Latin America and the CaribbeanTechnical Department

Regional Studies Program

Report No. 40

Sustainable Agriculture and Poverty Reductionin Latin America's Risk-Prone Areas:

Opportunities and Challenges

by

Francisco J. Pichon and Jorge E. Uquillas

Environment Unit

October 1996

Papers in this series are not formal publications of the World Bank. They present preliminary and unpolished results of country analysis or research

that is circulated to encourage discussion and comment; any citation and the use of this paper should take account of its provisional character. The

findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to

the World Bank, its affiliated organizations, members of its Board of Executive Directors or the countries they represent.

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Sustainable Agriculture and Poverty Reductionin Latin America's Risk-Prone Areas:

Opportunities and Challenges

Francisco J. Pich6n and Jorge E. Uquillas

The authors are social scientists at the World Bank, Washington, D.C. The findings,interpretations, and conclusions are the authors' own and should not be attributed to the WorldBank, its Board of Executive Directors, or any of its member countries. Many persons havecontributed to earlier revisions of this paper. Special thanks for their useful comments are dueto Nigel Smith, Jutta Blauert, Thomas Wiens, and William Partridge.

CONTENTS

Preface ......................................................... v

I. Introduction ....................................................... 1

II. Poverty, Agriculture, and Natural ResourceDegradation in Latin America and the Caribbean ...................... 3

III. Relationships Affecting Natural ResourceDegradation and Sustainability in Agricultural Areas ................... 6

IV. Agriculture, Technology, and Sustainable Livelihoods .................. 12

V. Traditional Farming and Innovation ............................... 17

VI. Conclusions: The Building Blocks for a NewStrategy Towards Traditional Risk-Prone Areas ......... ............. 23

Notes ......................................................... 28

Bibliography ........................................... ........ 31

111

Preface

In pursuing efforts to reduce poverty, increase agricultural productivity, and improvemanagement of natural resources, policymakers, governments, and donors such as theWorld Bank are confronted with the fundamental question of whether to focus their attentionon high-potential areas or low-potential, risk-prone areas. Agricultural growth can usuallybe best achieved through investments in the highest-potential regions, whereas rural povertyand resource degradation problems are often located in low-potential regions. Publicinvestment in traditional agricultural areas may have high opportunity costs relative toinvestment in areas of commercial agriculture where most agricultural growth has occurred.However, efficiency tradeoffs may be justified in terms of the greater impact on poverty andenvironmental degradation that can be achieved by investing in traditional areas.

In addition, many high-potential areas are now degraded or suffer from environmentalstress, and there is doubt whether high-potential areas will have the capacity to meet futurefood needs in a sustainable manner. On the other hand, a large proportion of the poor in LatinAmerica and the Caribbean (and the world) live in the highly vulnerable traditional areas, andwithout external assistance there is a high risk of destruction of the natural resource base as aresult of their survival strategies. In such areas the nexus of rural poverty, rapid populationgrowth, and unsustainable agriculture is leading to the degradation of land, water, and forestresources as well as the breakdown of indigenous institutions and their natural resourcemanagement systems, which are critical for sustaining the livelihood of the poor.

This study shows that poverty-driven environmental degradation is common throughoutLatin America and the Caribbean, with the most critical areas being in northeastern Brazil,southern Mexico, and the densely settled hillside areas of the Andes, Central America, andthe Caribbean. These areas pose an acute set of problems: limited potential for growth, highand often increasing population density, degradation of natural resources, and lack of skills,education, and health care due to persisting socioeconomic biases against traditional ruralcommunities. Under such conditions farmers are likely to continue sinking deeper intopoverty, either becoming slumdwellers in cities or turning to environmentally destructivecultivation of marginal lands and forest frontiers. Although in the long run out-migrationmay be an answer for some traditional areas, merely transferring poverty and populationpressures to urban areas and forest margins cannot be an acceptable solution to ruralproblems. Clearly, to achieve the overriding goal of development-alleviating poverty in anenvironmentally sustainable manner-policy will have to focus on the risk-proneenvironments where poverty is concentrated.

The study also suggests that to make poverty reduction interventions both welcome andsustainable in traditional areas, they must be based on a more participatory approach for

v

technology development and dissemination. Such an approach should include an agenda forformulating participatory natural resource management strategies using both farmer-basedtraditional innovations and selective external inputs to improve and diversify rural incomes,conserve water and soils, and expand labor opportunities in agricultural areas. Designingand implementing such an agenda will require the cooperation of governments, internationalagencies, nongovernmental organizations, the private sector, and the technical and scientificcommunities in the region.

vi

I. Introduction

Poverty-driven environmental degradation is common throughout Latin Americaand the Caribbean's heterogeneous natural resource base, with the most critical areas innortheastern Brazil, southern Mexico, and the densely settled hillside areas of the Andes,Central America, and the Caribbean. These areas contain the largest concentrations ofrural poor in Latin America and the Caribbean.

In the poorest areas of Latin America and elsewhere in the developing world, thecombination of rural poverty, rapid population growth, and unsustainable agriculture iscausing degradation of land, water, and forest resources, as well as the breakdown ofindigenous institutions and natural resource management systems critical to sustaining thelivelihood of the poor. Resource degradation in these lands also originates in the unequaldistribution of land and other natural resources, and the economic and politicalmarginalization of the people whose livelihood depends on these resources. Policymakerscannot afford to underestimate the poverty reduction impact that agricultural developmenthas in these degraded areas, especially among small-scale farmers. Since farmers and otherresource users are the custodians of natural resources, agricultural development has a directimpact on how these resources are managed.

Agricultural intensification that increases the productivity of scarce resources iscrucial to poverty reduction and improved natural resource management in these areas(Kevin and Schreiber 1992). Experience gained from the green revolution in parts of Asiahas shown that broad-based agricultural growth, involving small- and medium-size farmsand driven by productivity-enhancing technology (namely the breeding of early maturing,fertilizer-responsive seed varieties and their diffusion into environments enhanced byirrigation and agrochemicals) can offer a way to create productive employment and alleviatepoverty on the scale required (Pinstrup-Andersen and Pandya-Lorch 1995). However, wheninputs are mismanaged new problems are created, such as increased pest resistance and anarrowing genetic base as large numbers of traditional crop varieties are replaced byrelatively few modern varieties. When yields are raised by diverting more water forirrigation or by increasing the use of chemicals, there are also increased possibilities forwaterlogging and salinization of irrigated lands, reduction of organic life, and chemicalcontamination of soils and waters, with severe consequences for human health.

In addition to degradation of the productive capacity of the natural resource base,growing concern about the spread of rural poverty in some parts of the developing worldcoincided with the drama of the major biological breakthroughs in food production

2

associated with the green revolution. Various questions grew out of the realization thatresource-poor farmers stand to gain very little from the processes of development andtechnology transfer characteristic of the green revolution. Critics of the green revolutionhave pointed out that the new technologies were not scale-neutral, that the farmers withthe larger and better endowed lands gained the most while farmers with less resourcesoften lost, and that income disparities were often accentuated (Chambers and Pretty1994). While subsequent studies have shown the spread of high-yielding varieties andbenefits for smaller farmers in green revolution areas where they had access to irrigationand subsidized agrochemicals, some disparities remained. Perhaps even more significantis that the areas characterized by traditional agriculture remained poorly served by thetechnology transfer approach. In these areas, the top-down technology transfer approachhas favored scientific, "modem" knowledge, while neglecting local participation andtraditional knowledge.

This paper argues that if agricultural intensification is to be desirable andsustainable in poverty-driven, environmentally degraded areas, it must be based on amore participatory approach for technology development and dissemination. Moreattention to agriculture and natural resource management in traditional risk-prone areaswill help solve the problems of poverty, food insecurity, and environmental degradation.We argue that improvement of natural resources management is not only linked to thealleviation of poverty but is also essential for achieving sustained productivity increasesin traditional and ecologically vulnerable areas. In these regions, agriculturalintensification may be the only realistic strategy for addressing poverty andenvironmental problems, including on-site erosion and desertification, or off-site forestdestruction due to expulsion of population to the humid tropical lowlands.

Such an approach should include an agenda for formulating a participatory naturalresource management strategy-based on both farmer-based traditional innovations andselective external inputs-that improves and diversifies rural incomes, conserves waterand soils, and improves the labor-absorptive capacity of agricultural areas. To design andimplement such an agenda will require the cooperation of governments, internationalagencies, nongovernmental organizations (NGOs), the private sector, and the technicaland scientific communities.

The paper is organized as follows. Section II describes the problems of poverty,food insecurity, and natural resource degradation in Latin America's risk-prone areas, andhighlights the significance of agriculture for the development of these areas and theeconomies of the region. Section III develops a characterization of the agriculturalintensification process both in traditional risk-prone areas and in higher-potential,commercially-oriented areas. We discuss the possible linkages between poverty responses,population pressures, agricultural intensification, and natural resource management in theseareas, and their likely implications for the agricultural sustainability of the region. SectionIV discusses the achievements and limitations of modern green revolution technologies andthe adverse sociopolitical and ecological impacts resulting from mismanagement of externalinputs in green revolution areas. Section V reviews some examples of traditional methods

3

of natural resource management and discusses the process of combining local farmers'knowledge and skills with those of external agents to develop and/or adapt appropriatefarming techniques. Section VI concludes the paper with a discussion of the main issuesand building blocks to be considered in the development of a new strategy towardstraditional risk-prone areas.

II. Poverty, Agriculture, and Natural Resource Degradation

in Latin America and the Caribbean

The problem of persisting rural poverty

Poverty alleviation is one of the World Bank's overriding objectives. However,during the past twenty-five years little progress has been made in reducing the overall rateof poverty in Latin America and the Caribbean. More than 46 percent of the population isstill poor, and the number of poor has increased from 120 million in 1970 to 196 millionin 1996.

While the rural population is expected to remain stable at 125 million through theend of the century, the urban population is expected to be around 400 million by the year2000, an increase of 125 million since 1970. These trends improve prospects forachieving more sustainable agriculture, but do not solve the problem of persistent ruralpoverty. With the dramatic growth of cities, poverty is said to be assuming an urban face,but the 61 percent rural poverty rate is still much higher than the 39 percent rate in urbanareas (CEPAL 1995).

It is remarkable that advocates of the urban development strategy have ignored theimpact of rural poverty and the fact that many urban poor grew up in and remain linked torural areas. Much of the massive population transfer to cities-driven by rural poverty-has been into the informal sector and slum areas, directly increasing urban poverty.

Poverty also has turned many rural people toward environmentally destructivecultivation of marginal lands and forest frontiers, and poverty has becoming increasinglyentrenched in neglected areas that were once sustainably managed using traditionalagricultural methods. Rural poverty is also becoming more gender specific-as more menmigrate to forest margins and cities in search of work, they leave behind poor householdsheaded by women.

Yet, since 75 percent of the population in Latin America and the Caribbean nowlives in urban areas, and urban poverty is becoming widespread, it is assumed thatpoverty in the region is mainly an urban phenomenon. The aggregate picture is, however,misleading. Four countries-Brazil, Mexico, Argentina, and Colombia-are large andrelatively urbanized, thus dominating the regional statistics. Brazil and Mexico togetheraccount for 51 percent of the total area of Latin America, 54 percent of the population,

4

and 58 percent of agricultural gross domestic product (GDP). Combined with Colombiaand Argentina, they account for 71 percent of the area, 69 percent of the population, and77 percent of agricultural GDP (World Bank 1993).

Still, the rural population accounts for 50 percent or more of the total populationin Bolivia, Paraguay, Haiti, Guatemala, Honduras, El Salvador, and Costa Rica. Also, amuch higher proportion of the rural population is usually poor, even in some of the moreurbanized countries. For example, less than 42 percent of Colombia's population is rural,but 74 percent of the country's poor live in rural areas. In Brazil rural areas are home to26 percent of the total population, but account for more than 40 percent of the poor.Venezuela's rural population accounts for only 16 percent of the total, but 30 percent ofthe poor (Wiens 1994). In addition, while some countries, including Trinidad andTobago, Venezuela, Chile, Argentina, Brazil, and Uruguay have had declining ruralpopulations, others such as Guatemala, Paraguay, Honduras, El Salvador, and Costa Ricahave had an average annual growth rate of 2 percent in their rural population.

