TO 53 RETURN TO REPORTS DESK WITHIN ONE WEEK This report is not to be published nor may it be quoted as representing the Bank's views INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT IMPORTANT FACTORS IN THE WORLD COFFEE MARKET May 7, 1954 Department Of Technical Operations Prepared By Dr Maurice F Perkins IMPORTANT FACTORS IINl THE WORLD COFFEE MARKET Table of Contents Page I Summary and Conclusions 1 II Introduction 1- 3 III The Postwar Coffee Market 3 - 5 IV The Present Situation (1953-54) 5 - 6 V The Situation to 1957-58 6 - 10 VI Long-Run Prospects 10 - 12 Appendix A 13 - 14 Tables 15 - 23 Charts 1. Real Coffee Prices and Stock Change 2 2. Changes in Importance of Markets 24 3. Changes in Importance of Producing Areas 25 --A- / - &CA / - 1 - I. SUMARY AND CONCLUSIONS 1. The nistory of coffee production and prices has been turbulent. The slowness with which production can be increased has resulted in periods of scarcity and high coffee prices, overplanting, and periods of surpluses and low prices. 2. This boom and bust situation has been peculiar to Brazil owing to the large proportion of world production concentrated there and the relative ease with which production could be expanded. 3. The 1930's was a decade of coffee surpluses and low prices. The first half of the 1940's saw the loss of a large segment of the market through war. Hence, this 15-year period represents one of relatively stagnant con- ditions with low coffee prices and large accumulated stocks in Brazil. Some expansion in production took place in favored locations in the 1940's. 4. In the postwar period consumption expanded rapidly as compared with production. The deficit was covered largely by drawing down the Bra- zilian stocks. 5. Higher prices stimulated the rate of expansion in new planting after 1949 but did not reduce consumption to the extent necessary to equate it with production. Stocks continued to decline. 6. Adverse weather in Brazil (drought and tnen frost) has reduced the volume of new coffee available in 1953-54 and 1954-55. 7. Low production and low stocks has created a critically short world supply situation. This brought about another sharp advance in coffee prices late in 1953. 8. The outlook is for a continuation of shortages and high prices for the next few years. 9. Except for frost damage, however, it is believed no shortage in productive capacity would exist. In other words no great increase in ca- pacity is required to meet 1953 levels of consumption. 10. The important issue is whether or not high prices will engender another cycle of overplanting in Brazil and thus a recurrence of surpluses. 11. More data is required before an answer to this question can be attempted. 12. In the event high prices since 1949 have increased the rate of production coffee prices wall need to decline in order to encourage more consumption. 13. In the event of resort to price support measures, surpluses will likely occur particularly of Brazilian coffees sometime after 1957. - 1 - II. RITRODUCTION 1. The July, 1953 Brazilian frost altered radically the world coffee supply for 1954-55 and 1955-56. There is certain to be further repercus- sions. Its implications, therefore, are important for the economic and financial situation of all coffee exporting countries. Coffee Price History 2. Chart 1 (page 2) shows annual average real 1/ coffee prices since 1913. Although the long-term trend is practically flat, the range of annual variation is wide. This extreme variation is chiefly attributed to 1) the difficulty of adjusting supply to demand, and 2) government price support policies. 3. There is a lapse of four to five years between planting a new tree and the first commercial crop. Supply, therefore, cannot be quickly varied in response to price changes, an insufficient supply and high prices can exist for several years before new trees can be brought to bearing age. During this "gestation" period it is difficult to judge when sufficient trees have been planted. The problem is intensified by the lack of good informa- tion as to the number of bearing trees, their ages and yields and the rates of abandonment and new planting. 4. Periodically overplanting has occurred. High prices nave engen- dered a boom in new planting and eventually, an oversupply of coffee and low prices. 5. This boom and bust sequence has been peculiar to Brazil largely because the availability of resources (chiefly land and labor) has permitted rapid expansion and because a large proportion of world production is con- Ocentrated there. The conditions for rapid expansion were not as favorable in other coffee producing countries and their influence on supply has been less spectacular. 6. Chart 1 shows tnat in the years 1924-29 real coffee prices 2/ were relatively high (18.3 cents per pound on the average or 30 percent above the 1920-39 price of 14.25 cents) 31 7. The late 1920's were the hey-day of coffee expansion in the State of Sao Paulo, Brazil. High prices led to extensive new planting, whtch was prolonged by the success of Federal and State price support measures. The result, as shown in Chart 1, was an excess of exportable production over world imports which extended from the late 1920's through the 1930's. 1/ Prices expressed in terms of the average value of the dollar in the in- terwar period, 1920-39. 