Agricultural colonization and environmental degradation in frontier developing economies

!，、亂―〕―!&!&！一‘ . i - This paper has been prepared by Douglas Southgate and David Pearce, consultants to the Environment Department of the World Bank. Dr. Southgate is a resource economist and an associate professor in the Department of Agricultural Economics and Rural Sociology at Ohio State University. Dr. Pearce is a professor of economics at University College London. The authors gratefully acknowledge Ernst Lutz' review of an earlier draft. Departmental Working Papers are not formal publications of the World Bank. They present preliminary and unpolished results of country analysis or research that are circulated to encourage discussion and comment; citation and the use of such a paper should take account of its provisional character. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank. to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Because of the informality and to present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. . IL - Developed in this paper is a microeconomic model of agricultural colonists' choices among resource development options. We use the model to demonstrate that, under tenurial conditions typically found along a third world agricultural frontier, a cycle of excessive deforestation and exhaustive management of existing farmland is bound to emerge. We also show that changing economic incentives or introducing new farming or resource management techniques can result in mixed environmental impacts. Tor-example, because it can-diminish the scarcity value- of laborf, the introduction of conservation tillage can accelerate land clearing. Similarly, although it -might -render moe beti the environmental -impacts of any given level of migration to fragile lands, extending information on improved systems for managing those lands might also-aweeUerate migration to hillsides, tropical forests, and other areas traversed by agricultural frontiers. The implications of these and other results of our causal analysis of land degradation for the design of projects and the formulation of policy are discussed at the end of the paper. In particular, the linkage between small farmers' decisions regarding the management of existing agricultural land and their decisions regarding land clearing needs to be appreciated when evaluating existing and alternative land tenure regimes, agricultural pricing policies, "technical" solutions to resource degradation problems, and conservation projects. - LLL - INVIROMML = ==AA?O IN lmUill = =nEoN gmmNmIEs TABLE OF CONTENTS I. INTRODUCTION 1 II. THE MODEL 2 III. FACTORS INFLUENCING CONSERVATION AND DEFORESTATION 8 IV. SUMMARY AND CONCLUSIONS 11 FOOTNOTES 13 REFERENCES 15 LIST 0 1100888= Figure 1: Allocation of Labor Between Erosion Control and Deforestation 7 Figure 2: Impacts on Labor Allocation of Changing Price Signals and Technology 9 -1- AGRICULTURAL COLONIZATION AND ENVIRONMETAL DEGRADATION IN FRONTIER DEVELOPING ECONOMIES I. INTRODUCTION 1. The extent and consequences of deforestation and related environmental degradation in the developing world have, in recent years, become the subject of considerable debate and concern. There is disagreement about how rapidly primary forests near the equator are being cleared or otherwise disturbed (Sedjo and Clawson [19831; Allen and Barnes [19851; Myers [1980]). Likewise, tropical deforestation's social costs are diffi;ult to estimate, both because natural scientists cannot always describe very precisely deforestation's impacts on soil quality, climate, and biological diversity and because attaching monetary values to those impacts is rarely straightforward. 2. Compared to controversy over the extent and economic consequences of tree clearing in the tropics and subtropics, it is widely agreed that small farmers are primary agents of deforestation in the developing world. For example, between 17.6 and 19.2 million hectares of forest have been cleared or otherwise disturbed in Brazil's "Legal Amazonia", which comprises provinces that lie, wholly or partly, in the Amazon Basin. Of this total, commercial logging accounts for about 5.0 million hectares, roughly 2.5 million is now in pasture, fuelwood gathering accounts for perhaps another 2.0 million, and the rest is in small farms. In Rondonia province, which borders on Peru, agricultural colonization is estimated to have accounted for 93 percent of all forest area disturbance (Browder [1985]). 