Rural poverty in environmentally threatened and degraded areas

Country-level empirical information on the dimensions and characteristics ofpoverty in environmentally threatened or degraded areas is scarce, though it is recognizedthat poverty-induced environmental degradation is a significant and persistent problemthroughout the region. Worldwide, Wolf (1986) estimates that 1.4 billion people, or aboutone-quarter of the world's population, lack sufficient income or access to credit topurchase appropriate tools, materials, and technologies to practice environmentallysustainable agriculture, protect natural resources against degradation, or rehabilitatedegraded resources. These people have lost the capacity to support themselves in asustainable way and live in what Chambers, Pacey, and Thrupp (1989) call complex,diverse, and risk-prone areas. Leonard (1989) refers to those rural areas as areas of lowagricultural potential and high ecological vulnerability, where limited soil fertility,adverse climatic conditions, or other natural factors inhibit success of modern agriculture.

Although risk-prone areas are most widespread in Asia and sub-Saharan Africa,they are also found in northeastern Brazil, many parts of Central America, Mexico, andthe Andean region. It is estimated that I billion people live in risk-prone environments inAsia, 300 million in sub-Saharan Africa, and 100 million in Latin America and theCaribbean (Wolf 1986). At the same time, two-thirds of the world's 2 billion hectares ofdegraded land is located in Asia and Africa, but poverty-driven degradation as aproportion of total agricultural land is most severe in the Andean region of SouthAmerica, and in Central America and Mexico, where one-quarter of the vegetated land isdegraded (Oldeman, van Engelen, and Pulles 1990).

Because of deforestation, overgrazing, overexploitation, and other agriculturalactivities, nearly 200 million hectares of land in Latin America are now moderately orseverely degraded. It is estimated that severe soil degradation has affected 75 percent ofagricultural lands in Central America alone. It is also known that most of the degradation

5

in Latin America and the Caribbean is taking place in the "breadbasket" areas, denselypopulated rainfed farming areas, and other areas providing important environmentalservices. In addition, of the 17 million hectares of forest that are cut down worldwideeach year, 8 million hectares are in Latin America, two-thirds of which is for conversionto agricultural use by farmers. Latin America had the largest area of forest convertedduring the 1980s-nearly 7.4 million hectares per year (Garrett 1995; FAO 1991).

The importance of agriculture in the region

Of utmost importance to poverty and its relationship to agriculture and naturalresource management is the question of food security and nutrition. About 55 millionpeople in the region suffer from food insecurity, meaning that they do not get enoughfood to lead a healthy, active life. Six million of these are children. Of the twenty-eightcountries in the region for which information is available, five countries showed anegative trend, and in 1985 three countries (Ecuador, Haiti, and Antigua) could not havecovered food needs even assuming equal distribution (FAO 1992). Despite the morefavorable balance for the remaining twenty-five countries, the number of undernourishedpeople in the region was 55 million in 1985 and is projected to reach 62 million by 2000(FAO 1992).

Many have the false impression that the role of agriculture in the economies ofLatin America and the Caribbean has become less important, especially as the region hasbecome more urbanized. However, there is still much diversity between and within LatinAmerican countries which can be obscured by regional aggregates. Thus, althoughagriculture is only 10 percent of GDP for the region as a whole, it still accounts for morethan 25 percent of GDP in Bolivia, Ecuador, Paraguay, Haiti, and Nicaragua.

The vitality of the food and agricultural system is even more important to theeconomies of the poorest countries in the region, such as Bolivia, Guatemala, Honduras,and Paraguay, where pervasive poverty continues to be found mainly in the rural sector. Itis also important to the more urbanized economies. Urban dwellers depend on agriculturefor food and textiles, and a healthy agriculture sector generates employment in othersectors as rising rural incomes create demand for additional goods and services. It hasbeen estimated that in Latin America an increase of $1 in agricultural output raisesoverall economic output by almost $4 (Garrett 1995).

The agriculture sector is also a more important source of export growth in LatinAmerica than in other regions, accounting for 33 percent of merchandise exports. Thisshare is nearly on a par with (but more diversified than) Africa, and is significantly higherthan East Asia (19 percent) and South Asia (24 percent). Another distinct characteristic ofagriculture in Latin America is the relatively high significance of livestock, perennialcrops, and forestry as a share of the gross value of agricultural production (CGIAR 1990).

Varying agricultural performance in the face of an unfavorable policyenvironment throughout the region is another intriguing aspect that suggests that

6

differences and special circumstances, both between and within countries, must be takeninto account for allocating investment in the region. For example, the regional growthrate of agriculture during the 1980s was 1.9 percent per year. However, four countries hadnegative agricultural growth rates, nine countries had rates of less that 1 percent, four hadrates between I and 2.5 percent, eight had rates between 2.5 and 3 percent, and eight hadrates of 3 to 4 percent (Krueger, Schiff, and Valdes 1993).

Summary: The role of sustainable agriculture in reducing poverty and improvingnatural resource management

With only 8 percent of the world's population, Latin America has 23 percent ofthe world's potentially arable land, 12 percent of its cultivated land, 46 percent of itstropical forests, and 31 percent of its fresh water. The Amazon region alone containsabout half the world's species of plant and animals. The diversity and abundance of LatinAmerica's agriculture and natural resources provide the region with an enormouscomparative advantage with which to compete in world markets, and thereby generatebroad-based growth throughout the economy, for urban and rural people alike. The WorldBank and its government and nongovernment partners cannot afford to neglect the naturalresource management problems of traditional risk-prone areas, where the largestconcentrations of rural poverty exist. Neglecting agriculture will threaten the foodsecurity of vulnerable populations, irreversibly degrade crucial natural resources, andperpetuate poverty.

III. Relationships Affecting Natural Resource Degradationand Sustainability in Agricultural Areas

Sustainable agriculture is extremely important to Latin America and hasreciprocal effects on poverty and the environment. Agricultural activities, whetherintensive or extensive, affect the natural environment. Environmental degradation cancompromise current and future agricultural productivity, and perpetuate poverty.Poverty, in turn, accelerates environmental degradation and threatens agriculturalsustainability.

Figure I shows a series of possible relationships affecting natural resourcedegradation and sustainability in the region's agricultural areas. The relationshipsleading to natural resource degradation are not inevitable, and of course multipletrajectories exist for farmers facing a "sustainability" crisis that can result in betterforms of natural resource management. However, many areas of traditional agriculturein Latin America are caught in this downward spiral of economic, social, andenvironmental decline.

Figure 1. Characterization of Relationships Affecting Natural Resource

Management (NRM) and Sustainability in Risk-Prone Environments

Natural Resource Management (NRM) Unsustainable agricultural growth

based upon intensification through modem because of long-term environmental

Effects on areas technologies (e.g., fertilizers, irrigation, effects of modem technologies

of commercial improved seed varieties, mechanization, etc.) (e.g., soil impoverishment, soil and

agriculture water pollution, downstream effects,salinization, etc.)

Population GrowthRising Demand for Food Potential for a more participatory I Su0nable.land Rural labor expulsionIncreasing Consumer Aspirations NRM strategy based on both _.. intcoificathm (usigGreater Market Integration/Penetration (traditional) farmer-based innovations I traditinal and/orLand ScarcityL Iaand (modem) science-based inputs I modern techaques)Lack of Employment Alternatives-- ------------

NRM based on Expansion of

iNtniationt cultivated area/ i tional egh to nearby lands

increasing labor fforts. Declining Unsustainable

reduced fallow times, c.r) land sxtoi ificationEffects on armas soiW]~ Out-migraflon - wim

oftraditional quality kn a

agriculture

Land fiagmentallon

Out-migrationto urban areas

Risk-Prone Environments

8

Rural dualism: traditional low-potential agriculture versus modern high-potentialcommercially oriented agriculture

Perhaps the most significant pattern of agricultural systems in Latin America isthat of rural dualism in which a relatively prosperous, commercially oriented agriculturalsector based on large-scale farms coexists with impoverished, traditional, oftenindigenous, small-scale agriculture.'

Traditional farms are not only small and unable to support a family's increasingconsumption demands, they are usually located in densely populated marginal areas onland with poor or severely degraded soils, steep terrain, unfavorable climate, or acombination of these factors. Population growth also tends to be high in these areas, anddespite considerable out-migration the local populations seem destined to continueincreasing in many areas. In addition, off-farm employment and income opportunities arelimited in traditional risk-prone areas, primarily because these activities depend on localdemand for nonfood goods and services, which is low because local incomes fromagriculture are low. Furthermore, degradation of the resource base limits opportunities foron-farm investment and undermines incomes and assets in ways that make sustainedgrowth difficult. The upshot of this situation is an increasing number of poor peopleliving in these areas with a natural resource base further degraded by their own desperatequest for subsistence.

There are also striking differences between the services provided by both privateand public agencies to commercial and traditional agriculture. In general, conventionalpublic sector approaches to agricultural technology development have had difficultycoping with the wide range of agroecological and socioeconomic conditions andinstitutional constraints found in risk-prone areas (such as weak markets, insecure landtenure, poor access to capital and knowledge, etc.). There needs to be a rethinking ofwhere public expenditure for research on technology development and disseminationshould be directed and how it should be managed.

Degradation of the natural resource base and deterioration of the capacity fortechnology generation and dissemination have had particularly serious consequences in therisk-prone areas of the region-including northeastern Brazil, Central America, and theAndes-where small-scale agriculture is concentrated. This neglect is likely to exacerbatepoverty, keep agricultural productivity low, transport costs high, markets for inputs andoutputs weak, and limit access to institutional credit. In addition, indigenous communitiesare often the most vulnerable population in traditional risk-prone areas (Wiens 1994).2

Survival strategies and linkages between poverty, population pressure, agriculturalintensification, and natural resource degradation in traditional risk-prone areas

Poverty, combined with population pressures, land constraints, and lack ofappropriate production technology to intensify agriculture, is a major source of

9

environmental degradation in risk-prone areas in Latin America. Increased food demandscan be met by either increasing the area under cultivation or increasing the intensity ofcultivation on existing land. The literature suggests that families historically tend toexhaust their economic responses prior to their demographic responses (expansion of theagricultural frontier or out-migration) because of the greater psychic disutility of the latter(Bilsborrow 1992). The family first attempts to adapt by reducing leisure and increasinglabor effort, and by changing technology to the extent that alternative technologies areknown and available.

However, when more land is available locally, the more likely response is landextensification. Land extensification can be achieved by (a) clearing more of the farmer'sown land or, when such lands are no longer available and technological adaptations donot exist, appropriating nearby lands without migrating away from the family plot; or (b)migrating to other areas where trees and brush are cleared and people begin newcultivation. Depending on the location, this may involve the greater use of steep highlandslopes, clearing of brush in semi-arid regions, or cutting down trees in commonly heldforest areas. Unfortunately, as populations grow and there are few remaining unusedlands with high agricultural potential, the land extensification process becomes associatedwith the exploitation of marginal lands (including lowland tropical rainforests) in theabsence of environmental policy controls or remedial measures (Bilsborrow 1992).4

Until recently, geographic expansion of agriculture was a common option formany, if not most Latin American countries. For example, much of the agricultural landin southern Brazil was still covered with natural vegetation well after the turn of thiscentury. Under these conditions, yield-enhancing technologies were not critical sinceexpanding demands for crops and livestock could be satisfied by simply bringing moreland into production (Southgate 1992). As Hayami and Ruttan (1985) point out, efficientagricultural development should begin with outward shifts in agriculture's extensivemargin. Investments to improve crop and livestock yields (via changes in agriculturaltechnology) are necessary after opportunities for geographic expansion start to beforeclosed (Binswanger and von Braun 1991; Hochman and Zilberman 1986). In general,development of North American agriculture has been consistent with this pattern.

Although unevenly distributed, current prospects for frontier expansion ofagriculture in Latin America are becoming limited. The frontier is rapidly closing in areassuch as southern Mexico, eastern Paraguay, northern Guatemala, Panama, Colombia,Ecuador, and other countries and regions with forest margins. The remaining frontier isnow mainly in the Amazon. Nevertheless, in the Amazon region, as in many othermarginal frontier areas around the world, virtually all the soils that are suitable for cropand livestock production have been occupied by farmers and ranchers, and yields havetended to be modest at best on newly cleared lands (Fearnside 1986). It is remarkable thatduring the past twenty-five years, every country in Latin America except Argentina haslost significant percentages of their forests. Almost 50 percent of the Central Americanforests have been cleared since 1960.