2/ Actual coffee prices are given in Table 1 in the Appendix. / All prices refer to Santos 4's at New York unless otherwise stated. UNITED STATES: REAL COFFEE PRICES AND STOCK CHANGE" SANTOS NO 4, NEW YORK (CENTS PER POUND -EXPRESSED IN TERMS OF 1920-39 DOLLARS)* STOCK CHANGE EXPORTABLE PRODUCTION MINUS WORLD IMPORTS 40FEB +tIO _ __ Lu. :* :. 10-DESTRUCTION IN BRAZIL so REAL COFFEE PRICES 3 020 0 F 111 1 I1l l l i l l I a l l 0 STOCK HANGEAFTERALLOWNG FO 1913 1915 1920 1925 1930 1935 1940 1945 1950 1954 *Actual prices deflated by the BLS General Wholesale Price Index and " Stock change is difference between exportable production 0 expressed in terms of 1920-39 average coffee prices (14 24 cents per pound) and world imports IBRD- Economic Staff (R) No 888 - 3 - 8. Practically all of this surplus was located in Brazil. The volume became so great that large scale destruction of stocks was undertaken. The dotted line in Chart 1 shows the net stock change on a world basis after allowance for annual destruction in Brazil. Is Another Boom in the Making> 9. Since 1949 real coffee prices have been very much higher than in the late 1920ts (see Uhart 1). Tne 1953 average price was 120 percent above the 1920-39 average in real terms. As of February, 195 the real price was more than three times the 1920-39 average (see Chart 1 above). The average for 195I will be much higher than that for 1953. 10. High coffee prices since 1949 have already caused an increase in the rate of new planting. Has this rate been high enough to result in over- planting9 Will present very high prices produce another boom and bust se- quence? These questions cannot be answered at present with any degree of accuracy because there are too many variables of unknown magnitude. In the sections below it is attempted to outline the sequence of events leading up to the present situation and to indicate the gaps and inadequacies in the available information. III. TE POSTWAR COFFEE MARKET Surplus Stocks and Stagnation 11. Although the war isolated most coffee countries from the European market most Latin American exporters except Brazil were able to market their coffee in the United States. The bulk of the unused supplies, therefore, were located in Brazil. In 1945-46 these amounted to 16.8 million bags 1/ (see Table 2). 12. High Brazilian stocks and U.S. price ceilings acted as a drag on coffee prices during the war and early postwar years. 13. It was generally believed this set of circumstances made coffee a relatively unattractive crop. Little new planting prior to 19h9 was re- ported except in areas particularly favorable to coffee or where alternative crops were not attractive. During the war years some Central American coun- tries added to their productive capacity, in Brazil new planting occurred in remote areas of Sao Paulo and in Parana. In Africa and Colombia produc- tion increased steadily throughout the 1940's. 14. The net effect of these tendencies is shown in Chart 3 (see page 25). The long-term trend of exports and exportable production declined un- til 1945. Since then a reversal in trend has occurred. An upward trend in production, therefore, was taking place prior to the 1949 price increase. 1/ Each bag weighs 60 kilos or 132.2 pounds. -4 Consumption Exceeds Production 15. Postwar world consumption was maintained at a nigher level than exportable proauction 1/ by drawing down on the Brazilian stock (see Table 2 and Chart 1). Such a situation could not continue indefinitely; in the absence of an increase in production, consumption eventually had to decline to production levels. The 1949 Price Rise 16. By the 1949-50 season Brazilian stocks had been reduced to 7.5 million bags and were to fall further (see Table 2). Recognition of the changing situation causea coffee prices to rise sharply in the fall of 1949. Within a few weeks Santos h's rose from 27.5 to 55 cents per pound. This higher price level persisted with little variation from late 1949 to late 1953. Increasing Consumption 17. Higher prices resulted in a drop in U.S. per capita civilian con- sumption after 1949. The effect on the volume of U.S. net coffee imports, however, was largely offset by increased military usage and a larger popu- lation (see Table 4) 18. In Europe demand remained strong as postwar reconstruction pro- gressed. Gonsumption was still considerably below prelvar and imports con- tinued to increase in spite of high prices 2/ 19. The 1949 price rise, therefore, slowed down the rate of increase in world consumption but the check was insufficient to balance it with pro- duction. The drain on Brazilian stocks continued. New Planting After 1949 20. Higher prices stimulated new planting. In Brazil a significant spurt occurred in Parana, the newest coffee area. This plus attempts to maintain old trees in production and some new planting in other states presaged an expansion in Brazil's exportable capacity 3/ Private observers h.ad forecast a 3 million bag increase by 195-55ove_rthe 14-15 million bag average of the past few years. 21. In Other iestern Hemisphere countries a reviving interest in cof- fee nas occurred. The last three Colombian crops also show moderate increases, but this has not been reflected in the volume of exports due to adverse weather in 1949-50 and 1950-51 plus stock accumulations in 1952-53 1/ Lack of adequate data on domestic consumption in producing countries re- stricts the study to the export market. Hence throughout this report the production data refer to exportable surpluses and the consumption data to imports. 