Peru - Agricultural sector survey (Vol. 2 of 2) : Annexes

Report No. 549a-PE FILE COPY Agricultural Sector Survey Peru (In Two Volumes) Volume II: Annexes March 14, 1975 Latin America and Caribbean Regional Office Not for Public Use Document of the International Bank for Reconstruction and Development International Development Association I his report was prepared for otticial use only by the Bank Croup. It may not be published, quoted or cited without Bank Group authorization. The Bank Croup does not accept responsibility for the accuracy or completeness of the report.  CURRENCY EQUIVALENTS* Currency Unit - Sol (S/.) UIs$1 = S/.38.70 S/. = U$0.026 ,S/.1,000,000 = US$25,840 US$1,000,000 = S/.38,700,000 GOVERNMENT OF PERU FISCAL PERIOD January 1 to December 31 of following year (a biennium), starting from January 1, of uneven years. (*) Above rate applies to some 90 percent of total foreign exchange transactions (largely trade) and is used normally for conversions in this report. Other transactions are done at a draft market rate of US$1 = S/.h3.38  , T. 5C,'ITT 41- Sf:T0 SURVE' Y TATiLF! O' CE(CONTENTS Voluime I GENERAL REPORT Para No. GLOSSARY PREFACE . ..................................................... i - vii Slil*LkpY ...................................................... i x GENERAL REPORT ................................................ 1 - 118 A. Background ...................... ........................ 1 - 9 Supply/Demand Situation and Trade ................... 2 - 4 Employment and Resource Base.5 - 7 The 1968 Revolution ................................ 8 - 9 B. Objectives .. .......................................... 10 - 16 C. Recommendations ......................................... 17 - 113 Resource Development ............................... 18 - 60 General ....................................... 18 - 21 The Sierra .................................... 22 - 32 The Selva ..................................... 33 - 49 The Coast ..................................... 50 - 60 Public Administration .............................. 61 - 68 Land Reform . . . ..................................... 69 - 81 Marketing and Price Policies ............. .......... 82 - 92 Agro-industry ...................................... 93 - 100 Credit ............................................ 101 - 113 D. Projects Identified by the Mission ....................... 114 - 118 Maps: IBRD - 10841 IBRD. - 10838 IBRD - 10840 IBRD - 10837  TABLE OF CONTENTS (Continued) Volume II Annex 1 - Agriculture in the Economy Annex 2 - Natural Resources and Farming Systems. Annex 3 - Organization and Administration of the Public Sector Annex 4 - Agricultural Education and Research Annex 5 - Agrarian Reform Annex 6 - Agricultural Marketing Annex 7 - Agro-industry Annex 8 - Agricultural Credit Annex 9 - Forestry Statistical Annex Mission Contacts  GLOSSAF5Y OF APBREVT&TICHl BCR - Banco CenLral. de Re&cerxra - Central Reserve Batic BFA - Banco de Fomento Agropecuario - Agricultural Developnent Bank CECOAAP - Central de Cooperativas Azucareras delPeru - Central of Agricultural Sigar Production Cooperatives CENCIRA - Centro de Capacitacion e Investigaci6n para la Reforma Agraria - National Training and Research Ccnter for Agrarian Reforn CIDA - Comite Interarnericano de Desarrollo Agricola - Interamerican Coimmittee for Agricultural Development OQFIDE - Corporacion Financiera de Desarollo - Development Finance Corporation DGP2AykAR - Direccion General de Reforma Agraria y Asentamiento Rural - Directorate General for Agrarian Reform and Raral Settlement DG PA - Direccion General de Produccion Agraria - General Directorate for Agricultural Production IDGC - Direccion General de Counmercializacion - --Directorate General for I4arketing DGA - Direccion General de Aguas - Directorate General for Water Resources mFC - Direccion General Forestal y Caza - Directorate General for Forestry and Wildlife mIA - Direccion General de Investigacion Agraria - Directorate General for Agricultural Research EMDEPA1MA - Empresa para el Desarrollo y Explotacion de la Palma Aceitera, S.A. - Oil Palm Enterprise ENCA - Encuesta Nacional de Consumo de Alimentos - National Survey of Food Consumption EPSA , Empresa Pu.blica de Servicios Agropecuarios - Public Enterprise for Agricultural Services IA - Instituto de Investigaciones Agro-Industriales - Institute for Agro-industrial Investigations INP - Instituto Nacional de Planificacion - National Planning Institute  - ii - JURPAL ; Junta Reguladora de Precios de PruOdtctos Agricolas ;.wticios de Lifira 1*^etropolijtana Control Board of Food Prices for Liriia OGIP - Oficina General de Ingenieria y Proyectos - General Office of Engineering and Projects ONEC - Oficina Nacional de Estadisticas y Censos National Office of Statistics and Census ONER:N - Oficina Nacional de la E'valuaci6n d.e Recursos Naturales National Office of Dvaluation of Natural Resources ORDEZA - Organizacidn de Desarollo de las Zonas Afectadas - Regional Agency for Development of the Earthquake Zone OSPA - Oficina Sectorial de Planificaci6n Agraria - Agricultural Sector Planning Office PIAR - Proyecto Integrado de Asentamiento Rural Integrated Rural Settlement Project PID - Proyecto Integrado de Desarrollo - Integrated Regional Development Project SAP-CAP - Sistema de Asesoria y Fiscalizacion para las Cooperativas de . - Producci6n = Advisory Services for Sugar Cooperatives SAIS - Sociedad Agr{cola de Interes Social - Agricultural Society Interest SENAFER - Servicio Nacional de Fertilizantes - National Fertilizer Service SIMAP - Servicio de Informacion de Mercadeo Agropecuario - Agricultural Marketing Information SINAMOS - Sistema Nacional de Apoyo a la Mobilizacio;n Social - National System of Support for Social Mobilization UNA - Universidad Nacional Agraria - National Agriculture University ZA ,- Zona Agraria - Agrarian Zone  ANNEX 1 Page i AGRICULTURE IN THE ECONOMY Paragraph No. Rural Population and