10

The rural poor also tend to lose their capacity to sustainably support themselveswhen their access to resources is diminished or available resources are reduced. Theymay lose traditional access to resources if they are displaced by population pressure thatreduces their access to land; by misappropriation of common resources by otherclaimants; by events such as wars, social strife, and natural disasters; and by developmentactivities such as construction of dams, establishment of forest plantations, and creationof wildlife preserves that take land out of use by the poor. In response, the poor may beforced to migrate to urban areas or ecologically marginal lands. Those forced to the lattermay move higher and higher up hillsides or cut down forests for agricultural land,displacing indigenous inhabitants of these areas who may have developed sustainableresource-use techniques.5

Moreover, since agricultural expansion in the Latin American humid tropics iscarried out largely by those displaced from traditional areas by poverty and land scarcity(and by other social and political pressures), the new, most fragile lands are managed bythose with the fewest resources to devote to their management. Under suchcircumstances, the need for technological innovation in agriculture is greater, but themeans with which to innovate are lacking. Hence, declining productivity is compensatedprimarily by bringing even more land (which is usually of marginal quality) intoproduction rather than by maintaining or increasing productivity in the long term.Degradation of the natural resource base in these areas is thus an inevitable consequence,at least in the short term.

While Figure 1 and the discussion above suggest some of the possiblerelationships affecting natural resource degradation and sustainability in the region'straditional agricultural areas, the conditioning influence of the particular context withinwhich these relationships occur must be taken into account. That is, consideration mustbe given to major determinants of the extent to which (a) extensification or intensificationof agriculture takes place in a specific country or area during a given period, (b) the formsin which it occurs, and (c) the prevailing natural resource endowments (including land,geography, climate, rainfall), and institutional and attitudinal factors (which are derived

6ultimately from the country's history and contemporary government policies).

Agricultural intensification in high-potential agricultural areas

Agricultural intensification in high-potential commercial agricultural areas hasbeen based mainly on a combination of inputs such as fertilizers and pesticides, plant-breeding technology to develop high-yielding and pest- and disease-resistant varieties,and irrigation technology. Notwithstanding the substantial contributions to the regionalincrease in food production attributable to the modem technologies, there are seriousconcerns that the agricultural intensification process accruing in these areas is leading tosevere degradation of the natural resource base (Pinstrup-Andersen and Pandya-Lorch1995). Increased destruction of habitat and wildlife, soil erosion, desertification,waterlogging and salinization of soils, contamination of surface and groundwater,extinction of species, resurgence of pests, and other forms of natural resource degradation

11

in recent decades have been attributed to mismanagement in the use of external inputsand modem green revolution technologies.

Despite the wide array of life forms and ecosystems in Latin America, greenrevolution technologies were only introduced in the humid pampa of South America, theirrigated areas of Mexico, the Caribbean islands, the Pacific coast, and some tropicalareas. During the period between 1961 and 1990, 71 percent of the increased productionin Latin America and the Caribbean resulted from increased yields obtained mainly fromthese areas, while the remaining 29 percent was obtained from area expansion (WorldBank 1992). Nevertheless, past and present research and extension efforts have notproduced significant new technologies for the lower-potential and more marginal areas,mainly in Central America and the Andean region, where most traditional agriculture andrural poverty are concentrated. Extension, like research activities, has suffered frombudget cuts and salary caps, and the traditional areas have not benefited as hascommercial agriculture from the development of private suppliers of technical and farmmanagement information and counsel.

Government policy has tended to ignore resource-poor farmers in risk-proneenvironments, except as pools of labor or target zones for welfare transfers. Prematureshedding of labor from the commercial farm sector and failure to make alternativearrangements for the rural labor force in smallholder sectors cause the latter to sinkdeeper into poverty or become slumdwellers in cities. By neglecting the smallholdersector, the excess rural population is forced to migrate to the growing urban centers orfrontier areas such as rainforests in the lowland tropics, where they turn toenvironmentally destructive cultivation of marginal lands. The promise of elimination ofrural poverty via migration to the cities has not materialized (Binswanger 1994). Much ofthe massive population transfer to the cities has been into the informal sector and slumareas where they continue to be largely excluded from participation in the developmentprocess. As a result, many of the countries where massive out-migration from traditionalareas has occurred have experienced high rates of urban crime and rural violence,including land related uprisings and land invasions.

Summary: The need to address the sustainability crisis unfolding in the region

The overexploitation of the natural resource base in agricultural areas and theexpansion of subsistence agriculture to new, often marginal farming areas have often lednot only to rural populations becoming increasingly impoverished in both rural and urbanareas, but also to significant degradation of the agricultural resource base, with theresultant effects described above. Many areas of traditional agriculture in risk-proneenvironments can be found somewhere in the midst of this type of downward spiral ofnatural resource degradation and economic and social disintegration. At issue here is thatresource degradation in these areas may be deepening because technologies to intensifyland use appropriate to the conditions of such risk-prone environments have not beendeveloped or are not known to farmers. As a result, they are forced to exploit theiragricultural base beyond its capacity under available resource utilization methods.

12

Particularly, this occurs when farmers have been ousted or deprived from access to better-quality land reserved for commercial cropping or ranching under misguided agriculturalpolicies. The "do-nothing" policy response usually adopted regarding traditional areashas caused continued insecurity in such areas and continual migration of the destitute tourban and marginal rural areas, both responses entailing huge social costs (Wiens 1994).

IV. Agriculture, Technology, and Sustainable Livelihoods

Modern green revolution technologies and the mismanagement of the agriculturalintensification process

The introduction of high-input agriculture under the green revolution initiativechannelled scarce investment resources into capital-intensive agriculture which becamedependent on imported machinery, equipment, hybrid seeds, agrochemical inputs, andirrigation. The development of modem seed varieties of several food crops such as maize,wheat, rice, and other commercial crops played a key role in relation to the othercomponents of the farming system (i.e., fertilizer, irrigation, plant protection, andmechanization). By and large, it is undisputed that under otherwise unchanged conditionsAsia could not have escaped widespread famine without the new seed varieties. The gainsin productivity, total grain production, and associated income gains to producers and

7consumers resulting from the green revolution have been and continue to be enormous.

The new capital-intensive technologies worked well where ecological conditionswere relatively uniform (such as in irrigated areas) and where delivery, extension,marketing, and transport services were efficient. This made the new technologies moreattractive in areas which could produce a high return to the required investments.However, since the early productivity successes of the green revolution, it has not beenpossible to even remotely approach the yield increases of the order recorded during the1960s and 1970s (Rosegrant and Svendsen 1993). Failure to maintain past levels ofinvestment in agricultural research, technology, and irrigation, along with falling realagricultural prices and associated decreases in the use of fertilizers, are reportedlycontributing to falling yield growth rates for the major cereal crops of rice and wheat.8

More importantly, there is increasing anxiety that the goal of meeting currentand future food needs may be in conflict with the goal of protecting the productivecapacity of the natural resource base. The experience with the green revolution inseveral parts of the world has shown that when cultivation is intensified with the aid offertilizers and pesticides, new seed varieties, controlled irrigation, and mechanization,the benefits of short-term yield increases must be set against the costs and ecologicalrisks. Excessive use of fertilizers and plant protection agents (such as pesticides andfungicides) and overintensive mechanization are not only costly, they have also causedunwanted effects on the natural environment and human health (Pinstrup-Andersen andPandya-Lorch 1995).

13

Poor management of irrigation has led to considerable degradation of the naturalresource base in many green revolution areas. About 20 million to 30 million hectaresworldwide (or 10 percent of all irrigated land) suffer from severe salinity, which if nottreated can ultimately destroy the land, while another 60 million to 80 million hectaresexperience waterlogging and salt buildup (FAO 1993). In Mexico more than 50,000hectares have been abandoned due to salinity. Salinization removes about one milliontons a year from Mexico's grain harvest (Yudelman 1989). The salinity problem is growing:worldwide it causes losses of 1 million hectares to 1.5 million hectares of irrigated landeach year, equivalent to about half the rate at which new land is brought into irrigation.Besides environmental consequences, the increased prevalence of malaria andschistosomiasis in many irrigated areas has been noted for some time (WHO 1980, 1983).

Inappropriate or excessive pesticide use has significant environmental and socialconsequences. Repeated application of broad-spectrum pesticides has led to buildup ofresistance in target pest species. Due to long-term and nonselective use of pesticides,more than 1,600 insect species have developed significant resistance to pesticides, morethan 100 weeds have developed resistance to herbicides, and about 150 plant pathogensare exhibiting resistance to fungicides (FAO 1989; Hansen 1987; Weber 1992). Besidesthe resurgence of known pests, secondary pests are emerging from nontarget species thatwere not originally pests but whose natural enemies are unwittingly destroyed byrepeated applications of pesticides. For example, in Mexico pesticide treatment to controlthe cotton boll weevil and the pink bollworm also destroyed the natural enemies of otherharmful insects, thus negatively affecting the cotton industry. During the 1960s acreageunder cotton fell from 300,000 hectares to only 500 hectares (Pinstrup-Andersen andPandya-Lorch 1995; Hansen 1987).

Pesticide poisoning is also a serious problem in many parts of the world.Widespread and indiscriminate application of pesticides contaminates soils and polluteswater through runoff and leaching into groundwater. In Honduras, the greatest threat tothe Gulf of Fonseca is water pollution caused by the misuse of pesticides (DeWalt,Vergre, and Hardin 1993). Worldwide, some 20,000 deaths and 1 million illnesses eachyear are attributable to pesticide poisoning, and most of these cases occur in developingcountries (Conway and Pretty 1991). Improper pesticide use impairs farmers' health,which in turn affects farm household productivity. Exposure to pesticides can lead tocardiopulmonary disorders, neurological and hematological symptoms, and skin diseases(Rola and Pingali 1993). Effects of pesticide exposure also can be passed on to infantsthrough poisoned breast milk. In cotton-growing regions of Nicaragua, breast milksamples from women contained some of the highest levels of DDT ever measured inhuman beings (World Bank 1992).

Still, agricultural intensification, if managed properly, need not degrade theenvironment. There is strong evidence that agricultural productivity can be significantlyincreased b combining traditionally practiced farming systems with modem agriculturaltechniques. Alternative technologies and farming practices potentially profitable tofarmers include appropriate crop rotations, mixed farming systems with crops and

14

livestock, agroforestry, biological pest control, disease- and pest-resistant varieties,balanced application and correct timing and placement of fertilizer, selective use ofpesticides, minimum or zero tillage, and many others.

Later in this paper we will discuss the important role that these systems can playin intensifying agriculture and reducing poverty in risk-prone areas, but first we wouldlike to examine the sociopolitical impact of the modem technologies and the allegedeffects on resource-poor farmers. At the heart of this debate lies the recognition, by bothcritics and supporters of the new technologies, that the question is not whether but how tointensify agricultural production. Many analysts agree that the most seriousenvironmental problem in risk-prone areas is not so much inappropriate technologicalchange, but the many millions of people who live in absolute poverty. Poverty is asignificant force behind lack of incentives to apply appropriate technologies and access toother critical inputs required for agricultural intensification. We believe that if the povertyproblem is not systematically tackled in risk-prone areas, traditional agriculture willbecome even more marginalized with further natural resource degradation and increasingout-migration of resource-poor farmers to urban and frontier areas.

The achievements and limitations of modern technologies

The sociopolitical ramifications and ecological impacts of the green revolution indeveloping countries are complex and controversial. There is ongoing debate aboutwhether there was a causal relationship between the technologies of the green revolutionand the incidence of rural poverty (Mellor and Desai 1985; Rhoades 1988; Thrupp 1989;DeWalt 1995). Unfortunately, clear empirical information on the dimensions, causes, andeven location of problems is lacking in many developing countries. It is essential thatsuch information be collected to enable informed debate and decisionmaking. Reijntjes,Haverkort, and Waters-Bayer (1992) offer several interesting explanations to understandthe bias of green revolution technologies and modem agricultural research toward high-potential areas, export crops, and better-off farmers:

* The emphasis has been on maximizing production of particular commoditiesrather than on total farm production. Given the focus on single crops, the long-term effects on soil fertility and the regenerative capacity of natural vegetationand fauna have not usually been given sufficient consideration. This has alsohindered the study and improvement of positive interactions between differentplants and animals which, in addition to providing farmers' livelihoods, cancontribute to the continuity and stability of farming. The promotion of modemvarieties has also led to the disappearance of many indigenous varieties. This mayspell disaster for farmers who have to produce their crops with low external inputsunder highly variable and risk-prone conditions.