2/ In addition to higher green coffee prices increased import and excise taxes had raised the average European roasted coffee price considerably above U.S. levels (see Table 7). 3/ The term "capacity" is used here to mean the level of production under normal weather conditions given the number of trees and their condition. 22. In Africa a significant postwar increase in exports has taken place (see Table 5 and Chart 3 page 25). Exports have more than doubled as com- pared to prewar, in the past three years they have averaged 1 percent above the 1945-49 level. Increasing Capacity 23. The degree of increase in productive capacity following the 1949 price rise is not known with certainty because aata on new planting is in- adequate and the new trees have not had time to bear. From 1945 to 1953, however, the trend of proauction for export has been rising at an annual rate of 2.5 percent for Brazil, 3.6 percent for the Western Hemisphere, excluding Brazil and Colombia, and 2.5 percent for the world as a whole. There has been a persistent long-term trend increase of 2.3 percent per annum for Colombia and 6.2 percent for Africa (see Chart 3, page 25). 24. On the basis of these trends and an absence of bad weather a theo- retical increase in world production to about 33 million bags could be ex- pected for 1954-55. This is about equivalent to the present level of con- sumption. 25. The implication is that a reasonable balance between production and ponsumption would have been a possibility for a few years at a price level of between 50 and 60 cents. No considerable deficiency in productive capacity, therefore, is believed to have existed prior to tne 1953 Brazilian frost. IV. THE PiRESENT SITUATION (1953-54) Poor 1953-54 Brazilian Crop 26. During the 1952-53 growing season, whilst the 1953-54 Brazilian crop was on the treesl - drought conditions prevailed. Official estimates of the 1953-54 crop were progressively lowered fron 16.9 to 14.0 million bags.2/ 27. The effect of the drought in Brazil was to reduce the estimate of 1953-54 world exportable production to 31 million bags. It became evident, therefore, if world consumption did not decline to a similar level, a further reduction in stocks would occur. Tightening Supply 28. At the beginning of the present season (July 1, 1953) Brazilian port stocks were 3 million bags (see Table 2). Interior old crop stocks 1/ The first flowering occurs in August/Septemoer; the ripe cherries are harvested and processed from larch/April on and the crop moves to market after July 1. 2/ These estimates include Brazilian port and coast consurption of some 0.9 million bags which must be deducted to arrive at the exportable surplus. - 6 - were small. Since then Brazilian exports have exceeded exportable produc- tion. It is currently estimated that by June 30, 1954 port stocks will be down to 1.5 million bags. This means that there will be very little stock left in Brazil to bolster the 1954-55 supplies. Frost in Brazil 29. July 4, 1953 a bad frost was reported in Brazil. In Parana, the newest producing area, many trees were heavily damaged. Hopes for a substan- tial increase in production have now disappeared and a low 1954-55 crop is in prospect. The Rise in Prices 30. Prices began to rise sharply in late December, 1953 (the Noveiber average was 58 cents per pound). As of March 31st the market had gone up to 91.5 cents but by April 23rd it was down to 86 cents. 1/ 31. This price rise is the result of anticipatory market forces. Statis- tically there is no coffee shortage at present in the sense that levels of consumption have to fall. But there is a probability that this will happen in future. In anticipation of this sizuation, therefore, buyers are becoming more eager to buy and sellers more reluctant to sell. Higher prices, there- fore, are conditioning the market to an expected reduction in supply by forcing consumotion downwards. V. TfE SITUATIUi TO 1957-58 32. For the next four seasons the world's coffee capacity is unexpand- able in terms of the number of trees planted. The most important variables are the number of non-bearing, newly planted trees that will come into pro- duction each year less old trees which go out of production, the rate of recovery from the 1953 Brazilian frost and actual weather conditions. Yields per tree can be increased by using fertilizers and by adopting other known improved production techniques, but it is unlikely thet such measures w.11 be aaopted to the extent aecessary to make an appreciable difference in pro- duction in the short-run. A. 1954-55 Production Level 33. Until Parana and parts of Sao Paulo have recovered from the effects of the 1953 frost, Brazilian productive capacity will be below the 17 to 18 million bag pre-frost estimate given above (see Section III (20) page 4). It will take at least two years for substantial recovery. / Thus in 1954-55 1/ A peak of 96 cents was reached on the Futures Market April 3rd. 1/ This is an assumption requiring verification, the period may be longer. and 1955-56 at least crops well below nornal will be harvested. Private esti- mates for the 1954-55 marketing season are currently averaging around 13.5 to 14 million oags. 34. Production in other countries will approximate 17.5 million ba;s. Total 1954-55 exportable production, therefore, will be about 30.5 million bags. Total Supply 35. From a statistical point of view some further reduction in stock is possible as a means of supplementing the volume of coffee available for consumption in 1954-55. 