3. Small farmers become agricultural colonists for many reasons. They are frequently obliged to leave land they and their forebears have long cultivated as socially disruptive rural development strategies are pursued (Blaikie [1985]) or as that land is exhausted. Though most displaced peasants settle in urban areas, many move to forested hinterlands. The "push" to emigrate in the latter direction is often reinforced by the "pull" of attractions specifically designed to encourage agricultural colonization. Road construction and other types of infrastructure development reduce the cost of marketing commodities grown in areas formerly beyond agriculture's extensive margin. Also, agricultural colonists in many countries benefit from subsidies, land grants, and grace periods for development credit (Pearce and Myers [1988]). 4. Aside from push and pull factors that encourage agricultural colonization, the tenure regime typically encountered along a developing country's agricultural frontier promotes deforestation. Quite often, removing trees and other vegetation is a prerequisite for formal property rights. In Sudan, for example, the parastatal Mechanised Farming Corporation awards fifteen-year leases only to individuals who clear 85 percent of their respective holdings in three years. Alternatively, formal property law may make no mention of deforestation's being required for -2- tenure. However, developing country governments typically take a long time to adjudicate claims for formal property rights, particularly claims made by small farmers settling in hinterlands. Also, the mechanism for enforcing formal tenure against squatters' claims is weak in many developing countries (Pearce and Myers [1988]). Consequently, informal tenure regimes, in which agricultural use rights are a principal feature, tend to arise in newly settled areas.1 5. Although the factors inducing small farmers to migrate to agriculture's extensive margin and to clear land are widely acknowledged, no one has analyzed why a cycle of excessive deforestation Ad exhaustive use of existing farmland is usually observed along third world agricultural frontiers. That cycle is the focus o this paper. Using a simple microeconomic model designed to refle.t land tenure arrangements typically encountered in areas undergoing settlement, we undertake a causal analysis of agricultural colonists' choices among resource development options. We also discuss how altered economic incentives, the introduction of new farming techniques, and changes in clearing costs affect that group's decisions regarding land tlearivg and soil conservation. This analysis yields insights for the design of conservation projects and the formulation of policy that are summarized a the end of the paper along with topics that should be addressed in future research. II. THE MODEL 6. Most economic analysis of the use and management of land is undertaken in countries where agricultural frontiers are fairly stationary. Perhaps for this reason, studies of the selection of conservation measures, the timing of erosion control activities, and related issues (Burt [1981]; McConnell [1983]) rarely take into account how the management of existing farmland is influenced by the ready availability of land that can be cleared for crop or livestock production. 7. The option to bring "idle" land into production is incorporated in some models of natural resource development. For example, control theory has been used to characterize a socially optimal sequence of "stages" of growth in an agricultural economy. That sequence generally involves geographical expansion before major investments to increase yields, which can include the application of conservation measures, are made (Hochman and Zilberman [1986]). Regardloss of how well such models describe how land should be used and managed, however, they do not fully explain the resource development decisions made by small farmers living on or near third world agricultural frontiers because they are based on the premise that the rental value of natural resources is to be maximized. 8. The premise of rent maximization is invalid for a causal analysis of environmental degradation in third world hinterlands because of the tenure regime facing agricultural colonists. Because they must deforest land to acquire formal tenure or informal usufructuary rights, settlers are obliged to disregard non-agricultural rents. Furthermore, frontier tenure regimes circumscribe agricultural colonists' decisions regarding the timing of resource development activities, which is the primary focus of most -3- dynamic models used to describe how the present value of a stream of rents can be maximized. Any individual settler realizes that he would risk losing land not currently used for crop or livestock production to others if he ever acted on the basis of a judgment that forest conservation, or even future deforestation, is more profitable than present land clearing. Accordingly, agricultural colonists do not forego the opportunity to clear any parcel immediately if agricultural rents can be captured by doing so. Recognizing this pattern of behavior, one analyzing the causes of land degradation on or near an agricultural frontier need focus only on settlers' decisions regarding erosion control and land clearing made during a single time period. 