Agriculture for Developing Nations

Agriculture for Developing Nations The capital-intensive, highly mechanized Western model may not suit every developing region. Systems of intensive polyculture, exemplified by rice cultivation, may be better by Francesca Bray P eople in the rich industrial countries have Þxed ideas about the development of agriculture. Children at school learn about the technical progress from digging stick to hoe and from the cattle-drawn wooden ard to the tractor-driven, steel-shared plow. Economists and sociologists describe the shift from small family farms to large, eÛcient commercial enterprises. Human labor and skills yield to increasingly complicated machines. Although at times we feel pangs of nostalgia for the old ways, we know that the Western model traces the inevitable path of human progress. Or does it? Mounting frustration over attempts to plan agricultural development around the world has made it clear that the way farming developed in Europe and North America may not, after all, be the best model in the poor countries of Africa, Asia and Latin AmericaÑnor, indeed, for the survival of the biosphere. The Earth Summit held in 1992 in Rio de Janeiro marked the oÛcial endorsement of a new, critical approach to the worldÕs problems with resources. Its key words are not ÒgrowthÓ and ÒdevelopmentÓ but ÒconservationÓ and Òsustainability.Ó The basic philosophy of classical agricultural and economic developmentÑmore is better, for everybodyÑis now seriously in question. Yet the world still faces urgent problems of poverty, hunger and disease. Rural populations are especially deprived and vulnerable. The great question is whether agricultural policies based on conservation and sustainability can solve these acute problems. Or is conventional growth-driven development, for all its drawbacks, the only way to improve rural living standards? I shall argue here that the Western model may not be the ideal for every developing region. FRANCESCA BRAY is professor of anthropology at the University of California, Santa Barbara. Taking her bachelorÕs degree in Chinese studies and her Ph.D. in social anthropology at the University of Cambridge, she worked at the Needham Research Institute in Cambridge from 1973 to 1981. Her Þeld there was the history of Chinese agriculture, and she wrote Agriculture, a volume in the series Science and Civilisation in China, edited by Joseph Needham. From 1981 to 1983 she held a Leverhulme Research Fellowship to study the rice economies of Asia, and for the next four years she worked at the CNRS in Paris. She came to the U.S. in 1987, serving as professor of anthropology at the University of California, Los Angeles, until she transferred to Santa Barbara last year. Her books include The Rice Economies: Technology and Development in Asian Societies, published in 1986 and reissued this year, and the forthcoming Fabrics of Power, a study of the technologies that deÞned womenÕs lives in imperial China. Her next research project will be on innovation in Chinese medicine. 30 SCIENTIFIC AMERICAN July 1994 As critics of classical development policies have pointed out, the world now produces more than enough food for everyone, but development has often worsened the inequities of distribution. In fact, this trend is hardly surprising if one examines the criteria that deÞne development in agriculture. The ÒmodernizationÓ of agriculture, as generally understood, entails the application of science, technology and capital to increase the output of just a few crops that have world marketsÑamong them wheat and rice for human consumption, corn and soybeans for animal feed, and cotton for industry. T his approach gives rise to issues of equity and conservation. In terms of equity, the system favors rich farmers and puts poor ones at a disadvantage. In addition, specialization and economies of scale reduce economic diversity and employment opportunities in rural areas. The system entails three problems in conservation. Monoculture reduces biodiversity. The intensive use of fossil fuels and chemical inputs creates pollution; often the inputs of energy equal or even exceed the output of crops. Large-scale mechanized operations hasten soil erosion and other environmental degradation. For these reasons, the trend of Western agricultural development toward industrial farming has come under increasing challenge from conservationists and also from social groups that feel threatened by itÑamong them In- Copyright 1994 Scientific American, Inc. TRADITIONAL RICE FARMING entailed large amounts of hand labor. This scene of rice cultivation in the beautiful but poor dian rebels in the Mexican state of Chiapas and owners of small farms in France. Are there alternatives to the Western model, or do we need to invent new models? Environmentalists have identiÞed several apparently sustainable local farming traditionsÑall of them forms of polyculture. All farming systems were originally