* The recent stagnation in yields has also raised doubts about whether the long-termproductivity of modem technologies is secure. Modem seed varieties areessentially high response varieties, bred to respond to high doses of chemical

15

fertilizers. If they are sown under conditions of high nutrient and water supply andadequate pest control, modem varieties and hybrids can be and have been high-yielding. However, when these conditions cannot be guaranteed, such as in risk-prone environments, risks of yield losses may be higher than with local varieties.According to Dover and Talbot (1987), as much as 80 percent of agricultural landtoday is farmed with little or no use of modem irrigation, agrochemicals,machinery, or improved seeds.

* Farmers who are given access to credit may be required to engage in high capitalinvestments and production methods which demand that high levels of externalinputs be maintained or increased. However, when purchased inputs aresubsidized by the government or a development project, their use is feasible onlyfor a limited time. As soon as subsidies are removed and farmers are forced toabandon the inputs, it is unlikely that they would be able to adjust other aspects oftheir farming systems (such as reduced diversity of crop and livestock species orincreased nutritional dependence on crops like maize which require hih fertilizerinputs) back to the original conditions, without adverse consequences.

* The conventional top-down approach of technology development withinagricultural research institutions has given scientists little opportunity to becomewell acquainted with the irretrievable loss of local genetic resources andtraditional knowledge about ecologically oriented husbandry and local alternativesto purchased inputs. Similarly, the production conditions of experiment stationsseldom resemble those of farmers. Consequently, technology tested in a stationoften has not worked under farmers' conditions, while good qualities of localvarieties, which are adapted to local conditions, are not recognized under stationconditions. The "products" delivered for extension therefore have tended to beincomplete and designed without sufficient regard for household issues such asrisk spreading, labor allocation among other existing crops, affordability ofmodern inputs, and other crucial socioeconomic aspects. Also, until recently,conventional agricultural research has given little attention to important questionsof women's influence on decisionmaking and labor allocation in the design ofnew technologies and extension systems.

The implications of modern technologies for income distribution in ruralcommunities have also been the subject of much debate. Since the new technologies saveland by permitting more intensive use of labor and external inputs, they might beexpected to contribute to a more favorable income distribution among farm households.However, modern technologies have often been criticized for benefiting landlords at theexpense of tenants and laborers in high potential areas, on the grounds that land rentshave increased where modern crop varieties have been introduced. Nevertheless, evidencefrom some parts of the developing world does show that the adoption of modern varietiesresulted in increases in labor demand, even in areas where it has been accompanied byconcurrent progress in mechanization. Despite this evidence, many analysts havedocumented growing inequalities emerging in many green revolution areas. The extent to

16

which these inequalities can be attributed to the new technology itself and/or toinsufficient progress in its development and diffusion is, however, unclear. Someinteresting views in this debate are summarized below:

Some analysts have argued that critics of the new green revolution technologies lackspecialized knowledge and failed to differentiate between the risks inherent in a technologyand those that transcend it (Leisenger 1995). Because sociopolitical effects are seen asfalling within the sphere of the latter type of risks, increased inequalities, aggravation of theprosperity gap, and other adverse impacts experienced in green revolution areas are not seenas caused by technology as such and therefore could not be prevented by changing thetechnology. Rather than demonizing a particular technology, analysts contend, there is aneed to evaluate to what extent these risks spring from country-specific social, economic, orcultural configurations and how they can be altered. Where access to land, extensionservices, credit, marketing channels, and new technologies are governed by a sociopoliticalpower structure that favors only a small minority, technological progress cannot beexpected to break down inequalities or remain neutral in impact.

Others, notably Pingali and Binswanger (1987), have explained how societieshave achieved agricultural growth resulting from "farmer-based innovations" rather than"science-based inputs." They argue that prior to the institutionalization of research in thelast century, agricultural technology had evolved over millennia through farmers'selection of crop varieties and the local improvement of materials and methods.Although these solutions have been sufficient to cope with the relatively low historicrates of growth in demand from increasing populations, farmers' own methods oftechnology discovery and natural resource base investment have not been sufficient toaccommodate modem rates of growth in demand in the developing world. With fewexceptions, Pingali and Binswanger argue, such advances have rarely been rapid enoughto do more than slow the rate of decline in labor productivity and natural resourcedegradation. The authors conclude that developing countries have been able to achievehigh rates of agricultural growth only by incorporating modem green revolutiontechnologies into farmer-generated technologies.

Hayami and Ruttan (1987) have posed the question of whether the development ofgreen revolution technologies should have been withheld because of the alleged adverseeffect on income distribution. In the absence of the new technology, the authors argue,many developing countries would have moved several steps closer to the Ricardian trapof economic stagnation and even greater stress over the distribution of income. Aspopulation growth presses against limited land resources under existing technology, thecultivation frontier is forced onto more marginal land. This, in turn, means that greateramounts of labor must be applied per unit of cultivated land, with the result that the costof food production increases and food prices rise.

Another reason often cited for encouraging the development and diffusion ofmodem technologies, even in societies characterized by inequitable distribution of wealth

17

and power, is that the new income streams generated by technical change represent apowerful source of demand for institutional reform (Hayami and Ruttan 1987). The gainsfrom the new technology can be fully realized only if land tenure, water management, andcredit institutions perform effectively. Markets for the inputs that are required by the newtechnology (for example, seeds, fertilizer, pesticides, etc.) also must perform efficiently.Hayami and Ruttan (1987) argue that in a society in which technology is static andmarketable surpluses are not increasing, there are few gains to either producers orconsumers from the reform of market institutions. However, when rapid growth ofproduction and of productivity becomes possible such as in green revolution areas, thegains become larger, and the incentives that act to induce institutional reforms becomemore powerful. Similarly, unless the potential gains from land tenure and otherinstitutional reforms are enhanced by technological change, it will be difficult to generatethe effort needed to bring about the reforms.

Summary: Is there room for improved farm management through traditionalfarming systems?

Aware of the above debate between critics and supporters of the newtechnologies, the interest of this paper is to ask the question of whether advances intraditional or indigenous technology would be sufficient to sustain rising levels of percapita income and consumption in risk-prone areas or, at least, what role shouldindigenous technology and management systems play in helping attain such goals. Manytraditional farming systems were sustainable for centuries due to their ability to maintaina stable level of production (TAC/CGIAR 1988, 1993). A wide range of different farmingand animal husbandry systems evolved, each adapted to the local ecological conditionsand inextricably linked with the local culture. However, traditional systems have changedquickly, particularly over the last few decades, as a result of increasing populationpressure, market integration, indiscriminate promotion of modem inputs, and financialconstraints. The impact of population pressure on inducing indigenous improvements inagricultural technology in contemporary farmer societies has been documented by anincreasing literature led by Esther Boserup. Her insistence on the importance ofpopulation growth in inducing the development of intensive systems of agriculturalproduction is an important correction to the view that agricultural technology intraditional farming communities was essentially static.

V. Traditional Farming and Innovation

Some examples of traditional methods of natural resource management

There are no technical blueprints for exploring the possible contributions ofindigenous practices to new technology developments. Nevertheless, techniques holdingpromise and which are most likely to be applicable in the various contexts are thoseinvolving careful conservation of soil and water; use of complementary or symbiotic

18

genetic resources (intercropping, integrating trees and animals, etc.); taking advantage ofnitrogen fixation; and complementary and efficient use of external nutrient inputs (naturalor artificial). Many traditional practices, not all of which are yet known to formalagricultural science, represent at least the seeds of promising new technologies based oncomposting, green manuring, mulching, multiple cropping, contour farming with hedges,water and nutrient harvesting, and ways of controlling pests. If these indigenous practicesare understood well in formal scientific terms, it may be possible to improve them, forexample, by careful use of external inputs and modem technologies. Also, manyindigenous, sometimes unconventional crop and animal species and local varieties andbreeds may have great potential for new technologies in risk-prone areas.

A growing number of publications are now appearing about indigenousknowledge systems and the farming systems based on them (for example, Brokensha andothers 1980; Biggs and Clay 1981; Rhoades 1984; Richards 1985; Marten 1986; Wilken1987; Warren 1991; Warren and Cashman 1989; Reijntjes and others 1992; Chambersand others 1989; Scoones and Thompson 1994; Pretty 1994). Case studies of successfulexperiences include experimentation and/or innovation in food and tree crops, irrigationand other water harvesting techniques, gardening, seed distribution, field and seedpreparation, fertilization, livestock nutrition, rodent and weed control, natural resourcemanagement, food storage, food processing, and market products and outlets amongmany others. A common theme running throughout most of this work is that traditionalfarming systems are in constant change. An attempt is continually being made to adapt tothe new conditions imposed by population changes, greater aspirations, marketintegration, etc. These adaptations have not, however, always been adequate and entirecultures have disintegrated as a result. Many indigenous practices that in the pastsustained human populations for centuries have become obsolete as conditions have alsochanged. For example, several forms of shifting cultivation have proved to beunsustainable under increased population pressure and, as a result, they cannot bemaintained.

Not all traditional systems have reached a point of causing ecological damage,and those that are in the process of decline can benefit greatly from modern technologyinterventions to increase stability and productivity of the farm system while conservingthe natural resource base. There are literally hundreds of studies that have recorded theimportance of traditional knowledge systems in natural resource management in manycountries. For example, there are studies illustrating the growing interest by developingcountries and donor agencies in the role of indigenous knowledge in making developmentprojects more effective and efficient (Franke and Chasin 1980; Bheenick 1989; Verhelst1990; Barborak and Green 1987; Hoskin 1984; Wilcox 1995, 1991). Niamir (1990) hasrecently compiled a list of pastoral projects and programs that have included indigenousknowledge components. Titilola (1990) and Dommen (1988, 1989) have analyzed thecosts and benefits of adding indigenous knowledge components to development projects.Others have addressed the potential role of indigenous knowledge in internationalagricultural research (Cashman 1989), in forest management (Poffenberger 1990), ingender issues and development (Norem and others 1989), in sustainable approaches to

19

agriculture and development (Ascher and Healy 1990; Jodha 1990; Warren 1991; Warrenand Cashman 1989), and in the agricultural research and extension process (Butler andWaud 1990; Cernea and others 1985; Denning 1985; den Biggelaar 1991; Rolling andEngel 1989; Moris 1991; Fairhead 1990). Working with and through indigenousorganizations for development has been addressed by Cook and Grut (1989),Messerschmidt (1991), Rau (1991), Groenfeldt (1991), and Uphoff (1985). A wealth of"grey" literature (unpublished reports and articles in newsletters, circulars, disseminationnotes, and project reports) also reveals the experiences of farmers and supportivescientists in a wide range of nonconventional forms of agriculture such as agriculturewith low external inputs, organic farming, biodynamic farming, permaculture, etc. NGOshave been particularly active in disseminating this type of work.

Participatory technology development

The process of combining local farmers' knowledge and skills with those ofexternal agents to develop and/or adapt appropriate farming techniques has been giventhe name "participatory technology development." Farmers work together withprofessionals from outside their community such as researchers and extensionists inidentifying, generating, testing, and applying new technologies. Participatory technologydevelopment seeks to strengthen the existing experimental capacity of farmers andencourage continuation of the innovation process under local control (Haverkort andothers 1988).

Numerous case studies illustrate the range of initiatives in participatorytechnology development.12 From a review of conceptual approaches to participatorytechnology development and recent case studies, there are many instances of successfulparticipation of local farmers in problem identification, including substantial reorientationof initial objectives defined by researchers. Of particular importance for technologydevelopment is the capacity of farmers to understand the local biophysical and culturalenvironment and to predict and explain the outcome of experiments under localconditions (Ashby 1987; Ashby and others 1990, 1995). There are a substantial numberof cases in which farmers' evaluation of technology has provided researchers with newinsights and in which farmer-to-farmer dissemination has been successful. Whenresearchers already have a good knowledge of a technology, a common and cost-effectivetechnique is to offer farmers several technology options related to the problem oropportunity at hand, and leave it to them to experiment in an ad hoc fashion.

