Road Building and Environmental Preservation in Amazonia: Turning an Environmental Liability into an Environmental Asset James Randall Kahn a,b * Alexandre Rivas b,a Renata Mourao b a Washington and Lee University b Universidade Federal do Amazonas Corresponding author: James R. Kahn Environmental Studies Program Science AG-15 Washington and Lee University Lexington, VA 24450 kahnj@wlu.edu (540)-458-8036 Key words: deforestation, environmental preservation, economic incentives, environmental enforcement Abstract: One of the most contentious developmental issues surrounding the Brazilian Amazon is the federal government's proposal to dramatically augment its transportation system, with a particular emphasis on a network of new highways connecting remote areas of the Amazon to the existing highway network and major cities such as Manaus and Porto Velho. Opponents of the plan cite past studies that link the presence of roads to increased deforestation and argue that the plan will lead to dramatically increased deforestation. The federal government believes that the development of transportation infrastructure will increase economic activity and the quality of life and that the impact on deforestation will be much less than predicted because of changes in environmental law and enforcement mechanisms since the period of time upon which the deforestation studies are based. Our paper argues that there is no gain to be realized by trying to answer the question of whether, ceteris paribus, more road-building causes more deforestation. The appropriate question to ask is, how can policy be changed to reduce the deforestation potential of an expanded road system, both in terms of limiting activities that lead to deforestation and in terms of encouraging sustainable uses of forest resources? Our paper answers this question with a particular emphasis on the development of economic incentives to limit activities and consequences of activities that might be associated with road building. In particular, we outline a plan that utilizes the revenues associated with road use to create a more effective and more independent environmental enforcement agency. 1 Introduction: Over the last decade, a tremendous controversy has developed as outlined in the pages of Science and other outlets (Laurance et al, 2004, Laurance et al 2001, Carvalho et al 2004) concerning the development of the transportation system of the Legal Amazon of Brazil. Avança Brasil and more recent plans have called for the development of a greatly expanded system of roads and waterways to spur the economic development of the interior sections of the Amazon. This has caused great consternation among many who are concerned about the Amazon, because of the historical connection between the existence of roads and the level of deforestation. However, the causes of deforestation arise from a complex set of human activities, economic incentives and government policies. Although we do not discount the potential impact of roads on deforestation, we believe that it is counterproductive to focus the discussion on the hypothesis that more roads will cause more deforestation, with the corresponding conclusion that road construction should cease. Rather, we prefer to focus the discussion on how the quality of life of rainforest inhabitants can be improved, while limiting the potential deforestation associated with this economic and social gain. Part of this process is an overall orientation towards sustainable development, and part of this process is controlling deforestation associated with the building of roads and with other activities. A solitary focus on the prohibition or restriction of new roads will not prevent deforestation, as commercial products such as soy beans and timber can be transported on existing roads and waterways. In the absence of these activities, small scale farmers could cause commensurate levels of deforestation. More importantly, it is highly unlikely that the transportation improvements will be cancelled as they are associated with large potential 2 economic benefits. Consequentially, we focus on the development of policy to minimize the deforestation associated with road building and with other activities that may or may not be related to the existence of an augmented transportation system. Most importantly, we show how an augmented road system can be turned into a source of revenue that can be used to finance a more efficient and more independent environmental enforcement agency, neutralizing not only the threat of deforestation associated with new roads, but reducing the threat of deforestation from other pressures. In short, we suggest a method for turning a potential source of deforestation into a source of environmental preservation. The organization of our chapter proceeds as follows. Section 2 discusses the benefits of an augmented system of roads, with a focus on providing the market infrastructure for sustainably produced products, particularly non-timber forest products. Section 3 examines the debate about the association of roads and deforestation, as well as providing a discussion of other causes of deforestation. Section 4 discusses a transportation user-fee based environmental enforcement procedure designed to prevent the predicted negative environmental consequences of road building. Section 5 discusses improved environmental policies for reducing the environmental impact of timbering and soy bean production. Section 6 summarizes our discussion and makes discusses knowledge gaps that must be rectified before such as system can become a reality. Section 2: The Economic and Social Benefits of an Improved Transportation System in the Legal Amazon of Brazil The extension of the