polycultures providing a range of basic requirements for subsistence. In some Mediterranean areas even today, one can Þnd farmers planting wheat and barley around their olive trees. Much of the North American wheat belt used to support mixed grain and dairy farming. A form of polyculture that has recently attracted attention from agronomists because of its inherent sustainability is the system whereby corn, beans and squash are Copyright 1994 Scientific American, Inc. Malaysian state of Kelantan was photographed some 20 years ago by the author. The woman is transplanting seedlings. planted in the same hole. They complement rather than compete with one another because their root systems draw nutrition and moisture from diÝerent levels of the soil. In fact, the roots of the bean actually Þx nitrates and so furnish natural fertilizer for the corn. This highly intensive form of land use was Þrst developed well over 2,000 years ago. It sustained great civilizations such as the Maya and continues to support dense pockets of population all over Central America. Are there other types of agricultural systems that might support sustainable but intensive development on a scale large enough to address the dire problems of rural poverty faced by many developing nations? The answer requires Þrst a deÞnition of Òsustain- able.Ó To my mind, a sustainable agricultural system cannot be judged simply by the ecological soundness of its farming methods. It must also provide a living for all its population, farmers and nonfarmers alike. The majority of the worldÕs poor live in the countryside; rural populations are still growing, and urban services and industries now absorb less labor than they once did. A sustainable agricultural system must therefore be able to create employment as well as to produce food. It should be ßexible and diversiÞed, able to yield not only subsistence but also marketable surpluses, and it should sustain an internal rural exchange of goods and services instead of depending heavily on the external world for both inputs and markets. SCIENTIFIC AMERICAN July 1994 31 where a plowshare had to cut deeply to turn the soil over, as many as a dozen oxen might form a team. Where draft animals and heavy implements Þgure prominently in agricultural production, it is clear that large farms, which can aÝord more animals and equipment and can organize their use more eÛciently, will have a significant advantage over smaller holdings. The larger the farm in medieval Europe, the more likely it was to produce a surplus. Urban markets for food grew in the 12th and 13th centuries, and the old feudal systems, under which serfs worked both their own strips of land and the lordÕs domain, began to break down. Manorial lords started to consolidate and enclose large holdings and farm them with wage labor. The laborers were often peasants who had lost traditional rights to land as he view of ÒproperÓ its ownership became privaagricultural progress tized. If landowners let their that the West has inland to tenants, it was not to ßicted on the rest of the subsistence smallholders but world has its historical roots to better-oÝ farmersÑsmall in the development of farmcapitalists like the English ing in northwestern Europe yeomen, who could bear the and the grain belts of the risks of investment in aniNew WorldÑthe regions that mals and equipment. Capisupplied food for the urban talist relations in agriculture centers of the industrial revhad formed in many parts of olution. But this dynamic, in northwestern Europe before which labor is a scarce rethe 15th century. Markets in source and output is inland and labor were well decreased by substituting techveloped. The social relations nical innovations for mannecessary for the foundation power and animals, is not of a modern mechanized agSUPPLEMENTARY CROP of kabocha ( Japanese pumpkin inevitable; it is predicated on squash) is grown on a bund, or small dike, surrounding a riculture were thus in place, the conditions of production rice field in Japan. Such a concentrated use of land is characbut the necessary technical speciÞc to those regions. expertise was lacking. teristic of farming in East Asian polyculture. Northern Europe, where Development of this agthis dry-grain farming sysricultural system was driven tem evolved, has a short growing sea- fertilizer available was manure, land by the superior performance of large, son. The staple cerealsÑwheat, barley had to be left fallow often and could be centrally managed units of production. and ryeÑbear seed heads, or panicles, planted with cereal only once every The 18th century recorded improvewith relatively few grains, at best a few two or three years. In short, this farm- ments that included new crop varieties dozen compared with 100 or more ing system used land extensively and and breeds of animal, better plows and grains on a panicle of rice or millet. could not support high population den- drainage systems, and crop rotations