Agricultural Employment .1 - 4 Production and Productivity 5 - 8 Incomes 9 - 13 External Trade 14 - 17 Appendix 1: Definitional Problems   ANNEX 1 Page 1 Agriculture in the Economy Rural Population and Agricultural Employment 1. Peru is undergoing a process of rapid urbanization. The 1961 census found a majority of the population -- some 53% -- in the rural areas. How- ever, official estimates 1/ for 1972 predicted a fall to only 45% and pre- liminary data from the new census indicate an even lower share: only just over 40%. Appendix 1 discusses the problems of definition, which have an important bearing on the figures. While agriculture no longer employs a majority of the Peruvian labor force, it remains by far the largest generator of employment among the economic sectors. In comparison, all manufacturing activities employed only some 15% of the labor force in 1971, and of this only about one-third was in factory-scale industry; the rest represented cottage industry and artisans. It is also worth noting that the growth of the total labor force has been so rapid that agricultural occupation has continued to grow despite the relative decline: from 1.6 million in 1961 to 1.9 million in 1971. The latter estimate is of course also subject to uncertainties but the qualitative statement that the labor force in agriculture grew significantly seems beyond doubt. 2. To say that agriculture gives employment to some 44% of the labor force is not to say that all the people involved are fully employed. On the contrary, it is known that, while open unemployment is rare in the agricultural labor force, underemployment is widespread. It was estimated in 1971 that the full-time equivalent employment in the sector was only some 69Z of the nominal -- the lowest value of any of the economic sectors (Table A.3 in the Statistical Annex). 2/ This, however, is again not tantamount to saying that almost one-third of the labor force in agriculture -- almost 600,000 - is simply superflous; it merely means that the number of man-days actually worked is only 69% of the number available if the full labor force worked full time all year. An important question that remains is obviously whether this would be at all feasible, i.e., if the labor demand is reason- ably stable throughout the year. Not surprisingly, the answer is no. Sea- sonal unemployment is indeed the great problem and particularly so in the 1/ Made by ONEC (Oficina Nacional de Estadistica y Censos), the National Office of Statistics and Census. 2/ It should be noted that the full-time employment equivalent is calculated differently for agriculture than for the urban sectors. In the latter, the basis is the observed shortfall in worked hours or income gained (shortfall compared to the legal minimum wage), but in agriculture the figure is derived as a simple ratio between the theoretically necessary labor force and the actually employed one, where the theoretically necessary is calculated from the production figures by the use of simple labor absorption coefficients. It is not known whether the two methods really give results that are com- patible, or whether it could happen that a farmer working full-time with the poorest technology on the poorest lands could end up with an income below the legal minimum wage and thus be underemployed with the definition used in the other sectors.  ANNEX 1 Page 2 Sierra where usually only one crop per year is possible. In fact, some studies have concluded that there is no absolute surplus of labor at all in Peruvian agriculture in the sense that with the existing technologies no workers could be transferred permanently out of the sector without affecting output. These studies referred to the situation in the second half of the 1960s, but there is little reason to believe that any fundamental change in this respect has taken place. 3. Future trends in agricultural employment are virtually impossible to estimate, given the current lack of real information even about the current situation. What can be illustrated is the magnitude of the problem posed by the likely growth of the Peruvian labor force. It can safely be assumed that this will be at least 3% annually for the next several decades, irrespec- tive of population trends; after all, the entrants in the labor force during the next 15 years are already born. Assuming this, the total labor force in 1981 will reach about 6 million, compared to the 4.4 million in 1971. If the agricultural labor force were to grow at the same rate, i.e., if no further relative decline took place, it would reach some 2.6 million, an increase of about 700,000 from 1971. Now, in the decade from 1961 to 1971, non-agricultural (urban) occupation did grow more rapidly than the average and there was a relative shift out of agriculture. If one were also to assume the same higher growth rate for labor absorption in the urban sectors for the future, the increase in the labor force left in agriculture would be cut in half to some 350,000. To entirely avoid an increase in the agricultural labor force would require that the absorption in the other sectors reached an annual rate of somewhat over 5%, compared to a little over 4% in the past decade -- a rate 25% higher. 4. There is too little known also about likely developments in sectors other than agriculture to make possible a judgment as to the plausi- bility of the developments indicated above. It appeared rather unlikely to the mission, however, that further growth of the labor force to be given employment in agriculture can be entirely avoided. Thus, planning for agri- cultural development must face the problem, not only of reducing the current severe underemployment -- equivalent, as noted earlier, to almost 600,000 full-time jobs -- but also of providing useful employment for further sig- nificant numbers, perhaps as many as 35,000 annually. Production and Productivity 5. In 1972 agriculture contributed some 15% of total GDP in Peru. Production estimates are very weak (see Statistical Annex, comments on Section D), but indications are that this proportion has been falling quite rapidly in the past decade, more so than in the preceding one. The following table gives the average growth rates implied by the national accounts for various periods.  