9. The design of the model developed in this paper to describe agricultural colonists' choices among land use and soil management options reflects tenurial conditions typically encountered along third world agricultural frontiers. It is our premise that that group maximizes agricultural rents, neglecting returns associated with non-agricultural land uses. Also, although the impacts on future agricultural rents associated with erosion control and land clearing in the present are taken into account, our model is static, explaining only the intratemporal allocation of labor between erosion control and land clearing. 10. AssUons. In order to focus on erosion control and deforestation, three assumptions underpin our model. A. In addition to farming, working outside of agriculture, or working in other regions, settlers can choose either to clear land or to apply conservation measures to existing farmland. B. The region's farmers possess only one non-land input: labor. C. Farmland never reverts to a non-agricultural use. No significant insights would be gained by relaxing the first assumption. In particular, analyzing the option to clear wd to apply conservation measures to idle land is not important for our purposes. Also, saying that labor is small farmers' only non-land input is no great oversimplification insofar as that group has limited access to agricultural chemicals, machinery, and credit. The consequences of relaxing the third assumption are discussed in Southgate and Pearce [1987]. Taken together, the three assumptions simplify modeling of the value of additional crop production associated with soil conservation and deforestation as well as the opportunity cost of labor allocated to those two activities. 11. Soil Conservation and the Value of Crop Production. Erosion control can affect the value of crop production in two ways. On the negative side, output typically falls while conservation measures are being applied. For example, part or all of a growing season might be lost in order to install terraces. On the positive side, arresting erosion enhances the future productivity of agricultural land. . 4 - 12. We represent both impacts with a single expression, PVe.2 On fields where in situ topsoil resources are meager, failure to control erosion can result in immediate and major reductions in crop output. On those fields, the present value of future productivity impacts is relatively large. Alternatively, that present value is often small where topsoil layers are thick (Pimentel at al. [1976]). Recognizing that variations in soil quality influence PVc, we say that an increasing and concave relationship exists between it and the total area receiving conservation practices, which can in turn be stated by dividing labor allocated to erosion control, N., by the labor intensity of erosion control, c/hectare: PVc - PVc(Nc/c) . PVNc > 0 and PVc" < 0 . (1) 13. The Value of Output from Deforested Land. The present value, at farm gate, of future production in deforested areas equals the present market value of that production less the present value of transport costs. Because of variations in soil quality, the former component of farm-gate value, PVd,3 can be expressed as an increasing and concave function of the extent of deforestation. The latter, in turn, equals labor allocated to deforestation, Nd, divided by the labor intensity of clearing, d/hectare. PVd - d(Nd/d) , PVd' > 0 and PVd" < 0 . (2) 14. Transport costs vary according to a farm's location and its output. In general, the present value of those costs can be expressed as an increasing and convex function, T, of the extent of deforestation: T - T(Nd/d) , T' and T" > 0 . (3) 15. The Opportunity Cost of Labor. The scarcity value of labor allocated to erosion control and deforestation depends not only on Nc, Nd, and the intensity of farming in the frontier region. It is also a function of off-farm employment opportunities in the frontier region and the performance of other regions' labor markets, since one can leave farming or seek employment in other regions. In this model, however, Nc and Nd are the only variables stated explicitly in the convex function, W, describing the opportunity cost of labor. The first partial derivatives of V are positive. - 5 - 16. We can normalize units of measurement in order to say that one unit of labor is needed to farm a hectare where no