As they become familiar with new technology, farmers are also likely to changeother components of their farming system to exploit the advantages the new technologyoffers. Such changes can be complex and variable over time and space so that researchershave little prospect of predicting them on the basis of their own trials (Ashby and others1995). Observation by researchers of the evaluation criteria used by farmers can then befed into the next round of technology development for release to farmers (Farrington andMartin 1988). The search for new participatory methods has also led to efforts to meetboth researchers' and farmers' requirements in a single set of trials, usually via interfarm

20

instead of intrafarm replication. These have been particularly useful in accelerating therelease of new genetic material (Maurya and others 1988), although in other cases theyhave incurred both a substantial cost and a high risk of uninformative failure, prompting amove back to on-station trials.

Most of this work has implied participation at the individual farmer level, butother important opportunities for participation should perhaps not be neglected. Importantdivergences of obligations, rights, technical knowledge, and therefore acceptability oftechnologies have been found within farm households, especially implying a need toinvolve women in technology development. Other experiences have shown thecommunity to be the more appropriate level of participation, especially when dealingwith technologies concerning the exploitation of common resources. Other technologies,such as innovations in animal-drawn equipment, have traditionally developed throughinteraction between farmers and artisans, implying the need to build on these channels oflocal knowledge.

Regarding the types of technology being developed, there are many more fieldexperiences reported in the selection of genetic material than in any other application.Important but isolated examples have been recorded in the management of soil, water andforest resources, crops, and storage facilities. Examples of crop protection, fertilization,farming equipment, and livestock research are less numerous. The focus on geneticmaterial perhaps highlights the area of greatest complementarity between researchers andfarmers. Both have a vast range of material on which to draw and have developedbreeding methods that are very diverse both in scope and speed. It is remarkable thatmore than 80 percent of the crops cultivated in developing countries are planted withseeds saved from the preceding season and from informal farmer-to-farmer seed systems(for example, potato seed systems in Peru). For self-pollinated crops (such as rice andwheat) and for crops grown primarily for subsistence (such as dry beans, millet, andcassava), the proportion of farmer-saved seed is generally even higher. These informalsystems are strong even in countries which have relatively advanced seed industries(Cortes 1995; Jaffee and Srivastava 1992; Wiggins and Cromwell 1995).

In addition, a tendency throughout the literature is to describe the intention andrationale behind farmer involvement in technology development, but to give only a briefsummary of the procedures and problems, making it difficult to assess the success ofparticipation in practical terms. For example, much closer attention needs to be given tothe role of extension once researchers and farmers have been drawn closer together in aparticipatory approach. Similarly, the role of local organizations and of nonindigenousNGOs in articulating client demand for, and mediating participatory inputs intoagricultural research has received little attention, yet it appears to offer considerablepotential. There are numerous case studies of projects using innovative methods at theoutset, but which have not yet produced substantive evaluation of their experiences.Perhaps more importantly, the time and cost-related effectiveness of participatorymethods is poorly documented.' 3 Also, while field experiences all recognize thatparticipatory approaches can lead to greater cost effectiveness, not only in problem-

21

focused research but also in commodity and factor research, precisely how the results ofthe participatory methods influence the agenda for research in these areas is rarelyillustrated from empirical evidence.

Concerning the institutional framework, practically all participatory technologydevelopment experiences have been undertaken outside national agricultural researchprograms. Numerous examples have had a continuing institutional base through theCGIAR centers, NGOs, and universities. However, many more have emerged fromspecific research projects of limited duration with no apparent commitment to theireventual incorporation into any institutional framework. It is clear that many more casesare needed of instances where the incorporation of participatory technology developmentinto national agricultural research programs has been attempted.

As a result of decreasing levels of both government and international donorfunding, NGOs are moving to fill vacuums caused by the decline of extension services.They are assuming greater responsibility for identifying and distributing required inputssuch as suitable seed, and providing technical support services in risk-prone areas(Farrington and Bebbington 1994). Some funding agencies have supported collaborationbetween government and nongovernmental entities utilizing the latter's capacity as"brokers" between farmers and research services. The capabilities that NGOs bring tobear in doing this is derived from their close knowledge of the needs and opportunities ofthe rural poor in relation to agricultural change, not merely in the sense of crop or animaltechnology, but in the wider context of innovation.

Although in most cases NGOs have occupied service delivery roles, there is anincreasing number of experiences of NGO involvement in the development anddissemination of technologies and improved management systems (Farrington andBebbington 1994). However, as inevitably occurs with any new approaches such asparticipatory technology development, methods are being developed in a piecemealfashion. Therefore, information on the types of participatory technology developmentmethods most easily incorporated, the features of institutions likely to facilitate theirincorporation, and the types of institutional change most conducive to fuller incorporationof technology generation methods needs to be more carefully documented.

Summary: Building on traditional knowledge through farmer participation in thedevelopment and dissemination of technology

In the constant struggle to sustain their agricultural livelihoods, resource-poorhouseholds and communities in risk-prone environments have developed innumerableways of obtaining food and fiber from the natural environment. In many risk-prone areasthroughout the developing world, long-term concern for the sustainability of the naturalresource base has been, in fact, an important traditional management objective, ofteninstitutionalized in local regulations and cultural norms (Warren 1995; Rhoades 1988).However, when poverty has become so extreme that only day-to-day survival can beachieved, traditions of natural resource conservation may disappear. At issue here is that

22

there is no margin for reconciliation with ecological sustainability at the present levels ofpoverty found in risk-prone areas. Thus, it would be naive to argue that farmers living inareas of traditional agriculture would not be interested in incorporating modemagricultural technologies into their farming systems, if they had the means andopportunities to do so.

To ensure the continuity of their livelihoods, farmers must be able to adjust tochange driven by increasing demands (for example, population growth, greater marketintegration, desire for more consumer goods, etc.). Vital to such adaptability is thecapacity to manage farm development: to choose appropriate combinations of geneticresources and inputs; and to develop new technologies and fit innovations into the farmsystem to raise output in subsistence food crops and commercial crop production. In thiscontext, resource-enhancing techniques are particularly important, as they can be used notonly to rehabilitate degraded land but also to create new opportunities as new needs arise.

Notwithstanding the important role of modem technologies in traditional areas,agriculturalists and policymakers need to recognize that in many poor areas, especiallywhere farmers depend mainly on local resources, modem technologies may not be thefirst or only option to improve agriculture. Rather, improving the insight of farmers anddevelopment agents into the ecological principles behind farming and adding to theirknowledge of the available technical options is an important step in the process ofstrengthening farmers' capacity to develop and manage technology for sustainabledevelopment (Smith 1995). This also implies that the solutions to farmers' problems willbe as diverse, complex, and site-specific as their farming systems, but the principlesinvolved in finding the solutions will be of wider validity.

The crucial issue is not, therefore, whether one type of technology should replacethe other, but how-in terms of methods and institutions-the most relevant aspects of eachcan be brought to bear on the natural resource management issues of a particular area(DeWalt 1995). The argument is that agricultural sustainability in risk-prone areas requiresan understanding of the diverse and complex environments in which resource-poor farmersoperate so that developments in technology can be tailored to suit their circumstances,building where possible on farmers' indigenous technical knowledge. Innovationscombining both modem and traditional approaches to natural resource management, such ascombining chemical and organic fertilizers, appropriate forms of green manuring orintegrating new crops, can open up new windows of possibilities for farmers. Spreadingknowledge about these technical options and combining the forces of farmers, fieldworkers,and scientists in discovering the opportunities and limitations of these options definitelymust play a role in any sustainable agricultural development strategy.

Democratization of political processes, combined with trends towardsdecentralization, offer hope of greater government accountability and farmer participationin developing new approaches for facing the technology challenge in risk-prone areas.Existing centralist and top-down approaches to technology development anddissemination are challenged throughout most of the work reviewed, not merely in

23

technology design, but in support facilities (such as local gene banks and seedmultiplication) and in legislative provisions (for example, government certification andrelease of new varieties). It is evident that as the variability of agricultural and ecologicalconditions under which farmers in risk-prone areas operate becomes better recognized,pressure will increase for a paradigm shift away from central control and towards localcontrol and a blend of locally and centrally available support facilities. Thus, informationis required on the long-term incorporation of participatory technology developmentmethods into institutions outside national agricultural research programs such as NGOsand universities.

The wide range of biomass on which farmers in risk-prone areas must draw forsupporting their agricultural livelihood, and the diverse criteria by which they will assessthe increased production and welfare benefits offered by technological change, are ofcritical importance and significance to the technology development process. Scarcepublic and private resources simply will not be sufficient for scientists at national andinternational research centers to undertake this task, even if they could grasp theopportunities and constraints in all their diversity. This remains one of the mostcompelling reasons for promoting farmer participation in the development of technology.

VI. Conclusions: The Building Blocks for a New StrategyTowards Traditional Risk-Prone Areas

A new paradigm of sustainable development in risk-prone areas

In pursuing efforts to reduce poverty, increase agricultural productivity, andimprove management of natural resources, policymakers, governments, and donors such asthe World Bank are confronted with the fundamental question of whether to focus theirattention on high-potential areas or low-potential, risk-prone areas. Many high-potentialareas are now degraded or suffer from environmental stress, and there is doubt whetherhigh-potential areas will have the capacity to meet future food needs in a sustainablemanner. On the other hand, a large proportion of the region's (and the world's) poor live inthe highly vulnerable, traditional areas, and the risk of destruction of the natural resourcebase as a result of their survival strategies in the absence of external assistance is high.

As discussed in this paper, risk-prone rural areas pose an acute set of problems:limited growth potential, high and increasing population density, degradation of naturalresources, marginalization, and inadequate education and health care. Under theseconditions, farmers are likely to sink deeper into poverty, either as smallholders lackingsupport and excluded from participatory development processes, or by becomingdisplaced and joining the swelling ranks of urban slumdwellers. Clearly, to achieve theoverriding development goal of alleviating poverty in an environmentally sustainablemanner, efforts must focus on the areas most plagued by poverty, such as risk-prone ruralcommunities.

24

But difficult tradeoffs arise when investing in the development of risk-proneareas, at least in the short term. Agricultural growth can usually be best achieved throughinvestments in the regions of highest potential, whereas rural poverty and resourcedegradation problems are often located in regions of low potential. Public investment intraditional agricultural areas may have high opportunity costs relative to investment inareas of commercial agriculture where most agricultural growth has occurred. However,efficiency tradeoffs may be justified in terms of the greater impact on poverty andenvironmental degradation that can be achieved in traditional areas. Therefore, whilealleviating poverty and improving natural resource management may not necessarily belinked to promoting significant agricultural growth in traditional areas in the short term,in the long run poverty alleviation is necessary for achieving sustainability in these areas.

The discussion in this paper suggests that a new paradigm needs to be developedthat integrates risk-prone areas into rural growth (Delgado 1995). While there is notsignificant disagreement on what is needed to boost production in the areas of highest-potential, there is little consensus on the other 80 percent of cropped area worldwide.Experience suggests that attempts to make low-potential zones do the same things ashigh-potential ones are problematic. Some argue that the best policy to pursue invulnerable, low-potential areas is to relieve population pressures by encouraging massiveout-migration. Although in the long run out-migration may be an answer for sometraditional areas, merely transferring poverty and population pressures to urban areas andforest margins cannot be an acceptable solution to rural problems. Failure to addressproblems within the traditional areas themselves will only intensify or shift the problemsto cities as the rural poor exhaust natural resources and move to join the urban poor.

High-potential areas that have been degraded must be rehabilitated and theirproductivity restored to the extent feasible. Although more attention must be paid to areaswith fragile ecosystems and large numbers of poor people, further improvements alsomust be pursued in the agricultural productivity of high-potential areas to meet the needsof rapidly increasing urban populations. In a sense, the successes of the green revolutionhave bought us time in these areas to adjust the intensification approach to assure that it issustainable in the longer term. In addition, policymakers need to explore the scope forassisting low-potential areas to diversify production into goods demanded by adjacenthigh-potential areas and thereby benefit from their growth.

Intersectoral linkages

It is evident from the above discussion that the focus of agricultural developmentmust be broadened to include intersectoral linkages influencing employment generation,poverty reduction, and related issues of population growth and expulsion/attraction forcesdriving rural migration to the forest frontier and urban areas. The investment allocationprocedures of donors and governments generally fail to consider the broader questions ofhow investment in agricultural research and extension, health, education, urban services,etc. may influence decisions on out-migration and resource use in a particular region.