transportation system in the Amazon region is associated with four major types of benefits. They are 3 Reducing transportation costs for industrial agriculture and other extractive industries such as mining and forestry. Reducing transportation costs for sustainably produced products such as fruits, nuts, fiber, sustainably harvested timber and products using these materials as an input, such as cosmetics, toiletries, ice creams, frozen juices, furniture and pharmaceutical products. Increasing the ability of government to provide services to remote regions Increasing the quality of life of remote communities. The importance of the first type of benefit is fairly straightforward. Since export crops such as soy beans and timber compete in international markets at market determined prices, the competitiveness of these exports is primarily determined by their cost. Since many of the areas in which these crops are produced are quite distant from ports and associated with poor roads, transportation costs are a relatively high component of the total cost of delivery of the product to the market. A reduction in transportation cost increases both the market share and per unit net revenue associated with the extractive activities. The transportation cost issues associated with sustainably produced products are even greater than those associated with conventional extractive commodities such as soy beans and timber. For example, production costs are much higher for sustainably produced timber than conventionally produced timber for a variety of reasons, including that only a small percentage of the volume of wood (6 to 8%) can be removed from a given plot of forest, and then the area must remain undisturbed for 30 years. This implies a much lower amount of revenue per hectare for sustainable forestry than for other methods. If transportation costs remain extremely high because of the lack of good transportation infrastructure than the sustainably produced timber is at a competitive 4 disadvantage in comparison to both non-sustainably produced timber and certified timber from more accessible areas (such as the hardwoods from temperate forests). New types of products, such as cosmetics from rainforest fruits, foliage and seeds face an even greater cost barrier. Since these products are new, the firms are most likely producing at high per-unit cost, not having achieved the economies of scale that are associated with higher production levels. Cost disadvantages imposed by inferior transportation services make the development of markets for these types of projects even more problematic. Improving the transportation infrastructure can dramatically improve the profitability of these types of products and move production activities away from destructive extractive industries, encouraging the type of "sensitive" development advocated by Carvalho et al (2001) and others. Another important benefit of the presence of roads is a greater ability to provide social services in remote regions. Survey-based research in rainforest communities has shown that community residents regard improved health and educational services as more important than increased income opportunities1. One of the major obstacles in providing these services is the lack of transportation infrastructure. In addition to improved public services, rainforest communities benefit from better transportation facilities in a number of ways. For example, if the quality of roads exceeds a minimum standard, bus routes will be established along the roads. This connects remote regions with larger cities, allowing rainforest dwellers to go to larger cities to obtain access to services and markets. In addition, the improved transportation lowers the cost of goods that are imported from abroad or from other parts of Brazil. In particular, clothing, tools and household implements become dramatically cheaper with a 1 Casey. Kahn and Rivas (2005), Rivas, Kahn and Casey (2005) 5 better transportation system. Information also flows along the roads, giving remote communities access to newspapers and other sources of information. Section 3: Roads and Other Causes of Deforestation Starting with the construction of the Trans-Amazon Highway and BR-164 in the 1960s and 1970s, moving through the largely unimplemented Avança Brasil in the 1990s, the establishment of a network of improved highways has been a central feature of the development strategy for the Legal Amazon. Improving the transportation infrastructure in this region remains a central thrust of the development strategy of the Brazilian federal government. This development strategy has not been without controversy, as waves of peasant farmers and favelistas (shanty-town dwellers) from the Northeast and South followed the highways into the periphery states of Mato Grosso, Rondônia, Acre and Pará. These peasant farmers clear-cut the land and burned the slash, planting field crops but quickly destroying the productivity of the soil. Large scale agriculturalists, primarily cattle ranchers (in the 1970s and 1980s) and soy bean farmers (primarily in the 1990s and 2000s) also followed these roads into the interior, buying cleared land from small scale farmers, clear-cutting untouched forest and implementing their agricultural schemes. The evidence strongly suggests that roads had an impact on the level of deforestation. Casual observation of satellite and aerial photography shows clearing in proximity to roads (and hidrovias), and detailed spatial statistical analysis shows a significant and strong relationship between the amount of deforestation in an area and its proximity to roads. For example, much of the analysis was conducted with the municipality as the basic unit of observation. Ceteris