ANNEX 1 Page 3 Growth Rates for Value Added in Agriculture and in the Total Economy, 1951-72 (Cumulative annual averages, percent) 1951-60 1961-67 1968-72 GNP in constant 1963 prices 5.1 5.7 Agricultural value added 4.3 2.0 GDP in constant 1970 prices .. 6.3 4.2 Agricultural value added . 1.9 1.5 Source: 1963-price data from Cuentas Nacionales del Peru 1950-1965, 1970-price data from Statistical Annex, Table A.6. The exact values that result from estimates of cumulative averages of the kind shown above are sensitive to the choice of periods, and the figures should thus not be interpreted literally. However, there are two major qualitative features of the data that seem not to depend on the periods chosen, and which thus indicate real trends in the development: (a) the production performance of agriculture was poorer in the last decade than in the 1950s, and (b) growth in agricultural production 1/ since 1961 has not kept pace with population growth (annual average of 2,8% per annum). 6. For an analysis of the production performance of Peruvian agricul- ture, the difference between the export subsector (primarily cotton, sugar, and coffee) and the domestic consumption subsector is important. Much of the rapid growth in the 1950s was due to an expansion of export crops -- all three of them -- while the slower growth in the 1960s was associated with a sharp contraction in the cotton output. Production for domestic consumption has been much less dynamic throughout the period, and, in fact, earlier studies concluded that the growth of food output was already slower than population growth in the 1950s. Throughout, the statistics show livestock products as expanding more rapidly than crops, but there are some indications that the estimates are particularly weak in this case (see Statistical Annex, comments on Section D) and the mission would hesitate to draw any ffrmhoiluslons on the basis of them. 7. It follows from the fact that agriculture employs over 40% of the labor force but contributes only 15% to GDP that the average labor productivity (value added per worker) is much lower in this sector than in 1/ The table above shows growth in value added, not production. However, Peruvian national accounts assume that these two coincide for as long as the same base year prices are used in the accounts.  ANNEX 1 Page 4 the rest of the economy. For 1971, the data from the Statistical Annex (Tables A.3 and A.5) give a per capita figure of some S/. 20,200 (US$520) in agriculture and S/. 85,200 (US$2,200) as an average for all the other sectors, a relation of 1:4.2. Furthermore, the figures also indicate that the figure in agriculture has fallen slightly since 1961, while for the rest of the economy it increased some 30%. Even if the data must be used with care, the conclusion that the gap between agriculture and the other sectors on the average has been increasing rather than narrowing appears inescapable. 8. Increases in yields are an important potential source of production growth in agriculture, and it is notable that, if the Peruvian statistics are not altogether wrong there has been very little of this in the past decade. Globally, the increase registered for agricultural land (which in- cludes the fallow land) from 1961 to 1971 was about 20%, while value added shows only a 17% increase. 1/ Also, the statistics for yields of individual crops do not show a picture of any great sustained increases; rather it is one of annual fluctuations around a fairly level trend (Table D.6 of the Statistical Annex gives figures for some of the major crops). Incomes 9. From the figures in the preceding section, one can infer that aver- age incomes in the rest of the economy must be several times those in agri- culture and that the difference has been increasing. The simple averages are misleading in many ways, however, for there are great variations both within agriculture and outside it. The owner/workers of the coastal sugar cooperatives are generallly regarded as belonging to the highest quartile of income earners in Peru, and certainly many others in the more prosperous farms can derive incomes that are quite respectable for Peru. Likewise, it is clear that many of those registered in the statistics euphemistically as underemployed in the commerce or service sectors eke out a very meager living indeed. 10. After these reservations are made, however, there is little doubt that the great masses of the very poor in Peru are to be found in the agri- cultural sector. They exist in the Coast and in the Selva, too, but the archetype is the farm family of Indian descent in the higher parts of the Sierra, having 1 to 2 ha of very poor cropland and a small number of sheep. They produce hardly anything for cash sales, and in consequence their diet is extremely unbalanced and lacking in nutritional value. Harsh climatic conditions make agriculture a constant risk and make real improvement vir- tually impossible for the individual to attain. At the same time, Peruvian governments until now have very largely ignored the needs of this type of farmer, with public investment and services-having been directed to the Coast and parts of the Selva. 1/ This includes livestock products as well as crops, but as livestock products appear to have been the more dynamic, the rate for crops only is certainly no higher than the total average.  