36. The size of world coffee stocks July 1, 1954 is not knomn but a good guess would be 10 million bags.2/ The major proportion represents working stock which cannot be removed from the marketing pipeline. Sone quantity - perhaps 3 million o-s - can be made available for consumption. 37. For the 1954-55 season, therefore, a reasona ;e estimate of world supply available for consumption is 30,5 million bags W from the new crop and 3 million from stock or about 33.5 million bags. Consumption Levels 38. The recent sharp upward price movement, however, will undoubtedly reduce world consumption from the 33 million bag 1953 level. This may come about in several ways. There could be a switch to substitutes such as tea or cocoa. This iay not occur to any great extent because, for independent reasons, cocoa doubled in price in 1953 and tea prices are rising. Moreover, any significant degree of substitution for coffee would drive cocoa and tea prices still higher. 39. Other alternatives will be adopted, such as adulteration with chickory and similar extenders including a weaker brew. Finally greater use will undoubtedly be made of soluble coffee, which produces more cups of coffee per pound of green coffee than can be obtained by ordinary methods of preparation. A rapid expansion of soluble coffee consumption, however, depends upon how soon new manufacturing capacity can be created. 1/ From these estimates about 0.9 million bags must be deducted for Brazilian port and coastuise consumption. 2/ This assumes domestic consumption in coffee pioducing countries will not decline to the point where a larger proportion of the crop is exported. / This is made up of 1.5 million bags in Brazil, 0.75 in Colombia, 3.5 million in the United States, 2.0 million Western Europe and 2.25 million bags afloat and in other countries, W/ This estimate may be expected to change as oetter information becomes available. - 8 - 40. Higher coffee prices after 1949 reduced U.S. per capita consump- tion by some 10 percent or from 18 to 16 pounds (see Table 4). The increase since December,1953 has carried prices from 58 to about 90 cents or by 32 cents per pound as compared with 27 5 cents in 1949. A further shrinkage in per capita consumption may be expected; should it amount to another 10 per- cent per capita the effect would be about 2 million bags. On the other nand annual increases due to a larger population will continue. 41. In Europe the situation is complicated by the diversity of condi- tions. Import duties and excise taxes have raised coffee prices consider- ably above the U.S. level (see Table 7). The rise in green coffee prices, therefore, represents a smaller percentage increase to European than to Amer- ican consumers. Imports continued to rise after the 1949 price increase by roughly 0.7 million bags per annum. The total volume of imports remains somewhat lower than prewar and the per capita supply is lower still. As a guess higher prices will cause a smaller reduction in consumption in Europe than in the United States assuming no change in government import policies. 42 The basic point, however, is that, if the reduction in world con- sumption is insufficient to prevent a further reduction in stock in 1954-55 the situation in 1955-56 would become more vulnerable because tappable world stock available to meet possible emergencies would be further reduced Prices 43. Even if the effect of higher prices on the level of consumption were known, it would not be possible to name the equilibrium price for the next season without taking into consideration the attitude of the coffee trade. This largely hinges on 1) the actual adjustments in the levels of consumption which finally emerge, 2) expectations of the size of the next coffee crop (i.e. 1955-56), and 3) the trade's readiness to reduce stocks which are already at very low levels. 44. July and August is the season of frost danger in Brazil. To this may be added other unfavorable weather news droughts, insect damage, fungus diseases, etc. With the present short supply situation and jittery market atmosphere such events could have an exaggerated effect coff prices 45. Another important factor is the policy/e-cof e producing govern- ments, especially when prices show a tendency to fall. lould coffee coun- tries try to maintain high prices for an uneconomically long period by hold- ing a part of the supply off the market? 46. Under such conditions no accurate forecast can be made of the prospective level of coffee prices in 1954-55. Conditions necessary for a price increase from present levels would be 1) no substantial reduction in world consumption, 2) poor prospects for the 1954-55 crop outside Parana, and 3) price support programs in important producing countries. The reverse of these could cause prices to Lr mi ud . " 0&_CZ L. -9- B. 1955-56 47. In 1955-56 the Brazilian crop will still be below normal as a re- sult of the frost, although with no further adverse weather, production should increase over 195h-55. A level of consumption in 1954-55, which would further reduce stocks, plus adverse weather could cause a further deteriora- tion in supply conditions in 1955-56. C. 1956-57 and 1957-58 48. In these years Brazil should have recovered from the worst effects of the frost. It is reasonable to expect that by that time the level of world capacity would rise to the point where 33 million bags could be sup- plied out of annual production, i.e. the equivalent of 1953 consumption, and, perhaps, allow some addition to stock. 