Information, Finance and the New International Inequality: The Case of Coffee

This paper argues that a “new” international inequality has been superimposed over the “old” international inequality, and that this superimposition can help to explain the increasing degree of inequality in the world economy today. Th is argument is illustrated using the empirical example of the world coff ee market. First the paper identifi es the basic features of the old international inequality. Next it describes the basic elements of the new international inequality. Th en it illustrates how the combination of new and old forms of inequality further disadvantages coff ee producers located in peripheral and semiperipheral areas of the world economy. Th is is shown through comparison of the events following two severe frosts in Brazil, which signifi cantly disrupted the market, causing all participants to adjust to the new circumstances. Th e diff erences between these two series of events show how transnational corporations (TNCs) based in the core have gained further advantages over their suppliers in non-core areas.


INTRODUCTION
T his paper argues that a "new" international inequality has been superim- posed over the "old" international inequality, and that this superimposition can help to explain the increasing degree of inequality in the world economy today.Th is argument is illustrated using the empirical example of the world coff ee market.First the paper identifi es the basic features of the old international inequality.Next it describes the basic elements of the new international inequality.Th en it illustrates how the combination of new and old forms of inequality further disadvantages coff ee producers located in peripheral and semiperipheral areas of the world economy.Th is is shown through comparison of the events following two severe frosts in Brazil, which signifi cantly disrupted the market, causing all participants to adjust to the new circumstances.Th e diff erences between these two series of events show how transnational corporations (TNCs) based in the core have gained further advantages over their suppliers in non-core areas.
Th is paper argues that a "new" international inequality has been superimposed over the "old" international inequality, and that this superimposition can help to explain the increasing degree of inequality in the world economy today.Th e old international inequality was based on the colonial division of labor, in which the periphery provided raw materials to core-based industries.Th e new inequality is based on control over fl ows of information and fi nancial capital by core-based transnational corporations (TNCs).Th is argument is illustrated using the empirical example of the world coffee market, comparing the responses of market participants to two severe frosts in Brazil, which significantly disrupted the market.Following the first frost, in  under the "old" international inequality, TNCs responded gradually amidst uncertainty over the frost's impacts, allowing coffee-producing countries to reap windfall profits during an extended period of high prices.TNCs responded immediately to the second frost in , due to their access to information about the severity of the frost and their control over financial instruments used to set the world market price of coffee.This quick response enabled them to capture most of the excess profits resulting from a much shorter period of high prices.
abstract John M. Talbot  Information, Finance and the New International Inequality  inequality, and to identify the specifi c characteristics of the new form.Arrighi argued that the development of the capitalist world economy over the past  years can best be understood as consisting of four "systemic cycles of accumulation" with similar structures: the Genoese, Dutch, British and US.Th ese cycles are designated by the hegemonic power that served as the center of capital accumulation during each cycle.Each cycle begins with a period of crisis of the previous regime, during which there is instability and increasing competition for capital.Th en, a new regime is consolidated, initiating a period of material expansion, during which capital is rapidly accumulated by developing the means of production under the direction of the new hegemonic power.Th is material expansion leads to a crisis of overaccumulation, initiating a new period of fi nancial expansion, increasing instability, and renewed competition for capital, which establishes the conditions for the rise of a new hegemon.I argue that the old form of international inequality was established during the British cycle of accumulation, and was consolidated during the phase of material expansion of the US cycle.Th e establishment of a new form of international inequality dates from what Arrighi calls the "signal crisis" of the US regime, around , and it is defi ned by the characteristics of the current phase of fi nancial expansion.
Th e main achievement of the British cycle of accumulation was to draw the entire world into a single world market, or more precisely into a single social division of labor based on industrial production (Arrighi :-).Th e "discovery" of the rest of the world and the initiation of world trade in products such as spices, sugar, coff ee, tea, and cocoa took place during the Genoese and Dutch cycles.But a single, integrated system of international inequality was the creation of the British "cosmopolitan-imperial" regime which extended the division of labor to the areas that produced these products.As Arrighi states, "[u]nder the Genoese regime, the world was ' discovered,' under the British it was ' conquered' ."(p.) Th e essence of the old international inequality, then, is the roles assigned to the diff erent areas in the world division of labor.Th e colonized areas became suppliers of mass quantities of raw materials to feed the industrial machine, and the new working class that operated it, in Britain and the other core countries of that period, and also served as markets for its products.Th e stimulants, such as coff ee and tea, played an important role in the process, because they served to keep the working class sober and alert as they labored in the "satanic mills." In Arrighi' s framework, extensive cycles alternate with intensive cycles, so while the old international inequality was established during the British cycle of accumulation, it was consolidated under the US cycle.During the crisis period dividing these two cycles, the colonial empires of the British period were broken up, and under the US cycle the world economy was nominally divided into "national" economies.Th e newly independent former colonies received their own national states and nominal sovereignty within an interstate system; however, their roles in the international division of labor as suppliers of raw materials and markets for fi nished products remained essentially the same.Th e integrated national economy of the US served as a model for these newly independent nations; as Philip McMichael () has argued, this model was advanced as the "development project," carried out under US leadership during its hegemonic period, from the Second World War to the  crisis.Under this model the national economies were to be managed and "developed" by the states.But under the rubric of this model, US-based TNCs became, in Arrighi' s words, "so many 'Trojan horses' in the domestic markets of other states" (p.), reinforcing the positions of the former colonies in the consolidating world division of labor and international system of inequality.To be sure, there were innovations involved.One of the most important was the spinning off of routine, labor-intensive manufacturing operations to the semiperiphery.But this was carefully managed by the TNCs and the US state so as to avoid creating any serious competition with the operations of the TNCs.It is in this sense that the old international inequality was consolidated under the period of US hegemony.
Th e signal crisis of the US regime occurred around , marked by the US defeat in Vietnam, the breakdown of the Bretton Woods system of fi xed exchange rates, and the declining legitimacy of the US anti-communist crusade.Th e result was a period of fi nancial expansion.According to Arrighi, periods of fi nancial expansion are driven by overaccumulation crises.Th ere is excess capital in search of profi table investments in the expansion of material production and trade, driving down rates of profi t.In response to this development, capital is increasingly invested in various kinds of fi nancial deals and speculation, which yield higher profi ts than investments in production, as well as preserving the liquidity of capital, so that it can be shifted quickly to more profi table opportunities.One result of this fi nancial expansion is a concentration of capital, but another is increasing instability.Arrighi argues that the US government response to this crisis was to abandon the ideal of state management of the national economy and to put its faith in the "self-regulating market," in the hopes that this would preserve the competitive advantage of US-based TNCs on the world market, and maintain US hegemony.In McMichael' s terms, the "development project" gave way to the "globalization project." My argument is that this shift resulted in the creation of a new form of international inequality that is superimposed on the old.Th is new international inequality is based in the fi nancial expansion that has occurred since , because control over fi nancial capital is its foundation.But rather than just benefi ting US-based TNCs, the new ideology of the self-regulating market has John M. Talbot  Information, Finance and the New International Inequality  nies: the Dutch from Indonesia, the French from West Africa, the British from East Africa, and the US from its neo-colonies in Latin America.
During the crisis of British hegemony and into the period of US hegemony, national coff ee companies began to establish dominant positions in their national markets.In the US, General Foods, with its Maxwell House brand was the dominant company, but it was being challenged by the Folger Coff ee Company, which was acquired in  by Procter and Gamble.In France, Jacob Suchard was the largest roaster.In Holland, it was Douwe Egberts.Two large companies, Tchibo and Eduscho, held dominant positions in the German market.Zoegas and Gavalia were the major Scandinavian roasters.General Foods had begun to go international, with operations in Canada, and a large share of the British market.Th ere was one truly global TNC, the Swiss-based Nestle Corporation, with the leading market share in Britain, a major presence in France, and a dominant position in the US market for instant coff ee.
Th ese national roasters developed distinctive national blends, based on the types of coff ee the coff ee drinkers in that country were used to getting from their former colonies.Th us, in France, coff ee blends had a high proportion of robustas, the type grown in West Africa, while US blends were based on Brazilian coff ee.2 Th e Germans consumed the highest quality Central and South American arabicas, obtained through contacts with German immigrants who had gone into the coff ee business in those countries.Th e large roasters obtained some of their supplies directly from exporters in the producing countries and the rest from national coff ee importing companies.Th e use of blends enabled the roasters to substitute coff ees within the four broad types to maintain the overall taste of the blend while purchasing the cheapest coff ee available.For instance, US roasters blended Brazils with some higher quality arabicas.If arabicas from El Salvador were unavailable or too high priced, they could substitute coff ee from Guatemala or Costa Rica, and the blend would taste about the same.Th ey could cheapen the blend somewhat by replacing a small proportion of the Colombian milds with other milds, or replacing a small proportion of the Brazils with robustas, but they depended on at least some Brazilian coff ee.enabled all core-based TNCs to tighten their control over production located in the peripheral and semiperipheral regions of the world, through their control over fi nancial capital.Th is has enabled them to squeeze additional profi ts out of these regions to fuel their capital accumulation.Th is new form of international inequality is superimposed on the old-the old world division of labor that underlies the old inequality is still largely intact.Th e new inequality just increases the overall degree of inequality in the world.But it is also fundamentally diff erent from the old inequality.Th e old inequality was based on control over production processes, while the new inequality is based on control over fi nancial capital and closely related fl ows of information.Th is change suggests that Arrighi' s call for a new research agenda focused on non-territorial spaces-of-fl ows has the best chance of allowing us to discern the outlines of the next systemic cycle.
Th is argument is illustrated through an examination of the world coff ee market during the period of fi nancial expansion.Specifi cally, I compare the events following two severe frosts in Brazil, the largest producing country, which accounted for between a quarter and a third of total world coff ee production.Th ese frosts disrupted the market and forced all participants to deal with a new situation.Th e ways in which TNCs and actors in the producing countries dealt with these shocks to the market illustrate how the new international inequality works.Th e process of fi nancial expansion was at an early stage when the fi rst frost hit in , and it had minimal eff ects on the development of coff ee prices over the succeeding - years.By the time of the second frost in , the fi nancial expansion was fully developed, and it greatly accelerated the response of prices to the shortage, to the detriment of the producing countries.