ANNEX 1 Page 5 11. The foregoing should not be interpreted as saying that all agri- cultural activities in the Sierra are doomed to give poor results. There are certainly counterexamples to this: the production of vegetables and flowers for the Lima market in Tarma (about 3,000 m altitude) and some of the better sheep ranches (now turned into Agricultural Societies of Social Interest (SAIS)). But these are the exceptions today and many factors contribute to the much better results in the Coast. The soil is better, the climate is more favor- able, and all the land is irrigated (it has to be, but once it is, it is a great advantage). This gives higher yields of almost all crops (see Table D.7 of the Statistical Annex) and also permits the growing of more valuable crops -- there is very little production for export in the Sierra (apart from wool). The marketing system is much more developed in the Coast, which permits the producers to obtain better prices for their products. Finally, while a mini- fundio problem also certainly exists in the Coast, the average land/man ratio is more favorable there than in the Sierra. Table D.3 in the Statistical Annex shows how, with few exceptions, the ratio of rural population to cropped land is much higher in typical Sierra departments (Apurimac, Cajamarca, Cuzco, Huancavelica, and Puno) than in the typical ones in the Coast (Piura, La Libertad, Lambayeq, Ancash, Lima-Callao, and Ica). Similarly, Table C.3 shows with fi- gures from the 1961 census both that the great number of minifundios were in the Sierra and that, on the average, they were considerably smaller than the ones in the Coast. This, in conjunction with the lower productivity of the lands, is a very great disadvantage to the economic base. 12. There is little information about actual incomes of farmers in various circumstances in Peru. There are some case studies, however, from which the following two illustrative examples are taken: (a) In a study (unpublished) of the impact of a small farmer credit program in the Coast, it was found in a sample of small farms that the net cash income from the farms 1/ in 1969 varied from S/. 8,850 per year to S/. 62,260 per year. In part, this reflected farm size variations (all were under 10 ha), but also the varying crops pro- duced. A most interesting feature of the findings was that, in addition to the farm income, most families had substantial incomes from other sources. These other incomes tended to be more important in the cases of low farm income so that variations in total incomes were reduced to S/. 19,170 to S/. 76,170 per year. Possibly this indicates that farmers with lower income from their farms had more time to spend earning income elsewhere. (b) In 1970 a study 2/ was made of the socio-economic conditions in a group of districts near Cuzco in the Sierra. The area then had some 24,000 inhabitants, while the 1961 census had shown about 20,000. All the land was at altitudes of 3,400 meters or more, and 1/ These figures are in fact not individual incomes, but averages in each case for a number of farms in the same district. 2/ "Empleo Rural y Reforma Agraria," by a unit of the Ministry of Labor. Despite the general sounding title, this study refers only to a parti- cular area in the Department of Cuzco.  ANNEX 1 Page 6 the base for agriculture comprised 80 ha of land planted to per- manent crops, 1,630 ha with transitory crops (main crops: potatoes, barley, olluco -- and indigenous tuber -- and maize), and 132,000 ha of natural pasture. It is notable that some 3,880 ha were re- ported as fallow agricultural land. It was estimated that, with the given cropping pattern and usual yields, the net annual income of a ha of cropped land was about S/. 3,720, while 1 ha of grazing land would yield only SI. 110. Given the fact that no major occu- pations existed outside agriculture, and that there was only about 0.35 ha of cropped land and 27 ha of grazing land for each family in the area, this defined a situation of very low family incomes: the study calculated a total average, including all sources of incomes, of only Sf. 8,550 (US$220) per year. The area and the population exhibited all the social conditions associated with these low income levels. The 1961 census found a literacy rate of only 19.5%, and it was estimated that total school enrollment in 1970 represented only some 34% of the 5-14 year age groups. Working children were thus a common phenomenon, and, in general, activity rates were very high in the population: over 85% for both men and women aged 15 to 64. Not nearly enough gainful employment was available, however, and the study estimated that, even given the low technology used in agriculture, only about half of the number of man-days worked were really necessary, i.e., underemployment was very widespread. It is notable that the average labor income for a day's work was calculated at some SI. 10.5, while the official minimum wage in the area at the time was SI. 27. SI. 8,550 would be a very low annual income for a family of five, but the study further found that this was an average in a very uneven distribution. In the area, there was one very large and some medium-sized haciendas (now turned into a cooperative) and the incomes generated there totally dominated all others. When the study tried to estimate the distribution, the result was one where 10% of the population -- those on farms over 500 ha and those not dependent on agriculture -- had average family incomes of some S/. 50,000 per year while the remaining 90% -- those depending on small agriculture -- had an average family income of only about S/. 4,000 (UJS$100). While some of the assumptions used for this last calculation were quite loosely founded, it is indisputable that the poorest farmers in the area had family incomes significantly below even the low average of US$220. 