volume of co?ee traded with (or through) nonmember importing countries (at
lower prices), the continuing fragmentation of the geography of production, and the increasing
heterogeneity of development models––as Brazil and Indonesia moved toward a more export-oriented
industrial strategy (Daviron, 1993, 1996). Furthermore, quotas were relatively stable because they were
costly to negotiate. As a result, the mix of co?ee supplied by producers tended to remain stable, while
in the 1980s consumers in the United States progressively switched from soluble co?ees (that employ a
high proportion of Robusta) to ground co?ees (that use a higher proportion of Arabicas). The rigidity
on the supply side worried roasters, who feared that competitors could get access to cheaper co?ee from
nonmember countries. This undermined their cooperation within the ICA system. Finally, the Cold War
politics of the United States in relation to Latin America had changed in the 1980s. The United States
did not perceive the left in Brazil as a real threat anymore, and the rigidity of quotas meant that the
US administration could not punish its ‘‘enemies’’ in Central America (Bates, 1997, pp. 172–175). The
combined result of these changes led to the failed renewal of the ICA in 1989.
(b) The post-ICA regime
The end of the ICA regime has profoundly a?ected the balance of power in the co?ee chain. From a fairly
balanced contest between producers and consumers within the politics of the commodity agreement, market
relations shifted to a dominance of consuming countrybased operators (including their agents based in
producing countries) over farmers, local traders and producing country governments. This has been
accompanied by lower and more volatile co?ee prices, a higher proportion of the income generated in the
chain retained in consuming countries, and a declining level of producerheld stocks. In relation to
price levels, we can observe that the average real indicator price for 1990–93 was only 42% of the
average of the ?nal four years of ICA activity (1985–88). Even accounting for the price rise of 1994–97
(due to frost and drought in 1994–95 in Brazil, and the speculative hike of 1997), the average
composite price for 1994–97 was still 20% below 1985–88 (Gil
bert, 1998). In 1993, with the establishment of the Association of Co?ee Producer Countries (ACPC), 11
producing countries started again attempts to re-install some control over supply ?ows through an
export retention scheme. However, the process of liberalization of domestic co?ee marketing in
producing countries has made it more di?cult for them to control stocks and the ?ow of exports. In
addition, the scheme was lacking proper monitoring and punitive clauses. Some of the major producers
did not join the scheme, 12 and other member countries withdrew from it in 1998–99. Finally, during the
same season, Brazil exceeded its quota by six million bags. Chronic oversupply, due to technical
innovations and new planting, also contributed to the generally decreasing level of international co?ee
prices experienced in the last decade. Global production in 2000–01 was over 110 million bags, the
third consecutive year in which world output exceeded 100 million bags (see Table 1). Stocks in
consumer markets, the most obvious index of co?ee availability, have been rising (Prudential Securities
Futures Research: Co?ee, June 28, 2000). 13 In May 2000, the ACPC adopted a new retention plan that
started to be operative on October 1, 2000. The plan targeted the retention of 20% of total world
production as long as the 15-day moving average of the ICO composite price indicator was below 95 cents
per pound. Major nonmember producers provided their support to the plan. But, participation in the
retention plan by nonmembers was largely voluntary. Some of these countries stated that retention had
to be cost free. Mexico, for example, aimed at achieving ‘‘export retention’’ by increasing consumption
in government-controlled institutions. Forecasts also indicated a strong increase in production for
2001–02, which would have implicated a further increase in export retention levels. The retention plan
did not include provisions for destroying stocks; therefore, it did not address the fundamental problem
of overproduction. Even though year-to-year ?uctuations of the global production volume are inherent in
the world co?ee market, the longterm trend is generally perceived on the upward side. As a result of
these problems, the retention plan did not succeed in raising prices. The average ICO composite price
indicator went from 69.2 cents per pound in May 2000 (when the retention plan was signed) to 56.4 cents
in October 2000 (when the plan o?cially started). By October 2001 (when the plan was
‘‘LATTE REVOLUTION’’ 1105
abandoned), the average composite price had dropped to 42.2 cents per pound (ICO, 2001a). In the 1990s,
lower co?ee prices have also been accompanied by a higher level of price volatility. Price volatility
is not a new phenomenon in the co?ee market. A major ‘‘traditional’’ factor in volatility is that co?ee
yields are vulnerable to changes in temperature and rainfall, as well as disease. Frosts and drought in
Brazil have normally led to sudden upward movements in co?ee prices. The delay between new planting and
production can also contribute to magnifying the price movements in the co?ee cycle. However, something
qualitatively di?erent took place in the 1990s. The ?nal eight calendar years of ICO activity were
characterized by monthly nominal price variability of 14.8%. This indicator almost doubled to 37%
during 1990–97 (Gilbert, 1998) and then further increased to 43% during 1998–2000. 14 Higher price
volatility in the co?ee market is not only linked to the end of price stabilization mechanisms that
were built in the ICA quota system, but also to increased activity in the co?ee futures market. In
1980, the amount of co?ee traded in the futures market was only around four times the co?ee traded in
the physical market. By the early 1990s, the ratio had risen to 11 times (van Dijk et al., 1998, p.
45). Futures markets allow market transactors to ?x their prices in advance of delivery so that they
can hedge their price volatility risk. Yet, futures contracts lose much of their hedging function when
the price of futures contracts is too volatile. The volatility of futures prices is normally triggered
by market ‘‘fundamen
tals’’ (demand–supply-stock relationships), but is magni?ed by speculative activity. In the last
decade, investment funds have become increasingly active in commodity markets. Because managed funds
operate on the basis of trend-following, ‘‘trigger signals’’ (which may not necessarily be linked to
the actual conditions of supply and demand) tend to cause larger movements in and out of the market
than if the market was operated by the co?ee industry alone (Crowe, 1997). On the one hand, this
additional activity increases liquidity in the market. On the other hand, the increased price
volatility that ensues a?ects those actors who do not have access to hedging instruments–– farmers and
small-scale traders in producing countries (Gilbert, 1996). The collapse of the ICA regime and
increased consolidation in the co?ee industry (see Section 5) have also a?ected the distribution of
total income generated along the co?ee chain. 15 Talbot (1997a, pp. 65–67) estimates that, in the
1970s, an average of 20% of total income was retained by producers, while the average proportion
retained in consuming countries was almost 53% (see Figure 2). 16 Between 1980–81 and 1988–89,
producers still controlled almost 20% of total income; 55% was retained in consuming countries. After
the collapse of ICA in 1989, the situation changed dramatically. Between 1989–90 and 1994–95, the
proportion of total income gained by producers dropped to 13%; the proportion retained in consuming
countries surged to 78%. 17 This represents a substantial transfer of resources from producing to
consuming countries, irrespectively of
Figure 2. Distribution of co?ee income along the co?ee chain (1971–80 to 1989–95), in percentage.
Source: Adapted from Talbot (1997a, pp. 65-67). Note: Co?ee income¼weighted average of retail prices in
ICO member importing countries, expressed in green bean equivalents. Monetary values of total co?ee
income for the periods indicated in this ?gure: 1971–80 (262.6 US cts/lb); 1981–88 (363.5 US cts/lb);
1989–95 (435.8 US cts/lb) (calculated from Talbot, 1997a, pp. 65-67).
1106 WORLD DEVELOPMENT
price levels. The share of income retained by producers in the last two–three years is likely to have
dropped further due to the current situation of oversupply and low prices for green co?ee and the
ability of roasters to maintain retail prices at relatively stable levels. While green co?ee prices
almost halved between December 1999 and January 2001 (see Figure 3), average retail prices in the US
decreased by 4% (ICO, 2001a). This suggests that not only gross margins––but also pro?ts––have
increased for roasters. Finally, the end of the ICA regime meant that the bureaucracy that was needed
to monitor exports and ensure compliance with quota restrictions was no longer needed. This, coupled
with the general switch in economic thinking in the 1980s and 1990s away from public intervention in
markets, led to the dismantling of co?ee boards, institutes and other quasi-governmental bodies that
regulated export sales. As a result, the capability of producing countries to control exports and to
build up stocks has decreased. Producer-held stocks are roughly at the lowest level in 30 years. 18
5. MARKET POWER AND CORPORATE STRATEGIES
In Section 4 I have argued that there has been a general shift of power from producing to consuming
countries in the co?ee-marketing
chain following the end of the ICA regime. Power relations between producers and buyers have also
become more complex. Domestic market liberalization in producing countries entails that states as such
cannot be considered ‘‘market units’’ anymore (Daviron, 1996). Grower organizations have not been able
to substitute governments as organizers of co?ee exports. ‘‘Local’’ exporters have not been able to
raise necessary funds to compete with international traders, and have now either disappeared or allied
themselves with international traders. The general trend has been a strengthening of the position of
roasters vis  aa vis other actors. International traders went through considerable restructuring in
the last two decades. Mid-sized traders with unhedged positions su?ered major losses. They also found
themselves too small to compete with larger ones. As a result, they either went bankrupt, merged with
others, or were taken over by the majors (Prudential Securities Futures Research: Co?ee, June 28,
2000). 19 Therefore, the market has become more concentrated. In 1998, the two largest co?ee traders
(Neumann and Volcaf ee) controlled 29% of total market share, and the top six companies 50% (see
Figure 4). At the same time, prospects are good for smaller and specialized companies that trade in the
specialty co?ee market (high quality and speci?c origins). With some exceptions, there has been little
vertical integration between roasters and
Figure 3. New York co?ee futures prices; nearby contract (US cts/lb) 1994–2001. Source: CSCE (2001).