COFFEE: THE OLD INTERNATIONAL INEQUALITY
Th e origins of the coff ee trade go back to the Arab traders who brought it into Europe from what is now Yemen, in the  th Century.Th e Dutch began planting it in Java late in the  th Century, and the French brought it to the Caribbean in the  th , sowing the seeds of the colonial system of coff ee production.Coff ee is a tropical crop, so the Europeans, who quickly became addicted to it, had to produce it in their tropical colonies.Coff ee was just one of the tropical products through which many areas of the globe were drawn into a world division of labor.Sri Lanka, Java, and later, Brazil, became the main suppliers of coff ee.While Britain, during its hegemonic cycle, became the major consumer of tea brought from China, and later produced in India by the East India Company, the US, as a result of its revolution against British rule, symbolized by the Boston Tea Party, became the world' s largest coff ee consuming country.During the period of British hegemony, the European powers obtained their coff ee from their colo-John M. Talbot  Information, Finance and the New International Inequality  Th ese national roasters used extensive national advertising campaigns to establish their brand names, and engaged in oligopolistic competition with other major roasters in their national markets through brand diff erentiation and centsoff promotions.Given the huge promotional eff orts expended by roasters, supermarkets, particularly in the US, often used coff ee as a "loss leader," an item sold at or below cost in order to bring people into the store, where they would make other purchases.Once a company had established its brand as the dominant one in a particular market, it was hard for another brand, even one with heavy fi nancial backing for advertising and promotion, to break into that market.Th is was shown most clearly by the ferocious battles that took place when Folger' s attacked the eastern cities dominated by Maxwell House in the s.Procter and Gamble, the owner of the Folger' s brand, eventually succeeded in breaking the grip of Maxwell House, but it was a long and extremely costly battle (New York Times, January , , Section , p. ).
Given this situation, it was almost impossible for any coff ee processors located in the coff ee producing countries to break into the roasting, packing and selling of coff ee in the major consuming markets.For one thing, no single producing country could produce a blend comparable to those produced by the large national roasters.Each country grew one, or sometimes two, diff erent types of coff ee, and would have had to import coff ee to produce comparable blends.For another, roasted coff ee went stale quickly, although the vacuum can did keep it fresher somewhat longer.Th is put potential competitors in the producing countries at a further disadvantage, because they would have had to ship roasted coff ee over long distances to the consuming markets.Further, few coff ee processors in the producing countries had the market knowledge or the fi nancial clout to compete with the national roasters' brand advertising and promotional strategies.And as the attempts of Brazilian instant coff ee manufacturers to break into the US market showed, even if a manufacturer in a producing country managed to leap all of these hurdles, the major roasters were prepared to use political strategies to thwart the eff ort (Talbot a).
Th e colonial role of the peripheral and semiperipheral countries in the overall division of labor surrounding the production and processing of coff ee was thus fi rmly locked in during the period of US hegemony.Th e coff ee producing countries continued to supply green coff ee to the importers and roasters in the major consuming countries, and they processed it into fi nal form for consumption.Attempts by states and fi rms in the producing countries to break into the higher value-added segments of the commodity chain achieved only very limited success.However, collective action by producing countries did bring some success.Th ey began to organize in the s, and were successful in negotiating an International Coff ee Agreement (ICA) with the major consuming countries in .Th is agreement established an export quota system to limit the fl ow of coff ee to the world market, thereby stabilizing and propping up the price.Because of this, coff ee producers, while relegated to the lower segments of the commodity chain, did fare better than the producers of many other primary commodities, in terms of their share of the total income and profi t available from all of the operations along the commodity chain (Talbot b).

COFFEE: THE NEW INTERNATIONAL INEQUALITY
Th is situation began to change in the s, as the period of fi nancial expansion got underway and changed the economic conditions within which the coff ee trade operated.Th e fi nancial expansion was manifested in fi ve interrelated changes in the coff ee trade between the mid-s and the mid-s.First, the concentration of capital took the form of a major consolidation of both coff ee trading and coff ee manufacturing TNCs.Companies shifted from growth through investment in expanding production to growth by acquisition.Second, producing states' abilities to regulate the segments of the commodity chain within their own borders was weakened, also hampering their abilities to intervene in the world market.Th ird, there was an explosion of speculative trading in fi nancial derivatives based on coff ee: futures and options contracts.Th is growing speculative interest loosened the connection between changes in the supply of, and demand for, coff ee, and movements of coff ee futures prices, and increased instability in the futures market.Fourth, prices of physical coff ee became increasingly linked to futures prices, thereby increasing the uncertainty in the prices at which coff ee producers would be able to sell their coff ee.Fifth, these changes increased the need for detailed, instantaneous information about coff ee supplies and futures markets movements, creating a situation where a decisive advantage accrued to the giant, consolidated coff ee TNCs.Each of these changes is described in detail below.