paribus, the greater the linear 6 kilometers of roads in a jurisdiction, the greater the amount of deforestation. The 2002 study by Laurance et al is a good example of this type of study. They use GIS analysis and multivariate regression analysis to show that population density, physical accessibility (distance to paved roads, unpaved roads and navigable rivers) and factors affecting land-use suitability (severity of dry-season, soil fertility, etc.) are correlated with deforestation. An ordination analysis showed that 60% of the variation in deforestation can be explained by variation in two ordinates. The first ordinate axis discriminates between densely populated areas with heavy road presence and sparsely populated areas with no roads. The second axis discriminates between wet areas with low dry-season severity, many navigable rivers and few roads and those areas with opposing values in these attributes. Does this mean that implementation of plans to connect the disparate parts of the Legal Amazon by highways will lead to more deforestation, and if it does lead to more deforestation, how large will the increase be? In order to answer this question, it is important to look at all the causes of deforestation, because the statistical correlation between roads and deforestation may not be due to causation, as both increased roads and increased deforestation may be caused by other factors. If deforestation and road building are caused by other factors, then scaling back the creation of new roads may not reduce future deforestation, as it would not address the underlying factors that cause deforestation. One of the most interesting and important more recent studies is that of Alexander et al (2004). The most important innovation in their study is the use of census tract level data, rather than municipality level data as seen in most studies. Since the physical unit of 7 observation is smaller, there are many more data points in the statistical analysis, by a factor of 20. In addition, the smaller unit of observation allows better measurement of the characteristics of the spatial unit, such as soil type, length of dry season, existing level of deforestation, and so on. Their results indicate a correlation between roads and deforestation, with two major qualifications. First, they find that this relationship varies with the characteristics of the census tract. For example, the rate of deforestation in areas that are already heavily cleared has a significantly lower association with reductions in transportation costs. They also find that the impacts of roads is significantly lower that in past studies, because variation in exogenous variables may increase both the amount of roads and the amount of deforestation. López and Galinato look at both direct and indirect causes of deforestation using panel data from Brazil and several other countries. They find evidence that support the hypotheses of Alexander et al, indicating that economic variables such as export policies may impact deforestation directly, roads directly, and deforestation through roads. Various studies stress the importance of differing sets of causal factors. For example, many authors cite lack of clearly defined and enforceable property rights as the primary source of deforestation (Deacon, 1994, 1995.). Caviglia and Kahn show that the lack of knowledge about sustainable forestry alternatives is a major factor for small farmers to choose clear cutting and planting of field crops as their (non-sustainable) agricultural technology. Casey, Caviglia-Harris, Kahn and Rivas showed that although property rights were a necessary condition for the choice of sustainable alternatives, the access to information was also a necessary condition for farmers to move away from destructive technologies towards agricultural technologies that were conserving of the 8 forest cover. Kahn and McDonald show that macroeconomic factors, particularly debt, are important factors in deforestation. The discussion above indicates that although roads may be a source of deforestation, they are clearly not the only source. A look at past data suggests that an increase in road building activity may increase the level of deforestation, but a cessation in road building activity will not stop deforestation. As one looks at these past studies, based primarily on data from the 1970s, 1980s and 1990s, several observations can be made. During this period in many regions of the Legal Amazon, property rights were more loosely defined and difficult to enforce, information on sustainable agriculture was scarce, and both federal and state environmental laws were weak and relatively unenforced. In fact, much of this period of analysis is in the time before the new (1988) Brazilian Constitution more clearly specified environmental law. Since that time, federal laws have become stricter, enforcement has improved, and many states have added even tougher state laws. For example, the 1988 Constitution specified that a property owner could cut no more than 50% of the forest on his land holding. More recently, the allowable amount of forest conversion has declined to 20% of a land owner's holdings. If all of these factors have changed, is it reasonable to assume that past relationships between the construction of roads and the loss of forest will continue and that these past relationships are a good predictor of the future? We suggest that it is unproductive to waste time and energy on this debate. Rather than debating whether, ceteris paribus, more roads leads to deforestation, we should be asking how we can develop good policy to reduce deforestation from all sources, 9 including roads. The following section suggests how roads can be a