‘‘LATTE REVOLUTION’’ 1107
international traders (van Dijk et al., 1998, pp. 34–35). 20 The level of concentration in the roaster
market has reached a level even higher than for international traders. Figure 5 shows that the top two
groups combined (NestleeandPhilip Morris) control 49% of the world market share for roasted and
instant co?ees. The top ?ve
groups control 69% of the market. Nestleedominates the soluble market with a market share of 56% (van
Dijk et al., 1998, p. 34). International traders argue that roasters have gained increasing control of
the marketing chain in recent years because of oversupply, increased ?exibility in blending, and the
implementation of ‘‘supplier-managed inventory’’ (SMI).
Figure 4. Green co?ee market share by international trade company (1998), in percentage. Source: van
Dijk et al. (1998, p. 34).
Figure 5. Market share of roasting and instant manufacturing companies (1998), in percentage. Source:
van Dijk et al. (1998, p. 52).
1108 WORLD DEVELOPMENT
It is actually not clear what were the precise motivations behind the adoption of SMI systems by
roasters. One interpretation is that SMI allows roasters to minimize costs by transferring the working
capital costs of inventory holding to trading houses. However, successful management of SMI requires at
least two key conditions: (a) a close balance between supply and demand, or a supply surplus; and (b)
supply conditions of various types and origins of co?ee that do not force roasters to change blends in
ways that would not satisfy theirconsumers. 21 AccordingtoLodder(1997), these factors were not present
when roasters started to apply SMI in 1997. Therefore, they found themselves short of Arabica and
scrambled for co?ee purchases, triggering a panicbuying situation that led to a major price hike. In
later years, however, roasters seem to have been able to implement a more cautious SMI system
successfully. A second interpretation for the adoption of SMI is that roasting companies quoted in
stock markets need to contain the size of inventories and of circulating capital within ‘‘optimal’’
parameters set by ?nancial analysts––large inventories and a high ratio of circulating capital being
normally interpreted as indicators of ine?ciency. When roasters started carrying out SMI, the futures
market was in ‘‘backwardation.’’ In that situation, carrying stocks was costly because forward future
contracts were valued less than nearby positions. Therefore, applying SMI also made sense for roasters
in terms of ?nancial returns. However, the co?ee market has been ‘‘carrying’’ in more recent years,
which means that forward contracts are valued more than nearby contracts. In this situation, if the
costs of stocking (warehousing, ?nance, and insurance) are lower than the spread between positions, the
holder of stocks can make a pro?t just by holding inventory. In sum, outsourcing stock management
during a period of backwardation could be interpreted as an indicator of the increasing power of
roasters over international traders. Sticking to SMI in a carrying market should not however be seen as
a revival of international traders in their power relations with roasters, but rather as a sign of
captivity of quoted roasting companies to the logic of ?nancial markets. In any case, as a result of
the adoption of SMI by roasters––and in combination with market liberalization in producing
countries––international traders have strengthened their supply network. This has taken place through
co
ordination (mostly ?nancing) or vertical integration with local exporters. In some countries,
international traders have moved upstream 22 all the way to domestic trade and in some cases to estate
production (Akiyama, 2001; Losch, 1999; Ponte, 2002a). International traders are likely to continue
investing in operations in origin countries so that they can cater to the needs of major roasters.
Roasters seem to have little interest in vertical integration upstream in the current market
conditions. They seem better o? concentrating on marketing and branding, while leaving supply
management to a network of independent traders––even if, in periods of carrying markets, this means
foregoing a source of pro?t. Some roasters (such as Nestlee)aresaidtosourcenot only from a variety of
international traders, but also directly from some ‘‘local’’ exporters. The aim is to allow these
exporters to compete with international traders in strategic origins. This allows the roaster to be
less dependent on any actor, and especially on major traders. Furthermore, more ?exibility in
developing blending formulas has made roasters less vulnerable to shortages of particular types of co?
ee in recent years. Shortages of Colombian co?ee have been o?set by greater use of Central American
Milds. Another example of substitution is the greater use of Mexican beans in place of Brazilian. The
new technique of steam-cleaning Robusta allows roasters to improve its quality and to substitute poorer
Arabicas with premiumgrade Robustas. Another trend that seems to be emerging in the industry is one
toward the creation of a system of ?rst-line and second-line suppliers, subject to price premia and
discounts. Major roasters tend not to accept co?ee for their blends from countries that cannot
guarantee a reliable minimum amount of supply––in the case of Arabica, around 60,000 tons a year
(Raikes & Gibbon, 2000). As a result, on the one hand, minor producers may become increasingly
marginalized in the future––without necessarily increasing the bargaining power of major producers vis
 aa vis roasters. On the other hand, this has pushed some international traders to be (directly or
indirectly) involved in domestic trade in major producing countries even though these operations may
not be profitable (Uganda, for example), as long as they can satisfy their major roaster clients
(Ponte, 2002a). As a result of these factors, no signi?cant forms of coordination between international
‘‘LATTE REVOLUTION’’ 1109
traders and roasters have emerged so far. The ‘‘traditional’’ market, as long as there is oversupply
and roasters can manage SMI effectively, is likely to remain governed by armslength relationship and/or
by forward contracts of short duration (under 12 months). The next section will show that in the
specialty co?ee sector, where brand development in relation to a particular origin or estate requires
security of supply, roasters may be pushed toward closer forms of coordination with international
traders and exporters in the near future. 23
6. THE ‘‘LATTE REVOLUTION’’? SPECIALTY COFFEE AND THE CHANGING WORLD OF COFFEE CONSUMPTION
Globally, most co?ee for in-home consumption is purchased in supermarkets. The food retail sector is
highly concentrated in the United States, the United Kingdom and Northern Europe and plays a dominant
role in the food marketing chain (van Dijk et al., 1998). Yet, through consolidation and with massive
investment in advertising their brands, roasters have managed to keep control of the co?ee chain (van
Dijk et al., 1998). This happened in spite of the development of private co?ee labels by supermarkets.
As a result, supermarkets’ retail margins for co?ee have remained generally lower than for the average
food portfolio. In some countries, such as the United States, retailers sell co?ee even at a loss in
order to ‘‘generate tra?c.’’ Retailers need to stock co?ee because consumers expect them to do so. They
can attract customers with relatively cheap co?ee and entice them to buy other (higher-margin) items
during their visit (Dicum & Luttinger, 1999; Pendergrast, 2001; van Dijk et al., 1998). Furthermore,
co?ee sales have recently moved into even lower pro?t margin outlets, such as warehouse and discount
stores. In 1997, 10% of total retail co?ee purchases in the United States were made at Wal–Mart (Dicum
& Luttinger, 1999, pp. 114, 159). Does this mean that roasters will continue to dominate the co?ee
chain in the future? In Section 5, I have argued that entry barriers in the ‘‘traditional’’ co?ee-
marketing chain have increased in both trading and roasting, and that strategic choices made by
roasters in the last decade have shaped the reactions of all other actors upstream. Recent signals,
however,
suggest that a fragmentation of the market is taking place. The emergence of new consumption patterns,
with the growing importance of ‘‘conscious’’ consumption, 24 single origin co?ees, the proliferation of
cafeechainsand specialty shops, and increasing out of home consumption poses new challenges to
‘‘traditional’’ roasters (van Dijk et al., 1998). They are used to selling large quantities of
relatively homogeneous and undi?erentiated blends of mediocre to poor quality. According to co?ee
industry analysts, these roasters have been slow in changing long-established ways of carrying out
business and advertising. Major co?ee roasters lost their regional image and their focus on localized
taste preferences a long time ago. In the United States, regional roasters such as Folgers, Hills
Brothers, and Maxwell House became national in scope and then started being bought by food
conglomerates as early as the post-WWII period (Dicum & Luttinger, 1999; Pendergrast, 2001). 25 When
they became part of major industrial empires, co?ee roasters had to move away from a focus on quality
and locality. They started to concentrate on consistency in price, packaging and ?avor. As a result,
roasters homogenized blends. They started to use cheaper beans and cut down roasting times to reduce
weight loss and mask the poor quality of the beans. Overall co?ee quality decreased. As brand
competition took the fore in corporate strategies in the United States, the product itself became of
secondary importance (Dicum & Luttinger, 1999). Homogenization and mass marketing of co?ee further
increased with the gaining importance of instant co?ee after WWII. By competing almost exclusively on
advertising, the major roasters stripped o? co?ee of most of its charm and appeal even as per capita
consumption started to decline after 1962. On the contrary, in Europe co?ee standards remained higher
due to cultural factors and different patterns of consumption even after multinationals moved into the
co?ee market (Dicum & Luttinger, 1999, pp. 116–163). It is in the background of these changes that the
specialty co?ee industry emerged as an important player, ?rst in the United States and later in Europe.