Consolidation of Capital
Th e fl urry of mergers and acquisitions that was part of the fi nancial expansion, and that was given further impetus by the Reagan deregulation of the early s, was also felt in the coff ee trade.By the early s, four major manufacturers and about eight major trading companies controlled a majority of the coff ee fl owing into and being consumed in the major consuming markets in North America, Europe, Japan, and Australia.
Four TNCs now account for well over  of total coff ee sales across all major consuming markets. 3 Close behind Nestle is Philip Morris, which began to diversify out of tobacco and into food processing in the s.It had a huge amount of cash on hand from tobacco profi ts, but saw that it was no longer profi table to invest that capital in tobacco.Philip Morris is now the largest food processing company in the US, and second in the world to Nestle.In , Philip Morris acquired General Foods; GF' s Maxwell House division had been the largest US coff ee company for most of the post-war period, and number one in the market for R&G coff ee, until Folgers passed it in the late s.GF also already had signifi cant market shares in many of the major European markets, and Philip Morris further con-solidated its position there in , by acquiring Jacob Suchard, one of the largest roasters in France, with a large share of the EEC market; and Gavalia, a major Swedish roaster with large shares of Northern European markets.General Foods, in a joint venture with food processing giant Ajinomoto, is also the largest coff ee company in Japan.
Th e world' s third largest coff ee manufacturer is Sara Lee, the US clothing and food processing giant which owns Superior Coff ee in the US.In , it acquired Douwe Egberts, a Dutch roaster with large market shares throughout Northern Europe, which itself had previously merged with Van Nelle, another major Dutch roaster and food processing conglomerate.Sara Lee also has significant shares of the French and Spanish markets, and is the largest coff ee roaster in Brazil.In , Sara Lee moved into third position overall in the US market, by acquiring Chock Full O' Nuts, the fourth largest coff ee company in the US, including Tenco, the largest supplier of private label instant coff ee in the country.It also purchased Hills Bros., MJB, and Chase and Sanborn from Nestle.
Th e fourth coff ee TNC is Procter and Gamble.In , P&G acquired the Folger Coff ee Company, then a major regional roaster based in San Francisco.In the early s, P&G took the Folger' s brand name national, by going into the East Coast stronghold of Maxwell House and engaging in a series of brutal discount pricing wars, beginning in Cleveland in  (Hilke and Nelson ).By the early s, Folger' s passed Maxwell House to become the best-selling brand of R&G in the country.In , P&G bought Maryland Club Foods, producer of the Butter-Nut brand, with large market shares on the East Coast.In  it acquired Millstone Coff ee, and in , it bought the bankrupt Brothers Gourmet Coff ee, to gain a foothold in the growing specialty coff ee market.Procter and Gamble does not have large coff ee sales outside the US and Canada, but is still the world' s fourth largest coff ee company by virtue of being the largest overall in the US, by far the largest consuming market.
All four of these TNCs are multi-product conglomerates, and despite the fact that they are the largest coff ee manufacturers in the world, coff ee is not their main product.Th ese four companies control over  of coff ee sales in the major consuming markets, but this statistic actually underestimates the degree of TNC control.In some of the major markets, coff ee and food-processing TNCs of only slightly smaller scale also have signifi cant market shares, for instance Tchibo-Eduscho in Germany (these two large roasters merged in ), Lavazza in Italy, Paulig in Finland, and Ueshima and Key Coff ee in Japan.All of the second-tier European TNCs have expanded their operations since the unifi cation of the European market, and all of them plus the four major TNCs have moved rapidly into, and are competing vigorously for, the newly-opened Eastern European 3. Information on the consolidation of the coff ee manufacturing TNCs in the following paragraphs has been drawn from the following sources: Tea and Coff ee Trade Journal, March , p. ; August , p. ; September , pp.-; July , pp.-; January , pp.-; April , pp.-; July , pp.-; September , p ; December , p. ; December , p. ; January , p. ; November , pp.-; December , pp.-; April , p. ; June , p. ; World Coff ee & Tea March , pp.-; April , p. ; January , pp.-; January , pp.-; Landell Mills, April, ; Stopford (); Boletin Cafetera, May , ; Mattera (), New York Times, June , , p. ; March , , p. C; June , , p. C; Washington Post, September , , p.; Los Angeles Times, October , , p. C; Business Wire, June , ; December , ; December , ; PR Newswire, November , ; December , .

John M. Talbot 
Information, Finance and the New International Inequality  markets, particularly the more stable ones, such as Poland, Hungary, and the Czech Republic. 4h rough the s, the increasing concentration of coff ee roasters and instant coff ee producers in consuming markets began to lead to concentration of coff ee importing fi rms.At the beginning of the s, most commodity trading fi rms specialized in a single commodity, but during the s, the largest ones began to expand into other commodities related to their main specialization.Tropical commodities were prominent in this movement; thus Gill and Duff us, the largest cocoa trader, moved into coff ee and later into sugar, while Sucres et Denrees, the large French sugar trader, moved into cocoa and later coff ee.In the late s and early s, the largest manufacturing TNCs, which had been directly importing some of their own coff ee, particularly from the large suppliers like Brazil and Colombia, began to turn all of their importing operations over to the trading companies.High interest rates, fl uctuating exchange rates, fl uctuating prices in the world coff ee market, and political instability in some of the producing countries, all combined to increase the risks involved in importing, and the manufacturing TNCs preferred to transfer these risks to the importers.In addition, under pressure of high interest rates in the late s, manufacturers signifi cantly reduced the stocks of green coff ee they carried, relying on the importers and on improved transportation and communication systems to supply green coff ee for more fl exible, just-in-time production.
As the TNCs acquired signifi cant market shares in a number of consuming markets, they began to rationalize their operations, closing roasting or instant coff ee plants with small capacity or outdated equipment, and expanding their more modern plants or building new state-of-the-art facilities.Th ese new plants were strategically located near major coff ee ports (e.g., New Orleans, Hamburg, and Marseilles) and were often designed to produce for major markets in several diff erent consuming countries.Th e manufacturers were thus bringing in larger volumes of coff ee through a smaller number of ports, and preferred to deal with the largest importers, who could handle the volumes of coff ee they required.All of these developments favored the larger trading fi rms, particularly those that had established multi-commodity operations.As the manufacturing TNCs consolidated, they were also able to use their oligopsony positions to demand better deals from the importers, driving down their profi ts.Due to all of these added pressures on the importers, when there were sharp downward price movements, as happened in -, and again in -, some traders, even some of the largest ones, were driven out of business. 5John M. Talbot  Information, Finance and the New International Inequality  Volcafe is the former coff ee trading operation of Volkart Brothers, a large European multi-commodity trader and fi nancial company.In , the coff ee operation was spun off as a separate company to a management group, and was then acquired by the ERB Gruppe, a Swiss conglomerate that deals in everything from commodity trading to banking to auto importing and distribution. 7ED&F Man, already a major sugar and coff ee importer, became the world' s largest cocoa trader when it acquired Gill and Duff us.In , the company went public, but in , its futures trading business became a separate, publicly traded company, while the commodity trading business returned to being privately held. 8Cargill, the giant grain trader, instantly became one of the world' s largest coff ee importers when it purchased ACLI Coff ee in ; and J. Aron, the other major US coff ee importer, was taken over by Goldman, Sachs in .In addition to the  share held by these fi ve majors, the largest sogo shosha, C. Itoh, Marubeni, and Mitsubishi, control most coff ee imports into Japan, the third largest consuming market, and also import some coff ee into the US and European markets.All of these importers are large, multi-commodity TNC traders, and several of them specialize in a range of tropical commodities.As is true for the major coff ee manufacturing TNCs, although these fi rms are the world' s largest coff ee importers, coff ee is generally not their most important commodity. 9h is concentration of coff ee importing and processing TNCs has gone hand-in-hand with an increasing role for fi nancial capital.For the manufacturing TNCs, access to large amounts of capital is crucial, both for pursuit of their merger and acquisition strategies, and for fi nancing the purchases of the huge volumes of coff ee with which they deal.Because they are large multi-product conglomerates, they are better able to generate this capital in-house, and have more clout with the largest banks to be able to borrow what they need on the most favorable terms.For the trading TNCs, the line between banks and commodity traders became increasingly blurred as a result of the banking deregulation of the s.On the one hand, the large traders were increasingly participating on the commodity futures markets, both to protect themselves against losses on their purchases and sales of physical commodities, and also as part of an integrated trading strategy designed to maximize their profi ts.Th ey added fi nancial services to their range of commodity trading activities.On the other hand, while some banks got out of commodity fi nancing altogether, other banks began developing specialized commodity divisions to handle this aspect of their business, as it became more risky and complex.As banking was deregulated, some banks also began to trade in fi nancial instruments, including commodity futures, to protect their loans, and also to increase their profi ts.And some fi nancial services companies, such as Goldman, Sachs, became importers of physical commodities as well.Th e end result was a set of giant trading and fi nancial companies that had three important advantages in the world market conditions of the s and s.Th ey had the ability to shift funds from one commodity to another in response to price changes and profi t opportunities.Second, they had access to large amounts of capital, both in-house and from major banks.Th is not only allowed them to purchase the huge lots of coff ee demanded by the consolidated giant roasting TNCs, but also to be able to take quick advantage of opportunities to take over other coff ee traders who might fi nd themselves in fi nancial diffi culties.Finally, they also had the capital and the expertise to play the commodity futures markets, not only to hedge their coff ee purchases, but also to increase their profits.For some, the trading of fi nancial instruments became almost as important to their bottom lines as the trading of physical commodities. 10

Weakening of Producing States
Most states in the coff ee producing countries exerted some control over coff ee growing, processing and exporting that occurred within their own borders.Coff ee was an important source of foreign exchange for most of these countries, and as a major export, also a potentially large source of government revenues.In Latin America, most producing countries had state coff ee agencies that performed a variety of functions.Typically they had agricultural extension and research services for growers, but they also attempted to protect growers by set- export quotas.But after the ICA quotas ended in , many coff ee producing countries were pressured by the US and the international fi nancial institutions to reduce the roles of their state agencies in the coff ee sector.In particular, many of the state marketing boards were forced to end their monopolies on coff ee exporting, and open the trade up to private exporters.After the world market price crashed in , following the lifting of export quotas, the US and the international fi nancial institutions gained an unlikely ally: coff ee growers.World market coff ee prices remained at historically low levels for several years, and the coff ee agencies and marketing boards were forced to signifi cantly lower the prices paid to growers.Many growers then seized on the fact that they had only been receiving a percentage (sometimes very low) of the world market price, and began to actively campaign for reducing the power of the coff ee agencies and abolishing the marketing boards.In Brazil, President Collor, in a fi t of free market zeal, abolished the IBC, but given the central role of the IBC within the complex Brazilian coff ee sector, this move wreaked havoc.Within a couple of years, there were calls by all segments of the coff ee industry for renewed regulation, and a new National Coff ee Department was created.Th e Colombian FNC, which enjoyed a high degree of legitimacy with coff ee growers, kept its role in the sector, but even it was severely weakened by the period of low prices following the  crash, and was forced to sell off some of its assets.
Th us, at the same time that the coff ee TNCs were consolidating their control over the coff ee markets in the major consuming countries, the abilities of states in the producing countries to manage their own coff ee sectors, and to infl uence world market prices, was declining.Th e overall balance of power was shifted decisively in favor of the coff ee TNCs.