13. Neither of the two studies cited allows one to generalize on the income situation of farmers. The two studies are not quite comparable either, since the study in the Coast focussed on cash income only, while the one in the Sierra tried to include the value of self-consumed products, thus de- creasing the gap between incomes in the two regions. On the other hand, it is probably true that, while the Sierra study had selected one of the areas known as very poor, this was not so with the other one. On the contrary, one can reasonably assume that the very poorest farmers were not included in the small farmer credit program and thus not studied at all. However, as stated  ANNE.X 1 Page 7 above, even if there ace certainly poor faimers everywhere in Peru, the great masses are in the Sierra, and the mission is inclined to believe that the cited study gives a realistic picture of the conditions of large numbers in Peru. External Trade 14. Agricultural production for export has long been an important econo- mic activity in Peru, and the vast coastal sugar plantations (now cooperatives) were developed a hundred years ago. As recently as 1950, agriculture produced more than half of the total export value, with cotton, sugar and wool being the main items. The 1950s saw a further increase in production and export of both cotton and sugar, in addition to which coffee became an important crop. Despite an increase in export value by over 60% from 1950 to 1960, however, the share in the total fell to some 35% as the decade was also one of rapid growth in mineral exports and further saw the beginning of the fishmeal boom. With the subsequent reduction in cotton production, agricultural exports stagnated and then fell towards the end of the 1960s, at their lowest repre- senting only 16 to 17% of the total. During the last few years, however, better prices and the development of some new export products have contributed to a recuperation. Combined with the reduction in fishmeal exports that has taken place, this permitted the share of total exports to reach some 22 to 23% again in 1972 and 1973 (see Table E.1 of the Statistical Annex for details for the years 1960-72). 15. In judging the likely future development of agricultural exports, two important factors are the Government's general intention of furthering food import reduction rather than increased production of the traditional export crops (sugar and cotton) and the growing domestic consumption of both sugar and cotton which reduces the availability for exports. While current high prices make farmers want to increase the area planted to cotton, it is hardly likely that any major increases will inI fact result. For sugar, only limited expansion is possible without major new investment, and while the Central of Agricultural Sugar Production Cooperatives (CECOAAP) has a project idea for a new sugar plantation in the Selva, implementation will take several years. Coffee production could probably be expanded, particularly given the stimulus of current high prices, but there are serious limitations from the market side. Wiool and fiber prospects are perhaps better, as there are now strong efforts both from the various SAIS and cooperatives and from the Govern- ment to improve the rather run-down industry. However, there is reason to believe that the meat will be more interesting than the wool in this, and thus no major expansion of exports can be foreseen. Nontraditional exports, finally, can certainly increase, but will hardly make a major contribution in the near future. In 1973, their value was unusually high, but this was due to an exceptional sale of rice that had been accumulated over a number of years and this will certainly not be repeated soon. For the rest, it is worth noting that several of the major items among the smaller exports are natural products (cochinill and achiote) with limited demand. In sum, it is likely that export volumes from agriculture will continue to decline in the short and medium term. Better prices may reduce the fall in value, but it is to be expected that the relative importance of these exports will fall fairly rapidly as others grow (minerals, and possibly petroleum).  ANEX 1 Page 8 16. Existing import statistics are hard to use as classification prin- ciples and coverage have changed over time (see Tables E.2 to E.7 in the Statistical Annex). Some trends do seem to be relatively clear, however. Total imports of food and animal feeds increased only slowly in the 1950s, from some US$40 million in 1950 to about US$50 million in 1960. After that, they shot up to about US$110 million in 1965 after which they appear to have increased very little until the most recent years. 1/ With poor production performance for several of the major crops in 1972 and 1973, increased feed demand from the expanding poultry and pig raising industry, and probably also more rapid increases in consumer demand due to some redistributive effects of the recent reforms (e.g., the new member/owners of cooperatives may raise their own consumption standard), the overall demand pressure increased. The Government, which meanwhile effectively controlled all food imports, was forced to permit increases in import volumes (somewhat reluctantly, as indi- cated by the appearances of direct shortages of various consumer products in Lima and other cities). During 1973 the higher import prices also contributed to an increased value so that food imports alone substantially exceeded US$200 million. For 1974, prices are even higher for cereals and edible oils, and, as volumes will be larger rather than smaller for several of the major imports (feedgrain, dairy products and vegetable oils), a further increase is likely. 