One of the characteristics of specialty co?ee is that it means di?erent things to di?erent people.
Nowadays, the term covers basically all co?ees that are not traditional industrial blends, either
because of their high quality and/or limited availability on the producing side, or because of ?
avoring, packaging
1110 WORLD DEVELOPMENT
and/or ‘‘consumption experience’’ on the consumption side (ICO, ITC, & CFC, 2000). The evolution of
specialty co?ee cannot be appreciated without making a reference to the ‘‘Starbucks factor.’’ Starbucks
was founded in 1971 in Seattle, following the steps of Peet’s, another quality roaster based in
Berkeley. As other specialty operators, Starbucks spent most of the 1980s building a loyal customer
base and ‘‘educating’’ consumers on the qualities of ?ne co?ees. The breakthrough that made Starbucks a
stunning success was creating a cafeeatmosphere where customers could hang out and consume an
‘‘experience’’ at a place that was neither home nor work. This happened at the same time as other
consumer products moved from mass-production and marketing to being recast as more authentic, ?avorful
and healthy (micro-brewed beer, specialty breads, organic vegetables). By combining ‘‘ambience’’
consumption and the possibility for consumers to choose type, origin, roast, and grind, Starbucks
managed to de-commoditize co?ee. It sold co?ee ‘‘pre-packaged with lifestyle signi?ers’’ (Dicum &
Luttinger, 1999, p. 153). By 1997, Starbucks was operating 2,000 outlets (mostly directly owned) in six
countries. In 1998, it entered the European market through the acquisition of the London-based Seattle
Co?ee Company and plans the opening of 500 outlets in the continent by 2003 (Starbucks, 2001).
Accompanying the growth in cafeechains, there has also been an explosive increase in the number of
roasters in the United States, although the smallest 1,900 roasters still control only 20% of the
domestic market. As recently as 1987, the three major roasting companies in the United States held
almost 90% of the retail market. By 1993 they had lost 12% of the market share to Starbucks, other
regional caf ees and specialty roasters (Dicum & Luttinger, 1999). Specialty co?ee consumption is
growing rapidly in ‘‘traditional’’ consuming countries, whereas regular co?ee consumption is
stagnating. It is estimated that the number of Americans drinking specialty co?ees on a daily basis
grew from 20 to 27 million in 2001, up from only seven million in 1997 (Financial Times, April 27,
2001). Traditional roasters have been slow in responding to this new phenomenon. They have put darker
roasts in the market and created their own specialty brands, but consumer response has been poor so far
(Dicum & Luttinger, 1999). One interesting inroad that some industrial suppliers are experimenting with
is
o?ering ‘‘high-quality’’ co?ee roasted on the spot by computerized roasters in large discount stores.
In this case, it is not the intrinsic quality of co?ee that makes it ‘‘better.’’ These co?ees are
mediocre and are bought in bulk. Their ‘‘selling point’’ is that they are freshly roasted. They also
sell at much cheaper prices than in specialty stores. Another likely future strategy for the mainstream
roasters to conquer back market share will be acquisition of smaller specialty roasters and
cafeechains(Dicum& Luttinger, 1999). Starbucks, on its side, has adopted fairly mainstream corporate
strategies. It has acquired competing chains, and has opened outlets in neighborhoods with traditional
cafeesto drive them out of business (Wal–Mart style). It has also entered into joint marketing
programs with other corporate giants (PepsiCo, Barnes & Noble, Capitol Records, United Airlines). By
becoming another large corporation and by providing a homogenized retail experience with a consistent
but not exceptionally good product, Starbucks has in many ways become the opposite of what independent
co?ee houses perceive themselves to be (Dicum & Luttinger, 1999). Furthermore, as
cafeechainsconsolidate, quality per se may not be as important in the future. If chains get bigger,
they tend to (re)commoditize the supply chain and simplify business. Higher sales entail more
centralized buying requirements and more di?cult relations with smaller suppliers. They also entail
more prominence for blends rather than ‘‘straight origins’’ (ICO, ITC, & CFC, 2000). Therefore, more
consumption of specialty coffee may not entail increased use of high-quality co?ee. The ‘‘Starbucks
phenomenon’’ may have revitalized interest for co?ee in consuming countries and new (higher value
added) ways of consuming it. Still, it is unclear whether specialty co?ee will be successful in
permanently de-commoditizing co?ee and in breaking the oligopoly held by a few roasting companies. It
is also not certain whether the specialty co?ee industry holds much promise for co?ee producers, who
are facing the lowest prices for green co?ee in decades. What di?erence does it make to a smallholder
if a consumer can buy a ‘‘double tall decaf latte’’ for $4, or if specialty beans are sold at $12 per
pound in the United States if he/she gets less than 50 cents for the same pound of co?ee? Since the
co?ee content of new co?ee consumption experiences is very low (see Fitter & Kaplinsky, 2001), the
‘‘latte
‘‘LATTE REVOLUTION’’ 1111
revolution’’ may have more to do with milk (latte) than with co?ee.
7. COFFEE AND GCC ANALYSIS
In this section, I provide a reading of the restructuring of the global co?ee chain through the
analytical categories of GCC. I also assess the insights o?ered by the co?ee case study to wider
debates that are taking place in the GCC literature. As explained in Section 2, Gere? (1994, 1995)
identi?es four key dimensions of GCCs: the input–output structure, the geographical coverage, the
governance structure, and the institutional framework. Tables 3 and 4 summarize changes and
continuities within these dimensions in relation to two broad periods: the ICA regime (1962–89) and the
postICA regime (1989-present). These two periods were selected for the sake of simplifying the
analysis. However, even though the ICA ended in 1989, the regime shift did not occur overnight. Some of
the forces that led to its transformation were already at work. Others changes took place later (the
adoption of SMI, for example).
(a) Governance
The governance structure of the global co?ee chain has clearly been transformed in the transition
between the two regimes. During the ICA regime, the co?ee chain was not particularly driven by any
actor, nor was it possible to clearly state that producing or consuming countries controlled it. Entry
barriers in farming and in domestic trade were often mediated by governments. The international co?ee
trade was regulated by the commodity agreement. The establishment of quotas and their periodic
negotiation entailed that entry barriers for countries as producer units were also politically
negotiated within the ICA mechanisms. Yet, the rise of power of roasters over international traders had
already started to occur. This was re?ected in the leadership structures of the co?ee industry in
consuming countries––where roasters played a key role––and meant that the trading ?rms’ goal of maximum
pro?ts in the short term was being replaced by the search for an optimum expansion of activities on the
part of roasters (Daviron, 1996). Contrary to what claimed in another analysis of the co?ee value chain
(see Fitter & Kaplinsky, 2001, p. 78), I would argue that the post
ICA regime exhibits many of the characteristics of a buyer-driven chain. Strategic choices made by
roasters in the last 10 years have shaped entry barriers not only in the roaster segment of the chain,
but also in other segments upstream. Several indicators suggest an increase in the level of
‘‘drivenness.’’ First, new requirements set by roasters on minimum quantities needed from any
particular origin to be included in a major blend can be interpreted as setting entry barriers to
producing countries. These barriers used to be set by governments on the basis of political negotiation
under the ICA regime. Now, private ?rms set them on the basis of market requirements. Second, roasters
have been able to devise new technological solutions to be less dependent on any type or origin of co?
ee. It is not clear yet how roasters have combined the minimum supply quantity strategy with more ?
exibility in product substitution, and which one of the two has relatively more weight in their global
sourcing strategy. In any case, they both indicate a potential increase in the level of drivenness of
the chain by roasters. Third, roasters have been able to set the terms of co?ee supply with the
implementation of SMI. The adoption of SMI has added new requirements for international traders to be
part of the game. Guaranteeing a constant supply of a variety of origins and co?ee types has prompted
international traders to get even more involved in producing countries than they would have anyway as a
result of market liberalization. Fourth, the persistent ability of roasters to keep retailer margins at
low levels suggests that they are still the driving force in the chain even downstream. Countervailing
tendencies are arising in the specialty market. These may not, however, be as threatening to main
roasters as it seems because these large corporations always have the possibility of buying out signi?
cant specialty players. Moreover, as specialty co?ee actors grow, they tend to streamline operations
and homogenize products; therefore, they adopt some of the same supply strategies used by giant
conglomerates.