Increased Financial Speculation
Th e third way in which the period of fi nancial expansion was manifested in the coff ee trade was in the expansion of trading in fi nancial derivatives based on coff ee.Coff ee futures have been traded in New York since the founding of the New York Coff ee Exchange in .Sugar futures were added in , and in  it merged with the New York Cocoa Exchange to assume its present identity as the New York Coff ee, Sugar, and Cocoa Exchange (CSCE) (World Coff ee and Tea, March , pp.-).Th e coff ee futures contract traded on the New York exchange is called the "C" contract, based on Central American arabica coff ees.A futures contract for robusta coff ee futures began to be traded on the London Commodity Exchange in the s; this exchange has been reorganized several times and is now the London International Financial Futures Exchange (LIFFE).But until the s, the major participants in coff ee futures trading were import-ting minimum prices at which processors and exporters could buy coff ee from the growers.Most of these agencies also regulated exports by issuing export licenses to exporters and setting minimum export prices.African countries typically had state marketing boards that held a monopoly over coff ee exporting, in addition to providing agricultural extension and regulating the internal market.Brazil' s state coff ee agency, the Instituto Brazileiro do Café (IBC) was a typical Latin American agency, but it also performed several additional important functions.It regulated the coff ee roasting industry that produced for Brazil' s large internal coff ee market.It maintained the massive Brazilian coff ee stockpiles, using them to regulate the internal price of roasted coff ee and also to promote the Brazilian instant coff ee industry that produced for export, by selling coff ee from the stockpile cheaply to these industries.Th e Colombian agency, the Federacion Nacional de Cafeteros (FNC), was unique.It was an independent organization, jointly controlled by the state and the large coff ee growers, with broad responsibilities for regulating the coff ee sector.In addition, it also exported coff ee, in competition with the private exporters, and aggressively sought new markets for its coff ee.It is probably best known for its invention of Juan Valdez, used to promote the image of Colombian coff ee in the consuming countries.
Th e state agencies used their regulatory power to extract revenues from the coff ee sector, so that coff ee growers usually only received a percentage of the world market price.In Latin American countries this percentage was usually fairly high, but in the African and some Asian countries, the marketing boards extracted signifi cant revenues from the coff ee sector, and this percentage was often less than half.However, there were advantages to this arrangement for the growers.In a world market where prices tended to fl uctuate wildly, the state agencies could cushion the growers from these price swings by adjusting the percentage of the world market price that was returned to growers.Th us when world market prices were high, they could return a lower percentage to the growers and keep the additional revenue to allow them to maintain a steady price to growers even when the world market price dropped.
After the crisis of the s, the US government decided to abandon the ideal of national regulation of nationally-based economies that had governed its international economic policies in the post-war period.In its place, the US began a push to "free" markets, in order to open them up to US-based TNCs (Arrighi,  ).Th is new "Washington consensus" was forced on many peripheral and semiperipheral countries through structural adjustment programs during the debt crisis of the s (McMichael ).Th e eff ects of structural adjustment on the state coff ee agencies were delayed by the existence of the ICA, because the agencies needed to regulate their coff ee sectors in order to comply with the John M. Talbot  Information, Finance and the New International Inequality  ers and roasters, who used it mainly for hedging, or protecting themselves against sudden price changes. 11Th is is clearly shown by the relationship between the trading volume on the CSCE and the status of the ICA quotas; when quotas were in eff ect and prices were relatively predictable, trading volume went down.
Trading volume began falling in the early s, as the fi rst ICA was being negotiated.By , after the ICA had been in eff ect long enough to stabilize prices, the exchange was forced to close trading on the "C" contract because the trading volume was so low.Just as some members of the trade were beginning to suggest that the futures market might be unnecessary because of the stabilizing eff ects of the ICA, a frost hit Brazil in .Th is destabilized prices, despite the fact that quotas were still in eff ect, and trading picked up again.By , strains within the membership of the Agreement were making its renewal uncertain, and trading volume kept increasing, spurred on again by the suspension of the quotas in .By , trading volume had surpassed its late-s peak, and the  Brazilian frost drove it to record highs.Trading volume fell off somewhat as prices declined in the late s, and as quotas were reinstated in , it fell off again. 12rading was stimulated again in , when a drought in Brazil began to drive up prices and introduce instability into the market once again.But by this time, a number of other changes had occurred, which increased the centrality of the futures markets.Th e fi rst change was part of the general proliferation of fi nancial instruments and derivatives in the mid-s: the introduction of trading in options on coff ee futures contracts by the CSCE in .Since these options contracts were considerably cheaper than the futures contracts themselves, they allowed smaller traders and roasters (and speculators) to participate in the market. 13But they also gave the TNCs (roasters, importers, and fi nanciers) another instrument to juggle into their integrated trading strategies.Th e second change in the market in the mid-s was the rise of the commodity funds, huge conglomerations of fi nancial capital seeking the highest and most rapid profi ts available, by trading in fi nancial, oil, metals, and agricultural futures markets.Th e funds were another way in which smaller speculators, who found trading in coff ee futures and options alone too risky, but who could not aff ord to invest in a diversifi ed portfolio of commodity futures, were drawn into the fi nancial markets.Due to these changes, futures trading remained heavy through the late s, despite the reinstatement of quotas in -.And since the end of quotas in , the volume of futures and options trading has taken off , posting new record highs each year. 14h ese developments decisively shifted the balance of trading on the coff ee futures exchanges, from hedgers who were involved in the coff ee trade to speculators who were in it only to make a profi t. 15 Th is is demonstrated in Table , which 11.A coff ee futures contract is an agreement to deliver a lot of coff ee (, pounds on the New York Exchange, fi ve metric tons in London) at a specifi ed future date at a set price.A coff ee importer who had just purchased , pounds of coff ee from a producing country, at a fi xed price, for delivery in, say, three months, would hedge, or protect himself against the price of coff ee going down during that three month period (and thus having to resell the coff ee at a loss), by selling one futures contract.Th is contract would obligate someone else to take delivery of the coff ee in three months, at a price high enough to cover the fi xed price plus his costs for the transaction.If the price of coff ee went down, he could still resell the coff ee without losing any money.If the price of coff ee went up, he could make a profi t on selling the coff ee, but would have to use some of that profi t to buy back the futures contract he had sold, at a higher price.Th e importer thus trades off the possibility of making a profi t on a price increase for insurance against a loss caused by a price fall.Th e transaction works exactly the same way, but in reverse, for someone who contracts to sell coff ee and wants to protect himself against a price rise.Note: the male pronoun is used here because the vast majority of coff ee traders are still male.
13.An options contract is the right (but not the obligation) to buy or sell one futures contract at a set price.As a hedging instrument, it works in basically the same way as a futures contract.Th us an importer who has purchased physical coff ee and wants to protect himself against a price decline, would purchase a "put" option, the right to sell a futures contract at a given price at some future date.If the price of coff ee falls below that price, the importer could exercise the option and sell the futures contract at an abovemarket price to recoup his losses on the physical coff ee.If the price rises, the importer would not exercise the option and it would expire; he would lose the cost of the option, or the premium, which is analogous to a premium paid to buy insurance.See ITC () for more details.However, one futures contract is a contract for delivery of , pounds of coff ee; at . to . per pound in the late s, this was a very expensive contract (although the contracts are actually purchased on margins, for a small percentage of this total).In contrast, a coff ee option was selling for around  cents per pound at the same time.