Each plant usually has no more than sities. The typical 11th-century English that combined cereals with fodder crops three or four stems, or tillers. In princi- holding, as recorded in the Domesday such as clover and turnips. All the exple, one seed could produce some 200 Book, was 30 acres (12 hectares). perts agreed that only large farms were oÝspring, but the biblical parable reDraft animals played a crucial role in suitable for these Òhigh farmingÓ methminds us that many seeds die where this farming system. Yields were so low ods. Economies of scale dictated who they fall. Farmers in medieval Europe that it was impossible to till enough land could aÝord such improvements. had to keep as much as a third of their for subsistence by manpower alone. Before mechanization, many highcrop for the next yearÕs seed; another Some plow teams consisted only of a farming innovations required increased large portion went to feeding draft ani- pair or two of oxen, but in the heavy labor as well as capital. In northwestern mals over the winter. Because the only clay soils typical of northern Europe, Europe, farmers had to compete with I want to propose that it is easier to plan development toward sustainable rural economies if we take as our model not the farming systems of the West, which inherently tend toward systems of monoculture and economies of scale, but systems of polyculture that use land intensively and oÝer a basis for economic diversiÞcation. Some food staples lend themselves more easily than do others to intensive polyculture. In this article, I use wet-rice farming in East Asia as my case, because its historical record is suÛciently rich to demonstrate a coherent pattern of technical and economic evolution. I do not suggest, however, that the worldÕs problems will be solved if every region switches to wet rice. Almost any combination of food staples that uses land intensively will do. T 32 SCIENTIFIC AMERICAN July 1994 Copyright 1994 Scientific American, Inc. the new and expanding industries for workers; in the sparsely populated New World, labor was simply very scarce. Inventors had been tinkering with farm machinery as early as the 16th century, but without much success. By the early 19th century the need for such machines was felt acutely. That was the time when engineers could at last draw on materials and expertise from the industrial sphereÑ steel, steam power and chemicalsÑto develop labor substitutes for agriculture. The Þrst successful mechanical threshers came on the British market in the 1830s (provoking riots by agricultural laborers as they saw their precarious livelihoods threatened ). Horsedrawn reapers, harvesters and mechanical drills followed, and eventually in the 20th century the tractor replaced the horse. Chemical fertilizer eliminated the necessity for crop rotations and facilitated monoculture. Herbicides and pesticides further reduced the need for labor. The amount of agricultural land per agricultural worker in the U.S. today is 137 hectares, and a medium-size farm of the type usually run by a single family ranges between 20 and 100 hectares. T his is the historical experience from which our image of ÒnormalÓ agricultural progress derives. Just as Western patterns of industrialism spread from nation to nation, deÞning our notions of a modern economy, so, too, after World War II the characteristics of the Western agricultural revolution deÞned the worldwide agenda of agricultural modernization. Such progress seemed normal and inevitable to the postwar agricultural economists and scientists, mostly from or trained in the U.S., who worked out a package of technical and economic aid to modernize agriculture in the poorer nations. The new technology they developed gave such impressive initial results that it quickly came to be called the green revolution. The technology centers on the use of high-yielding varieties of wheat, corn and rice. These varieties are hybrids that farmers cannot breed themselves and that need chemical fertilizers and herbicides to thrive. In experimental stations the hybrids produced such high yields that they were soon called miracle seeds. As Indian economist Vandana Shiva points out, however, comparisons between old and new varieties measure only the output of that one crop, not of the whole mixed cropping system that it often displaces, so the overall gains may be much less than claimed. Because of the emphasis on monoculture, the agricultural agencies that Copyright 1994 Scientific American, Inc. supply technical information, seed and credit to farmers usually advocate largescale cultivation and the consolidation of holdings to make mechanization feasible. Under these conditions, salable surpluses and proÞt margins (but not necessarily yields) are generally proportional to the size of the farm, and small farms lose their viability. The primary aim of the green revolution policies of the 1960s and 1970s was the eradication of world hunger: the