(b) The institutional framework
The institutional framework within which the co?ee chain operates has changed dramatically as well. The
inherent stabilization forces of the ICAs and regulated markets in producing countries created a
relatively stable institutional environment where rules were relatively clear,
1112 WORLD DEVELOPMENT
change politically negotiated, and proportions of generated income fairly distributed between consuming
and producing countries. The relatively homogeneous form of trade limited the possibilities of product
upgrading, but producing countries ensured product valorization through higher prices generated by the
ICA. In the post-ICA regime, market relations have substituted political negotiation over quotas.
Producing countries have disappeared as actors in these interactions, with the exception of not
so-successful retention attempts under the ACPC umbrella. The ICO has become a relatively empty
institutional shell. Domestic regulation of co?ee markets plays an increasingly weaker role. Relatively
stable producer-negotiated and product-based quality conventions are increasingly giving way to
conventions that are generally buyer established. As concerns ‘‘conscious’’ co?ees, these conventions
are based on (buyer-de?ned) process monitoring––as well as product speci?cation. Product upgrading
Table 3. Characteristics of co?ee chain restructuring (input–output structure and geographies of
production and consumption)
ICA regime (1962–89) Post-ICA regime (1989-present)
Geography of production At ?rst concentrated in few large producing countries (Brazil, Colombia);
later, increasingly dispersed with the emergence of new producers
Fragmentation continues
Entry barriers to production Low, due to government intervention (input and credit supply, extension,
co?ee cultivation campaigns, price stabilization)
Increased, due to government withdrawal from the provision of services to farmers (end of input supply
schemes, breakdown of research and extension networks, end of price stabilization mechanisms)
Characteristics of internationally traded product
Relatively homogeneous, but distinguished by physical and intrinsic qualities (the latter especially
for Mild Arabica)
Bifurcated trend: increased homogenization of lower quality co?ees, especially Robusta (bulk export in
containers without bags); at the same time, increased trade of small quantities of speci?c high-end-
quality beans (Mild Arabica)
Entry barriers to trade Domestic trade and export: high barriers due to monopoly of marketing or
politically set domestic trade quotas
Domestic trade and export: ?rst, decreased entry barriers due to liberalization; later, increased
barriers following the strengthening of international trader operations in producing countries
International trade: increasing due to consolidation
International trade: increasing entry barriers in ‘‘fair-average-quality’’ market due to further
consolidation and requirements set by roasters through SMI; decreasing in the specialty market due to
fragmentation and the growing importance of e-commerce sales
Distribution of total income generated along the chain
Relatively stable, with farmers getting around 20% of the total, and consuming country operators around
50%
Shifted to the advantage of consuming country operators
Geography of consumption Concentrated in North America, Western Europe and Japan
Emergence of new markets (Eastern Europe, China, East Asia)
Typology of consumption Segmented by group of countries (di?erent co?ee types and blends catering for
the USA/UK markets, Southern Europe, Scandinavia, Central Europe, Japan), but relatively homogeneous
consumption within these geographical areas
Increased fragmentation: multiplication of types of product and blurring of distinctive lines of
preference between di?erent groups of countries; increasing importance of ‘‘single origin’’ co?ees
‘‘LATTE REVOLUTION’’ 1113
possibilities have increased through the fragmentation of consumption patterns, marketing of specialty
co?ee and e-commerce sales. Yet, openings in specialty markets so far have been more suitable to
estates than smallholders.
(c) The insights of co?ee to GCC analysis
The co?ee case study provides a number of insights to GCC analysis. Here, I will examine three key
aspects: (i) the signi?cance of the externalization of noncore functions that is manifested in several
buyer-driven chains; (ii) the importance of including regulation in
the analysis of any chain; and (iii) the signi?cance of di?erent levels of drivenness and di?erent
forms of coordination within buyerdriven chains. In the early GCC literature, outsourcing of supply
management and/or manufacturing was often interpreted as an instance of externalization of low-pro?t
and noncore functions upstream that is peculiar to many buyer-driven chains––although increasingly
relevant in some producer-driven chains as well. More recently, Sturgeon (1999, 2001) questioned this
interpretation. He argued that the functions externalized by brand-name ?rms to contract
Table 4. Characteristics of co?ee chain restructuring (governance structure and institutional
framework)
ICA regime (1962–89) Post-ICA regime (1989-present)
Governance structure of the chain
Low level of ‘‘drivenness;’’ increasing concentration in roasting and trading segments raises entry
barriers, but roasters are neither in the position to dictate the terms of the trade to traders, nor to
set inclusion/exclusion thresholds; control over the chain by any actor is limited
‘‘Buyer-driven’’ (speci?cally, roasterdriven); further consolidation in roasting; oversupply; adoption
of SMI by roasters forces traders to integrate upstream; vertical integration by traders made easier by
market liberalization in producing countries
Vertical integration Not common; sometimes occurring in export/international trade links; more rarely
into domestic trade and processing
Increasing; international traders integrate into export, processing, domestic trade and sometimes even
estate production; vertical integration much more limited in the roaster-international trader link
Producer–consumer country relations
In relative equilibrium; mediated through the ICAs
Absence of formalized relations; consuming country domination
Institutional framework (international)
Strong: international trade regulated by ICAs
Weak: end of ICA; producing country cartels fail to set up e?ective quota or retention schemes; futures
market increasingly de-linked from market fundamentals
Institutional framework (domestic)
Strong: markets monopolized by marketing boards, or regulated by stabilization funds and
quasigovernmental producer associations
Weak: government and quasi-government institutions retreat into oversight functions or are eliminated
altogether; trade associations ?ll only part of the formal institutional vacuum
Quality conventions International-level: product-based; set in negotiation with producing-country
sellers (and/or marketing boards) and maintained via instrument-based testing and inspection, cup
testing, and certi?cation of the product; in general, quality assessed by the buyer ex-post
International-level: increasing importance of conventions de?ned by buyers; process monitoring (in
addition to product testing) becomes important for fair trade, organic, shade-grown co?ees; quality
increasingly assessed by buyers ex-ante
Domestic-level: set by a regulatory agency; includes speci?c quality control procedures along the chain
Domestic-level: increasingly set by buyers; formal rules of quality control remain but are increasingly
disregarded
Upgrading possibilities Limited; undi?erentiated trade; however, producing countries achieve product
valorization through higher international prices provided by the ICA
Potentially increasing through marketing of ‘‘conscious’’ co?ee and direct e-commerce sales; openings
in specialty markets more suitable to estates than smallholders
1114 WORLD DEVELOPMENT
manufacturers in ‘‘turn-key’’ production networks are not necessarily low pro?t and that they do not
entail a ‘‘captive’’ position of suppliers. ‘‘Turn-key’’ systems are common in electronic products, but
also emerging in the auto parts industry, food processing and pharmaceuticals. In the agricultural
sector, they seem to be emerging in the cocoa-chocolate complex, where branded chocolate manufacturers
are increasingly outsourcing the supply of cocoa intermediate products (Fold, 2001). This has not
happened in the co?ee chain, where roasters have maintained the processing (roasting) function.