positive factor in preventing deforestation rather than a source of more deforestation. Section 4: A User-fee based environmental monitoring and enforcement system Independence of environmental protection It is widely recognized that Brazil has strong and well-formulated environmental law, but that it has been inadequately enforced. Inadequate enforcement has three distinct dimensions that must be addressed for enforcement to be effective. The first dimension revolves around the lack of resources for monitoring and enforcement. The second is related to the lack of an adequate baseline against which to measure changes in environmental quality and with which to enforce public property rights. The third dimension has its origins in the lack of independence between the environmental enforcement authorities2 and those parts of government charged with stimulating economic growth, responding to International Monetary Fund directives and promoting primary production activities such as agriculture, forestry and mining. All three of these impediments to successful implementation of environmental regulations can be ameliorated with more funding and more independence of funds, both of which can be generated from a transportation user tax. A tax charged for the commercial (and possibly household) use of roads has the advantage of taking a social change (the development of more roads) and turning it from something with potential to harm the environment into something with the potential to contribute to the effective protection of the environment, while stimulating economic development at the same time. 2 We are not implying that the hard-working staff of IBAMA and other enforcement agencies lack integrity or motivation. Quite the contrary, they do an amazing job given budgets and other constraints. 10 Although the exact magnitude of a potential user tax would need to be determined through a scenario analysis, a user-tax on the order of R$10 per 100 kilometers could be used to finance the environmental enforcement system. Although such a tax would only have a marginal impact on the cost of transporting goods to markets, it would generate an impressive amount of money. For example, Table 1 contains projections of the number of commercial vehicles traversing the existing highway BR174, which goes north from Manaus, through Boa Vista, to Caracas traverses more than 1000 km. in Brazil and has a volume of traffic that was predicted to include 183 commercial freight trucks per day in 2006 (see Table 1). Using this volume of traffic as a lower bound example of potential revenue, a fee of R$100 per 1000 km would then generate more than R$ 6,679,500 (US$ 3,180,714) over a 1000 km stretch of highway in 2006, with higher levels as the volume of traffic increases over time. Cars and buses could also pay a user-fee. . Estimated Commercial Vehicle Traffic on Highway BR174 and potential revenue Year Predicted daily commercial freight traffic 2006 2007 2008 2009 2010 183 192 201 211 222 Fee (R$) per 1000 km 100 100 100 100 100 Annual revenue from fee on commercial trucks 6,679,500 7,008,000 7,336,500 7,701,500 8,103,000 It is important to stress that this user fee is not designed as an economic incentive with a direct effect to modify behavior to reduce the environmental impacts of road-building. In this sense, it is not analogous to a carbon tax. In the same vein, the fee is not designed as a typical highway toll, with the purpose to provide for the maintenance of the road itself. 11 The toll is designed as an independent source of funding for environmental protection. As such, the fund should be collected and administered by either the environmental agency itself, or by a non-governmental organization (or university of national laboratory) acting on behalf of the environmental agency. The user-fees should not be collected by the transportation or economic development ministries, or by the highway authority, or the money will be diverted away from environmental protection purposes. A separate toll could be collected for the maintenance of the highway, but this would be distinct from the environmental user fee. Expenditure of the revenue on environmental protection: As mentioned earlier, the user-fee does not provide a direct incentive for environmental compliance, but it does provide the funding to ensure compliance with other environmental regulations and policy. Several important policy areas urgently need to be addressed. The first of these is the problem associated with "land-grabbing." Much of the proposed road networks traverse federal and state-owned land, where private ownership of land and activities such as agriculture and forestry are prohibited. However, many people are making false prior-use claims and going to the courts to establish ownership rights. This type of activity is a major source of deforestation, as people holding false property rights have no incentive to manage their land for the long-term, instead focusing on activities that yield high income in the short-run and result in the destruction of the rainforest. (see Mendelsohn) Those who make false land claims construct structures on the land that appear to be decades old, claiming a long history of use of the land, and 12 thereby claiming the rights to it. The SIPAM system3 can be used to inventory land in the vicinity of proposed transportation projects and note the existence or absence of humanmade structures. Such structures would have a distinct radar signature that would show their presence even if hidden from view by the canopy. This would preempt those who tried to create new "old" structure to establish a prior use of the land, when none actually existed. The radar readings of the land could be plotted