modernization of underproductive farming systems would increase the world output of staple grains. In this respect, the green revolution has been a great success. The worldÕs produc- tion of the main staple grains (wheat, corn and rice) would today be more than adequate to feed the worldÕs population if it were not for problems of maldistribution. But as farmers have been encouraged to concentrate on monoculture, they have become more vulnerable to crop pests and price ßuctuations. The variety of local diets has been drastically reduced, as have employment opportunities. The new technology uses enormous amounts of chemicals and fossil fuels. In energy terms, it is less eÛcient than many traditional farming systems. Monoculture, mechanical plowing, the extension of crops into woodlands and Agricultural Productivity E conomic calculations of agricultural productivity usually take into account only the yield of a particular crop per unit of land and overlook other uses to which the land may be put. The drawings show the result of such a calculation comparing a polyculture (growing several different crops on a hectare of land) with a monoculture in which only rice—a dwarf, high-yielding variety common in green revolution agriculture—is grown. In the polyculture (top) one hectare of land is used for several crops in a year, producing as the main crop 1.1 tons of a cereal grain (rice) and 1.6 tons of straw used for fodder and fuel, but also producing as secondary crops quantities of oil, beans and fiber. The monoculture (bottom) produces four tons of rice and two tons of straw. Because the typical calculation of productivity applies only to yields of a single crop, the comparison puts the polyculture in an unfavorable light—1.1 tons of grain per hectare as against four tons for the monoculture crop. The other yields of the polyculture are ignored. TRADITIONAL POLYCULTURE OIL CROPS BEANS AND PULSES FIBER CROPS CEREALS GRAIN = 1.1 TONS STRAW = 1.6 TONS GREEN REVOLUTION MONOCULTURE DWARF HIGH-YIELD VARIETIES OF CEREALS GRAIN = 4 TONS STRAW = 2 TONS SCIENTIFIC AMERICAN July 1994 33 pastures, and the use of chemical products all contribute to environmental degradation. The second aim of green revolution policies was to generate rural prosperity through the production of marketable surpluses. It seemed clear that the application of science and capital would yield more eÛcient and productive farming practices. Theories in vogue at the time recognized that the capital requirements of this kind of modernization would initially favor wealthier farmers but assumed that soon the beneÞts would trickle down to the entire population. In fact, many regions have experienced a severe economic polarization. Rich farmers add to their holdings while poor ones are edged out of farming into a dependent wage-labor force. The people who can aÝord to farm rely increasingly on the urban economy for goods, services and markets. Opportunities for work in the countryside diminish, but urban industry cannot generate enough jobs, and the unemployed congregate in city slums. The parallels between the green revolution and the 18th-century modernization of Western farming are clear. If advocates of the green revolution neglected to consider the negative social and ecological consequences of their plans, it was largely because this style of development, with its reliance on capital and machinery, seems to represent the inevitable path to modernization. I t was in 1976, during a year spent studying farmersÕ reactions to the green revolution in the beautiful but poor Malaysian state of Kelantan, that I began to think about alternative models of agricultural development. Before I went to Kelantan, I had spent several years researching the history of rice cultivation in China. As I read more about agricultural development, I realized that many Japanese experts had reached conclusions similar to mine based on their historical experience. They, too, saw a logic in the historical intensiÞcation of Asian rice cultivation that was quite diÝerent from what had happened in the West. They also felt that the introduction of green revolution technology often represented a disastrous break with the past, and they suggested that there would be many advantages to adopting the ÒJapanese model.