However, the experience of SMI in the co?ee industry seems to lend weight to Sturgeon’s argument in
that outsourcing is not necessarily implemented to externalize lowpro?t functions––in a period of
carrying co?ee futures market, supply management can be highly pro?table. Yet, roasters have maintained
the control of key decisions in supply management, therefore keeping international traders in a
‘‘captive’’ position. The co?ee case study also shows that the end of the commodity agreement and
market liberalization in producing countries were among the factors shaping key transformations in the
chain. 26 GCC analysis has so far focused on how business (particularly sourcing) strategies in?uence
governance and institutional structures of commodity chains––relatively neglecting the role of
regulation. Finally, the restructuring of the co?ee chain suggests (in line with Gibbon, 2001a) that
di?erent typologies of ‘‘buyers’’ need to be identi?ed within the buyerdriven category. This is because
lead actors who are in di?erent positions in the chain may apply di?erent forms of coordination. Forms
of coordination, however, should not be confused with ‘‘levels of drivenness,’’ which refer to the
degree of power of lead actors in setting modalities and thresholds of inclusion and exclusion. Levels
of drivenness tend to be higher in chains led by supermarket chains (fresh fruit and vegetables), 27
retailers and branded marketers (apparel, footwear), 28 and industrial processors (co?ee,
cocoa/chocolate) 29 than in those led by international traders (cotton, ?sh, cashew nuts). 30 They can
also change in time, as the co?ee study shows. Yet, forms of coordination may be di?erent within highly
driven chains, even in the agricultural sector alone. Strategies centered around blending and branding
characterize the modalities of chain coordination by roasters in the co?ee industry. On the contrary,
the fresh fruit and vegetables
chain is organized along a system of preferred suppliers that need to match the phytosanitary, process,
timing and product quality standards required by supermarket chains (Dolan & Humphrey, 2000). Standards
also play a key role in the rubber chain, which is, however, coordinated by industrial end-users
(Daviron, 2002). Vertical integration seems to be the dominant form of coordination in the banana chain
(and to some extent in tea and sugar), where global branders who are also producers play a key role
(see Gibbon, 2001a, p. 350). In order to be able to o?er proper policy and strategic advice to
developing countries in selecting involvement (or upgrading) in one chain or another, future GCC
research should further analyze the signi?cance of di?erent modes of coordination and the relationship
between these and levels of drivenness. It should also examine more systematically the role played by
standards, quality conventions and regulation.
8. CONCLUSIONS AND POLICY OPTIONS
The GCC approach provides useful tools for the analysis of commodity markets. It examines how key
agents build, co-ordinate and control the linkages and ?ow of produce between producers and consumers,
and the roles played in this process by contractual forms, the coordination of ?nance and business
services, and––increasingly––the wider regulatory framework. It pays attention to the organizational
aspects of the chain, to the whole range of activities from primary production to ?nal consumption, and
to the linkages binding them. GCC analysis also pursues the implications of economic power––in the form
of strategic behavior a?ecting up- and down-stream activities and agents. These aspects are almost
entirely ignored in other approaches to the study of commodity trade. 31 GCC studies have been able to
indicate trends in commodity markets that were previously unknown. They have shown that ‘‘buyers’’ of
various kinds (supermarket chains, processors and international traders) are increasingly dominating
several commodity chains. GCC studies have also highlighted that these buyers use a variety of
mechanisms of chain coordination––such as determination or control of standards and quality
conventions, control of market and consumer information, vertical integration, and branding.
Furthermore, they
‘‘LATTE REVOLUTION’’ 1115
have underlined that the end of commodity agreements and market liberalization––with the consequent
weakening of domestic regulatory powers, including quality control and stock management––have
contributed to transferring power from producers (based in developing countries) to consumers (based in
industrialized countries). Finally, GCC studies have suggested that policy advice should be based on
the speci?city of individual commodity chains rather than on the application of a general economic
model. 32 As concerns the co?ee industry, this article suggests that the present crisis faced by
farmers and producing countries is not only one of overproduction, but also one relating to changes in
the governance structure and the institutional framework of the chain. In the global co?ee chain, the
institutional framework has moved away from a formal and relatively stable system where producers had
an established ‘‘voice’’ toward one that is more informal and buyer dominated. In the process, a
substantial proportion of total income generated in the co?ee chain has been transferred from farmers
to consuming country operators. Furthermore, if roasters had provided stability to the ICA regime in
their search for an optimum expansion of activities, they are now one of the destabilizing forces in
the co?ee market. Increased corporate ?nancialization of giant roasting ?rms entails that their more
pressing goal is not expansion of activity per se anymore. 33 Their goal is rather the maximization of
pro?ts in the short term to increase the value of shares, even if it means disposing of noncore
functions. In this system, inherent instability is not a major problem for equity holders of roasting
?rms as investment fund managers can diversify risk for them. International traders, themselves
increasingly falling under the same corporate model and its pressures, have either upgraded their
functional roles and invested in new logistics systems, restructured their organization, and become
more involved in producing countries, or have disappeared. Those trading ?rms that have survived are
hedging increased risk through futures market operations. Local actors in producing countries do not
have the same ease of access to hedging instruments. Therefore, they have either allied themselves with
international traders or have disappeared. In most cases, they are losing control of processing,
domestic trade and export functions. Further consolidation seems inevitable throughout the industry.
Smallholder farmers,
however, do not have easy ‘‘consolidation’’ options. Their cooperatives ?nd it di?cult to compete with
local subsidiaries of international trading ?rms. As governments retreat from the regulation of
domestic co?ee markets, farmer organizations lose a political forum of negotiation. The weakness and
inherent instability of the institutional framework falls straight on the shoulders of co?ee farmers in
developing countries. The policy and strategic advice that follows is based on these observations. Co?
ee-producing countries are slowly realizing that the revival of the ICA system with quotas and price
bands does not seem to be possible in the short term. There is no public or political support for
quotas in consuming countries nor––with the end of the Cold War––is there a foreign policy reason for
it. Retention schemes through producer cartels, such as the recent e?ort organized by the ACPC, have
not been able to in?uence markets in the presence of a fundamental excess of supply. A second option
that has been proposed in the co?ee sector is the establishment of quotas on production. This could be,
in theory, a better solution but is opposite to what governments have been promoting in the past in
their own countries, that is higher––not lower––production. A third and more promising option, at least
in the short term, is the withdrawal of low-quality co?ee beans from the international market. This
option has been discussed within the ICO and has found some support from consuming country governments.
An ICO Quality Committee is presently discussing a minimum co?ee quality standard for export. The basic
idea is to reduce supply in the short term and raise the overall quality, therefore value, of co?ee
exports. Whether the ICO ‘‘quality initiative’’ succeeds or not, donors and producing country
governments should also increase their e?orts in promoting ‘‘conscious consumption’’ for it can provide
an extra channel for small producers in recapturing a higher proportion of the total income generated
in the co?ee chain. One way is through increased promotion of fair trade. Fair trade operators pay a
minimum ?oor price to registered producer organizations and cooperatives. They also o?er ?nancial and
technical support. The relative success of fair trade in Europe in the 1990s has shown that some
consumers are willing to pay a premium for co?ee so that farmers receive a just payment for their e?
ort. Other forms of conscious consumption are consumption of organic, shade
1116 WORLD DEVELOPMENT
grown and bird-friendly co?ees. The transition to organic farming is relatively easy in Robusta co?ee
cultivation, especially in Africa where chemical input use is low. Many producers are already growing
organic co?ee, but are being paid prices for nonorganic co?ee. They lack information on certi?cation
processes and on how to approach certi?cation agencies. The development of sun-resistant large-scale
co?ee plantations has led to the uprooting of trees and loss of biodiversity––except where co?ee is
cultivated in areas of natural savannah. These trees used to provide shade to co?ee bushes and a
natural habitat for birds in more ‘‘traditional’’ co?ee farming systems. Again, smallholders cultivate
co?ee under shade trees already, but consumers are not paying a premium for it. While the markets for
‘‘conscious’’ co?ees are growing and constitute an important development channel, they are likely to
remain niche markets in the near future. In addition, the ICO idea of a ban on exports of low-quality
co?ee is unlikely to be supported by consuming country governments in the long term. Therefore, solving
the current imbalances in the global co?ee chain also requires initiatives aimed at improving co?ee
quality in producing countries and the appreciation of quality in consuming countries. Producing
countries need to raise the reputation of individual origins and re?ne marketing skills. The key for
would-be producers of high-quality co?ees is to know how to sell the right co?ee to the right people.
They need to know which quality characteristics are appreciated where, what kind of premium will
be paid, and what are the motivations that are needed for consumers to take a product seriously.