into a GIS system that could then become an enforceable part of the land tenure records. The second major problem has to do with the funding for monitoring and enforcement of environmental law. Not only are there insufficient resources to engage in these activities in such a large land area, but the environmental agencies are under pressure from other government agencies whose primary mission is to support development in general, or specific economic activities such as agriculture or mining. In addition to increasing the funding for these monitoring and enforcement activities, a funding mechanism should be established to ensure the independence of environmental enforcement agencies, both at the state and federal level. Monitoring and enforcement activities should be focused at three levels, 1. Ensuring the existing land-use and environmental regulations are followed. 2. Preventing the illegal construction of local roads connected to the proposed system of major highways. This type of connector road increases the amount of land vulnerable to deforestation. 3. Solidifying land ownership records and ensuring property rights, with particular attention to preventing the surreptitious conversion of public land to privately owned land. 3 System for Integration and Protection of Amazonia—a satellite and radar surveillance system of the Brazilian Amazon. See xxxx for a discussion of this system. 13 Section 5: Policies for controlling agriculture, timbering and migration The development of an adequately funded, independent environmental agency is necessary condition for protection of the rainforest in the face of an augmented highway and hidrovia network. In addition, new policies must be created to generate economic incentives for agriculture and forestry, and the problem of interstate migration (particularly from the northeast of Brazil) must be addressed. We will illustrate this process with forestry, and note that similar types of policies can be developed for mining and industrial agriculture. The ecology of sustainable forestry is now well understood.4 First and foremost, the disturbances generated by harvesting activity must mimic natural disturbances. This implies no clear-cutting and that all openings in the forest are small, with a high ratio of edge of the disturbed area to surface area of the disturbed area. Also, only small amounts of harvesting are acceptable. As Thomas Lovejoy puts it, the only types of activities from which the rainforest can recover are "islands of development" in a "sea of forest." These general guidelines can be translated into specific restrictions that include the following. 1. The volume of wood removed from a sustainably harvested track should be kept low, on the order of 6 to 8% of the volume of wood. 2. Harvesting must be spread over at least 40 different species to prevent highgrading and changing the species composition of the forest. 3. Collateral damage to non-harvested trees must be avoided. 4. Care must be taken to avoid soil damage. 5. Sufficient number of trees of each species must be left in the area to ensure continued propagation. 4 This section is based on Kahn (2002) 14 6. After initial harvesting under these restrictive conditions, the area must be left undisturbed for 30 years. Kahn (2002) illustrates how economic incentives such as contractual leasing systems and bonds can be used to ensure that the above conditions are met and the forest is left in a condition in which the forest quickly recovers to its original state. Such as system of incentives, combined with stricter enforcement against illegal harvesting activities can significantly reduce the negative impacts associated with harvesting activity, eliminating an important source of deforestation that may be associated with the construction of roads. As mentioned above, similar types of policies can be generated for agriculture and mining and are discussed in Kahn and Rivas. Small scale farming associated with internal migration requires radically different types of policies to prevent deforestation from these sources. Low income urban residents as well as poor rural residents from the dry Northeast of Brazil move into the Amazon region to try to improve their quality of life. Since they do not have the appropriate knowledge (Caviglia, Caviglia-Harris and Kahn, Casey, Caviglia-Harris, Kahn and Rivas) to implement sustainable agriculture, their activities result in substantial deforestation in their new local. The easiest solution would be to bar internal migration, but mobility is a cornerstone of democracy. Internal migration is not only constitutionally protected, but an important element of democratic freedom. Therefore, the problem of internal migration most be addressed in other ways, chiefly by changing incentives. 15 The best solution, but the one that is most difficult to implement is to improve the quality of life in the regions from which the immigrants originate. This requires acknowledgement that solutions to the environmental problems in Amazonia require nation-wide solutions, in particular, improving educational and economic opportunities in the areas of origin, such as the coastal cities and the interior of the North East of Brazil. Since generating such improvements is a long-term solution, more immediate solution to reduce internal migration and the impacts of internal migration on the rainforest must also be found. The first step in this process is to prevent land-grabbing, which has been discussed previously in this paper. If it is more difficult to obtain land, the expected benefits of migration will be lower, lessening the demand to migrate. The second step is to direct those migrants who do arrive in Amazonia onto previously degraded