Ó Looking at the conditions of production and the consequences of development, one Þnds that the Japanese (or, better, East Asian) model, which centers on the production of wet rice, diÝers radically from the dry-wheat model of northern Europe. In China, Japan, Vietnam and Korea, the use of land was in- tensiÞed over the centuries because of the increasing availability of skilled labor. There were few economies of scale, smallholdings predominated and intensive cropping patterns sustained a mixed farming system and a highly diversiÞed rural economy that could provide a living for large populations. Water is a crucial factor in shaping the development of rice cultivation. Rice is a monsoon crop; it can be grown in dry Þelds, but water is its natural habitat. The earliest Þnd of domesticated rice so far is in a Neolithic Chinese village near Shanghai, situated at the edge of a shallow marsh and dated to approximately 5000 B.C. Other early sites dotted around southeastern continental Asia are also close to marshes or other natural water supplies. A good rice Þeld or paddy is one in which the water supply can be accurately regulated and drained. As a result, paddies are usually quite small by Western standards: a Þeld 20 yards square would be considered large in China. Young rice seedlings need damp soil but rot in standing water; once they are about a foot tall, they like to have several inches of standing water through the period of ßowering and ripening, after which the Þeld should be drained for several days before harvesting. Rainwater can easily be impounded in a Þeld surrounded by bunds (small TRADITIONAL RICE-BASED POLYCULTURE (16TH-CENTURY SOUTHEASTERN CHINA) HOUSEHOLD ECONOMIC ACTIVITIES OUTPUT = 100 FODDER FOR LIVESTOCK INPUT = 5 (EXCLUDING LABOR) TRANSPLANTING IRRIGATION SEED FROM FARM HARVEST WEAVING COMMERCIAL FERTILIZERS (BEAN FIBER, NIGHT SOIL) FALL FUEL RAW MATERIALS FOR COMMODITY PRODUCTION WINTER RICE OTHER CROPS HANDICRAFTS, ETC. SILKWORM SEASONS 34 FOOD CAPITAL GOODS (SIMPLE EQUIPMENT) MANURE FROM FARM SPRING SUMMER SEED GRAIN SCIENTIFIC AMERICAN July 1994 FAMILY LABOR INPUTS AND OUTPUTS are compared for a traditional rice-based polyculture in 16th-century southeastern China and a modern green revolution rice mono- Copyright 1994 Scientific American, Inc. dikes), but it may evaporate before the rice is fully grown. Rice farmers in some regions therefore adopted rain-fed tank irrigation systems very early. Other forms of irrigation include the channeling of small streams into hillside terraces and the construction of diversion channels from larger riversÑin which case the water usually has to be pumped up into the Þelds. All these forms were common in China and Japan by medieval times and allowed rice farming to spread from small river valleys up mountainsides and down into the deltaic ßoodplains. Constructing bunds, irrigation networks, tanks or terraced Þelds requires large initial investments of labor, but thereafter maintenance is relatively cheap and easy. So it is not surprising that rice farmers have often preferred intensifying production in their existing Þelds to extending the cultivated area. Water enhances the sustainability of rice systems. Unlike dry Þelds, rice paddies gain rather than lose fertility over the years. Whatever the original structure and fertility of the soil, over several years of continuous wet-rice cultivation the top few inches of soil turn to a Þne, gray, low-acidity mud with a layer of hardpan below that retains the water. Nitrogen-Þxing organisms that occur naturally in the water serve as a manure. Traditional rice varieties usually respond well to organic fertilizers; lime and soybean waste were widely used in both China and Japan by the 17th century, giving annual yields of up to six tons per hectare in some double-cropping areas. Rice plants have several seed-bearing stems, and each seed head contains on average about 100 grains. The technique of transplanting rice seedlings augments these traits. A small patch of fertile land is meticulously tilled, manured and sowed with carefully selected pregerminated seed. Meanwhile the main Þeld is soaked, plowed and harINPUT = 300 (EXCLUDING LABOR) PURCHASED HYBRID SEEDS IRRIGATION FEES CHEMICAL FERTILIZERS HERBICIDES rowed to create a Þne silky mud. After a month or so, the seedlings are pulled up, the sickly ones are discarded and the tops of the leaves of the healthy ones are chopped oÝ. Then the seedlings are replanted in shallow water in the main Þeld. This procedure is labor intensive, but it permits the careful selection of healthy plants and the eÛcient use of small amounts of manure. Moreover, the plant responds to the transplanting process by growing more tillers. By the time the seedlings are transplanted, they need only a few weeks in the main Þeld. Hence, the land can be used for other crops in the oÝ-season. W et-rice cultivation has enormous potential for expanding the uses of land. Tanks, channels and bunds may occupy as much as a Þfth of the land, but no space need be wasted. Fish nibble the weeds in the tanks or eat snails in the paddy, and ducks feed on the Þsh. Narrow bunds serve for growing vegetables, and broad bunds may be planted with mulberries to feed silkworms, whose droppings are used as manure. After the rice is harvested, the Þeld can be drained to grow barley, vegetables, sugarcane or tobacco. The alternation of winter rice with summer wheat became common in the lower Yangtze region of China 1,000 GREEN REVOLUTION HIGH-TECH RICE MONOCULTURE (CONTEMPORARY JAPAN) FOSSIL FUELS HOUSEHOLD ECONOMIC ACTIVITIES OUTPUT = 100 TRANSPLANTING HARVEST