Selling a ‘‘story’’ is particularly important. Farmer groups and/or co-operatives could be helped to
become better at exploiting their stories than they are doing now. Market failures in agricultural
input and credit markets should also be tackled because they are making it di?cult for producers to
improve quality (see also Ponte, 2001, 2002c). Furthermore, new initiatives should be aimed at
‘‘cultivating’’ consumers rather than more co?ee. A consumer who knows how to discern the intrinsic
qualities of co?ee will look for particular kinds of co?ee and be willing to pay more for its speci?
city. More informed consumers are also a market-based guarantee for higher demand of better quality
co?ee. Finally, they can address power imbalances in the global co?ee chain by facilitating market
fragmentation. 34 Even if the ‘‘latte revolution’’ and initiatives aimed at ‘‘cultivating’’ consumers
worked in permanently fragmenting and upgrading co?ee consumption, the developmental impact in
producing countries will not appear unless donors, the ICO, NGOs and producing country governments
ensure that value added is transferred to producers. This can be done by (a) facilitating the
establishment of farmer groups and producer associations and of direct links between them and
consumers; (b) promoting regulation requiring co?ee buyers in producing countries to pay producers
higher prices for higher quality co?ee; and (c) developing systems of appellation similar to the ones
used in the wine industry.
NOTES
1. Fitter and Kaplinsky (2001, p. 76) estimate that the co?ee content of the cost of a cappuccino
bought in a co?ee bar in the UK is less than 4%.
2. This term was ?rst used in Dicum and Luttinger (1999).
3. Gere? himself has mainly applied the GCC framework to analyzing exports of apparel from East Asian
countries, Mexico and the Caribbean to the United States (Appelbaum & Gere?, 1994; Gere?, 1994, 1999a),
exports of footwear (Gere? & Korzeniewicz, 1990) and e-commerce (Gere?, 2001a,b). Other GCC and related
studies have analyzed: services (Clancy, 1998; Pedersen, 2000; Rabach & Kim, 1994); footwear (Schmitz,
1999); electronics and semiconductors (Borrus, 1994; Henderson, 1989; Humphrey, 2000; Lee &
Cason, 1994; Kenney & Florida, 1994); furniture (Kaplinsky & Readman, 2000); automobiles and auto
components (Barnes & Kaplinsky, 1999; Doner, 1991; Hill, 1989; Kaplinsky & Morris, 1999; Sturgeon,
1999, 2001), illicit commodities (Wilson & Zambrano, 1994), and apparel/garments (Bonacich, Cheng,
Chinchilla, Hamilton, & Ong, 1994; Gibbon, 2000, 2001b; Kessler, 1999).
4. See, among others, Gibbon (1999) and Larsen (2001, 2002) on cotton; Raikes and Gibbon (2000) on
African export crops; Gibbon (1997) on ?sh; Kaplan and Kaplinsky (1999) on fruit canning; Barrett,
Ilbery, Browne, and Binns (1999), Calvin and Barrios (2000), Dolan, Humphrey, and Harris-Pascal (1999),
Dolan and Humphrey (2000) and Raynolds (1994) on fresh fruit
‘‘LATTE REVOLUTION’’ 1117
and/or vegetables; Fitter and Kaplinsky (2001), Ponte (2002a) and Talbot (1997a,b) on co?ee; and Fold
(2001, 2002) on cocoa.
5. For exceptions to this rule, see Gibbon (2001a) and Raikes and Gibbon (2000).
6. The major exception is Brazil, which is the top producer and also one of the main consuming
countries in the world.
7. This has changed recently. In 1996–97, co?ee ranked only ?fth among internationally traded
commodities after oil, aluminum, wheat and coal.
8. In Africa, for example, co?ee exports in 1996–98 represented more than 50% of agricultural export
earnings in ?ve countries, and more than 20% in nine countries. In three of these countries, co?ee
exports represented more than 50% of total merchandise exports, and in eight countries more than 10%
(see Ponte, 2002a).
9. The ICO classi?cation does not take into consideration that some countries produce di?erent types of
co?ee: Brazil, for example, produces Robusta as well as Hard Arabica. Cameroon, India, Papua New
Guinea, Tanzania and Uganda also produce both Robusta and Arabica. These countries are classi?ed in
accordance to the main type of co?ee they produce.
10. Co?ee is also traded in its instant and roasted forms. Trade between producing and consuming
countries consists mostly of green co?ee and bulk instant co?ee. Bulk instant co?ee imported from
producing countries is usually blended and re-packaged in consuming countries. The roasted co?ee trade
takes place almost exclusively between consuming countries. This pattern of trade comes from the fact
that green and instant co?ees can be stored for a long period of time, while roasted co?ee loses its
freshness much more quickly.
11. Currently, the ACPC has 14 rati?ed members: Angola, Brazil, Colombia, Costa Rica, C^ooted’Ivoire,
DR Congo, El Salvador, India, Indonesia, Kenya, Tanzania, Togo, Uganda and Venezuela. Together, they
make up nearly 85% of world co?ee supply.
12. Vietnam (No. 2 world producer, ranked by volume of 1999–2000 crop), Mexico (No. 4), and Guatemala
(No. 8).
13. Co?ee stocks in the United States have risen from 2.7 million bags in May 1999 to over ?ve million
bags in January 2001. May 2000 was the ?rst time since 1994 that stocks topped ?ve million bags.
14. Calculated from CSCE (2001) data. Fitter and Kaplinsky (2001, p. 77) show a similar trend using a
di?erent data set.
15. Talbot (1997a, p. 63) de?nes the total income generated along the co?ee chain as ‘‘equal to the
total amount of money spent by consumers to purchase co?ee products for ?nal consumption.’’
16. The remaining shares of total co?ee income are (a) transport costs and weight losses; and (b) value
added in producing countries.
17. Talbot’s (1997a) calculations are based on weighted average prices for all ICO member countries at
various nodes of the chain. An alternative approach is to calculate the distribution of value along
speci?c producer–consumer country chains. Pelupessy (1999) has applied this method to the
C^ooted’Ivoire–Franceand the Costa Rica–Germany chains. The results are fairly similar to Talbot’s. In
1994, the grower’s share of total retail price was 13.8% in C^ooted’Ivoireand14.6%in Costa Rica.
18. Producer-held stocks were estimated at 21.2 m bags in 2000–01 (Prudential Securities Futures
Research: Co?ee, June 28, 2000).
19. Recent takeover instances include Rothfos by Neumann, SICAFE by Bolloree,andACLIIbyCargill
(Daviron, 1996). In 2000, Cargill sold its co?ee interests to ECOM.
20. Exceptions are represented by Decotrade, the trading arm of Sara Lee/Douwe Egberts, and Taloca,
which is owned by the Jacobs Suchard/Kraft group (Philip Morris). Tchibo has a trading arm that is very
active in Kenya and Tanzania. Roasters/traders, however, do not rely on their trading arms alone for
their supply needs. They source from a variety of other international traders as well.
21. Roasters producing high-quality blends need to have greater cover (store a larger number of
varieties and origins) than roasters that produce ‘‘traditional’’ blends. The latter are able to
substitute co?ee types more readily than the former.
22. ‘‘Upstream’’ means movement toward producers. ‘‘Downstream’’ means movement toward consumers.
23. Vertical integration issues are more complex in the case of instant co?ee, where a number of
manufacturers have installed plants in co?ee-producing countries. For
1118 WORLD DEVELOPMENT
an exhaustive treatment of this subsector of the co?ee industry, see Talbot (1997b).
24. By ‘‘conscious’’ consumption, I mean consumption of fair trade, organic, shade-grown and bird-
friendly co?ees.
25. Maxwell House was bought by General Foods back in 1928. Folgers was taken over by Procter & Gamble
in 1963. General Foods was eventually taken over by Philip Morris in 1985 and merged with Kraft in
1995.
26. Similarly, Dolan and Humphrey (2000) emphasized how the UK 1990 Food Safety Act has in?uenced the
governance structure of the fresh fruit and vegetables chain and the possibilities and modalities of
upgrading. The Act requires retailers to be able to trace their products all the way to the farm level.
27. See Dolan et al. (1999) and Dolan and Humphrey (2000).
28. See Gere? (1994, 1999a) and Gibbon (2000).
29. It is not completely clear to what extent the cocoa/ chocolate chain ?ts other buyer-driven
commodity chains. Fold (2002) characterizes its mode of governance as ‘‘bi-polar,’’ where the two lead
actors are cocoa grinders and chocolate branders.