erosion control measures are in place and that no farming occurs where such measures are being applied. While decisions about erosion control and land clearing are being made, then, the scarcity value of labor is: Uo - Uo [(1-1/c)Nc+Nd] . (4) 17. To model Nc's and Nd's impacts on the discounted scarcity value of human effort in the future, one must recognize both that farm labor will be employed on deforested land and that applying erosion control measures affects the labor-intensity of farming the original stock of agricultural land. Given the assumption that agricultural land is never abandoned, the former employment impact equals Nd/d per anm. To address the latter impact, we specify a coefficient, a, equal to a persisting difference between labor needed to farm a hectare where erosion is not being controlled and labor used on a hectare where conservation measures are in place. Many farmers who practice conservation tillage, for instance, find that coefficient to be positive because they spend less time cultivating their fields. Alternatively, if an erosion control measure (e.g., a terrace) requires substantial maintenance, then the coefficient can be negative. Given these definitions, the present value of future employment impacts of Nc and Nd equal the following indefinite integral: j - J Wt [Nd/d-(a)(Ne/c)] e-rt dt (5) where r is th% real discount rate. 18. Efficient Tradeoffs for Agricultural Colonists. Identified in equations (1) through (5) are the elements of agricultural rents associated with N. and Nd. Adding up those elements, we obtain an objective function for the model describing efficient intratemporal allocation of settlers' labor: PV (N /c) + PVd(Nd/d) - T(Nd/d) - Wo[(1-1/c)Ne+Nd] - W. (6) 19. We will say that the optimal values of the choice variables, Nc* and N*, are positive. Also, so that economic incentives are the sole determinant of Nd*, the potential land constraint on that variable is assumed to be inactive, as is the land constraint on N,*: Nc/c < existing farmland within original frontier. (7) -6- Accordingly, the Kuhn-Tucker conditions describing Nc* and Nd* are: Wo'[(1-1/c)N,* + Nd -PV'/c + a/cf Wt'e-rtdt - Rc/c . (8) Wo1(1-1/c)N * + Nd*1 - PVd'/d - T'/d - /df t'e-rt dt - R /d . (9) Equations (8) and (9) indicate that a settlet should increase labor allocated to erosion control or to deforestation up to the point where the wage, Wo', equals the remaining arguments of marginal agricultural rents associated with Nc and Nd: Re/c and Rd/d, respectively. 20. The difference between such a guideline and efficient trade-offs made under alternative tenurial environments is clearest in the case of Nd*. Where all land is privately owned, then anyone wishing to clear a parcel for crop or livestock production must compensate that parcel's owner. Compensation, which would enter negatively on the right-hand side of equation (9), would equal the parcel's agricultural rental value (Rd/d minus Wo in this model) or its non-agricultural rental value, whichever is greater. Where they are obliged to make such compensation, farmers push agriculture's extensive margin only up to the point where the two rental values are equal. By contrast, where farmers need not pay anyone before clearing a parcel, they will follow the allocational guideline stated in equation (9), setting Nd* too high in order to push the extensive margin out to the point where the agricultural rental value of land is zero. 21. As agricultural colonists, neglecting all non-agricultural rents, allocate too much effort to land clearing, the scarcity value of labor rises. In turn, the latter impact leads to a reduction in erosion control. This linkage between Nc and Nd is illustrated by referring to a four quadrant diagram (Figure 1), the northeastern and southwestern quadrants of which show Rc/c and Rd/d, respectively. Also indicated in the southwestern quadrant is the difference between Rd/d and the marginal rental value of tree-covered land, C/d. The marginal opportunity cost of labor allocated to soil conservation and land clearing, W', is shown in the northwestern quadrant and the sum of Nc* and Nd* is represented in the southeastern quadrant. Note that, if the tenure regime were to change so that farmers began to internalize C/d, not only would Nd decline, from Nd* to NO*, but the associated decrease in Wo' would induce an increase in conservation effort, from Nc* to Nc*1. 