17. The Government's policy is very decidedly to try to arrest the growth in food imports as the balance of payments situation is going to be difficult for the next few years. To this end, it has quite successfully so far reduced the import of meat through a combination of restrictions on consumption and promotion of alternative foods. With the other products, however, this is more difficult. Growing imports of dairy products and oils are likely for a number of years until possibly the recent plans for increasing production of such products have been realized. Given the restrictions on area, it is also hard to see how both production of feedgrains and of wheat can be increased without a substantial gain in yields. There remains the possibility of substituting, e.g., cassava or other meals for wheat flour, but this also appears to be a prospect for the medium-term future at best. A realistic prediction must then be for increased import volumes, and the best to be hoped for is that prices will fall somewhat from the current very high levels. In the longer term, the question is probably more open as possibilities certainly appear to exist for sharply increasing average yields and also for making use of some currently underutilized production potential, both in the Coast and in the Selva. However, Peru is not richly endowed for agri- culture, and a general strategy of making room for some food imports by increased exports of minerals appears the only reasonable one. Continuing efforts to improve the performance of agriculture will in all likelihood be 1/ The quality of the available data is particularly low for the recent years (see Statistical Annex, comments on Section E), but it is temDting to correlate the rapid increase in imports in the early 1960s with the con- tinuation of the large cotton production in those years, and the lower rate later with the sharp reduction in the cotton area that followed. By the end of the decade, this area was again more stable.  ANNEX 1 Page 9 necessary, however, even under optimistic assumptions about the export potential of the other sectors. Population increases at a rapid rate and any stagnation in domestic output has a sharp immediate impact on imports as witnessed in the past few years. In addition, there is of course the great need to improve the economic conditions for the farming population, and this need will be present for many years to come; if ways can be found to increase productivity and thus production of the small farmers, it would be doubly beneficial.  ANNEX 1 Appendix 1 Page 1 Definitional Problems 1. There are some definitional problems that make the validity of a direct comparison between the 40% figure for rural population indicated by preliminary data from the new census and the 55% figure from the 1961 census somewhat uncertain (see Statistical Anlnex, comments on Section I, para. 2), but there is little doubt that growth of the urban areas has continued in the past decades at a rate even higher than that expected from the trend occurring between the 1940 and 1961 censuses. 2. The above figures refer to the census definition of what an urban area is. It has been argued that that definition is too broad for a realistic discussion of the rural-urban migration and the shift of the labor force from agricultural 1/ to urban occupations. Under it, many small agglomerations are classified as urban areas when actually they are little more than groups of farmers' houses which have very limited urban economic activity. The mission believes this to be a valid criticism and prefers an alternative definition that counts as urban only centers with over 2o000 inhabitants. This delimita- tion, which has been used in Peru 2/ leads to a substantial increase in the population referred to as rural: from about 5.45 million (53%) to 6.37 million (62%). There are as yet no figures available from the 1972 census to make the corresponding comparison, but the expectation is that the absolute difference -- about 1 million -- will have remained or even increased between the two concepts of rural population. 2/ Roughly, one can thus expect to find, that Peru's population is currently half urban, half' rural. 3. The alternative definition increases the absolute size of the rural population, but does not alter the fact that it is clecreasing in relative terms. In parallel, the proportion of the Peruvian labor force that finds its employ- ment in agriculture is rapidly falling. Official estimates gave the figure for 1971 as 44%, while the 1961 census had 49%7 and an estimate for 1950 was 58%. 3/ Considerable uncertainty adheres to the estimate for 1971, however. On the one hand, it is derived from the assumption that the rural population would develop according to the earlier official population projection, on which the 1/ Agriculture here and in most other statistical references includes forestry and hunting. Neither of these two has been a major activity in Peru until now, however. 2/ See "'Diagnostico Socio-Economico Preliminar del Area Rural Peruana," by the Direccion General de Organizaciones Rurales of the National System of Support for Social Mobilization (SINAMOS) (Lima, 1972). 3/ The 1961 and 1971 figures are from Table A.3 in the Statistical Annex. The estimate for 1950 comes from "Cuentas Nacionales del Peru 1950 - 1965," and may not be quite comparable with the other two as the same source gave a share of 52% for agriculture in 1961, i.e., 3 percentage points higher than the more recent source. Possibly, then, one could use 55% as a more comparable estimate for 1950.  ANNEX 1 Appendix 1 Page 2 preliminary census information casts doubt. On the other hand, if one accepts the argument in the preceding paragraph, the development of the rural popula- tion as defined in the census is not directly relevant for the question of the size of the agricultural labor force. Until full employment data from the census are released and have been analyzed, it is not possible to reduce this uncertainty. It appears likely however, that the 44% estimate will turn out to have been not too much off the mark: the "true" figure is certainly a bit lower than the 1961 one, but hardly as low as 40%.   