49. It is unlikely, however, that coffee prices would weaken to any great extent before 1957-58 because 1) over a four year period further ad- verse weather could be encountered in one or more years 2) even a substan- tial surplus due to better than average weather conditions could be absorbed into stocks 3) coffee governments are likely to be tempted into price sup- porting measures and 4) Brazilian producers may allow a large number of existing old trees to go out of production. D. Summary of Intermediate Term Prospects 50. In order to clarify and summarize the exposition to this point the table below has been constructed. It shows the statistical balance be- tween stocks, production and consumption by crop years from 1951-52 to 1957- 58 under the various assumptions made above. These data are presented for illustrative purposes rather than as firm estimates. Prospective World Coffee Supply and Distribution 1951-52 1952-53 1953-54 1954-55 1955-56 1956-57 1957-58 (m i 1 1 1 o n b a gs Supply Opening Stocks 13.2 11.9 11.0 10.0 9.5 9.5 11.0 Exportable Production: Brazil 14.1 15.1 14.0 13.0 14.0 16.0a/ 17.5b/ All other 16.7 17.3 17.0 17.5b/ 18.0b/ 18.5/ 19.0b Total 30.5 32.h 31.0 30.5 32.0 3 . 36. Total Supply W W 2.0 70.5 7 L3 7 Distribution Consumption 32.1 33.3 32.0 31.0 32.0 33.0 34.0 Closing Stocks 11.9 11.0 10.0 9.5 9.5 11.0 13.5 Total Distribution Z 7 720 70.5 41.5 Z 0 173 Change in Stock (-1.3) (-0.9) (-1.0) (-0.5) (0.0) (A1.5) (/2.5) This estimate has been adjusted downwards to account for some linger- ing effects of the 1953 frost. b/ Rounded estimates of the extrapolated trends in Chart 3, page 25. - 10 - 51. Further reductions in Txorld stocks are a possibility up to and in- cluding the 1955-56 season because expected reductions in consumption will be insufficient to provide a balance with expected production. From 1956-57 on, however, production should increase to the extent necessary to meet a larger volume of consumption and still permit sorely needed additions to stock. 52. It should be reiterated that these conclusions are based on two major assumptions: 1) that Brazil will recover from the 1953 frost by 1956-57 and 2) that no further adverse weather will occur. Ib the extent that these assump- tions are incorrect the period of short supply will be extended. VI, LONG-RUN PROSPECTS 53. The period from 1958-59 on is differentiated from the previous four years by the ability of producers to expand capac2ty through new planting. This has been the traditional way in which output has increased in the past and it could play an important role in the future, 54. Long-run prospects for the coffee market are here considered in terms of the extrapolation of trends as modified by recent events. This involves first an estimate of the likely volume of consuription against which the trend in production may be compared. 55. In the long-run the world demand for coffee will increase. World trade has expanded from a preJTorldThar I level of 18 million bags to an average of 31 million oags over the past five years. 56. Data on import trade since 1930 are shown in Chart 2 (page 24) for various areas. The fitted lines of trend extrapolated to 1960 show the general direction of imports in each area, assuming no great variation in the underlying conditions influencing consumption, 57. The trend in U.S. imports has risen steaoily with the exception of the early war years. The annual rate of .nciease of the fitted line is 3.0 percent or about 0,7 riallion bags per annum. Since 1949, however, the actual level of imports has flattened out. In 1954 an actual decline may be expected due to high prices. After 1955 it is reasonable to expect a continuation in the long-term trend in growth as population increases. 58. In Europe the trend during the 1930's was fairly flat. During the war years imports declined drastically. Recently the volume of imports has increased rapidly and is now near prewar levels. It can be expected that higher prices will check the rate of increase out not eliminate it. In lieu of a fitted trend line, therefore, a freehand line having rate of increase of about 4 percent per annum has been drawn in. After 1960, how- ever, the rate of change may decline. 59. In Western Hemisphere countries excluding the United States the long-term trend has been upwards at an annual rate of 4 percent or about 0.06 million bags. Since 1949, however, the level of imports has ceased to increase. The effect of higher prices, therefore, is similar to that - 11 - in the United States, an absolute decline in imports may occur over the next few years. 60. For Africa, Asia and Oceania some decline in the level of imports is assumed. 61. The long-term trend of total world imports from 1930 to date reflects the absolute decline in European consumption during the war, It is thus abnor- mally low and requires an upward adjustment for purposes of estimating future consumption. Two such estimates have been made in order to illustrate the method of adjustment. They are both plotted in Chart 2 (page 24). 