22. Of course, the degree to which tenurial conditions influence agricultural colonists' behavior depends on the elasticity of functions describing PVc, Pd, and other components of agricultural rents. In addition, distortions in Nc* and Nd* will be great only if marginal non- agricultural rents, C, are large. A decade ago, most observers reckoned that, outside of plantations, forestry rents in the tropics were negligible (Leslie [1987]). Understanding of tropical forest management has grown, - 7 - Figure 1: Efficient Land Clearing and Erosion Control under Alternative Tenure Regimes WO' $R 0c .....t i.... N NCN $ N O~~-- e N N Ig i I RdC 1mI dt Nd deNeC + Nd N + Nd Nd Nd -8- however, leading to wider acceptance of the proposition that rents obtained from tree-covered land near the equator exceed agricultural rents obtained from the same land after clearing (Schmidt [1987]). Indeed, decisions made in Malaysia, the Ivory Coast, and other countries to arrest settlement in forested areas verify the claim that tropical forest management is feasible (Maitre (19873; Schmidt [19871). III. FACTORS INPWIENCING CONSERVATION AND DEFORESTATION 23. The model presented in the preceding section yields insights into why a cycle of excessive deforestation and exhaustion of agricultural land emerges in the vicinity of a third world agricultural frontier. That cycle would undoubtedly be less pronounced if settlers could acquire property rights needed to capture non-agricultural rents. 24. In this section, we use the model to demonstrate how the tenurial conditions typically encountered along agriculture's extensive margin influence the magnitude of variations in Nc* and Nd* provoked by changing economic incentives and technology. In particular, we investigate the consequences of changes in four market signals: wages, marketing costs, coumodity prices, and interest rates. We then identify the environmental consequrnces of changes in land clearing costs and improvements both in erosion control techniques and in farming systems for marginal hinterUinds. 25. Wgs. In general, a wage increase discourages both deforestation and erosion control. The extensive margin, at which the rents captured by an agricultural colonist reach zero, is close to markets if wages are high. The influence of wages on deforestation is especially important in Africa, where land clearing is more labor-intensive because logging is less predominant than in Asia or Latin America (Ruthenberg (1980]). At the same time, Rc equals labor's current scarcity value, Vol, at a lower level of Nc. In the context of our model, both impacts are illustrated In Figure 2A. 26. The consequences of a wage increase might be different if erosion control reduces the labor intensity of agriculture. This is the case with reduced tillage, which allows a farmer to spend less time plowing fields, often with no appreciable sacrifice in yields but almost always with increased applications of agricultural chemicals. All else remaining the same, an increase in wages will encourage the adoption of that technique. 27. Marketing Costs. The expense of moving crops and livestock to market has been shown to be an important determinant of agricultural colonization. For example, In his analysis of deforestation and settlement in lowland, humid regions in Ecuador, Bromley [1981] identified road construction and other infrastructure development, which lower marketing costs, as a primary cause of an expanding agricultural frontier. The Kuhn- Tucker conditions in our model also suggest that management of land within the original frontier is also affected by a change in marketing costs. As indicated in Figure 2B, a higher Nd, elicited by increases in the rental value of land on or beyond the existing frontier, pulls up Vol. Responding to the latter effect, farmers cut back Nc.4 Figure 2: Impacts on Land Clearing and Erosion Control of Changes In Prices and Technology A. Wage Increase B. Lower Marketing Costs, Reduced Clearing Costs, or Improved Farming Systems for Fragile Lands .... .... ... ... ... .b .. .. . .. . . . .. . . . . . . ... ....6 a 6 6 6 fie .. . . . . ------6 ' * 6 9 a a/ -- - - - - .. . / .. .. . .. .. . .. . .......... * a 6 9 * a04 C. Hihe Comdt arce D. Inraean Prdctvt or Loe IneetRtsAscaedwt rso oto ada -4 .. .... . ...6. . . ...a.............. -- -- -- - - -- -- - r ol a 4* a* - 10 - 28. Commodity Prices. Along or near an agricultural frontier, the environmental impacts of a change in commodity prices are mixed. A price rise, for example, enhances the net returns to soil conservation, thereby improving the management of land previously cleared for crop or livestock production. However, the net returns captured by those who claim land through deforestation increase as well, resulting in an outward shift in the extensive margin (Figure 2C). 