politics of sustainable develpment: Environmental policy making in four Brazilian states, The

Barry Ames

The term sustainable development has become a catch phrase of the 1990s, a handy slogan for politicians, bureaucrats, environmental activists, multinational aid officials, and even business leaders. Implementing sustainable development policy, however, is no mere technical problem. Indeed, environmental policy making is classically political: a competition among multiple interests with differing goals, resources, tactics, information, and time horizons. Who "sustains" what, for whom, why, and how? These questions underpin any analysis of the politics of environmental policy.

Scholars have paid little attention to the political side of environmental policy making in developing countries.' Although environmental policy making is often understood as a case of "diffusion," in which ideas flowed from Western Europe and the United States to the developing world, the acceptance of new ideas is always mediated by local institutions and cultures (Sikkink, 1991). Furthermore, as international linkages have come to involve more and more actors outside foreign ministries, the form of diffusion differs from classic examples like social security policy (Collier and Messick, 1975). While social security adoption followed a state-centric pattern, the adoption of environmental policies has been polycentric, engaging from the beginning both governmental and nongovernmental actors and involving international, national, and subnational interactions.

This essay examines environmental politics in four Brazilian states. Brazil is a country of continental dimensions, encompassing a wide variety of ecosystems and environmental challenges. Its regions differ in economic base, wealth, settlement patterns, population density and diversity, and social characteristics. In the last several decades, however, the country has become increasingly integrated commercially. Beginning in the 1970s, moreover, the pattern of internal migration that had begun in the 1930s-toward the interior of the country and the major cities of the South, especially Sao Paulo-broke down. Economic activity has become less concentrated, moving to the interior of states like Sao Paulo and to other regions of the country, including the Northeast (Guimaraes Neto,1995). Population flows have correspondingly shifted; those directed at the settlement frontier, for example, are now increasingly toward frontier cities. During the 1980s, the fastestgrowing regions were the North and Center-West. For the first time since the 1940s, the Southeast grew at a slower rate than the Northeast (Martine, 1995). Under these circumstances, it is more important than ever for scholars to abandon the tendency to see Brazil through the lenses of Rio de Janeiro and Sao Paulo.

Because the nation's federal system gives considerable power to states, Brazil is an ideal place to study variation in the political factors affecting decision making on environmental policy. Our cases are not designed to test a single hypothesis. In an area involving many potential decisional sites and actors, where the political parameters remain fluid, a more inductive research strategy is appropriate, taking into account the contingent and multicausal nature of policy making. We examine states with different levels of economic development, different kinds of political struggles, and different environmental problems. In all four cases, we look at the institutional setting, the immediate political context, economic conditions, local bureaucratic capacity, and the role of nongovernmental organizations (NGOs).

We pay particular attention to the political process, arguing that explanations for patterns of environmental policy making must be sought in the political characteristics of the actors and arenas in which such policy is made and implemented. Our research is based on more than sixty interviews conducted in the four states between June 1992 and August 1993. The results show that environmental policy is a function neither of the nature of the environmental problem nor of the relative wealth of the region.

Brazil's formal experience with environmental legislation goes back about twenty years. After the 1972 United Nations Conference on the Human Environment, the Brazilian government created an environmental secretariat. The first broad federal environmental legislation, approved in 1981, established a National Environmental Council (CONAMA) with members drawn from federal ministries, state agencies, business and professional associations, and several environmental organizations. The national environmental secretariat shifted among ministries through the 1980s, finally attaining ministerial level in 1994 under President Itamar Franco. The federal environmental agency IBAMA was established in 1989, consolidating previously dispersed resource-management agencies. Most states established their own environmental organs in the early 1980s (Guimaraes, 1991).