62. The first estimate is represented by the upper line labelled "pre-1953 trend". This shows the lixely trend in world imports (and con- sumption) at the 50 to 60 cent price level prevailing from laie 1949 to late 1950. The rate of annual increase is about 3 percent 63. The second estimate is represented by the lower line labelled "trend adjusted for higher post-1953 prices". It is more realisticAin that the absolute level has been shifted downwar s. The rate of annual increase, horever, remains the same, 1.e. 3 percent,1/ as there is no evidence to in- dicate that higher prices will affect the rate of growth. Supply 64. During periods of low stocks the )evel of trade is largely deter- mined by levels of production. Moreover, supply has fluctuated nuch more than consumption and is largely responsible for the alternate periods of scarcity and surplus. The key to market conditions, therefore, is to be found largely in future production movements. / 65. In Chart 3 (page 25) an extrapolation of expor has been made of total production since 1945. The obvious reason for no using the long-term trend since 1930 is that a fundamental change in the direction of trend occurred in 1945 and that the new trend is likely to continue. The rate of increase is roughly equ-Lvalent to the rate of inciease in the consumption trend i.e. about 3 percent per annum. 66. It is important to note, hoi,ever, that the individual production trends have been extrapolated from product2on data up to 1953 whereas new trees planted after the 1949 price rise would not come into the bearing stage until 1954 and trees planted after the 1953 price rise will not bear until 1958 or later. Hence it may be necessary to adjust upwards the extrapolated pro- duction trends to account for the effect of the two price increases on the rate of new planting. 11 1,te11i e eat E jeustment after 960fEurpea-constEption lvisout. The data in Chart 3 represent exports for all areas except Brazil which are exportable production. As most of these areas do not maintain stocks of any size the level of exports is a fair guide to exportable surpluses. - 12 - 67. To the extent that the price increases do stimulate new planting, the future rate of annual increase in production will exceed consumption. The result on a free market will be the appearance of surpluses and falling prices. 68. In the past the stimulation of high prices has had the greatest effect on production in Brazil. Due to the large proportion of production concentrated in Brazil the effects of the increases has been a predominant factor. Hence in order to judge the effects of the 1949 and 1953 price in- creases attention must be focused primarily on Brazil. 69. Whether or not a new coffee boom and bust is in the making in Brazil is debatable. Some observers believe that the basic conditions for another big boom no longer exist. They argue the remaining good coffee land is limited, it is high priced, and there is an insufficiency of cheap labor. If so, then the future growth trend may well be quite different to that in the past. A steady and moderate increase in output in Brazil and elsewhere through expansion in new areas and the gradual adoption of new techniques with increased yields per tree is less likely to breed over- expansion and subsequent surpluses than are gold rush conditions on the frontiers of Brazil. 70. Others claim this view to be erroneous. Brazil is a large country with vast frontiers about which little is known. Prices are now high enough to permit producers to overcome transport and other obstacles associated with new areas. Tne stage, therefore, is set for another boom. 71. There is inadequate information on prospective production condi- tions in Brazil to express an opinion on this question. The situation needs further study. One of the main difficulties is the paucity of published information. This subject is treated more fully in Appendix A. The Problem of Coffee Policy 72. The present situation holds both an extraordinary opportunity and a grave responsibility for coffee producing countries. 73. In the first place the fruits of the recent price rise must be considered as a windfall which cannot be expected to continue indefinitely, as coffee production may be expected to increase. Meanwhile there is op- portunity to utilize the extra earnings from high coffee prices for purposes of economic development. 74. Secondly, producing countries will need to resist the temptation to hold coffee prices at high levels. The adoption of high price support measures will encourage overplanting to an unnecessary degree. The in- evitable resalt would be a surplus disposal problem and low prices. - 13 - Appendix A Available Data on Brazilian Coffee Productive Capacity 1. Specifically the Brazilian situation contains four important un- knowns: 1) Annual Weather Variability. The extreme southern location of the new producing areas and the concentration of production in Sao Paulo and Parana make the weather factor extremely im- portant, particularly frost and drought. This is more true of the future due to the increasing importance of Parana and its vulnerability to frost. The weather factor, however, is not susceptible to accurate annalysis and forecasting. It can only be hoped that alternate good and bad crops will cancel out in the long run. 2) The Present Situation. The lack of available and reliable data on rates of planting and abandonment, tree numbers, ages and yields precludes an accurate estimate of present capacity on the basis of published information. The trend of new planting and production is upwards but the degree is uncertain. 