29. At least as important as a general increase or decrease in crop and livestock prices is a change in relative prices. Veloz it 1. found, for example, that currency overvaluation in the Dominican Republic eracerbates land degradation because that policy reduces the incentive to produce less erosive crops (e.g., citrus fruit and coffee), which are sold abroad, rather than maize and beans, which are not traded. By contrast, the same policy might arrest land degradation in western Africa, where a highly erosive crop, groundnuts, is a principal export. 30. Interest Rates. A decrease in real interest rates enhances both the present value of future increments in crop production associated with soil conservation, PVc, as well as the present value of crops harvested from deforested land, PVd. Accordingly, both Nc and Nd increase (Figure 20). 31. Land Clearing Costs. To a settler, the expense of removing trees and other vegetation from a parcel so that crop or livestock production can begin depends on a number of factors. Along some agricultural frontiers, timber harvesting is intense, resulting in large expanses of open land (almost regardless of whether clear-cutting or more selective techniques are practiced). Furthermore, governments and logging companies often turn a "blind eye" toward small farmers occupying recently abandoned logging sites (Myers [1984]). Under these circumstances, the labor inputs to land clearing, d, are relatively low. Those inputs are also reduced, of course, by the introduction of chain saws and other land clearing equipment. 32. A reduction in d has the same impact as a reduction in marketing costs. That is, not only does Nd increase, causing the agricultural frontier to expand, but Nc declines because an increase in effort devoted to deforestation drives up wages, which in turn reduces the net returns to erosion control (Figure 2B). 33. An increase in forest product prices can have exactly the same effect. Agricultural colonists facing a frontier tenure regime are obliged to treat revenues from the sale of high-valued timber, which is usually a small portion of the biomass cleared from a tropical or subtropical parcel, solely as a negative component of clearing costs and not as an economic signal to manage a forest for its timber values. Higher forest product prices, then, can accelerate deforestation. - 11 - 34. Improved Erosion Control Techniques. An improvement in erosion control techniques either enhances the future productivity of existing farmland or diminishes the labor-intensity of arresting soil loss. Either way, Rc rises, which induces an increase in Nc. 35. Certainly, if the impacts on future yields associated with current erosion control effort (PVc in our model) are greater, then effort devoted to land clearing must fall, all else remaining the same (Figure 2D). However, a reduction in the labor-intensity of erosion control, represented in this model by a reduction in c or an increase in a, might actually reduce the opportunity cost of labor allocated to land clearing. Indeed, as inspection of equation (9) shows, Nd will expand if, for example, Ne[-l/c] falls. 36. The latter possibility deserves serious attention inasmuch as North American farmers practice reduced tillage primarily to contain labor costs and not so much to arrest soil loss. If farmers in the developing world behave likewise, then some labor will likely be redirected to land clearing. 37. Im2roved Farming Systems for Fragile Lands. Introducing better farming systems for fragile lands can likewise alleviate some environmental problems while exacerbating others. Introducing agroforestry, for example, diminishes land degradation associated with any g level of agricultural colonization. However, that intervention enhances the net returns to Ne, which in turn induces an increase in settlement (Figure 2B). IV. SUMNARY AND CONCLUSIONS 38. Land degradaticn is often severe in the vicinity of an expanding agricultural frontier. The microeconomic model developed in this paper yields insights into why this is the case. Using the model, we have shown that one's being able to acquire property rights in "idle" land by converting it to an agricultural use not only induces rational settlers to clear too much land but also discourages them from applying conservation measures to existing farmland. We have also shown that changing economic incentives or introducing new farming or resource management techniques can result in mixed environmental impacts in a region straddling agriculture's extensive margin in a developing country. 