25

Agricultural growth creates powerful multiplier effects on the rural nonfarmeconomy that enables farm households in degraded lands to mobilize capital and labor forfarm investment and rehabilitation, income diversification, evolution of property rights, andinfrastructure investments leading to additional employment and earning opportunities forthe rural poor (Scherr and Hazell 1993). In addition, off-farm employment (includingincome-earning activities in rural industry, services, and marketing) is an importantcomponent of the survival strategies of the rural poor, and hence should be an importantfocus of antipoverty efforts. Given its crucial importance for overall sustainability, it is thusessential that the policy framework for agricultural development be broadened toincorporate the relationship between geographic and intersectoral allocation of publicexpenditures influencing employment, poverty, and migration.

Technology development

Even with the benefits to be gained from economic liberalization, one must beaware that the increasing emphasis on structural adjustment is, in the short run, likely toincrease the number of vulnerable and at-risk people in the region. As the state withdrawsfrom organizing rural economic life, a vacuum is being left that it is not entirely filled bythe private sector, as is generally believed. Hence, interactions between public sectoragricultural research systems, farmers, private companies that conduct research, privateenterprises in food processing and distribution, and NGOs will need to be strengthened toassure relevance of research and appropriate distribution of responsibilities. The ability toengage local government, the private sector, and NGOs in addressing problems oftechnology development and dissemination will largely define the future scenario in theregion. This will make government policies, research priorities, and public sectorspending more effective and responsive to the resource constraints and local ecologicalconditions faced by farmers while providing a better foundation for interaction betweengovernment and civil society.

To ensure participation in technology development, agricultural research, andextension, more resources need to be allocated by governments and their donor partnersto promote intensive consultations with prospective beneficiaries. Technologydevelopment must be not only on-farm and farmer-managed, but also participatory inorder to draw on local knowledge and meet farmers' needs, opportunities, constraints,and aspirations (Farrington and Bebbington 1994; Bebbington and others 1993;Bebbington 1995; Chambers and Pretty 1994). Thus, enhancing the role of farmers inlocal analysis, setting priorities, experimentation, and other research and extensionactivities should be important elements of a demand-responsive strategy for dealing withrural poverty and natural resource management problems in risk-prone areas.

Supervision in the field also can yield insights on many issues, such as the desiredproperties of small irrigation works, desired features of soil and water conservationmethods, improved research on diverse cropping systems and long-term viability offarming systems, better integration of livestock and green manures into farming systems,

26

and desired training in activities performed by technicians. Such field knowledge, ifreflected in a project's design, can go a long way towards ensuring a convergence ofinterests around the project and hence produce commitment and the sense of ownershipby stakeholders without which project success is unlikely.

Infrastructure development, growth of the nonfarm economy, and decentralizationof financial and administrative decisions

The relative lack of public and private sector investment in the infrastructure andhuman resources of many traditional areas in the region has inhibited the fulldevelopment of the potential of agriculture and natural resources. This lack of attentionhas left these areas impoverished, isolated, and economically depressed. To reverse thisneglect, governments in the region will need to allocate to traditional areas the resourcesnecessary to develop and maintain productive infrastructure and services, including cropstorage and processing facilities, irrigation, roads, telecommunications, financial services,and the supply channels for inputs.

In many countries the potential multiplier benefits of agriculture are constrainedby investment codes and related legislation that discriminate against small, rural, nonfarmfirms. These policies need to be corrected. The rural nonfarm economy (as well as rural-urban linkages generally) can be strengthened by removing institutional barriers to thecreation and expansion of small-scale credit and savings institutions, and making themavailable to small traders, transporters, and processing enterprises. The experience withWorld Bank-supported projects is that such reforms have been effective in many countriesin helping the poor to face risk and to generate more income (Binswanger 1994).

Governments in the region must also improve the quality of education, healthcare, and sanitation in traditional areas. Productive and social sectors are synergistic, notcompeting. Just as investment in health and education can increase productivity and helpachieve economic goals in traditional areas, investment in the development of agricultureand natural resources can help achieve social goals in these areas. Governments will needto revitalize local governments in traditional areas and create an enabling environment forlocal institutions to identify, develop, and maintain new infrastructure and services. Toimprove efficiency, governments will need to recover costs through user fees, identifyand select projects with full involvement of farmer communities based on carefulevaluation of potential demand for services, and involve NGOs and private contractors inexecuting projects.

Correction of market and policy distortions through incentives and regulatorymeasures

Incentives and regulatory policies will need to be strengthened to compensate forexternalities related to natural resources. The nature of such policy measures will varyacross countries over time, but they are likely to include policy reforms dealing withwater allocation mechanisms and watershed management, exploitation of lands and

27

forests resulting from free access, determining and enforcing property rights (includingland, water, and forests), and correcting distortions in input and output markets, assetownership, and other institutional and market distortions adverse to the poor. Regulationswill be necessary where incentives are unlikely to achieve social objectives. However,policymakers need to be aware that regulations that contradict the survival strategies ofpoor farmers are unlikely to be successful simply because they will be difficult orimpossible to enforce in traditional areas. Finally, serious commitment to promotingequitable access to land, water, and capital in traditional areas will be crucial forachieving sustainable development and poverty reduction in the region.14

Political leadership and empowerment of farmers in risk-prone areas

An important aspect in the development of traditional areas is the strengthening ofpolitical organizations such as indigenous, farmer, and women's groups. These haveendured historical patterns of exclusion which persist to this day. The violence endemicto the region compounds these difficulties. Although in many cases this violence is bornof long-standing social, economic, and political inequalities, in others, such as in the caseof the Andean region, it reflects the rising influence of the illegal drug trade. Thisinfluence has infiltrated the societies of almost every country in the region, distorting theeconomic and social development of rural areas, and undermining the confidence ofcitizens in public officials and, indeed, the entire justice and political system (Garrett 1995).

After decades of advocacy and support of the smallholder sector by scholars,bilateral donors, NGOs, private voluntary organizations, and human right organizations,there is wide recognition of the need for farmer empowerment. As long as small farmersin risk-prone environments continue to be at the periphery of the political process andgovernments continue to favor the politically powerful urban population, these poverty-stricken and ecologically vulnerable rural areas will remain ignored, except as pools oflabor or target zones for welfare transfers. Consistent with human rights and democraticgovernance, it should be made legally and institutionally easier for resource-poor farmersin traditional areas to group together in order to have their own political voice and clout.

The weak legitimacy of many governments in the region also hinders policymakersfrom formulating and implementing rural strategies capable of mobilizing the population.Hence, mobilizing effective support for Latin American countries that are moving in theright direction and that can serve as models to other countries in the region is just as urgentas reducing support for those that have favored large farms and corporations to the neglectof small farmers and the rural poor. In addition, improvement in income distribution,although officially recognized as an essential means of fostering economic development andsocioeconomic stability, is still not operationally relevant to policy decisions in mostcountries in the region. The difficult choices and decisions based on a long-term view thataddresses the costs and tradeoffs of resolving these conflicts still have not been made.Nevertheless, we are convinced that, at least for now, the World Bank is one of the fewinstitutions that have the influence, experience, and resources to stimulate action in theregion and carry through a consistent strategy for the risk-prone environments.

28

NOTES

1. Rural dualism was long defined as "functional dualism" by De Janvry (1981) and others. The presenceof such rural dualism is recognized in two recent World Bank documents delineating regional strate-gies for the agricultural sector and rural poverty alleviation in Latin America (World Bank 1993;Wiens 1994), though its extent and various forms are not documented. Nevertheless, it is recognizedthat this phenomena is "at the root of many economic and political problems, threatening social har-mony and posing hard choices for agricultural strategy" (World Bank 1993:3).

2. Wiens (1994) reports that there is also a close correlation between ethnicity and rural poverty in risk-prone areas. Indigenous people are usually landless farm laborers in commercial farming areas, orsmallholders in areas of marginal quality with low levels of commercialization. A recent World Bankstudy estimates the indigenous population of Latin America to be between 19 million and 34 millionpeople, with the great majority living in Bolivia, Ecuador, Guatemala, Mexico, and Peru(Psacharopoulos and Patrinos 1994). The same study calculates (using the lower population estimates)that 80 percent of the indigenous population is poor, and more than half is extremely poor.

3. Referring to the effects of population growth on agriculture and technological innovation, Boserup(1965, 1981) challenges the (Ricardian-Malthusian) assumption of constant technology, and postulatesthat as rural land becomes scarce relative to population (because of population growth), land will beused more intensively in order to produce greater yields. Boserup sees agricultural intensification as anecological imperative based on a new human-land ratio: as long as the number of people continues torise on a given amount of land, or the amount of land available continues to fall for a constant numberof people, the need for increasing labor investment is expected to continue as well. She then proposedthat "the growth of population is a major determinant of technological change in agriculture," whichstems from the need to raise output per unit of area to offset the increase in labor requirements associ-ated with more intensive land use (Boserup 1965:56). According to Boserup, population becomes theindependent variable, and the dependent variables becomes agrotechnology, the intensiveness withwhich labor inputs are applied and hence the capacity of the system to support people.

4. There are additional effects other than those coming from the demand side. These occur directly on thesupply side and involve certain time lags. For example, there are more children who survive to enterthe labor force (because of declining mortality) and they seek employment 12 years to 15 years later.To the extent that farm families have insufficient plots to provide productive employment, and to theextent that off-farm employment opportunities are not available in the vicinity, young children willtend to migrate away. Secondly, having a larger number of children survive to adulthood means thatexisting family plots need to be divided among more children, leading to fragmentation of landhold-ings from one generation to the next (a period of about thirty years). Eventually land plots reach sub-sistence sizes that can no longer support the larger families, which again results in out-migration ofeither individuals or entire family units. When they move to other rural areas, this again results in landextensification, with the increased use of marginal lands and the effects described above. When theymigrate to urban areas instead, their effects on land use are indirect, due to increased urban consump-tion demands (Bilsborrow 1992).

5. In the process of extensification and resulting natural resource degradation, property rights to land,forests, and water have become particularly prone to externalities. Resources with open access are par-ticularly prone to exploitation because exploiters may benefit without paying the costs associated withreduced future productive capacity. In addition, indigenous institutions for managing common prop-erty resources are breaking down, partly through misguided efforts by government and internationalinstitutions to privatize common property without a thorough understanding of these common property

29

rights. For example, it is inappropriate to privatize property rights where people depend on key spa-tially concentrated resources, or when it could lead to "parcelization" of resources.

6. For example, the greater the availability of potentially arable land, the more likely there is to beextensification rather than intensification (Pingali and Binswanger 1987). The country's level of de-velopment, government policy priorities, and ability to finance roads and infrastructure to "open up"new lands are also important. Institutional factors and government policies regulating access to land(for example, land tenure, concentration of landholdings, privatization, etc.) and land use (price, taxa-tion and import/export policies, agricultural research and extension services, credit availability, etc.)can play major roles as well.

7. Substantial contributions to the global increase in food production are well recognized. The Collabo-rative Group on International Agricultural Research (CGIAR) estimates that during the past twenty-five years, the adoption of the new technologies increased food production by an amount sufficient tofeed I billion additional people. According to Alexandratos (1988), from 1961 to 1985, on a globalaverage, yields of the major food crops rose impressively: by 41 percent in rice, 45 percent in maize,and 70 percent in wheat. As a result, world cereal production doubled to 1.9 million tons between1961 and 1990. Green revolution wheat and rice varieties now cover approximately 24 million hec-tares and 45 million hectares, respectively, or almost 60 percent of the total planted area of each crop(Hazell and Ramasamy 1989; Pinstrup-Andersen and Pandya-Lorch 1995).

8. For example, the annual growth rate of wheat yields in some of the most productive areas of cultiva-tion in Asia declined from 6.2 percent in the early 1960s to 2.7 percent in the 1980s. The annualgrowth rate of rice yields has also declined from 2.9 percent in the mid- to late 1970s to 1.9 percent inthe 1980s (Rosegrant and Svendsen 1993).

9. Although the term "indigenous knowledge systems" has become standardized in the literature, in thisarticle we use the terms "traditional," "local," and "indigenous" systems which encompass the combi-nation of knowledge, productive resources, inputs, and services applied systematically by rural peoplesto produce desired outputs. They include physical forms of technology (for example, tools, seeds, etc.)as well as methods, practices, and strategies, including forns of social organization. Finally, ruralpeoples include not only "native peoples" but also peasant farmers, settlers, and other resource users.