John M. Talbot 
Information, Finance and the New International Inequality  compares the volume of futures contracts traded on the New York and London exchanges to the volume of physical coff ee traded on the world market.If futures contracts were being traded simply to hedge purchases of physical coff ee, then total futures volume would be expected to be about two times the volume of physical coff ee traded, assuming that the buyer and the seller in each purchase fully hedged their positions.Table  shows that the total volume of futures traded exploded from fi ve times the volume of physical coff ee in , to nearly  times the volume in .If options contracts, which were not traded in , are added in, the total volume of futures and options traded in  was the equivalent of . million tons of coff ee, or almost  times the volume of physical coff ee (ITC, ; ).Th us, by the mid-s, the vast majority of trades made on the coff ee futures markets were made for purely speculative purposes, and were not connected to sales of physical coff ee. 16h e shift from hedging to speculation was also accompanied by a shift in the type of speculation, from that based on fundamental analysis to that based on technical analysis.Speculators who play the coff ee futures market based on fundamental analysis rely on projections of future supply and demand, to forecast whether coff ee prices are likely to rise or fall in the coming months, and buy or sell futures accordingly, hoping to profi t when the futures prices rise or fall.Technical analysis, in contrast, attempts to predict market movements the future solely on the basis of past market movements, independent of supply and demand conditions.Technical analysts look at the combination of moving averages of prices, trends in total volume, and trends in open interest (the total number of outstanding futures contracts at a given time), to predict whether the market is likely to move up or down.Since they rely on charts of these indicators to make their forecasts, they are often referred to as "chart" traders.Th e development and refi nement of chart trading during the s meant that many small speculators could engage in commodity speculation without knowing a great deal about the commodities they were speculating in.And while the commodity funds use both kinds of analysis, they tend to rely more heavily on charting.Since the funds are also invested in many diff erent fi nancial instruments, they may sometimes move capital into or out of the coff ee futures markets because of their judgements of the profi tably of coff ee futures relative to other instruments.All of these developments meant that large amounts of money were being shifted in and out of the coff ee futures markets, for reasons that were often only marginally related to the actual global situation of supplies of, and demand for, coff ee. 17inally, the increased volume of trading on the futures exchanges, and the changing nature of the trading, also increased the volatility of futures prices.Speculators followed developments in the coff ee market hour-by-hour, if not minute-by-minute.A forecast of cold weather in the coff ee growing regions of Brazil, possibly portending a damaging frost, might set off a wave of buying by fundamental analysts, raising the price.Th e surge in volume and price could trigger a wave of buy orders from the technical analysts, who often had their computers set up to issue an automatic buy or sell order if market trends met certain conditions.Th en, a couple of large speculators who decided to sell their contracts to take a quick profi t might trigger a wave of sell orders, driving the price back down.A market movement like this could easily take place in the course of one trading day, a  hour and  minute period on the New York exchange, without any change at all occurring in the overall world supply and demand for coff ee, simply in response to speculation that there might be a frost in Brazil.

Linking of Physical Coffee Prices to the Futures Market
Another major development linked to the expansion of trading in fi nancial derivatives was the computerization of trading, and this ultimately revolutionized the business.First, reporting of futures trading was completely computer- 16.Ibid. 17. Ibid.; ITC, , Chapter .Th e older "coff ee men," who have been involved in the trade for decades, often look down on the younger speculators, who, they say, don't even know what a coff ee tree looks like, and couldn't fi nd Colombia on a map.

John M. Talbot 
Information, Finance and the New International Inequality  speculators on the futures exchanges, the price of physical coff ee drove the price of futures contracts.By the mid-s, when speculators dominated on the exchanges, the price of paper contracts drove the price of physical coff ee.Of course, the price of futures contracts was constrained by the fact that they were contracts for the actual delivery of physical coff ee at some time in the future.Th erefore, a speculator holding a futures contract had to either liquidate it before the start of the delivery month, or be faced with the prospect of having to take delivery of , pounds of coff ee.Th is insured that the price of a futures contract always converged on the price physical coff ee as the delivery month approached, and kept a linkage between futures and physicals prices.But the developments over this period signifi cantly weakened the linkage of futures prices to the underlying supply and demand conditions for coff ee.Th e combination of pegging the price of coff ee to the futures markets and the increased weight of the commodity funds in these markets has probably increased the overall instability of world market prices for coff ee. 21

Increased Need for Information
By the mid-s, anyone who was trading in physical coff ee needed to have access to up-to-the-minute information from all over the world.On any day that the futures exchanges were open, they needed to keep an eye on market movements, so that they would not be surprised by sudden price movements that could aff ect their business.Th ey needed information about weather conditions in several major producing countries, where severe weather that damaged the crop could shift overall world supply conditions.Th ey needed crop forecasts from these countries, because even in the absence of severe weather, an unusually large or small crop could change conditions (coff ee trees tend to produce in twoyear cycles, where a heavy crop one year is followed by a lighter one the next).Th ey needed information about political conditions and government policies in producing and consuming countries that could change tariff s or interrupt the fl ow of coff ee.Th ey needed information about economic conditions in consuming countries and exchange rate fl uctuations that could change the demand for coff ee or its import price.Many of these information needs were also present in the mid-s.But under the market conditions of the mid-s, they were much more pressing.Th e market was more unstable; it was moving much faster ized, so that the details of each transaction made on the trading fl oor could be fl ashed to computer screens around the world almost instantaneously.All of the major roasters and importers were linked into this system and kept continuous watch on the movements of the market.Second, deals in physical coff ee were increasingly transacted by computer.In the early s, off ers to sell coff ee were made by exporters through cables sent to the offi ces of importing companies, with replies expected by the end of the day or by the next morning.By the mid-s, these off ers were mostly made by computer messages, with replies expected within the hour. 18h ird, since the current futures prices were immediately available to the traders on their computer screens, and they received off ers to buy coff ee in the same way, they began to use futures prices to set the prices for their sales and purchases of physical green coff ee.Since the "C" contract specifi ed a generally accepted quality standard, prices would be set at a diff erential to the "C" contract, depending on whether the coff ee was of higher or lower quality than the standard Central American coff ees on which the "C" contract was based.Th is linkage made it easier for traders to agree on green coff ee contract terms and then immediately hedge the transaction, by buying or selling the appropriate futures contracts.But it also increased the uncertainty involved in the transaction.Exporters could agree to sell a certain amount of coff ee at a fi xed diff erential, and then watch the futures price, waiting for what they thought was a peak in the futures price to contact the importer and fi x the actual price for the coff ee.Of course, neither party to the deal knew whether the futures price would go up or down after the price of the physical coff ee was fi xed.But the importer did not really care, as long as he had hedged his purchase.Usually, he had already contracted to sell this coff ee to a roaster, with price also to be fi xed against the exchange, and made his profi t by charging the roaster a higher (or lower) diff erential that he was paying the exporter. 19In the mid-s, almost all coff ee was sold at prices fi xed when the sale was made; by the mid-s, as much as  of coff ee was sold at a fi xed diff erential to the futures exchanges, with the actual price to be fi xed later, by either the buyer or the seller. 20h is development made the futures market the key price determination mechanism for the entire industry.In the mid-s, when hedgers dominated John M. Talbot  Information, Finance and the New International Inequality  in response to news as well as to rumor.It had a tendency to overreact in one direction, and then overcompensate in the other.Because coff ee trading fi rms were operating under integrated strategies involving buying physical coff ee, hedging, and speculating, anyone missing out on a major market move stood to potentially lose a lot of money.Under these conditions, information itself has become a commodity in the coff ee trade, as it has in most other sectors of the economy.News services and wire services provide a wide variety of political and economic news from around the world, as as weather reports.One daily publication, Complete Coff ee Coverage provides news specifi cally related to the coff ee trade.Other newsletters giving more in-depth analysis are produced by the largest trading houses, such as ED&F Man in the UK, or by fi rms specializing in commodity analysis, such as F.O. Licht in Germany.All of these services are available on a subscription basis, and most subscriptions are quite expensive.Th e business of providing statistical data, market analysis, and charting programs for commodity market speculation has itself become a growth industry.Th ere are several diff erent statistical packages for performing chart analysis available on the market, and other services providing the raw data from commodity futures markets to input into these packages.Keeping on top of all of this information requires money and timemoney to buy access to it, and time to digest it.In this situation, the largest trading houses that dealt in multiple commodities and combined physical purchases with fi nancial speculation were clearly in the most advantageous position.Th ey had the capital to access the information, the manpower and expertise to analyze it, and they also had the capacity to develop their own in-house fundamental and statistical analyses of the market (ITC ).
But the most important advantage held by the largest trading houses was the capacity to develop their own in-house information systems.By the mid-s, the major trading houses had established their own exporting subsidiaries in the major exporting countries.Two important developments created the opportunity for this to happen.Th e fi rst was the wave of structural adjustment and market liberalization programs forced on developing countries during the debt crisis of the s.Th is reduced or did away with many restrictions on foreign ownership and control of trading fi rms within the producing countries.It also led to the ending of the coff ee export monopolies of many state marketing boards in African countries.Th is opened up coff ee exporting opportunities to privately owned fi rms for the fi rst time, and in some of these countries, there were few capitalists or private fi rms with the capital and expertise to move in and take advantage of this opening.Th e second development was the price crash following the lifting of export quotas in .In addition to driving many green coff ee importers in the consuming countries out of business, this crash also put many large exporting companies in the producing countries into fi nancial diffi culties, leaving them ripe for takeover by the major importers.For example, Neumann now has export companies in Brazil, Colombia, Peru, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Burundi, Cameroon, Ivory Coast, Kenya, Rwanda, Tanzania, Uganda, Indonesia, Papua New Guinea, and Vietnam.Volcafe has export companies in Brazil, Colombia, Peru, Costa Rica, Guatemala, Honduras, Nicaragua, Mexico, Kenya, Tanzania, Uganda, Indonesia, and Papua New Guinea.Th e large trading houses could thus get up-to-the-minute information from their own operatives inside these countries, who were familiar with the situation on the ground there, without having it fi ltered through some external service.