39. The model could be extended. For example, the consequences of failing to structure property rights so as to force individuals to consider downstream and other external benefits of forest conservation and erosion control (e.g., reduced eutrophication and sedimentation as well as enhanced biological diversity), which is pervasive both in the third world and in affluent countriis, could be investigated. However, such modification would not alter rhe basic approach to be taken in a causal analysis of land degradation near agricultural frontiers. That is, one would recognize that small farmers are rational, allocating scarce inputs among competing activities according to relative net returns that are conditioned by the - 12 - structure of property rights. Choices between erosion control and land clearing in particular would be understood in terms of agricultural colonists' not being able to acquire the rights needed to capture the full benefits of at least one resource development option. 40. Regardless of its simplicity and although data limitations impede its empirical validation, our analysis yields insights for the design of conservation projects. For example, the use of reduced tillage not only improves the management of existing farmland. It also lowers the opportunity cost of land clearing by reducing the demand for labor. Projects that encourage the adoption of those systems, then, might accelerate deforestation. Similarly, while it might reduce the marginal environmental impacts of settlement, introduction of improved farming systems for degradable hinterlands might encourage additional agricultural colonization. 41. The link between erosion control and land clearing in an expanding agricultural economy must also be remembered when policy affecting rural areas is being formulated. In particular, increased reliance on the price mechanism in markets for agricultural commodities, which is a centerpiece of many donor agencies' recommendations for the developing world, can encourage excessive deforestation, given tenurial conditions prevailing along many countries' agricultural frontiers. 42. Finally, our model, like other microeconomic models, is "functionalist". That is, it focuses on individual decision makers, describing their rational response to price signals under a given structure of property rights. However, it would be naive to ignore the economic interests with sufficient political power to bring into being the decision- making environment responsible for land degradation in the third world. Many developing country governments sponsor emigration to strengthen national territorial claims or to dissipate political unrest (Goodland [1985]). Also, agricultural development strategies that favor rural elites while limiting economic opportunity for small farmers are often pursued (Blaikie [1985]). In many countries, then, changing price and tenurial incentives so as to encourage resource conservation in the hinterlands can involve a realignment of political power. 43. That empowerment is far from a straightforward matter. The benefits of forest and soil conservation (e.g., climatic stability and biological diversity) are sometimes collected by foreigners. At present, mechanisms for compensating tropical countries that export those benefits are at a very early stage of development. Until they are effective, national policies and initiatives will continue to be designed without little account taken of the costs foreigners associate with land degradation. In addition, once indigenous tribes and others who dwell in tropical forests have combined their knowledge of flora and fauna with some familiarity with the outside world's technology and marketing system, they are often better prepared than anyone else to manage that resource on a sustainable basis. However, those groups are regarded by national governments as politically marginal, at best, and quite often as outside the national economy. Under these circumstances, the benefits they would associate with forest and soil conservativn, like the costs foreigners associate with land degradation, are not assigned full weight when national policies and initiatives influencing development of the agricultural frontier are determined. - 13 - 1. Where formal property rights are awarded to those who establish farms on land formerly covered with trees, emergence of an informal tenure regime in which people who conserve forests are allowed to retain property rights is possible. However, where agricultural colonization is intense, this outcome is unlikely. Game theoretic analysis consistently shows that agreement on any set of rules (e.g., those making up an informal tenure system in which property rights in forested land are recognized) is facilitated by affected parties being able to monitor easily others' compliance with those rules (Axelrod [19811). In areas undergoing rapid agricultural settlement, the cost to any individual of monitoring the behavior of all who might potentially violate an informal rule proscribing deforestation is high. Consequently, such a rule is likely to be honored in the breach, if indeed it