ANNEX 2 Page i NATURAL RESOURCES AND FARMING SYSTEMS Paragraph No. A. RESOURCE BASE 1 - 26 General 1 - 5 Soils 6- 11 Climate 12 - 15 Hydrology 16 - 21 Natural Vegetation 22 - 26 B. FRODUCTICN IN THE THREE ECOLOGICAL REGIONS 27 - 59 The Sierra 27 - 53 The Coast 54 - 59 Table 1: Typical 1.5 ha Cropland Farm in the Sierra Table 2: Possible Output Increases Table 3: Sheep Cooperative - 400 ha, Flock Development Project Table 4: Sheep Cooperative - 400 ha, Projectinn of Investment Costs Table 5: Costs and Benefits of Selected Annual Crops Table 6: Costs and Benefits of Selected Permanent Crops Table 7: Economic Costs and Benefits   ANNEX 2 Page 1 Natural Resources and Farming Systems A. Resource Base General 2 1. Peru's total land area of 1,285,216 km lies between the equator and 180 21' south latitude, with a coastline of about 3,000 km along the Pacific Ocean. It is a country of extreme contrasts, ranging from deserts on the coast, to the Andean chain extending the length of the north-south axis with its perpetually snow-covered peaks, to the Amazonian jungle, which makes up the eastern frontier with Colombia, Brazil and Bolivia. These three areas, constitute the three basic ecological zones known as the Sierra, the Selva and the Coast (Costa). Different combinations of soil, climate and altitude result in the fact that most of the ecological conditions known in the world are represented in Peru, which, in turn, makes for a large variety of cultivation possibilities, many of which, however, are feasible only in limited areas. 2. The Sierra is made up of the three ranges of the Andes -- the Western Cordillera, the high Cordillera Central and the lower Cordillera Oriental -- at elevations above 2,000 meters. It is bordered on the east by the Selva and on the west by the Costa. These ranges are narrowest in the North at the Ecuadorian border, having a width of approximately 100 km but widening to nearly 300 km at Lake Titicaca on the southern border with Bolivia. 3. Between these ranges are high valleys and areas of relatively flat plains, the broadest and highest area of which is the Altiplano in the south. Areas below 3,500 meters are intensely cultivated, in many instances under irrigation, and support an extremely dense population. Above 3,500 meters, production is confined to livestock, cattle and sheep in lower areas and sheep and cameloids on the higher slopes. The Sierra occupies 30.6%, or 393,950 km2, of Peru. It has 1.5 million ha (3%) of agricultural land, about a quarter of which is irrigated, much of it poorly, while 24 million ha (61%) are natural grazing lands. 4. The Selva, is an eastern tropical lowland area below an elevation of 2,000 meters east of the Cordillera Oriental. It is made up of two sub-zones -- the Piedmont, or Ceja de Selva, and the low Selva (below 600 to 700 m). The Ceja de Selva, runs from 80 to 150 km in width and contains areas of the best soils in the Selva concentrated primarily in four major valleys. In the lower valleys of Ceja de Selva where the valley slopes are manageable, permanent plantations of coffee and fruit trees have been made. Elsewhere, slashburn agriculture is carried on. As the valleys widen in the lower elevations, permanent plantations of fruit, citrus, avocado, papaya, and pineapples exist where access permits movement to coastal markets. In various areas of the Ceja  ANNEX 2 Page 2 de Selva, livestock production is being undertaken in cleared areas on pastures planted to tronical grasses. Agriculture of the low Selva is confined almost entirely to seasonal crops such as rice, cassava and beans, growm on the mud flats exposed by the receding annual flood waters of the Amazon and its major tributaries. The Selva occupies 58.8%, or 755,945 km2 of Peru. 5. The Coastal zone is a narrow belt, bounded by the Pacific Ocean on the west and the Piedmont of the Western Cordillera of the Andes on the east. It is divided into two sub-zones -- the sand dunes and dry plains ad- jacent to the coast below 300 meters and the equally dry Andean foothills up to an elevation of 2,000 meters sharply dissected and eroded by the rivers rising on the Pacific side of the Western Cordillera. These two sub-zones are referred to, respectively, as the Low and High Costa. The Coastal zone is broadest in the north, narrowing progressively towards the south. South of Ica, the Low Costa almost completely disappears as the Andean foothills reach to the Pacific. The zone occupies 10.6%, or 135,950 km2, of Peru, with 650,000 ha, or 5%, under cultivation in varying degrees of intensity depending upon availability of irrigation water. Soils 6. The study of soils in Peru began in 1919 and in 1947, the standards used by the Soil Conservation Service of USDA were adopted. Studies have been carried out by various Government agencies, but lately mainly by National Office of Evaluation of Natural Resources (ONERN). ONERN concentrates on agrological investigations on the regional level to support development programs and therefore categorizes soils on the basis of agricultural potential as wqell as natural and scientific classifications. 7. The latest overall survey by ONERN 1/ shows the following breakdown as to the areas and percentages of the various soil classes. Class Suitable for lia(000) % I - IV Intensive Agriculture 3,500 2.7 V - VI Permanent Agriculture 16,000 13.0 Fruits, Pasture and Silviculture VII Marginal Agriculture, usable 58,500 45.6 for Pastures and Silviculture VIII N4ot suitable for any form of 49,300 38.7 Agriculture, Silviculture or Pasturage 1/ ONERN. Capacidad de Uso de los Suelos del Peru (Tercera Aproximacion), Junio 1971.  