land rather than having them settle in pristine forest. This is particularly true for large land redistribution projects and movements, such as the MST (Movement for Workers without Lands). Of course, if immigrants are directed into previously degraded and currently unproductive land, they must be compensated for their efforts to restore the land, provided appropriate knowledge to be successful, and helped to gain access to markets for their eventual outputs, which could include energy crops such as castor bean or palm oil, or could include agroforestry products such as a combination of shade grown coffee, fruits, and eventually, sustainably produced timber. Finally, for those migrant who somehow do become established in existing forest land, they must be educated in sustainable agroforestry techniques, given access to markets and held to the letter of the law in terms of the clearing of forest land. 16 Section 6: Conclusions Although the preponderance of evidence from empirical studies suggests that the increased presence of roads is positively correlated with deforestation, this relationship is very complex. It is likely that the magnitude of the effect is overstated, because a set of exogenous variables lead to increases in both the presence of roads and deforestation. In addition, the strength and even the sign of the relationship are likely to change spatially due to the characteristics of the land, particularly with differences related to the productivity of the land and to the amount of deforestation that has already taken place. Most importantly, roads are not the only cause of deforestation, nor even the most important cause. Consequently, preventing the development of new transportation infrastructure may not slow deforestation and the important social and economic benefits associated with improved infrastructure would be lost. Rather than argue about whether the deforestation damages are great or less than the social and economic benefits of the road, we propose that the roads be used as an agent for environmental preservation. User fees associated with the roads can be used to finance an independent environmental authority and rectify the crucial problem associated with the lack of revenue for monitoring and enforcing compliance with environmental law. This revenue could be used to verify the validity of land ownership, to monitor activities that could lead to deforestation and to enforce existing law. Policies such as performance bonds can also be developed to create economic incentives to limit the negative impact associated with agriculture, mining and forestry. This type of integrated economic development and environmental preservation program has much better prospects for improving the quality 17 of life of rainforest residents and for preserving the rainforests than a narrow policy of seeking to prevent the development of new roads. 18 References: Alexander, P, E. Reis, J.P. Robalino and C. Bohrer, 2004, Effects of Roads on Deforestation in the Brazilian Amazon: new evidence using census-tract data and addressing endogeneity, American Geophysical Union, Fall meeting 2004, abstract #b42A-05 Carvalho, Georgia, A.C. Barros, P. Moutinho, D. Nepsted, 2001. Sensitive development could protect Amazonia instead of destroying it. Nature 409(131) 11 January 2001. Casey, James, Jill Caviglia, James Kahn and Alex Rivas, 2002. Information and the Subsistence Farmer's Decision to Deforest International Journal of Sustainable Development, 4:392-414, 2002. Deacon, Robert, 1994. "Deforestation and the Rule of Law in a Cross-Section of Countries," Land Economics, November, pp. 414-30; Deacon, Robert "Assessing the Relationship between Government Policy and Deforestation," Journal of Environmental Economics and Management, January, 1995, pp. 1-18 Franceschi, Dina and James Kahn, 2003. Strong Sustainability International Journal of Sustainable Development and World Ecology, 10(2003) 211-200. Instituto de Pesquisa da Económia Aplicado (IPEA), 2005. Brasil - O Estado de uma Nação (http://en.ipea.gov.br), 2005. The Development of Markets and Market Incentives for Sustainable Forestry: Application to the Brazilian Amazon, Organization for Economic Cooperation and Development, Environment Directorate, ENV/EPOC/GSP/BIO/(2001/6)/FINAL, 46 pages, http://www.olis.oecd.org/olis/2001doc.nsf/c5ce8ffa41835d64c125685d005300b0/ abc91129a37b59b9c1256b8a00456d70/$FILE/JT00123345.PDF. 28 March 2002. Kahn, James and Judith McDonald, 1995. "Third World Debt and Tropical Deforestation" Ecological Economics (vol. 12, pp. 107-123, 1995) Kahn, James and Jill Caviglia, 2001. Diffusion of Sustainable Agriculture In the Brazilian Tropical Rain Forest: A Discrete Choice Analysis Economic Development and Cultural Change, 49:2, pp 311-334, 2001 Laurance, W.F., A.K.M. Alnernaz, P.M. Fearnside, H.L. Vasconcelos, and L.V. Ferreira, 2004. Deforestation in Amazonia Science 304(5674)"1109-1111, 21 May 2004. 19 Laurance, W.F., M. A. Cochran, S. Bergen, P.M. Fearnside, P. Delamônica, C. Barber, S. D'Angelo and T. Fernandes, 2001. The Future of the Brazilian Amazon, Science 291 (5503): 438-39. 19 January 2001. López, Ramón and Gregmar I. Galinato, 2004 Trade Policies, Economic Growth and the Direct Causes of Deforestation. University of Maryland working paper, http://www.arec.umd.edu/rlopez/GovDev/Land%20economics%20final-rev.pdf. Lovejoy Mendelsohn Schaeffer, Roberto, and Ricardo L.V. Rodrigues, 2005, Underlying causes of Deforestation, Science 307(5712): 1046-47 18 February 2005. Young, Carlos Eduardo, undated, Public Policies and Deforestation in the Brazilian Amazon, NIEAD, Universidade Federal do Rio de Janeiro, http://www.niead.ufrj.br/artigocadu.htm. MT/DNER 2002 two PIATAM papers 20