30. See Cramer (1999) and Gibbon (2001a).
31. A notable exception is the French ?li eere approach, which is a loosely knit set of studies with
the common characteristic that they use the ?li eere (or chain) of activities and exchanges as a tool
and to delimit the scope of their analysis. This approach is seen by most of its practitioners as a
neutral, practical tool of analysis for use in ‘‘down-to-earth’’ applied research (see Raikes et al.,
2000). Another exception is business economics, where the notion of chains of activities linked by
complex networks of contracts and subcontracts is widely accepted. Porter’s (1990) ‘‘value chains’’ are
somewhat similar to GCCs, and the concept of supply chain management has become increasingly important
in recent years. There is also some convergence with Whitley’s (1992, 1999) notion of business systems,
although Whitley (1996) is critical of several aspects of the GCC approach.
32. For example, the experience of African agricultural commodity trade suggests that improved market
e?ciency is bene?cial to farmers and producer countries only when their main ‘‘insertion’’ point in a
GCC is volume rather than quality. Therefore, market liberalization may be the best option for some
countries, while highly regulated markets may be the best for others–– even within the framework of the
same commodity (see Friis-Hansen, 2000; Ponte, 2002a).
33. On corporate ?nancialization, see Grahl (2001) and Froud, Haslam, Johal, and Williams (2000).
34. For more details on speci?c policy options, see Ponte (2002b).
REFERENCES
Akiyama, T. (2001). Co?ee market liberalization since 1990. In T. Akiyama, J. Ba?es, D. F. Larson, & P.
Varangis (Eds.), Commodity market reforms: lessons of two decades. Washington, DC: World Bank. Akiyama,
T., & Varangis, P. (1990). The impact of the international co?ee agreement on producing countries.
World Bank Economic Review, 4(2), 157–173. Appelbaum, R. P., & Gere?, G. (1994). Power and pro?ts in
the apparel commodity chain. In E. Bonacich, L. Cheng, N. Chinchilla, N. Hamilton, & P. Ong (Eds.),
Global production: the apparel industry in the Paci?c Rim. Philadelphia: Temple University Press.
Barnes, J., & Kaplinsky, R. (1999). Globalization and trade policy reform: whither the automobile
components sector in South Africa. Mimeo, Institute of Development Studies, University of Sussex,
Brighton. Barrett, H. R., Ilbery, B. W., Browne, A. W., & Binns, T. (1999). Globalization and the
changing networks
of food supply: the importation of fresh horticultural produce from Kenya into the UK. Transactions of
the Institute of British Geographers, 24, 159–174. Bates, R. H. (1997). Open-economy politics: the
political economy of the world co?ee trade. Princeton: Princeton University Press. Bonacich, E., Cheng,
L., Chinchilla, N., Hamilton, N., & Ong, P. (Eds.). (1994). Global production: the apparel industry in
the Paci?c Rim. Philadelphia: Temple University Press. Borrus, M. (1994). Left for dead: Asian
production networks and the revival of U.S. electronics. In E. M. Doherty (Ed.), Japanese investment in
Asia: international production strategies in a rapidly changing world. San Francisco: The Asia
Foundation and the University of California at Berkeley. Calvin, L., & Barrios, V. (2000). Marketing
winter vegetables from Mexico. Paper presented at the IDS/ Rockefeller Foundation Meeting on Global
Value Chains, Bellagio, Italy, September 25–October 2.
‘‘LATTE REVOLUTION’’ 1119
Clancy, M. (1998). Commodity chains, services and development: theory and preliminary evidence from the
tourism industry. Review of International Political Economy, 5(1), 122–148. Co?ee, Sugar and Cocoa
Exchange––New York (CSCE) (2001). Available: http://www.csce.com. Cramer, C. (1999). Can Africa
industrialize by processing primary commodities? The case of Mozambican cashew nuts. World Development,
27(7), 1247–1266. Crowe, T. (1997). Co?ee futures price behavior and fund investment. ICO Co?ee
Newsletter, 4. Daviron, B. (1993). Con?ict et Cooperation sur le March ee International du Caf ee:
Une Analyse de Longue Periode. Th  eesedeDoctorat,EcoleNationale
SupeerieureAgronomiquedeMontpellier,Montpellier. Daviron, B. (1996). The rise and fall of governmental
power on the international co?ee market. In M. Gri?on, & P. Guillaumont (Eds.), Economics of
agricultural policies in developing countries. Paris: Editions de la Revue Franc
aise d’Economie.
Daviron, B. (2002). Small farm production and the standardization of tropical products. Journal of
Agrarian Change, 2(2), 137–161. Dicum, G., & Luttinger, N. (1999). The co?ee book: anatomy of an
industry from the crop to the last drop. New York: The New Press. Dolan, C., Humphrey, J., & Harris-
Pascal, C. (1999). Horticulture commodity chains: the impact of the UK market on the African fresh
vegetable Industry. Mimeo, Institute of Development Studies, University of Sussex, Brighton. Dolan, C.,
& Humphrey, J. (2000). Governance and trade in fresh vegetables: the impact of UK supermarkets on the
African horticulture industry. Journal of Development Studies, 37(2), 147–176. Doner, R. F. (1991).
Driving a bargain: automobile industrialization and Japanese ?rms in Southeast Asia. Berkeley:
University of California Press. Financial Times, London, April 27, 2001. Fitter, R., & Kaplinsky, R.
(2001). Who gains from product rents as the co?ee market becomes more di?erentiated? A value-chain
analysis. IDS Bulletin, 32(3), 69–82. Fold, N. (2001). Restructuring of the European chocolate industry
and its impact on cocoa production in West Africa. Journal of Economic Geography, 1(3), 405–420. Fold,
N. (2002). Lead ?rms and competition in ‘bi-polar’ commodity chains: grinders and branders in the
global cocoa-chocolate industry. Journal of Agrarian Change, 2(2), 228–247. Friis-Hansen, E. (Ed.).
(2000). CDR Policy Paper Series, Agricultural policy in Africa after adjustment. Copenhagen: Centre for
Development Research. Froud, J., Haslam, C., Johal, S., & Williams, K. (Eds.). (2000). Shareholder
value and the political economy of late capitalism [Special issue]. Economy and Society, 29(1). Gere?,
G. (1994). The organization of buyer-driven global commodity chains: how US retailers shape overseas
production networks. In G. Gere?, & M.
Korzeniewicz (Eds.), Commodity chains and global capitalism. Westport: Greenwood Press. Gere?, G.
(1995). Global production systems and Third World development. In B. Stallings (Ed.), Global change,
regional response: the new international context of development. Cambridge: Cambridge University Press.
Gere?, G. (1999a). International trade and industrial up-grading in the apparel commodity chain.
Journal of International Economics, 48(1), 37–70. Gere?, G. (1999b). A commodity chains framework for
analyzing global industries. Mimeo, Department of Sociology, Duke University, Durham. Gere?, G.
(2001a). Beyond the producer-driven/buyerdriven dichotomy. The evolution of global value chains in the
internet era. IDS Bulletin, 32(3), 30–40. Gere?, G. (2001b). Shifting governance structures in global
commodity chains, with special reference to the internet. American Behavioral Scientist, 44(10), 1616–
1637. Gere?, G., & Korzeniewicz, M. (1990). Commodity chains and footwear exports in the semiperiphery.
In W. G. Martin (Ed.), Semiperipheral states in the world economy. Westport: Greenwood Press. Gere?,
G., Korzeniewicz, K., & Korzeniewicz, R. (1994). Introduction: global commodity chains. In G. Gere?, &
M. Korzeniewicz (Eds.), Commodity chains and global capitalism. Westport: Greenwood Press. Gibbon, P.
(1997). Of saviours and punks: the political economy of the Nile perch marketing chain in Tanzania. CDR
Working Papers 97.3, Centre for Development Research, Copenhagen. Gibbon, P. (1999). Free competition
without sustainable development? Tanzanian cotton sector liberalization 1994/95–1997/98. Journal of
Development Studies, 36(1), 128–150. Gibbon, P. (2000). ‘‘Back to the basics’’ through delocalisation:
the Mauritian garment industry at the end of the twentieth century. CDR Working Papers 00.7, Centre for
Development Research, Copenhagen. Gibbon, P. (2001a). Upgrading primary production: a global commodity
chain approach. World Development, 29(2), 345–363. Gibbon, P. (2001b). At the cutting edge: UK clothing
retailers and global sourcing. CDR Working Papers 01.4, Centre for Development Research, Copenhagen.
Gilbert, C. L. (1996). International commodity agreements: an obituary. World Development, 24(1), 1–19.