CAPITAL EQUIPMENT (FARM MACHINERY) FOOD GRAIN OFF-FARM WAGE WORK RICE SPRING SUMMER FALL WINTER FAMILY LABOR culture in Japan. The height of the labeled bars reßects the relative amount of that input or output. The curves at the bottom left of each diagram indicate how the people of the farm household apportion their productive time. In the polyculture Copyright 1994 Scientific American, Inc. economy the women do little work in the Þelds but are heavily involved in handicrafts such as silk production. In the monoculture economy, women do more of the Þeldwork because many of the men have off-site jobs. SCIENTIFIC AMERICAN July 1994 35 years ago. A judicious choice of fastmaturing varieties and the abundance of water aÝorded 17th-century farmers in the Canton region two or even three crops per year plus a few side crops of vegetables; yearly yields totaled as much as seven tons per hectare. Because Þelds were small, farm implements were small, light and cheap. A single water buÝalo served the needs of a typical farm; if production was really intensive, the farmer might give up plowing altogether in favor of hoeing. In general, rice farming did not require much capital outlay compared with dry-wheat farming, and there were few economies of scale to be practiced. Although a landlord in south China might own as much land as his English counterpart, his home farm would be of modest size, and the rest would be let out in small parcels to many tenants, chosen not for their capital assets but for their skills and experience. The system did not polarize rural society and drive poor people out. The relative advantage of smallholdings guaranteed access to land for large numbers of peasants, even if it was through the exploitative relation of tenancy. The labor requirements of wet-rice farming are high but intermittent. Peasants in medieval China and Japan could therefore use rice farming as the basis for the commercial production of vegetables, sugar, silk or tea or for the household manufacture of textiles, liquor, bean curd or handicrafts. Rice served as the foundation of a rural economy that both required and absorbed the labor of a dense population. Economic historians have often equated this system with Òagricultural involution,Ó by which individuals work harder and harder for ever decreasing returns. The assertion might be true if calculations were based only on rice yields, as if one were dealing with a monoculture. But when all the other goods produced in such an economy are taken into account, the system appears in a much more favorable light. Although its capacities for expansion are not inÞnite, they are considerable. During several centuries of population growth, ChinaÕs rice regions established the foundation for a rural economy in which many of the people made salable goods at home. Only after 1800 did rural living standards begin sharply decliningÑa trend that was exacerbated by the eÝects of multiple wars. A similar process of rural development took place in Japan, creating the basis for the building of the modern state. This achievement is one reason Comparison between the U.S. and Japan AVERAGE FARM SIZE (HECTARES) 185 0.97 JAPAN U.S. AGRICULTURAL LAND PER AGRICULTURAL WORKER (HECTARES) 137 1.15 AGRICULTURAL LAND PER CAPITA (HECTARES) 1.77 0.04 AGRICULTURAL LABOR AS PERCENT OF TOTAL WORKFORCE 2.6 7.6 RICE YIELDS (KILOGRAMS PER HECTARE) 6,246 6,112 PRICE AS RICE LEAVES FARM (U.S. = 100) 100 688 RICE PRODUCTION COSTS (U.S. = 100) 100 1,150 LABOR PRODUCTIVITY (KILOGRAMS OF BROWN RICE PRODUCED BY ONE WORKER IN 10 HOURS) 2,435 106 36 SCIENTIFIC AMERICAN July 1994 Japanese agronomists see their system as an exportable model. Yet in Japan as in the West, industrialization was achieved through ruthless patterns of exploitation. Between 1600 and 1800 the rural economy expanded in conjunction with the growth of trade and cities. Techniques for growing rice were improved, and land became so productive that the Meiji government of 1868Ð 1912 was able to fund the construction of a modern industrial state mostly through raising agricultural taxes. But this level of extraction left tenant farmers in a state of near destitution that the state did not feel obliged to address until the introduction of universal suffrage in 1945. The new regime set out to guarantee rice self-suÛciency and to eliminate rural poverty. Land reforms were enacted to do away with tenancy and set stringent limits on the purchase of land. This policy institutionalized the tiny but independent family rice farm, supplying a framework for the successful long-term balancing and integration of rural and urban development. Y et Japanese agriculture today is in a state of crisis. Except for the aberrant poor harvest of 1993, caused by bad weather, rice is overproduced and wastefully produced, in large part because of heavy subsidies and price support paid by the government since the 1950s. The strategy of increasing rural incomes by raising