1 The first extensive multilateral coffee agreement was the Inter-American Coffee Agreement which stabilized coffee trade in the Western Hemisphere during World War II. This agreement was an unusual case since it was motivated primarily by the market disruptions of the war and a United States’ desire to guarantee Latin America's support for the Allied cause.

2 While most countries adhered to the provisions of the 1962 agreement as of October 1, 1962, it did not become official until December 27, 1963, when U.S. ratification was received by the United Nations. It was still another two years before the U.S. Congress passed the legislation enabling the United States to carry out all its obligations under the 1962 agreement. On February 19, 1968, the 66-nation International Coffee Council approved the 1968 agreement, which has maintained continuity in multilateral coffee price supports despite a delay in ratification by the member governments.

3 The uniting force behind the agreements has been a common desire to prevent major price declines. Beyond this, however, there are major differences among producers regarding the long-run division of the market. On one side, Brazil and Colombia want to maintain their relative market shares. On the other side many small producers, particularly in Africa, do not feel that historical market shares should be the sole or major basis for setting export quotas. On the basis of production potential, changing market conditions and economic dependence on coffee, some African nations feel their exports should be allowed to grow faster than the 2 to 3 percent growth in world coffee consumption. The 1962 agreement essentially called for a sacrifice of the rapid growth of African exports. In practice, however, the granting of emergency quota increases when one of the small producers had an unusually large crop, the failure of some exporters to adhere to their quotas, and the maintenance of minimum export prices by Brazil and Colombia have resulted in further reductions in the market shares of Brazil and Colombia. Some of these reductions are incorporated in the 1968 agreement that reduces Brazil's and Colombia's market shares by 1.9 and 0.6 percent respectively from their 1962 levels.

4 Production control policies have generally been inadequate and their adoption seems to have had little relation to commitments under the 1962 agreement. Prior to the enactment of the International Coffee Agreement, 1962, Brazil, Mexico, and the Ivory Coast had adopted schemes to reduce productive capacity. These were unilateral steps bearing little relation to any international agreement other than their mutual motivation by the world coffee glut. Passage of the 1962 agreement did not lead to a general adoption of production controls and, in fact, concern with such controls decreased when prices rose during the second year of the agreement.

5 All references to bags in this study refer to 60 kg. bags of green coffee.

6 Up to 20 percent of a country's contribution can be used to finance an approved project in any country. Chapter XVI, article 54 of the Internatinal Coffee Agreement, 1968, explains the fund.

7 Examples of this sentiment can be found in Jornal do Brazil, November 1, 1966, p. 13; O Estado de São Paulo, October 29, 1966, p. 16; and The New York Times, April 9, 1967, Section 3, p. 1.

8 Most of the information presented in this section is from two GERCA publications: Programa de Trabalho, December 1967, and Relatório 1967.

9 While the cost per hectare was higher for the new techniques, the cost per bag of coffee was considerably less. To ensure that farmers had access to the latest information regarding coffee production, in 1962 GERCA conducted a course on this subject attended by all IBC agronomists plus representatives of the Bank of Brazil, the Ministry of Agriculture, and state agricultural agencies.

10 Initially the planting, not to exceed 40,000 trees, was to be done as closely as possible in four equal annual installments. The planting could be completed in less than four years as long as no more than 4000 trees were planted in one year. The annual limitation on new plantings was subsequently abolished as long as the total did not exceed 40,000 trees.

11 Unfortunately, from the view of the social repercussions of eradication, about 44 percent of the liberated lands were turned into pasture, which uses only about one-tenth of the labor per hectare as coffee. Since coffee is one of the more labor-intensive crops in Brazil, the labor displacement from coffee diversification has been a serious problem.

12 GERCA, Programa de Trabalho, December 1967, p. 25.Google Scholar GERCA, through authorized banks, financed 60 percent of the value of the substitute crops planted on the eradicated lands. The loan did not come due until 60 days after harvest and the interest was 6 percent plus a 1 percent inspection fee with no monetary correction. Since these loans were for a much shorter time period than the loans for coffee replanting, the subsidization element arising from inflation was less for diversification.

13 Actual production figures are not given in view of the discrepancy among various series. For example, the Foreign Agricultural Service of the U.S. Department of Agriculture lists the 1964/65 crop (i.e., the crop for the marketing year that begins in July in Brazil) at 10 compared to the IBC's 18 million bags. One of the major reasons for the different estimates is the extent to which various sources try to account for market withholdings. Sales of prior crops were particularly large in 1964/65, which helps account for the unusually large differences in the estimates for that year. While this may cause sharp differences in the estimate of any one year, they tend to average out over a two to three year period.

14 This money was to be allocated through 1970. The minimum allocation was equivalent to about NCr$72 million at 1964 prices and compares with NCr$24 million applied to all aspects of the GERCA program from June 1962 to June 1966. The contribution quota is the difference between the export price and the return to the exporter, which is fixed by the government. The contribution quota is retained by the government and has recently amounted to about half the export price.

15 The major difference from the first stage was that the lands could not be replanted to coffee or turned into pasture. Pasture was excluded in order to minimize the social disruptions from diversification.

16 As Table 2 indicates there were also geographical payment differences. These apparently reflect the IBC's desire to eliminate most coffee from Zone A which is particularly susceptible to frost and a belief that the desired eradication could be attained with lower payments in Zone C, which is primarily Espirito Santo, because of the very low yields in that region.

17 For example, in 1964 the premature belief that excess coffee production had been eliminated led to the adoption of policies highly favorable to coffee planting.

18 This is based on data provided by Divisão de Economia Rural, State of São Paulo, March 1967.

19 GERCA, Relatório, 1967, p. 12.Google Scholar

20 Ibid., pp. 11-12. GERCA estimates an additional 350 million trees were eradicated outside the programs but there is no evidence to support this figure which appears to be grossly inflated.