10. Many governments have promoted fertilizer and pesticide use by subsidizing, explicitly or implicitly,their prices. Mechanisms such as access to foreign exchange on favorable terms, tax exemptions orreduced rates, easy credit, and sales below cost are used to promote pesticide use (Repetto 1985).

11. Evidence from several parts of the world shows that most agricultural technologies in use in the worldtoday were developed by farmers, not by scientists at international and national research institutes(Roling 1988; Biggs 1989; Chambers and others 1989).

12. The case studies themselves will not be discussed here since they are being compiled in a forthcomingvolume edited by the authors. A few of these experiences were presented in a recent World Bank-sponsored Workshop on Traditional and Modern Approaches to Natural Resource Management inLatin America and the Caribbean, April 25-26, 1995, in Washington, D.C.

13. Examples of detailed economic analyses include a long-term case study in western Honduras by Felberand Foletti (1989) which found that green revolution corn technology offered a lower economic returnthan traditional growing practices. Mausolff and Farber (1994) also compared the economic costs and

30

benefits of chemical technologies with those of ecological technologies using low levels of purchasedinputs in two Honduran rural development projects. They found that traditional practices based oncover cropping with velvet bean (Macuna pruriens) have tripled average corn yields, from roughly700 kilograms per hectare to about 2,000 kilograms per hectare, while using only one-fifth of thecommonly applied chemical fertilizer.

14. Wiens (1994) argues that irrigation and water rights are as unequally distributed as landownership inLatin America, and that the effects on efficiency may be as serious as the equity issue. For example,restoring water rights usurped from the highlands in the Andes, and rehabilitating traditional conser-vation structures and management institutions, could increase farm productivity by an estimated 40percent.

31

BIBLIOGRAPHY

Alexandratos, N., ed. 1988. World Agriculture Toward 2000: A FAO Study. Rome and London:FAO and Belhaven Press.

Ascher, W., and R. Healy. 1990. Natural Resource Policymaking in Developing Countries:Environment, Economic Growth, and Income Distribution. Durham: Duke University Press.

Ashby, J. 1987. "The Effects of Different Types of Farmer Participation on the Management ofOn-Farm Trials." Agricultural Administration and Extension 24:234-252.

Ashby, J., T. Gracia, M. Guerrero, C. Quir6s, J. Roa, and J. Beltrin. 1995. "OrganizingExperimenting Farmers for Participation in Agricultural Research and TechnologyDevelopment." Paper presented at the Workshop on Traditional and Modem Approaches toNatural Resource Management in Latin America and the Caribbean, April 25-26, 1995,World Bank, Washington, D.C.

Ashby, J., C. Quir6s, T. Gracia, M. Guerrero, and J. Roa. 1990. "Farmer Participation Early inthe Evaluation of Agricultural Technologies." Paper Presented at the Seminar on RevivingLocal Self-Reliance: Challenges for Rural/Regional Development in Eastern and SouthernAfrica, Arusha, Tanzania, February 21-24, 1990.

Barborak, J., and G. Green. 1987. "Implementing the World Conservation Strategy: SuccessStories from Central America and Colombia." In Southgate and Disinger, eds., SustainableResource Development in the Third World. Boulder: Westview Press.

Bebbington, A. 1995. "Organizing for Change-Organizing for Modernization? CampesinoFederations and Technical Change in Andean and Amazonian Resource Management."Paper presented at the Workshop on Traditional and Modern Approaches to NaturalResource Management in Latin America and the Caribbean, April 25-26, 1995, World Bank,Washington, D.C.

Bebbington, A., G. Thiele, P. Davies, M. Prager, and H. Riveros, eds. 1993. Non-GovernmentalOrganizations and the State in Latin America. London: Routledge.

Bheenick, R. 1989. Successful Development in Africa: Case Studies of Projects, Programs, andPolicies. EDI Development Policy Case Series, Analytical Case Studies, No. 1. Washington,D.C.: World Bank.

Biggs, S. 1989. A Multiple Source of Innovation Model ofAgricultural Research and TechnologyPromotion. Agricultural Administration (Research and Extension) Network Paper No. 6.London: ODI.

Biggs, S., and E. Clay. 1981. "Sources of Innovation in Agricultural Technology." WorldDevelopment 9(4):321-336.

Bilsborrow, R. 1992. "Population Growth, Internal Migration, and Environmental Degradation inRural Areas of Developing Countries." European Journal of Population 8:125-148.

32

Binswanger, H. 1994. "Agricultural and Rural Development: Painful Lessons." Simon BrandtAddress delivered at the 32nd annual meeting of the Agricultural Economics Association ofSouth Africa, Pretoria, South Africa, September 21, 1994.

Binswanger, H., and J. von Braun. 1991. "Technological Change and Commercialization inAgriculture: Impact on the Poor." World Bank Research Observer 6(1).

Boserup, E. 1965. The Conditions ofAgricultural Growth. Chicago: Aldine.

. 1981. Population and Technology. Oxford: Blackwell.

Brokensha, D., D. Warren, and 0. Werner, eds. 1980. Indigenous Knowledge Systems andDevelopment. Lanham: University Press of America.

Butler, L., and J. Waud. 1990. "Strengthening Extension through the Concepts of FarmingSystems Research and Extension (FSRE) and Sustainability." Journal of Farming SystemsResearch-Extension 1(1):77-92.

Cashman, K. 1989. "Agricultural Research Centers and Indigenous Knowledge Systems in aWorldwide Perspective: Where Do We Go from Here?" In Warren, Slikkerveer, and Titilola,eds., Indigenous Knowledge Systems: Implications for Agriculture and InternationalDevelopment. Technology and Social Change Program, Studies in Technology and SocialChange, No. 11. Ames: Iowa State University.

CEPAL (Comisi6n Econ6mica para Am6rica Latina). 1995. Panorama Social de AmdricaLatina. Santiago: CEPAL.

Cemea, M., J. Coulter, and J. Russell. 1985. "Building the Research-Extension-FarmerContinuum: Some Current Issues." In Cernea, Coulter, and Russell, eds., Research-Extension-Farmer: A Two-Way Continuum for Agricultural Development. Washington,D.C.: World Bank.

CGIAR (Consultative Group on International Agricultural Research). 1990. A PossibleExpansion of CGIAR. Technical Advisory Committee, Washington, D.C.

Chambers, R., and J. Pretty. 1994. "Are the International Agricultural Research Centres Tacklingthe Crucial Issues of Poverty and Sustainability?" International Agricultural Development(November/December).

Chambers, R., A. Pacey, and L. Thrupp, eds. 1989. Farmer First: Farmer Innovation andAgricultural Research. London: ITP.

Conway, G., and J. Pretty. 1991. Unwelcome Harvest: Agriculture and Pollution. London:Earthscan.

Cook, C., and M. Grut. 1989. Agroforestry in Sub-Saharan Africa: A Farmers' Perspective.World Bank Technical Paper No. 112. Washington, D.C.: World Bank.

Cortes, J. 1995. "Seed Systems Development in Peru." Paper presented at the Workshop onTraditional and Modern Approaches to Natural Resource Management in Latin America andthe Caribbean, April 25-26, 1995, World Bank, Washington, D.C.

33

DeJanvry, A. 198 1. The Agrarian Question and Reformism in Latin America. Baltimore: JohnsHopkins University Press.

Delgado, C. 1995. "Africa's Changing Agricultural Development Strategies: Past and PresentParadigms as a Guide to the Future." International Food Policy Research Institute, IFPRIDiscussion Paper No. 3. Washington, D.C.: IFPRI.

DenBiggelaar, C. 1991. "Farming Systems Development: Synthesizing Indigenous and ScientificKnowledge Systems." Agriculture and Human Values 8(1):25-36.

Denning, G. 1985. "Integrating Agricultural Extension Programs with Framing SystemsResearch." In Cemea, Coulter, and Russell, eds., Research-Extension-Farmer: A Two-WayContinuum for Agricultural Development. Washington, D.C.: World Bank.

DeWalt, B. 1995. "Using Indigenous Knowledge to Improve Agriculture and Natural ResourceManagement." Paper presented at the Workshop on Traditional and Modern Approaches toNatural Resource Management in Latin America and the Caribbean, April 25-26, 1995,World Bank, Washington, D.C.

DeWalt, B., P. Vergne, and M. Hardin. 1993. "Population, Aquaculture and EnvironmentalDeterioration: The Gulf of Fonseca, Honduras." Paper prepared for the Rene Dubos CenterForum on Population, Environment and Development, New York Academy of Medicine,September 22-23, 1993.

Dommen, A. 1988. Innovation in African Agriculture. Boulder: Westview Press.

_. 1989. "A Rationale for African Low-Resource Agriculture in Terms of EconomicTheory." In Warren, Slikkerveer, and Titilola, eds., Indigenous Knowledge Systems:Implications for Agriculture and International Development. Technology and Social ChangeProgram, Studies in Technology and Social Change, No. I 1. Ames: Iowa State University.

Dover, M., and L. Talbot. 1987. To Feed the Earth: Agroecology for Sustainable Development.Washington, D.C.: World Resources Institute.

Fairhead, J. 1990. "Fields of Struggle: Towards a Social History of Farming Knowledge andPractice in a Bwisha Community, Kivu, Zaire." Ph.D. dissertation. London: SOAS,University of London.

FAO (Food and Agriculture Organization). 1989. The State of Food and Agriculture. Rome.

_. 1991. Sustainable Agriculture and Rural Development in Latin America and theCaribbean. Regional Document No. 3. Rome.

_. 1992. FAO Agrotast-PC, Population, Production, and Food Balance Sheets Domains.Rome.

_. 1993. "Water Policies and Agricultural Development." Mimeo. Rome.

Farrington, J., and A. Bebbington. 1994. From Research to Innovation: Getting the Most fromInteraction with NGOs in Farming Systems Research and Extension. Gatekeeper SeriesPaper No. 43. London: IIED.

34

Farrington, J., and A. Martin. 1988. Farmer Participation in Agricultural Research: A Review ofConcepts and Practices. Agricultural Administration Unit Occasional Paper No. 9. London:ODI.

Fearnside, P. 1986. Human Carrying Capacity of the Brazilian Rainforest. New York: ColumbiaUniversity Press.

Felber, R., and C. Foletti. 1989. "Estudio sobre Agricultura Migratoria en la Zona de Guajiquiro,Opatoro." Programa Marcala-Goascoran, Secretaria de Recursos Naturales y Cooperaci6nSuiza del Desarrollo (COSUDE), Marcala, Honduras.

Franke, R., and B. Chasin. 1980. Seeds of Famine: Ecological Destruction and the DevelopmentDilemma in the West African Sahel. Montclair, N.J.: Allanheld, Osmun and Co.

Garrett, J. 1995. "A 2020 Vision for Food, Agriculture, and the Environment in Latin America."International Food Policy Research Institute, IFPRI Discussion Paper No. 6. Washington,D.C.: IFPRI.

Gips, T. 1986. "What Is Sustainable Agriculture?" In Allen and van Dusen, eds., GlobalPerspectives on Agroecology and Sustainable Agricultural Systems. Santa Cruz:Agroecology Program, University of California.

Groenfeldt, D. 1991. "Building on Tradition: Indigenous Irrigation Knowledge and SustainableDevelopment in Asia." Agriculture and Human Values 8(1):114-120.

Hansen, M. 1987. Escape from the Pesticide Treadmill: Alternative to Pesticides in DevelopingCountries." Mount Vernon, N.Y.: Institute for Consumer Policy Research.

Haverkort, B., W. Hiemstra, C. Reijntjes, and S. Essers. 1988. "Strengthening Farmers' Capacityfor Technology Development." IIEA Newsletter 4(3):3-7.

Hayami, Y., and V. Ruttan. 1985. Agricultural Development: An International Perspective.Baltimore: Johns Hopkins University Press.

. 1987. "Population Growth and Agricultural Productivity." In Johnson and Lee, eds.,Population Growth and Economic Development: Issues and Evidence. Madison: Universityof Wisconsin Press.

Hazell, P., and C. Ramasamy. 1989. The Green Revolution Reconsidered: The Impact of theHigh-Yielding Rice Varieties in South India. Baltimore: Johns Hopkins University Press.