RESPONSES TO THE BRAZILIAN FROSTS
With all of these developments laid out, we are ready to turn to an examination of the events following the two severe Brazilian frosts of  and .Th e fi rst one took place near the beginning of the period of fi nancial expansion, at a time when the old international inequality was fi rmly in place, but the new one was just beginning to become established.Th us the responses to the fi rst frost refl ect primarily the eff ects of the old form of inequality.Th e second frost, almost  years later, came at a time when the new international inequality was fi rmly established, and thus the responses to the second frost illustrate the ways in which the new international inequality is superimposed on the old.
Two key pieces of information that aff ect the underlying fundamentals of the coff ee market are crop forecasts and information on the amount of coff ee being held in warehouses in the producing countries.Th ese are two pieces of information that are much less accessible to traders and roasters located in the consuming countries than many of the other kinds of information discussed above.And this information is much more readily available to offi cials in the producing countries, giving them a potential advantage over the TNCs.Comparison of the events surrounding the two frosts shows how this advantage was eroded by the changes in the coff ee trade from the mid-s to the mid-s.

The 1975 Frost
On the night of July , , a killer frost struck the coff ee growing regions of Brazil, hitting hardest in the more southern areas of Parana, Minas Gerais, and Sao Paulo states.Most of the frosts that strike this area are called "white frosts," which kill the leaves of the coff ee trees and the fl owers that will become the next year' s coff ee crop.But this was a "black frost," one that turns the sap black and kills the entire tree.Even so, Brazil' s - crop was not severely damaged; the main harvest had begun in April and was about ⅔ completed, and many of John M. Talbot  Information, Finance and the New International Inequality  the coff ee cherries remaining on the trees were mature enough to survive the frost.But the dead trees would produce no crop in the - harvest year, and even if new trees were planted immediately, they would not begin to bear for another four years.Brazil at this time accounted for about a third of total world coff ee production; its production in the - season had been about  million bags of  kg.each.Th e frost had struck at a particularly bad time.Th ere were civil wars raging in Ethiopia, another producer of arabica coff ees of a similar quality to Brazil' s, and in Angola, the second-largest producer of robusta coff ee.Uganda, another large robusta producer, was also in chaos under Idi Amin.Traders and roasters in the US had been betting on a bumper crop from Brazil that would lower prices, and they had held off buying, so had relatively low stocks on hand. 22h e IBC, the state agency that regulated the coff ee industry, temporarily suspended exports immediately after the frost.Th e IBC regulated exports by setting a minimum registration price-all exporters had to register their exports, and the IBC would refuse registration for any coff ee to be exported at less than its minimum price.In the wake of the frost, prices would obviously be higher, and the IBC didn't want to sell any coff ee at too low a price, and needed to assess the situation before raising the registration price.On July , a week after the frost, the IBC announced that more than  of Brazil' s - crop had been wiped out by the frost.It also announced that it had  million bags of coff ee in stocks, and that growers and exporters held an additional  million bags.Th ese stocks could be used to replace some of the output lost to the frost.On August , it lifted the temporary ban, and raised the minimum export price from  to  cents a pound.Other coff ee exporting countries had already posted similar large increases. 23h e main independent US source for information on crop damage and stock levels was the US Department of Agriculture' s Foreign Agricultural Service (FAS).Th ey produced regular crop forecasts for a variety of tropical crops produced in Th ird World countries, including coff ee.Th ere was an agricultural offi cer stationed in the US Embassy in each of the major producing countries to keep tabs on the agricultural situation.Immediately after the frost, an offi cial from the FAS fl ew to Brazil, and with the agricultural offi cer stationed there, began a fi eld survey to assess the damage.On August , the FAS reported that the - crop had been more than  damaged.Th eir preliminary forecast for the crop was between  and  million bags.Th e IBC' s forecast was for a maximum of  million bags. 24Th ere were also questions about the amount of stocks held in Brazil.While the export quotas under the fi rst two ICAs were in eff ect, from -, Brazil had built up massive stocks, probably amounting to well over  million bags, more than one year' s total world consumption of coff ee.But after a series of minor frosts in the late s and early s aff ected production, Brazil had drawn down those stocks dramatically.Coff ee can be stored for - years under optimal conditions, but after that, its quality has deteriorated to the point where it is no loner usable.So even if the total of  million bags in stocks that the IBC had announced was accurate, it was unclear how much of it was of exportable quality. 25eanwhile, the situation set off a buying frenzy that some traders labeled "frost panic," as traders and roasters bought up whatever coff ee became available, even though the real supply shortage was not likely to be felt for another year.Everyone expected the price to go higher, and wanted to buy as much as they could before it did.And roasters began to raise their prices.Th e upper line in Figure  shows the average retail price of coff ee, which followed the wholesale price closely, because retailers' margins on coff ee tended to be very small.Th e fi rst to announce an increase was General Foods (Maxwell House), the largest roaster.It increased its wholesale price by  cents a pound on July , and the other major roasters soon followed suit. 26As prices continued to rise through , retailers and consumers began to hoard coff ee, following the same logic as the traders and roasters-buy it now, before it gets more expensive.Coff ee roastings in the US were up . in the fi rst half of , compared with the same period a year before.By October , the national average retail price of a pound of coff ee had risen to ., up from . in June , before the frost, John M. Talbot  Information, Finance and the New International Inequality  and there appeared to be no end in sight.By the end of , Folger' s had broken the previously unimaginable . barrier, raising its wholesale price to .. 27y early , consumers were fed up after almost a year and a half of steadily increasing prices, and there was talk of a coff ee boycott.One of the most prominent organizers was Elinor Guggenheimer, New York City Consumer Aff airs Commissioner, a self-proclaimed -cup a day addict.Th ere were also calls for a Congressional investigation into the soaring prices. 28able  below shows the eff ect of the frost on Brazilian production and exports.If - is taken as a baseline, the frost wiped out about threefourths of Brazilian production in -.Th is production shortfall in Brazil' s - crop year began to be manifested in world production for the - coff ee year (which overlapped the Brazilian crop year for the six months of April-September ), causing a decrease of about .Brazil maintained its export level by drawing down stocks, but by the - coff ee year, that was no longer possible, and its exports fell below  million bags, leading to a world export shortage of about . 29Th e response of prices of both physical coff ee and coff ee futures to this situation followed a similar trajectory; the lower line in Figure  shows the trend of the ICO indicator price, an average of the diff erent grades of physical coff ee in major importing ports.Before the frost, the ICO indicator price had been hovering in the low -cent range, and the "C" futures contract in New York was trading in the - cent range, for the fi rst half of .Th e ICO indicator jumped to the mid-s and the futures price to around  cents immediately after the frost, and prices stayed there through .Both rose steadily through , the ICO indicator from  cents in January to . in December, and the "C" contract from around  cents at the start of the year up to about . by the end.Both prices continued to rise in early , with the ICO indicator price peaking at . in April, and the futures price topping out at . on April .From that point, it was all downhill.Folger' s was the fi rst to respond, dropping its wholesale price on May  from . to . per pound.General Foods followed with a  cent cut the next day.Physical and futures prices continued to fall until hitting a low of about . in February . 30h e producing countries tried to stop the decline of prices during  and , and pulled off a temporarily successful manipulation of the futures market, by executing a series of "short squeezes." 31Th e fi rst occurred in July  and worked as follows.Th e main players were the IBC and the Compania John M. Talbot  Information, Finance and the New International Inequality  Salvadorena de Café.Th ey bought July futures contracts during June,  (in market parlance, they "went long").Th is entitled them, if they so chose, to take delivery of specifi c grades of Central American or Colombian coff ee in New York at the end of July, from the "shorts," traders who had sold July futures contracts.But they knew that there was very little coff ee available in New York of the quality certifi ed by the CSCE as deliverable against the New York "C" contract, and they prevented any more from arriving, by buying it and shipping it to Europe.Th e "shorts" had sold futures contracts for hedging or speculative purposes, and had no intention of actually delivering coff ee.Th ey had intended to liquidate their positions, by buying futures contracts back from the "longs."But the "longs" were not selling.Th e only other option the "shorts" had was to fi nd suitable coff ee somewhere and somehow get it to New York, to fulfi ll their obligations under the futures contracts.But the "longs" already held this coff ee.In this situation, the "longs" could practically name their price for allowing the "shorts" to liquidate their positions, and make a handsome profi t on the deal.Similar operations were carried out on the London exchange, and the operation was repeated for the December  futures contract.At that point, the Commodity Futures Trading Commission (CFTC) stepped in and ordered the "longs" to liquidate their positions in an orderly fashion (New York Times, December , , p. ).
In , eight Latin American producers formed the Bogota Group (Brazil, Colombia, El Salvador, Costa Rica, Guatemala, Honduras, Mexico, and Venezuela), and established a fund of  million to carry out similar operations during .In , they made a killing on the July futures contract because of an early frost in Brazil.Th e Group coordinated its buying of the futures contract with the IBC' s announcement of its estimate of the damage caused by the frost, which was probably artifi cially infl ated.Th ey were holding futures contracts bought at low prices before the announcement, and when the market reacted to the news, futures prices jumped.Th ey reportedly made over  million on this operation.In May , the group incorporated as Pancafe, but at that point, there was too much surplus coff ee fl oating around to make such a short squeeze profi table.Brazilian production had recovered, and new trees planted in other countries as prices began to increase in - were beginning to produce.Pancafe lost money and was disbanded later in the year, as part of the agreement to reinstate export quotas under the ICA.