Gilbert, C. L. (1998). What kind of economic clauses could generally be acceptable in a new
international co?ee agreement? ICO Co?ee Newsletter, 7. Grahl, J. (2001). Globalized ?nance: the
challenge to the Euro. New Left Review, 8, 23–47. Henderson, J. (1989). The globalization of high
technology production: society, space and semiconductors in the restructuring of the modern world.
London and New York: Routledge. Hill, R. C. (1989). Comparing transnational production systems: the
automobile industry in the USA and
1120 WORLD DEVELOPMENT
Japan. International Journal of Urban and Regional Research, 13(3), 462–480. Hopkins, T. K., &
Wallerstein, I. (1986). Commodity chains in the world economy prior to 1800. Review, 10(1), 157–170.
Humphrey, J. (2000). Globalization and supply chain networks: the auto industry in Brazil and India.
Paper presented at the IDS/Rockefeller Foundation Meeting on Global Value Chains, Bellagio, Italy,
September 25–October 2. Humphrey, J., & Schmitz, H. (2000). Governance and upgrading in global value
chains. Paper presented at the IDS/Rockefeller Foundation Meeting on Global Value Chains, Bellagio,
Italy, September 25–October 2. International Co?ee Organization (ICO), International Trade Centre
UNCTAD/WTO (ITC), & Common Fund for Commodities (CFC) (2000). The government co?ee project: adding
value to green co?ee (2 volumes). London, Geneva and Amsterdam: ICO, ITC and CFC. International Co?ee
Organization (ICO) (2001a). Available: http://www.ico.org/asp/statschoice2.htm. International Co?ee
Organization (ICO) (2001b). Available: http://www.ico.org/statist/po2.htm. Kaplan, D., & Kaplinsky, R.
(1999). Trade and industrial policy on an uneven playing ?eld: the case of the deciduous fruit canning
industry in South Africa. World Development, 27(10), 1787–1801. Kaplinsky, R., & Morris, M. (1999).
Trade policy reform and the competitive response in Kwazulu Natal Province, South Africa. World
Development, 27(4), 717–737. Kaplinsky, R., & Readman, J. (2000). Globalization and upgrading: what can
(and cannot) be learnt from international trade statistics in the wood furniture sector. Paper
presented at the IDS/Rockefeller Foundation Meeting on Global Value Chains, Bellagio, Italy, September
25–October 2. Kenney, M., & Florida, R. (1994). Japanese maquiladoras: production organization and
global commodity chains. World Development, 22(1), 27–44. Kessler, J. A. (1999). The North American
Free Trade Agreement, emerging apparel production networks and industrial upgrading: the Southern
California/ Mexico connection. Review of International Political Economy, 6(4), 565–608. Larsen, M. N.
(2001). Zimbabwean cotton sector liberalization: a case of successful private coordination? CDR Working
Papers 01.1, Centre for Development Research, Copenhagen. Larsen, M. N. (2002). Is oligopoly a
condition of successful privatization? The case of cotton in Zimbabwe. Journal of Agrarian Change, 2
(2), 185– 205. Lee, N., & Cason, J. (1994). Automobile commodity chains in the NICs: a comparison of
South Korea, Mexico and Brazil. In G. Gere?, & M. Korzeniewicz (Eds.), Commodity chains and global
capitalism. Westport: Greenwood Press. Lodder, C. A. (1997). The just-in-time policy: some conditions
for success. ICO Co?ee Newsletter, 5.
Losch, B. (1999). Le complexe caf ee-cacao de la C^ oote d’Ivoire (two volumes).
TheesedeDoctorat.UniversiteeMontpellierI,Montpellier. Lucier, R. L. (1988). The international
political economy of co?ee: from Juan Valdez to Yank’s diner. New York and London: Praeger. McClumpha,
A. D. (1988). The trading of green co?ee. In R. J. Clarke, & R. Macrae (Eds.), Co?ee (volume 6):
commercial and technico-legal aspects. London and New York: Elsevier Applied Science. Palm, F. C., &
Vogelvang, B. (1991). The e?ectiveness of the world co?ee agreement: a simulation study using a
quarterly model of the world co?ee market. In O. Guvenen, W. C. Labys, & J. B. Lesourd (Eds.),
International commodity market models. London: Chapman and Hall. Pedersen, P. O. (2000). The changing
structure of transport under trade liberalization and globalization and its impact on African rural
development. CDR Working Papers 00.1, Centre for Development Research, Copenhagen. Pelupessy, W.
(1999). Co?ee in Cote d’Ivoire and Costa Rica: national and global aspects of competitiveness. In H. L.
van der Laan, T. Dijkstra, & A. van Tilburg (Eds.), Leiden, African Studies Centre Research Series
15/1999, Agricultural marketing in tropical Africa: contributions from the Netherlands. Ashgate:
Aldershot. Pendergrast, M. (2001). Uncommon grounds: the history of co?ee and how it transformed our
world. New York: Basic Books. Ponte, S. (2001). Policy reforms, market failures and input use in
african smallholder agriculture. European Journal of Development Research, 13(1), 1–29. Ponte, S.
(2002a). Brewing a bitter cup? Deregulation, quality and the re-organization of the co?ee marketing
chain in East Africa. Journal of Agrarian Change, 2(2), 248–272. Ponte, S. (2002b). The ‘co?ee crisis’:
quantity vs. quality in production and consumption. CDR Aid Policy and Practice Issue Paper, Centre for
Development Research, Copenhagen. Ponte, S. (2002c). Farmers and markets in Tanzania: how market
reforms a?ect rural livelihoods in Africa. Oxford and Portsmouth: James Currey and Heinemann. Porter,
M. (1990). The competitive advantage of nations. Boston: Harvard Business School Press. Prudential
Securities Futures Research––Co?ee, New York, June 28, 2000. Rabach, E., & Kim, E. M. (1994). Where is
the chain in commodity chains? The service sector nexus. In G. Gere?, & M. Korzeniewicz (Eds.),
Commodity chains and global capitalism. Westport: Greenwood Press. Raikes, P., & Gibbon, P. (2000).
‘‘Globalization’’ and African export crop agriculture. Journal of Peasant Studies, 27(2), 50–93.
Raikes, P., Jensen, M. F., & Ponte, S. (2000). Global commodity chain analysis and the French ?li eere
approach: comparison and critique. Economy and Society, 29(3), 390–417.
‘‘LATTE REVOLUTION’’ 1121
Raynolds, L. T. (1994). Institutionalizing ?exibility: a comparative analysis of fordist and post-
fordist models of Third World agro-export production. In G. Gere?, & M. Korzeniewicz (Eds.), Commodity
chains and global capitalism. Westport: Greenwood Press. Schmitz, H. (1999). Global competition and
local cooperation: success and failures in the Sinos Valley, Brazil. World Development, 27(9), 1627–
1650. Starbucks (2001). Available: http://www.starbucks.com. Sturgeon, T. J. (1999). Turn-key
production networks: industry organization, economic development, and the globalization of electronics
contract manufacturing. Ph.D. dissertation. Department of Geography, University of California at
Berkeley, Berkeley. Sturgeon, T. J. (2001). How do we de?ne value chains and production networks? IDS
Bulletin, 32(3), 9–18. Talbot, J. M. (1997a). Where does your co?ee dollar go? The division of income
and surplus along the co?ee commodity chain. Studies in Comparative International Development, 32(1),
56–91. Talbot, J. M. (1997b). The struggle for control of a commodity chain: instant co?ee from Latin
America.
Latin American Research Review, 32(2), 117– 135. Tam, T., & Gere?, G. (1999). Industrial upgrading and
organizational chains. Mimeo, Department of Sociology, Duke University, Durham. van Dijk, J. B., van
Doesburg, D. H. M., Heijbroek, A. M. A., Wazir, M. R. I. A., & de Wol?, G. S. M. (1998). The world co?
ee market. Utrecht: Rabobank International. Whitley, R. (1992). Business systems in East Asia: ?rms,
markets and societies. London: Sage. Whitley, R. (1996). Business systems and global commodity chains:
competing or complementary forms of economic organisation. Competition and Change, 1, 411–425. Whitley,
R. (1999). Divergent capitalisms: the social structuring and change of business systems. New York and
Oxford: Oxford University Press. Wilson, S., & Zambrano, M. (1994). Cocaine, commodity chains and drug
policies: a transnational approach. In G. Gere?, & M. Korzeniewicz (Eds.), Commodity chains and global
capitalism. Westport: Greenwood Press.
1122 WORLD DEVELOPMENT
TO YOUR ASSIGNMENTS DONE AT A CHEAPER PRICE PLACE THIS ORDER OR A SIMILAR ORDER WITH US NOW.