3) Frost Recovery. Although it has been assumed above that Brazil will recover substantially from the 1953 frost in two years the rate and degree of recovery is dependent upon a number of variables of unknown magnitude. 4) The Effect of High Prices on New Planting. The level of Brazil- ian production after 195d will be largely dependent upon planters' reactions to the now extremely profitable coffee prices. Changes in the underlying economic situation in Brazil are such that it is not possible to state from published information 1) the exact extent of new planting to date, 2) the present rate of new planting, 3) the potential for an extensive increase in capacity, and 4) the probability of an intensive increase in capacity through better production techniques. 2. What is Needed. In order to gauge accurately the likely level of coffee productive capacity in Brazil over the next few years under nonal weather conditions the following information is necessary* 1) Number of bearing trees, their ages by groups and yields by age groups on a regional basis. 2) Annual numbers of new plantings and retirement of old trees. 3) The effect of the 1953 frost and the time period required for recovery. 4) The rate of adoption of new production techniques and high yielding varieties of trees. 3. For long-term estimates additional data is necessary covering the following: 5) Availability of new coffee land. 6) Attempts to replant on old coffee land. 7) General economic factors such as the relationship of coffee and competing cropst prices and costs, availability of labor etc. 4. What is Available. The information available to the Bank falls short of the minimum required for a sound judgment. Much of the required data is not amenable to collection by statistical methods and therefore depend on knowledge and experience. This is largely true for items 3) to 7) above. In the case of items 1) and 2) the published data is largely in- adequate because: 1) It is available only for scattered years. 2) Only portions of the necessary information are collected. 3) The annual returns are not always comparable due to varia- tions in the type of information gathered. - 15 - Table 1 Coffee Prices Basis - Green Coffee spot Prices in New York (cents per pound) Annual Brazilian Colombian Average Santos 4 Manizalez 1930 12.9 17.2 1931 8.8 15.6 1932 10.6 11.4 1933 9.1 10.5 1934 11.1 13.7 1935 8.9 10.3 1936 9.5 11.3 1937 11.0 11.6 1938 7.7 11.0 1939 7.4 11.6 1940 7.1 8.3 1941 11.3 15.0 1942 a/ 13.4 15.9 1943 a/ 13.4 15.9 1944 a/ 13.4 15.9 1945 a/ 13.4 15.9 1946 18.5 21.8 1947 26.7 30.1 1948 27.1 32.5 1949 31.8 37.4 1950 50.5 53.3 1951 54.2 58.7 1952 54.0 57.0 1953 57.9 59.8 a/ 0.P.A. ceiling prices. - 16 - Table 2 Brazil* Coffee Supply and Distribution (thousand bags of 132.3 pounds) Opening Stocks Exportable Total Stock July 1 Production Supply Exports Change 1945-46 16,800 11,394 28,194 16,007 -3,800 1946-47 13,000 12,608 25,608 14,373 - 00 1947-48 12,600 12,253 24,853 16,125 -2,600 1948-49 10,000 15,750 25,750 17,745 -2,450 1949-50 7,550 14,945 22,495 16,935 -1,800 1950-51 5,750 15,716 21,466 16,593 - 821 1951-52 4,929 14,120 19,09 16,333 -1,973 1952-53 2,956 15,120 18,076 14,968 - 4 1953-54 2,950 13,200 16,150 (14,650)1/ -1,500) 1/ This represents a balancing item rather than an estimate. Note* Small discrepancies occur between estimates on tnis and other tables due to rounding and the use of different sources. - 17 - Table 3 'horld Coffee Excortable Production as Related to Exports (thousand bags of 132.3 pounds) Crop-Year Calendar ExDortable Year Year Production 1/ Exports Difference 1945 20,634 27,393 $6,759 1946 23,695 29,277 $5,582 1947 27,266 28,608 -1,3 2 1948 26,271 32,297 $6,026 1949 30,650 34,737 4,o87 1950 29,306 29,288 - 18 1951 30,039 31,840 $1,801 1952 30,977 32,133 /1,156 1953 32,204 3,328 2,124 1954 30,800 Crop years July-June ending with year shown. These differences are illustrative but not statistically accurate due to differences between crop years and calendar years and inaccuracies of the procuction data Note: Small discrepancies occur between es- timates on this and other tables due to rounding and the use of different sources. - 18 - Table 4 United States' Coffee Supply and Distribution (thousand oags of 132.3 pounds) 1947 1948 1949 1950 1951 1952 1953 St;;ksJanuary 1 3,631 2,981 3,231 4,030 3,h63 3,380 3,302 Imports 18,867 20,947 22,060 18,424 20 301 20,266 21,017 Total Supply 22,495 23,925 25,291 22,54 23_76_ 2 24,319 Distribution Reexports 1/ 448 281 194 117 150 149 175 Military (381)a/ 248 589 482 1,254 509 533 Civilian Disappearance 19,450 20,168 20,478 18,392 18,980 19,686 19,601 Stocks December 31 2,981 3,231 4,030 3,463 3,80 3302 4,010 Total Distribution 22,498 23,928 25,291 22,454 23,764 23,646 24,319 Civilian Population (millions) 2/ 144.6 147.2 149.6 152.3 153.2 155.5 158.3 Per Capita Disappearance (lbs.) 17.8 18.1 18.1 16.0 16.3 16.7 16.4 Note: Small discrepancies occur between estimates on this and other tables due to rounding and the use of different sources* 1/ Green and roasted. 