The 1994 Frost
On the night of June , , a severe frost struck the southern coff ee regions of Brazil.It was immediately described as the worst since the  frost.Th en, about two weeks later, on July , another frost hit.It killed additional coff ee trees that had only been weakened by the fi rst frost, as well as striking new areas not hit by the fi rst frost.Once again, the main eff ect was not on the - harvest, which was well underway, but on the following year' s crop.Brazil' s production in the year preceding the frost had been about  million bags, similar to its level of production before the  frost, but by this time, Brazil accounted for only about a quarter of total world production.Th e frost had struck at a particularly bad time.When the ICA quotas had been lifted in , world market prices had crashed, hitting historic lows in .Some growers in many countries had switched from coff ee to other crops, and those who continued to grow it had cut back on maintenance and fertilizers.Th e Brazilian crop was already suff ering because of a prolonged dry spell preceding the frost, and this probably increased the frost damage.Th e - crop was expected to be smaller than that of -, and exportable production was expected to be below total world demand for coff ee for the third straight year, resulting in a further drawdown of stocks.Prices had already begun to turn upward in  because of the expected shortage and because of a coff ee retention plan announced by the newly formed Association of Coff ee Producing Countries (ACPC), in an attempt to raise the world market price. 32peculation about the amount of damage to Brazil' s crop abounded.Early estimates from sources in Brazil began to circulate almost immediately at the CSCE, and they put the damage from the fi rst frost at  million bags.Th e earliest offi cial estimate was by the private forecasting organization Accu-Weather, which estimated the loss from the fi rst frost at - of the - crop, with the second frost destroying an additional -.Th e Brazilian National Coff ee Department (NCD), the successor to the IBC, released its estimate of a production decline of  on July .Th e FAS representatives were in the fi eld assessing the damage caused by the fi rst frost when the second one hit.Th e FAS did not release its estimate until August , and it was for a decline of - in the - crop.Th e NCD forecast was for a harvest of about  million 31.off ee prices responded to the frosts immediately.On Monday, June , the fi rst day of trading after the frost, coff ee futures rose .Th e next day, the coff ee roasters reacted.Folger' s raised the price of its -ounce cans by  cents, and Maxwell House' s went up by  cents.Th ere was a second price increase in early July, even before the second frost hit, and a third one shortly after the second frost.Th e upper line in Figure  shows the trajectory of retail coff ee prices in the US.Th e July increase in the wholesale price of coff ee was a record-breaking ., driving the US Department of Labor' s Producer Price Index up by . for the month, and raising fears of renewed infl ation.Once again, these price increases generated consumer protests, and Richard Kessel, Executive Director of the New York State Consumer Protection Board wrote two letters to Attorney General Janet Reno urging an investigation of the manufacturers' price increases.Th ey had come so rapidly that they could not possibly refl ect actual cost increases, he argued. 34off ee futures prices were extremely erratic.After climbing for more than a week after news of the fi rst frost, they began to fall as traders second-guessed the initial reports from Brazil about the extent of the damage.Th en they rose on a forecast of more cold weather in Brazil, and fell when the Brazilian government announced an auction of some of its coff ee stocks.Th en the second frost hit, prompting a new round of increases.After another week of increases, GNI, a London broker, in its "International Futures and Options Briefi ng" newsletter, said that the reports of frost damage had been exaggerated, and a CFTC report showed that large speculators were buying heavily, betting on further increases.
Th ese reports sent prices down again for several days.Th en, when Brazil' s NCD released its offi cial estimate of the damage, prices soared again, because it was worse than anyone had expected.As the date approached for the release of the FAS offi cial estimate a couple of weeks later, prices fell again, as traders expected it to be more optimistic than the offi cial Brazilian estimate.When it was not much more optimistic, prices rose again.And so on. 35able  above shows Brazilian and total world production and exports for this period.Th e fi nal production fi gure for - turned out to be closer to the Brazilian NCD estimate than to FAS' s.Brazil made up some of the diff erence from stocks, and its exports were only about  million bags lower than in -.Total world exports were actually higher in - than in -, as other countries, particularly Colombia and the Central American countries increased their exports.Th ey had been building up stocks since the ACPC retention plan had gone into eff ect in late , and they were able to take advantage of the higher prices and sell off these stocks.By -, world coff ee production had recovered from the period of low prices in -, and was adequate to meet world demand for the fi rst time in four years.Th e lower line in Figure  shows the ICO indicator price during this period.After a rapid run-up following the frosts, it slowly drifted downward over the next two years.