rice prices has backÞred. Until the 1960s, Japanese farmers used moderate inputs and simple machinery. Since the 1960s, mechanization has taken over in rice production with small-scale tractors, transplanters and harvesters. Almost all farmers own a full range of expensive machinery, and the average use of fertilizer per hectare is 1,110 kilograms (compared with 160 in the U.S. and 48 in Thailand ). As long ago as 1977, the Japanese economist Taketoshi Udagawa calculated that energy inputs amounted to three times the food energy of the rice. It costs 15 times as much to produce a kilogram of rice in Japan as in Thailand and 11 times as much as in the U.S. No one in Japan today would call this policy economically sustainable. Nor is it any longer conservationally sound. Although the irrigated Þelds and channels still protect JapanÕs narrow river valleys from ßoods, the channels and soil are saturated with chemicals. No Þsh or frogs swim in the paddies now. JapanÕs current crisis makes it clear that the East Asian model of agriculture, too, can go awry. Yet it would be tragic if the Japanese gave in tamely to Copyright 1994 Scientific American, Inc. the advice they are hearing to adopt the Western style rather than seeking creative endogenous solutions that might be ecologically and socially more rewarding. Such solutions may be already at work. Through recent economic reforms in Japan, Taiwan and China, the patterns of land use and economic diversiÞcation based on rice cultivation have brought about a modernization characterized by an unusual degree of balance between rural and urban development. The rising ratio of farm-household income to the household income of industrial workers in Japan shows the trend: 69 percent in 1960, 92 in 1970, 115 in 1980 and 113 in 1988. W hat are the implications for sustainable rural development elsewhere? This is a problem faced not only by nations with large and impoverished rural populations, such as Mexico and India, but also by wealthy nations, such as France, that want to avoid further rural depopulation. Monoculture is not an irreversible trend, but in todayÕs global economy, rural diversiÞcation does require structured support and fair prices for agricultural products. In Japan, where consumers will pay high prices for fruits and vegetables, large numbers of rice farmers have been persuaded to switch part of their land to orchards and truck farms. In China, the state abandoned the Maoist policy of Òputting grain ÞrstÓ in the late 1970s. It allowed farmers to combine a basic level of grain farming with all kinds of other crops and livestock. At the same time, farm prices were increased to a realistic level. Agricultural production shot up overnight. Farmers produced not just food but also the raw materials for the development of rural industry. Moreover, they became wealthy enough to consume a wide range of industrial goods. ChinaÕs current spectacular growth rates can be understood only against this background of rural revitalization. The examples of premodern China and Japan show that intensive polyculture, precisely because it does not depend on expensive inputs, can yield a livelihood for poorer farmers, oÝer widespread access to land and generate other employment opportunities. Ideally, polyculture should not only support rural diversiÞcation but also lessen dependence on industrial inputs. Mayan peasants can grow corn without buying chemicals because beans naturally manufacture nitrates. But a farmer does not have to operate at the scale of peasant subsistence to do without chemicals. In California, organic vegetable growers Copyright 1994 Scientific American, Inc. MODERN RICE FARMING is increasingly done with machines designed for the small scale of rice Þelds. Here a farmer in Japan operates a machine that transplants rice seedlings in a wet paddy after they have grown to a length of about a foot. and wine producers are developing interplanting techniques (another form of polyculture) to substitute for chemical pesticides. They grow more kinds of plants, hire more workers and buy fewer chemicalsÑand they are doing a big business. The examples I have cited should be a stimulus to look closely at other nonWestern agricultural systems. If we are to Þnd long-term solutions to the truly modern problem of feeding the world without destroying it, we have much to learn from such systems. FURTHER READING JAPANESE AGRICULTURE: PATTERNS OF RURAL DEVELOPMENT. Richard H. Moore. Westview Press, 1990. THE VIOLENCE OF THE GREEN REVOLUTION: THIRD WORLD AGRICULTURE, ECOLOGY AND POLITICS. Vandana Shiva. Zed Books and Third World Network, 1991. JAPANESE AND AMERICAN AGRICULTURE: TRADITION AND PROGRESS IN CONFLICT. Luther Tweeten, Cynthia L. Dishon, Wen S. Chern, Naraomi Imamura and Masaru Morishima. Westview Press, 1993. THE RICE ECONOMIES: TECHNOLOGY AND DEVELOPMENT IN ASIAN SOCIETIES. Francesca Bray. University of California Press, 1994. SCIENTIFIC AMERICAN July 1994 37