2/ Population eating out of civilian supplies. a/ Army surpluses made available for civilian use. - 19 - Table 5 World Coffee Exuorts (million bags of 60 kilos) Other Latin All Period Total Brazil Colombia America Africa Other 1909-13 18.12 12.64 0 77 3.70 0.18 0.82 1914-18 17.70 11.88 1.12 3.70 0.20 0.81 1919-23 20.56 12.80 1.86 3.91 0.42 1.57 1924-28 22.95 1y.09 2.33 4.28 0.65 1.60 1929-33 25.30 15.00 3.11 4.52 1.15 1.55 1934-38 27.02 14.58 3.83 4.91 2.03 1.66 1940-44 22.0M 10.81 4.37 4.18 2.52 0.16 1945-49 30.40 16.27 5.43 4.56 3.95 0.20 1950 29.23 14A 5.7 4.8d 4.65 0.39 1951 31.84 16.36 4.79 4.87 5.27 0.55 1952 32.13 15.82 5.03 5.69 5.12 0.47 1953 34.33 15.56 6.63 6.38 5.03 0.73 Note. Small discreDancies occur between estimates on this and other zables due to rounding and the use of dif- ferent sources. - 20 - Table 6 United States Per Capita Income, Amount Sent on Coffee and Coffee Consumption Per Capita Average Retail Percent of Annual Coffee Price of Roasted Cost per Per Capita Income Spent Average Consumption Coffee per Pound Consumer Income on Coffee (pounds of green coffee) 1925-29 11.9 48.8 cents $ 4.90 $ 651 0.75 1930-34 12.6 31.0 1 3.30 410 0.80 1935-39 14.2 24 2 n 2.87 507 0.57 1940-44 14.8 26.6 " 3.29 896 0.38 1945-49 18.1 43.7 * 6.66 1 334 0.49 19149 0. 1950 16 5 79.4 " 11.03 1,482 0.74 1951 16.8 86.8 " 12.23 1,581 0.77 1952 16.9 86.7 " 12.32 1,639 0.75 1953 17.1 88.1 " 12.68 1,703 0.75 Note: Small discrepancies occur between estimates on this and other tables due to rounding and the use of different sources. - 21 - Table 7 Retail Coffee Prices and Per Capita Consumption in Various Countries Retail Prices of Per Capita Consumption Ropstea Coffee of Green Coffee (dollars per pound) (poinds) 1/ 1938 1953 1938 1951 1952 1953 United States 0.23 0.88 16.5 16.8 16.9 17.0 Canada 0.36 1.03 3.3 6.1 6.8 7.1 Argentina 0.15 1.03 4.5 3.5 3.8 4.o France 0.64 1.60 10.0 7.8 8.3 8.5 Italy 0.83 2.20 1.8 2.2 2.9 3.1 Germany 0.67 2.62 6.3 1.8 2.6 2.1 Belg./Lux. 0.53 1,71 13.6 13.8 13.1 12.7 Sveden 0.39 1.22 18.4 13.0 1).8 15.7 Norway 0.37 1.19 14.6 10.2 14.5 11.5 Finland 0.64 2.37 15.8 7.9 11.7 12.1 Denmark 0.35 1.1f 20.2 8.2 10.2 14.7 Netherlands 0.73 2.40 13.2 3.6 4.2 6.6 Switzerland 0.45 1.30 9.1 9.1 8.3 9.2 Spain 0.66 2.23 0.5 0.5 0.5 0.7 Europe 0.55 1.74 5.7 n.a. n.a. 4.2 1/ Strictly per capita imports except for the United States. Note* Small discre-ancies occur between estimates on this and other taoles due to rounding and the use of dif- ferent sources. - 22 - Table 8 Annual Average World Coffee Imports (thousand bags) Country 1909-13 1925-29 1930-34 1935-39 1940-44 1945-49 1951 1952 1953 United States 6,828 10,659 12,025 13,900 16,366 20,618 20,357 20,274 Appa, Europe France 1,862 2,721 3,108 3,036 692 1,262 2,522 2,771 R&4L Scandinavia 1/ 1,253 1,632 1,806 2,034 461 1,191 1,474 1,861 995' Benelux 1,307 1,245 1,461 1,462 128 1,505 1,195 1,194 3VAA- Italy 438 754 695 568 72 463 889 1,016 1/1 Germany 2/ 3,016 2,006 2,398 2,663 191 118 677 942 /3// United Kingdom 174 291 299 273 546 711 721 720 5-O7 Switzerland 190 210 247 289 165 280 331 304 3/9 Other 1,550 1,343 1 328 1,185 368 460 825 832 748 Total 9,790 10,202 11,382 11,510 2,623 ,9906T3T 7,67- ,"7 Other Countries Argentina 213 391 359 4ol 482 542 476 518 *7/ Canada 106 188 243 302 448 569 669 738 S/3 South Africa 199 214 213 253 403 381 198 201 196 Other 617 1,022 950 1,230 1,338 1 571 1,324 1,174 /.z14 Total 1,3 7 1)763 2,)176 2,671 3,6 1261P~i World Total 17,753 22,676 25,132 27,596 21,660 29,671 31,658 32,545 ,'iot? 1/ Finland, Norway, Sweden, Denmark. 2/ Western Germany After 1945. Note: Small discrepancies occur between estimates on this and other tables due to rounding and the use of different sources - 23 - Table 9 Annual Average World Coffee Exports (thousand bags) Country of Origin 1925-29 1930-3k 193c-39 1040-4 1945-49 29-1 19 /95- West. qerisrhere Brazil 14,102 14,936 15,095 10,812 16,266 16,358 15,821 ,aiL. Colombia 2,451 3,149 3,972 4,370 5,429 4,794 5,032 g4 631L El Salvador 729 863 918 891 1,026 1,059 1,098 4 Guatemala 760 720 759 762 877 848 1,007 is Mexico 443 524 599 486 608 865 870 ; Haiti 549 532 436 342 459 417 541 372. Venezuela 835 774 733 527 521 308 498 1 Costa Rica 263 323 396 338 321 309 333 0! Domin Reaublic 70 126 187 162 253 290 442 ?7 Ecuador 108 157 223 197 175 273 340 A// Nicaragua 231 217 259 218 186 268 303 13 Honduras 25 27 29 30 57 137 138 0 Peru 11 37 47 26 20 27 45 /R Cuba -- 12 89 94 -- -- -- - Other 182 129 136 106 57 62 67 10 Total 20,759 22,526 23,878 19,361 26,255 26,015 26,535 291- Africa Fr Africa.1/ 66 230 726 861 1,413 1,830 1,947 2,049 Br. East Africa 304 489 752 712 848 1,213 1,325 / /02. Nngola 169 190 279 320 742 1,074 794 7/? Belgian Congo 7 102 289 L16 528 590 515 SS6' Ethiopia 210 277 209 130 291 458 424 #41 Other _ 9 _8 82 128 104 115 Total 761 1,297 2,313 2,521 3,950 5,269 5,120 So5! Asia and Oceania Indonesia 1,395 1,305 1,349 -- 40 a/ 402 313 XF IncLia 135 161 156 51 33 a/ 14 40 33 Other 146 132 111 106 124 140 125 /5O Total 1,676 1,603 1,616 157 197 556 478 :3P World Total 23,196 25,426 27,807 22,039 30,402 31,840 32,133 f V-L- 1/ French "est and -quatorial Africa, French Cameroons and Madagascar, a/ Four-ye-r average. Note There may be slight discrenancies in these estimates as compared vith other tables due to the necessity of using different sources. CHART 2 COFFEE CHANGES IN IMPORTANCE OF MARKETS (IMPORTS IN MILLIONS OF BAGS OF 60 KILOGRAMS) 50 0 1T-T 5 0 0 PRE 1955 TREND 40 0 40 0 3 0 0 ~TO T A L H G3 30 -TREND ADJUSTED FOR 30 0 200 Oo 20 0 UNITED STATES 10 0 10 0 %.. EUROPE 80 - 80 60 60 50 - 50 40 1t40 30 -30 S I AFRIA ASIA AND OCEANIA 10 I0 8 8 OTH