Comparison
Th ere are several reasons why the reaction of prices to the  frost was more drastic than to the  frost.One was the severity of the frost.Th e  frost caused more lasting damage to production in Brazil than did the  frost; after the latter one, production recovered to previous levels within a year.Another reason was that Brazil accounted for a lower percentage of total world production in  than in , so the impact of the frost on world coff ee supplies was smaller.In addition, there was panic buying at all levels following the  frost: importers, roasters, retailers, and consumers were all hoarding coff ee, and that prolonged the run-up of prices.Th is also caused a steeper decline after prices peaked, because everyone used up their extra supplies before buying more.Th e Congressional Research Service estimated that this panic buying was the primary factor that drove prices up about twice as high as they would have John M. Talbot  Information, Finance and the New International Inequality  been if they had been based solely on availability of supplies (New York Times, November , , p. ).A major reason for the panic buying was uncertaintyno one knew how much coff ee would be available, and after many months of price increases, people started to expect the worst.Importers were willing to pay more to be sure of obtaining coff ee, and passed the higher prices up along the rest of the commodity chain.As importers and roasters paid the higher prices, they drove futures prices up by hedging their purchases.Th e dynamic following the  frost was completely diff erent.Futures prices jumped immediately, and roasters followed with immediate wholesale price increases that would cover their increased costs of buying green coff ee in the future.Traders and roasters had their sources in Brazil, who were closely monitoring the situation.Th ey had a better idea of the extent of the frost damage and of the size and condition of Brazil' s coff ee stocks than they had been in .In fact, there was almost too much information about conditions in Brazil available in .In addition to the NCD and FAS, there were numerous estimates of the frost damage fl oating around in .Commodity traders like ED&F Man and GNI published their estimates; analysts such as F.O. Licht published theirs; and even Accu-Weather released one.Th is plethora of information, and the fact that commodity funds and other speculators carried much more weight in the futures exchanges in , caused a highly volatile situation in the futures prices.And this was translated into volatility of actual green coff ee prices, because by that time the futures price determined the physical price.Despite this volatility, there was no prolonged run-up of prices to cause hoarding.Th e market was moving much more rapidly; after quickly climbing to a peak only a few months after the frost, futures and green coff ee prices quickly began drifting downward.However, wholesale and retail prices remained high.
Another major diff erence was that there was no attempt by producers to manipulate the market once prices began to decline in , as they had done in -.Th e volume of trading on the CSCE was around -, contracts per month in ; by , it was around , per month.Th e expense of trying to squeeze the market in  would have been astronomical.And once again, information systems were much better in .Brazil and El Salvador were able to pull off their operation in  in relative secrecy.Th ere were rumors fl oating around, but no one really knew what was happening at the time.Th e details only came out a few years later.Such a level of secrecy would have been impossible in , particularly regarding the buying of large quantities of physical coff ee and shipping them from one place to another.In fact, by , the largest traders were in a better position to manipulate the market than states or fi rms in the producing countries.Th eir integrated strategies involved both buying physical coff ee and speculating, and these could be combined in a variety of ways to their advantage.For example, they might sell futures contracts to lower the price slightly, and then quickly fi x the price of some physical coff ee that they had bought at the lower price.With the huge amounts of coff ee they dealt with, even a movement of a fraction of a cent could generate a signifi cant profi t.
But the most striking diff erence between  and  is revealed by comparing Figures  and , showing the trajectories of prices surrounding the two frosts.In Figure  shows retail coff ee prices generally following the trajectory of physical green coff ee prices, with a lag.Th e gap between the two lines narrows from the beginning of  through the point where green coff ee prices are roughly equal to retail prices, in March .Th is shows that, while retail prices were increasing, coff ee producing countries were able to increase the share of these prices that they retained.Since it took several months for green coff ee to get from the producing countries to the supermarket shelves, the coff ee manufacturers were never losing money, but they were being squeezed.However, the gap between the two lines is considerably wider after the prices peaked than it was before.Th is shows that the roasters more than made up for it on the downside of the price spike.Th ey lowered their wholesale prices, but by a smaller percentage than green coff ee prices were falling.Th e wider gap after the frost already begins John M. Talbot  Information, Finance and the New International Inequality  to show the increasing market power of the TNC roasters, who were able to increase their margins even in a falling market.
Figure  shows a diff erent pattern.First, the gap between green and retail coff ee prices is much larger before the frost in this Figure .Th is means that producing countries were receiving a much lower percentage of the total income available from coff ee sales in the early s than they had in the s.Second, roasters' responses to the increase in green coff ee prices were much faster in .In , the fi rst increase in wholesale price was announced by General Foods, eleven days after the frost.In , Folger' s increased its prices three days after the fi rst frost struck, and the only reason it took that long was that the frost hit on a Saturday night.Th ere were no futures traded for more than a day after news of the frost was fi rst reported.Futures prices jumped on Monday, the next trading day, and the roasters raised their prices on Tuesday.Th e phenomenal . increase in wholesale prices for the month of July stands out in Figure , so that the lower line briefl y approaches, but comes nowhere near, the upper.Even as green coff ee prices were increasing in , producing countries were not able to signifi cantly increase their share of total income.Part of the reason for this was that green coff ee and futures prices reached their peak so quickly after the frost, and began to trend down.And, as was the case following the peak in , the gap between the two lines only grows larger again after prices peak.Th e diff erences between these two Figures are literally a graphic illustration of the way the balance of power had shifted away from states and fi rms in the coff ee producing countries and toward the giant coff ee TNCs.

CONCLUSION
I have argued that a new form of international inequality has been established, superimposed on the old form.Th e old form was established during the period of British hegemony in the late  th Century.It was based in a global division of labor that assigned diff erent areas of the world to diff erent roles in an industrial production system.Th e position of colonial, semiperipheral, and peripheral regions of the world in this division of labor was primarily one of suppliers of raw materials to the industries of the core countries, and of consumers of their output.Th is old international inequality was consolidated during the period of US hegemony from World War II up to about .It was updated and modernized for a world that had undergone decolonization.Some routine, labor-intensive manufacturing processes were spun off from the core to the semiperiphery, but they were under the control of core-based TNCs and integrated into their global production systems.Most of the periphery, and sections of the semiperiphery, remained suppliers of raw materials to the core.Th e old system of international inequality was still fi rmly in place.After the crisis of the US regime of accumulation, around , a period of fi nancial expansion was initiated.I have argued that during this period, a new form of international inequality has come into being.While the old form relied on control of production processes and the fl ow of goods, the new form is based on control of capital and fl ows of information.Th is new form has been superimposed on the old form, thereby increasing the overall degree of inequality in the world.I have illustrated this argument by considering the case of coff ee.
Coff ee was one of the central commodities involved in the establishment of the old inequality.It was one of the fi rst commodities produced in mass quantities in the colonies for consumption in the industrial core.A large number of peripheral countries came to depend, and still depend, on the money earned in their roles as suppliers of green coff ee to the major consuming markets in the core.Although coff ee producers were able to organize, and fared better than the suppliers of many other raw materials, they were still in a disadvantaged, unequal, position in the world division of labor, and were kept there by the economic and political power of the TNCs.Th e coff ee TNCs controlled the fl ow of coff ee into the core markets, its manufacturing, and its distribution.Th ey used this position to maintain control over a large share of the income and profi ts generated by the global production system that provided the coff ee.After the fi rst frost, producers were able to take advantage of the shortage and the uncertainty to increase their shares of these income and profi ts.Th is proved to be only a short-term advantage, and the TNCs were able to more than make up for their losses in the longer term, after prices had begun to come down.But even in a declining market, the producers were able to play the futures markets in order to make a profi t and slow the decline somewhat.Developments during the fi nancial expansion have exacerbated the situation of coff ee producers.Th e TNCs are larger and more powerful, and they have consolidated their hold over the core markets.Th e states in producing countries are weaker, less able to control the growing and processing of coff ee that are carried out within their borders.While a few fi rms based in the semiperiphery have risen to the status of global TNCs, able to compete with the core-based TNCs, they are the exceptions, and the latest evidence suggests that even they are being squeezed out (Anderson and Cavanagh ).During the fi nancial expansion, control over capital and information has become an even more important source of power than control over production and fl ows of goods.And the coff ee TNCs have used this control to further extend their advantage.Th e price that producers get for their coff ee is now tied to the value of paper contracts that depend as much, if not more, on rumor and speculation than on the actual supply and demand conditions for coff ee.And with their vast networks of capital and information fl ows, the TNCs are in a much better position to capitalize on this situation than are coff ee producers.During the second frost, the coff ee TNCs were able to pull off a pre-emptive strike, raising their prices so quickly that not only did producers not even derive any temporary benefi ts from the shortage, but their share of overall income and profi ts declined even further.Th is is the eff ect of the new international inequality.

Total Volume of Coffee Futures Trading, New York and London, and Total World Imports of Green Coffee, 1980-1995, In Millions of Tons
Source: ITC (1996), Table14, p. 72.
32.. Tea and Coff ee Trade Journal, August , p. ; World Coff ee and Tea, August , p. ; September , p. ; October , p. ; New York Times, June , , p. D; June , , p. D; July , , p. D; Financial Times, June , , p. ; July , , p. ; USDA, FAS, Tropical Products: World Markets and Trade, June .Finance and the New International Inequality  bags, while the FAS estimated - million.Th e Brazilian government offi cially disputed the FAS forecast, saying that it was underestimating the frost damage.Th en the German commodity analysts F.O. Licht weighed in with their estimate of  million bags on August .

per pound) Figure 2 -ICO Indicator Price for Green Coffee and Average Retail Price of Roasted and Ground Coffee, Monthly Average, January 1973 - December 1996
John M. Talbot  Information, Finance and the New International Inequality 