ANNEX 2 Page 3 According to this breakdown, which incidentally reveals considerably less favorable picture than has previous ONERN studies, 1/ Peru is endowed with an extremely limited base for agricultural development, with only 2.7% of the total area suitable for intensive agriculture. A further breakdown of the first four classes to indicate area and location follows: (a) Class I soils with no limitation as to agricultural use are estimated to cover 177,000 ha, or 0.1% of the national territory. These are found primarily in the central part of the irrigated valleys of the coast, which may, however, be salinized, and alluvial terraces of recent formation above flood level on the banks of the large Selva rivers including the Huallaga, Maranon, Ucoyali and Amazon. (b) Class II soils make up the greater part of the irrigated valleys of the coast and include intermediate alluvial terraces of the major tributaries of the Amazon and terraces in the intermountain valleys of the Sierra, including the Mantaro, Huanaco, and the Urubamba. This class covers 532,500 ha, or 0.4% of the country. (c) Class III soils include potentially irrigable lands in the coastal valleys where unused water resources exist. In the Selva, they occur along the rivers and the slopes of the valley sides in the Ceja de Selva. In the Sierra, it includes rolling mesas at elevations between 3,500 and 3,800 meters and the sides of some valleys. This class involves 1.1 million ha, or 0.8% of the national territory. (d) Class IV, the lowest quality of the soils suitable for intensive use, occurs in the potentially irrigable areas in the Coast and in the Selva in the department of Madre de Dios in the dry tropical forest zone. A total of 1.8 million ha are involved, or 1.4% of the total. 8. According to production statistics, some 3.1 million ha are in pro- duction (0.8 million ha in the Coast, 1.7 million ha in the Sierra, and 0.6 million ha in the Selva). If one were to assume that all this land falls into Classes I to IV, an additional 0.4 million ha would seem to be available for potential development. The latter figure may even be higher, as part of the land used for crop production in the Sierra belongs reportedly to Class V or higher. There are, however, severe limitations to bringing this "reserve" into cultivation. Most of the unused Class I - IV lands are in Class IV and in the Selva. Apart from the high cost of gaining access to such lands and clearing them, sustained cultivation would in many cases be hampered by excessive rain- fall. For all practical purposes, not more than 0.3 million ha of the high Selva are still available to be opened up. 1/ A 1969 ONERN study estimated the total area suitable for intensive agri- culture at 11 million ha, i.e. three times higher than in its 1972 study.  ANNEX 2 Page 4 9. The unused land in the Coast involves two categories -- land that can be brought into cultivation if provided with high cost (US$6,000/ha) irrigation infrastructure, 1/ and land that is already irrigable but that has become so saline that it has been taken out of production. The latter category, which concerns some 150,000 ha, can be brought back into production relatively quickly and at reasonable cost (US$800 to US$1,000/ha). Apart from these possibilities, production increases will have to come from more intensive use of land already under cultivation. 10. With respect to the area suitable for grazing and planted pastures, ONERi classifies 15.7 million ha as Classes V and VI suitable for planted pastures and silviculture and 12.9 million ha of Class VII as suitable for extensive grazing: Class V - VI Class VII Total --million ha--------- Sierra 5.3 4.5 9.8 Selva 8.1 6.6 14.7 Coast 2.3 1.8 4.1 11. Regarding the Sierra total of 9.8 million ha suitable for permanent pastures or extensive grazing, it is estimated that up to 24 million ha are currently used for grazing, 21 million ha 2/ for sheep alone. This indicates that 50% to 60% of grazing area of the Sierra is on Class VIII land i.e., land actually unsuitable for any form of agriculture or grazing. This confirms field observations that the Sierra is heavily overstocked. The greatest possibility for livestock expansion is in the Selva where 8.1 million ha of Class V and VI land is available plus 6.6 million hla of Class VII land, virtually all of which is so far unused and is likely to remain so for many years to come due to inaccessibility. Climate 12. Peru's three major zones, the Sierra, the Selva and the Coast have their own distinct climates. Both the Selva and the Coastal areas, the latter under the influence of the von Humboldt current, have high barometric pressure, while the Sierra has low barometric pressure. The movement from the two highs, particularly from the Selva, into the Sierra low causes year-round rainfall on the Eastern Andean slopes. The Coast is continually covered by a cloud layer (la Garua) at 400 to 800 meters altitude from April to December causing low temperatures with high humidities, but no rain. During the December to April _peorid,_with the southern movement of tlle sun, the cloud layer disappears, 1/ New irrigation at lower cost would be irrigation through groundwater develop- ment; there are some distinct possibilities for this, such places as the raciia area. 2/ Lvaluation of the meat problem in Peru - Summary 1973 p. 36.  ANNEX 2 Page 5 bringing sunshine and a lowering of the high pressure. The Sierra is then primarily filled by the Selva high, causing rainfall throughout the Sierra, including the Western slopes facing the Pacific. Any changes in the relation- ship of the two highs causes strong fluctuations in this pattern. 13. Northern Peru is under the influence of the equatorial system which causes rainfall on the Sierra and some in the Coast. Due to the influence of the variations of the two coastal currents, the rainfall in the northern coastal area can in some years be considerable; e.g., in 1972 rainfall in the Northern Coast was over 1,000 mm but normally it is close to zero, as elsewhere in the Coast. Due to this general pattern, the rivers emptying into the Pacific through the coastal desert area carry much more water in the middle and northern part of Peru than they do in the South. This heavier flow contributes to the broader northern valleys with wide areas of flat land contrasted to the longer but narrow valleys in the south. 14. The Sierra climate varies widely with arid western slopes, having rain up to 200 mm in the December-April period, little cloud cover, and temperatures varying from 13 to 20