Hochman, E., and D. Zilberman. 1986. "Optimal Strategies of Development Processes ofFrontier Environments." The Science of the Total Environment 55(l):111-120.

Hoskins, M. 1984. "Observations on Indigenous and Modem Agro-Forestry Activities in WestAfrica." In Jackson, ed., Social, Economic, and Institutional Aspects ofAgroforestry. Tokyo:United Nations University.

Jaffee, S., and J. Srivastava. 1992. Seed System Development: The Appropriate Roles of thePrivate and Public Sectors. Discussion Paper No. 167. Washington, D.C.: World Bank.

35

Jodha, N. 1990. "Mountain Agriculture: The Search for Sustainability." Journal of FarmingSystems Research-Extension 1(1):55-75.

Kevin, C., and G. Schreiber. 1992. The Population, Agriculture, Environmental Nexus in Sub-Saharan Africa. Agriculture and Rural Development Series No. 1, Africa TechnicalDepartment, Agriculture Division. Washington, D.C.: World Bank.

Krueger, A., M. Schiff, and A. Valdes. 1993. The Political Economy of Agricultural PriceIntervention in Latin America. Washington, D.C.: World Bank.

Leisinger, K. 1995. "Sociopolitical Effects of New Biotechnologies in Developing Countries."International Food Policy Research Institute, IFPRI Discussion Paper No. 2. Washington,D.C.: IFPRI.

Leonard, H., ed. 1989. Environment and the Poor: Development Strategies for a CommonAgenda. New Brunswick and Oxford: Transaction Books.

Marten, G., ed. 1986. Traditional Agriculture in Southeast Asia: A Human Ecology Perspective.Boulder: Westview Press.

Maurya, D., A. Bottrall, and J. Farrington. 1988. "Improved Livelihoods, Genetic Diversity andFarmer Participation: A Strategy for Rice Breeding in Rainfed Areas of India." ExperimentalAgriculture 24(3).

Mausolff, C., and S. Farber. 1995. "An Economic Analysis of Ecological AgriculturalTechnologies among Peasant Farmers in Honduras." Ecological Economics 12:237-248.

Mellor, J., and G. Desai, eds. 1985. Agricultural Change and Rural Poverty. Baltimore: JohnsHopkins University Press.

Messerschmidt, D. 1991. "Community Forestry Management and the Opportunities of LocalTraditions: A View from Nepal." In Warren, Brokensha, and Slikkerveer, eds., IndigenousKnowledge Systems: The Cultural Dimension of Development. London: Kegan PaulInternational.

Moris, J. 199 1. Extension Alternatives in Tropical Africa. London: ODI.

Niamir, M. 1990. Community Forestry: Herders' Decision-Making in Natural Resources

Management in Arid and Semi-arid Africa. Community Forest Note No. 4. Rome: FAO.

Norem, R., R. Yoder, and Y. Martin. 1989. "Indigenous Agricultural Knowledge and GenderIssues in Third World Agricultural Development." In Warren, Slikkerveer, and Titilola, eds.,Indigenous Knowledge Systems: Implications for Agriculture and InternationalDevelopment. Technology and Social Change Program, Studies in Technology and SocialChange, No. 11. Ames: Iowa State University.

Oldeman, L., V. van Engelen, and J. Pulles. 1990. "The Extent of Human-Induced SoilDegradation." In Oldeman, Hakkeling, and Sombroek, eds., World Map of the Status ofHuman-Induced Soil Degradation: An Explanatory Note. Wageningen, The Netherlands:International Soil Reference and Information Centre.

36

Pingali, P., and H. Binswanger. 1987. "Population Density and Agricultural Intensification: AStudy of the Evolution of Technologies in Tropical Agriculture." In Johnson and Lee, eds.,Population Growth and Economic Development: Issues and Evidence. Madison: Universityof Wisconsin Press.

Pinstrup-Andersen, P., and R. Pandya-Lorch. 1995. "Alleviating Poverty, IntensifyingAgriculture, and Effectively Managing Natural Resources." International Food PolicyResearch Institute, IFPRI Discussion Paper No. 1. Washington, D.C.: IFPRI.

Poffenberger, M., ed. 1990. Keepers of the Forest: Land Management Alternatives in SoutheastAsia. West Harfort: Kumarian Press.

Pretty, J. 1994. Regenerating Agriculture: Policies and Practices for Sustainable Growth andSelf-Reliance. London: Earthscan.

Psacharopoulos, G., and H. Patrinos, eds. 1994. Indigenous People and Poverty in LatinAmerica: An Empirical Analysis. World Bank Regional and Sectoral Studies. Washington,D.C.: World Bank.

Rau, B. 1991. From Feast to Famine: Official Cures and Grassroots Remedies to Africa's FoodCrisis. London: Zed Books.

Reijntjes, C., Haverkort, B. and A. Waters-Bayer. 1992. Farming for the Future: An Introductionto Low-External-Input and Sustainable Agriculture, London: MacMillan Press.

Repetto, R. 1985. "Paying the Price: Pesticide Subsidies in Developing Countries." WorldResources Institute, Research Report 2. Washington, D.C.: WRI.

Rhoades, R. 1984. Breaking New Ground: Agricultural Anthropology. Lima: CIP.

-. 1988. "Changing Perceptions of Farmers and the Expanding Challenges of InternationalAgricultural Research." Paper presented at the Conference of Farmers and Food Systems,Lima, Peru, September 26-30, 1988.

Richards, P. 1985. Indigenous Agricultural Revolution: Ecology and Food Production in WestAfrica. London: Hutchinson.

Rola, A., and P. Pingali. 1993. Pesticides, Rice Productivity, and Farmers' Health: An EconomicAssessment. Manila and Washington, D.C.: International Rice Research Institute and WorldResources Institute.

Roling, H. 1988. Extension Science: Information Systems in Agricultural Development.Cambridge: Cambridge University Press.

Rolling, N., and P. Engel. 1989. "IKS and Knowledge Management: Utilizing IndigenousKnowledge in Institutional Knowledge Systems." In Warren, Slikkerveer, and Titilola, eds.,Indigenous Knowledge Systems: Implications for Agriculture and InternationalDevelopment. Technology and Social Change Program, Studies in Technology and SocialChange, No. 11. Ames: Iowa State University.

Rosegrant, M., and M. Svendsen. 1993. "Asian Food Production in the 1990s: IrrigationInvestment and Management Policy." Food Policy 18 (February): 13-32.

37

Scherr, S., and P. Hazell. 1993. "Sustainable Agricultural Development Strategies in FragileLands." Paper prepared for the American Agricultural Economics Association 1993International Pre-Conference on Post-Green Revolution Agricultural Development Strategiesin the Third World: What Next?, Orlando, Florida, July 30-31, 1993.

Scoones, I., and J. Thompson, eds. 1994. Beyond Farmer First: Rural People's Knowledge,Agricultural Research and Extension Practice. London: Intermediate TechnologyPublications and IlED.

Smith, N. 1995. "Biodiversity and Agroforestry along the Amazon Floodplain." Paper presentedat the Workshop on Traditional and Modem Approaches to Natural Resource Managementin Latin America and the Caribbean, April 25-26, 1995, World Bank, Washington, D.C.

Southgate, D. 1992. "Promoting Resource Degradation in Latin America: Tropical Deforestation,Soil Erosion, and Coastal Ecosystem Disturbance in Ecuador." Economic Development andCultural Change 40(4):787-807.

TAC/CGIAR (Technical Advisory Committee-Consultative Group on InternationalAgricultural Research). 1988. Sustainable Agricultural Production: Implications forInternational Agricultural Research. Rome: FAO.

-. 1993. The Ecoregional Approach to Research in the CGIAR: Report of the TAC/CenterDirectors Working Group. TAC Secretariat of the CGIAR. Rome: FAO.

Thrupp, L. 1989. "Legitimizing Local Knowledge: From Displacement to Empowerment forThird World People." Agriculture and Human Values 3:13-25.

Titilola, S. 1990. The Economics of Incorporating Indigenous Knowledge Systems intoAgricultural Development: A Model and Analytical Framework. Technology and SocialChange Program, Studies in Technology and Social Change, No. 17. Ames: Iowa StateUniversity.

Uphoff, N. 1985. "Fitting Projects to People." In Cemea, ed., Putting People First: SociologicalVariables in Rural Development. New York: Oxford University Press.

Verhelst, T. 1990. No Life Without Roots: Culture and Development. London: Zed Books.

Warren, D. 1991. Using Indigenous Knowledge in Agricultural Development. World BankDiscussion Paper No. 127. Washington, D.C.: World Bank.

-. 1995. "Using Indigenous Knowledge in Agricultural Development." Paper presented atthe Workshop on Traditional and Modem Approaches to Natural Resource Management inLatin America and the Caribbean, April 25-26, 1995, World Bank, Washington, D.C.

Warren, D., and K. Cashman. 1989. Indigenous Knowledge for Sustainable Agricultural andRural Development. Gatekeeper Series Paper No. 10. London: IIED.

Weber, P. 1992. "A Place for Pesticides?" Worldwatch 15(3): 22-23.

WHO (World Health Organization). 1980. Disease Prevention and Control in WaterDevelopment Schemes. Geneva.

38

1983. Environmental Health Impact Assessment of Irrigated Agricultural DevelopmentProjects. Geneva.

Wiens, T. 1994. Rural Poverty, Sustainable Natural Resource Management, and Overall RuralDevelopment in the LAC Region: The World Bank's Strategy. Latin America and theCaribbean Technical Department, Background Note. Washington, D.C.: World Bank.

Wiggins, S., and E. Cromwell. 1995. "NGOs and Seed Provision to Smallholders in DevelopingCountries." World Development 23(3):413-422.

Wilcox, B. 1991. "In Situ Conservation of Genetic Resources: Institutional and TechnicalRequirements for a Global System and the Need for Strengthening Capacity at the NationalLevel." Report Prepared for the Forestry Division of the Food and Agriculture Organizationof the United Nations. Rome: FAO.

. 1995. "Rural Development and Indigenous Resources: Toward a Geographic-basedAssessment Framework." Paper presented at the Workshop on Traditional and ModemApproaches to Natural Resource Management in Latin America and the Caribbean, April 25-26, 1995, World Bank, Washington, D.C.

Wilken, G. 1987. Good Farmers: Traditional Agricultural Resource Management in Mexico andCentral America. Berkeley: University of California Press.

Wolf, E. 1986. Beyond the Green Revolution: New Approaches for Third World Agriculture.Washington, D.C.: Worldwatch Institute.

World Bank. 1992. World Development Report 1992. New York: Oxford University Press.

. 1993. Towards a Bank Strategy for Agriculture in the Latin America and CaribbeanRegion. Latin America and the Caribbean Technical Department, Washington, D.C.

Other Reports in the Regional Studies Program Series:

No. 30: Indigenous People and Poverty in Latin America: An Empirical Analysis,LATAD, August 1993.

No. 3 1: Labor Market Legislation in Latin America and the Caribbean, December 1993.

No. 33: Reforms and Private Participation in the Power Sector of Selected Latin Americanand Caribbean Industrialized Countries, Vols. I and II, LATAD, March 1994.

No. 34: Government and the Economy on the Amazon Frontier, LATEN, May 1994.

No. 35: The Power Sector in LAC: Current Status and Evolving Issues, LATAD, June1995.

No. 36: Infrastructure and Growth: The Latin American Case, LATEA, January 1996.

No. 37: Sustaining Safety and Soundness: Supervision, Regulation, and FinancialReform, LATEA, January 1996.

No. 38: Effective Financing of Environmentally Sustainable Development in LatinAmerica and the Caribbean, LATEN, January 1996.

No. 39a: Argentina: Mutual Fund Regulation, LATAD, October 1996.

No. 39b: Argentina: Mutual Fund Industry, LATAD, October 1996.

No. 39c: Brazil: Securities Portfolio and Investment Fund Regulation, LATAD, October1996.

No. 39d: Chile: La Regulaci6n de los Fondos Mutuos, Fondos de Inversi6n, y Fondos deInversi6n Extrajera, LATAD, October 1996.

No. 39e: Chile: Los Fondos Mutuos, LATAD, October 1996.

No. 39f: Colombia: La Regulaci6n de los Fondos de Valores, LATAD, October 1996.

No. 39g: Mexico: Mutual Fund Regulation, LATAD, October 1996.

No. 39h: Peru: Mutual Fund Regulation, LATAD, October 1996.