ever em-rges at all. 2. Among alternative mathematical descriptions of the present value of soil conservation's effects on a single hectare's productivity is one pertaining to the case where (a) agricultural land is never abandoned, (b) the application of conservation measures results in slower temporal decline to the land productivity level observed after all topsoil has been lost, and (c) the real prices, p, of crops grown on the land in question does not vary over time (Southgate and Pearce [19871). Let the geometric rate of yield decline for fields where erosion control measures are in place be kc and let the (higher) rate for fields without such measures be kf. In addition, it is assumed that any field's productivity eventually declines to a minimal level, Q (for instance, after all topsoil has been washed or blown away). Under these conditions, the present value of increased crop production on a hectare where conservation measures have just been applied, Pvc, equals: PVc -(p[c(O)Q1)/(kc+r) - (p[Qf(O)-Q])/(kt+r) where r is the real discount rate, Qc(O) is the initial yield from a hectare if conservation measures are in place, and Qf(O) represents the same hectare's initial productivity if no such measures have been applied. Typically, Qe(O) exceeds Qf(O). 3. Consistent with Footnote 2's description of the present value of soil conservation's impacts on future land productivity, the present value of crops produced on a hectare of deforested land, PVd, can be expressed as follows (Southgate and Pearce [1987]): PVd - p[Qd(O)-Q]/[kd + r] - pQ/r - 14 - where Qd(O) represents initial yield from a newly cleared hectare and kd is the rate at which yields decline on a newly cleared hectare. 4. Environmental degradation, itself, can drive up marketing costs. In some parts of the Sahel, for example, desertification makes livestock trekking more risky. As a result, the drive to colonize hinterlands is weakened and the incentive to manage better whatever grasslands remain is strengthened. Land conflicts also emerge in the Sahel as individuals who would otherwise be colonizing hinterlands choose to remain close to market centers. - 15 - REFERENCES Allen, J. and D. Barnes. 1985. "The Causes of Deforestation in Developing Countries." Annals of the Association of American Geogtahers 75:2, 163-184. Axelrod, R. 1981. "The Emergence of Cooperation Among Egoists." American Political Science Review 75:2, 308-318. Blaikie, P. 1985. The Political Economv of Soil Erosion. London: Longman. Bromley, R. 1981. "The Colonization of Humid Tropical Areas in Ecuador." Singaore Journal of Tropical GeograRhy 2:1, 15-26. Browder, J. 1985. Subsidies. Deforestation. and the Forest Sector of the Brazilian Amazon. World Resources Institute. Burt, 0. 1981. "Farm Level Economics of Soil Conservation in the Palouse Area of the Northwest." American Journal of Agricultural EconomLn 63:1, 83-92. Goodland, R. 1985. "Brazil's Environmental Progress in Amazonian Development." Man's Imoact on Forests and Rivers, ed. J. Hemming. Manchester: Manchester University Press. Hochman, E. and D. Zilberman. 1986. "Optimal Strategies of Development Processes of Frontier Environments." The Science of the Total Enviroment 55:1, 111-120. Leslie, A. 1987. "A Second Look at the Economics of Natural Management Systems in Tropical Forests." Unasvia 39:1, 46-58. Maitre, H. 1987. "Natural Forest Management in Cote d'Ivoire." Unasyva 39:3-4, 000-000. McConnell, K. 1983. "An Economic Model of Soil Conservation." Americ Journal of Agricultural Economics 65:1, 83-89, Myers, N. 1980. Conservation of Tropical Moist Forests. Washington: National Academy of Sciences. Myers, N. 1984. The Primary Source. New York: Norton. Pearce, D. and N. Myers. 1988. "Economic Values and the Environment in Amasonia." The Future of Amazonia: Destruction or Sustainable Development?. ed. G. Goodman and A. Hall. London: Macmillan. Pimentel, D., E. Terbune, R. Dyson-Hudson, S. Rochereau, R. Samis, E. Smith, D. Denman, D. Reifschneider, and M. Shepard. 1976. "Land Degradation: Effects on Food and Energy Resources." Science 194, 149-155. * 16 - Ruthenberg, H. 1980. Farming Systems in the Trapica. Oxford: Oxford University Press. Schmidt, R. 1987. "Tropical Rain Forest Management." UnasyElva 39:2, 2.17. Sedjo, R. and M. Clawson. 1983. "Tropical Deforestation, How Serious?" Journal of Forestry 81:12, 792-794. Southgate, D. and D. Pearce. 1987. "Natural Resource Degradation in Developing Countries: A Causal Analysis of Agricultural Colonisation" (discussion paper 87-26). Department of Economics, University College London. Velos, A., D. Southgate, F. Hitzhusen, and R. Macgregor. 1985. "The Economics of Erosion Control in a Subtropical Watershed." Ladg enomics 61:2, 145-15