21 The data on coffee area by state are from Anuário Estatistico do Brazil, IBGE and the eradication data are from the 1965 and 1967 GERCA Relatório(s). The data on coffee area are very unreliable and in some years negative new planting was indicated for some states. Their use here is not to provide an accurate estimate of new plantings but to provide one more bit of evidence regarding the inconsistency of GERCA's conclusion that supply has been reduced to demand. If we were to accept GERCA's figure of 350 million additional trees eradicated outside their programs, the implied new coffee planting would be about 650 million trees. Use of GERCA's figure that on the average one new tree offsets four old trees would imply that productive capacity actually increased between 1962 and 1966, which is not likely.

22 Several knowledgeable people, including one IBC official, have indicated to this author that they feel 28 million bags is a more accurate estimate of Brazil's current average productive capacity than the IBC figure.

23 Broca damage results from a small insect which attacks the coffee cherry during its early stages of development.

24 The expansion of coffee into western Paraná has concentrated much more coffee in frost-prone areas. However, the frosts in August of 1963 and 1966 represent only the second and third major frosts in these areas since the late 1940s when large scale coffee planting started in western Paraná. Unless the IBC expects the frequency of major frost damage to have increased, there is little justification for using the 1964/65 to 1967/68 period without some adjustment to indicate average productive capacity. And if frequent frosts are now expected in some coffee areas, the social and economic disruptions caused by a frost should be sufficient cause to continue to strengthen eradication-diversification efforts in these areas.

25 This proposal is presented in GERCA, Programa de Trabalho, December 1967, pp. 25–30.Google Scholar This proposal was to be subject to revision pending interpretation of the 1965 aerial survey of the coffee areas. However, while this survey should clear up some of the uncertainty regarding the number of coffee trees in Brazil, it will contribute nothing definitive regarding productive capacity. At best this will have to await the results of the more detailed aerial survey currently being undertaken. It is very unlikely, therefore, that in the near future GERCA will find sufficient cause from the surveys to reinstitute their eradication-diversification efforts. On July 3, 1968, a senior official of the Brazilian Ministry of Finance indicated to the author that Brazil would probably not adopt this scheme because of its high cost. No concern was expressed over the likely effects of the program on coffee capacity.

26 GERCA, Programa de Trabalho, December 1967, p. 27.Google Scholar

27 Data relating yields and production costs for a group of farmers in Paraná rindicate increases from 6 to 24 and 6 to 48 bags per 1000 trees would reduce production costs per bag by about 47 and 77 percent respectively. See a study entitled “Brazil's Coffee Policy” which was done by this author for the Agency for International Development. It appeared as an unclassified airgram numbered TOAID A-180, Rio de Janeiro, July 28, 1967.

28 This is based on a conversation the author had with an IBC official in July 1968. Compared to the author's own analysis this seems to be a conservative estimate of the relative attractiveness of coffee. When allowance is made for differences in marketing conditions, credit availability, and risk, which have been important factors enhancing coffee's relative appeal, in mid-1968 coffee with average yields was one of the more profitable agricultural investments. See the author's study referenced in note 27, plus another study the author did as a consultant for AID, entitled “Coffee Eradication and Diversification in Brazil: an Evaluation of Past Accomplishments and Future Requirements,” July 1968.

29 This conclusion is supported by the author's studies referred to in notes 27 and 28 and a recent study entitled “Erradicação do Café e Diversificação da Agricultura Brasileira,” May 1968, by Stahis Panagides of IPEA, which is a division of Brazil's Ministry of Planning. This latter study is an application of linear programming techniques to determine the optimal investment pattern of a farm in the Mogiana region of the State of São Paulo. The study concludes that coffee planting is the most profitable investment on this farm and the validity of this conclusion is attested to in a letter from the farmer indicating his intention to increase coffee planting during 1968 in view of coffee's excellent prospective returns.

30 On July 9 and 10, 1969, (eight months after completion of this study) a frost, described by the Foreign Agricultural Service of the U.S. Department of Agriculture as “the most extensive … in Parana's coffee history,” struck the Northwestern part of the State of Parana where the bulk of Parana's coffee is grown. The Foreign Agricultural Service estimated that the frost would reduce the 1970/71 Paraná crop from 15.2 to 3 to 5 million bags with a negative but much smaller impact on the 1971/72 crop. Clearly for the next several years average coffee production in Brazil will be well below the 28 million bag level discussed in the text and probably substantially below the 25 million bag estimate of average productive capacity made by the IBC after termination of the coffee eradication and diversification program in 1967. In addition, the evidence is increasing that substantial frost damage to coffee in northwest Paraná can be expected to be much more frequent than indicated by the experience of the forties and fifties. From a longer-run perspective, events following the frost appear to be enhancing the likelihood that tremendous excess capacity at prices satisfactory to exporters may emerge in the mid to late 1970s. For example, the governor of the state of Paraná responded to the frost by asking the president of the Republic for higher prices for coffee growers, a policy that would benefit farmers inversely in proportion to the damage rendered by the frost and encourage new planting in areas outside the frost zone. By December 1969 coffee export prices had risen about 30 percent above their pre-frost levels despite the fact that world productive capacity and existing stocks are sufficient to meet demand at pre-frost price levels for several years to come. The combination of lower production and higher world prices will generate strong pressures within Brazil and other coffee producing areas for higher prices to farmers. With coffee planting being one of the most pofitable agricultural investments in most of the world's coffee areas prior to the frost, significant increases in coffee growers' prices would bring substantial new plantings and, with a four seven-year lag, substantial increases in production.