The four states in our study all have serious environmental problems. Parana, in Brazil's more developed South, has a high degree of bureaucratic capacity, highly developed nongovernmental organizations, and a low degree of political polarization. Parana's agricultural economy was threatened by massive soil erosion, but the state has been relatively successful in controlling the problem. Its capital, Curitiba, has become a showcase of innovative urban environmental programs.2 Pernambuco, in the Northeast, faces such basic environmental problems as inadequate water supplies and sanitation, and the state's poverty limits the volume of resources available to deal with them. Bureaucratic capacity, while below that of Parana, is at middle levels, but NGOs and governmental agencies are both highly politicized.

Mato Grosso, a Center-West state, combines sociocultural, geographic, economic, and ecological diversity with a remarkably entrenched political class. Its moderately politicized bureaucracy suffers primarily from a shortage of resources with which to attract and retain skilled personnel. Nongovernmental organizations are small, but many NGO leaders have unusually good contacts both in government and abroad. Rondonia, in the frontier Northwest, is characterized by a group-based patronage politics that makes the state vulnerable to the play of influence among political allies. Promoted from a territory to a state in 1981, with a fragile economy based on resource extraction, Rondonia depends on central government transfers to finance state functions; distributive questions are thus an important component of any Rondonia development project. Not surprisingly, institutions and bureaucratic capacity are weak. NGOs, profiting from international connections, wield a greater influence than might be expected.

This study consists of three sections. It begins with a discussion of the structure of Brazilian politics, stressing the factors affecting environmental policy making and the actors and institutions directly involved in environmental policy. Section 2 sketches the central features of the policy process in the four states and offers a series of cases of environmental policy conflict. Section 3 compares the cases and draws conclusions.

Brazilian politics and Environmental Policy

What enduring aspects of Brazilian politics affect environmental policy making? Our discussion emphasizes fourfactors: federalism, changing forms of representation, patterns of bureaucratic behavior, and the weakness of judicial and other oversight bodies. To these we add the impact of pervasive poverty and inequality, along with the persistent effects of the long economic crisis Brazil experienced during much of the 1980s and early 1990s.

Federalism

Since independence, states and municipalities have been important actors in Brazilian politics. The 1988 Constitution reinforced that importance by transferring new resources to subnational governments. Governors of economically important states immediately become major players in national politics (Ames, 1995).

In some states, politics has traditionally been a lucrative business monopolized by a small number of families who are themselves supported by such major economic groups as sugar planters or cattle ranchers. In other states, principally in the more developed South and Southeast, economic interests are more heterogeneous, the population is better informed about politics, and politicians are less of a class unto themselves.

State governments have historically been responsible for certain aspects of environmental policy, including pollution control, coastal management, and some management of forests and other reserve areas. Some attributions are shared between states and municipalities or between states and the federal government, and responsibilities can also be reassigned by formal agreement.

Jurisdictional confusion between state, municipal, and federal attributions is common. States, for example, construct sewage treatment plants, but municipalities often lack the money to install the sewer systems. River basin management can involve state water and sanitation authorities, federal energy agencies, and land-use planning by dozens of municipalities. Jurisdictional confusion also occurs frequently in the relationship between the federal environmental organ, IBAMA, and state environmental agencies. Disputes often arise in frontier regions over the relationship between the federal government's land agency, INCRA, and state institutions both land-related and environmental.

Representation

Brazil's electoral rules, based on open-list proportional representation, combine with the traditional strength of local and state-level political forces to produce a system of weak, ideologically incoherent parties and personalized politics (Ames, 1995; Mainwaring, 1995; Lamounier and Nohlen, 1993). Governors and presidents struggle to obtain solid legislative backing. Without strong majorities at the outset, they are often forced to build them vote by vote by making deals with individual legislators. The majority of legislators at the federal level represent state and local interests; at the state level, purely local (or family or economic group) interests predominate. Other legislators represent geographically dispersed religious or occupational groups, such as evangelicals or civil servants. Because whole states serve as atlarge electoral districts, at both the federal and state levels, money makes a big difference in campaigns for the federal Chamber of Deputies and the state assemblies. In effect, legislators are responsible neither to communities nor to parties, and their need for campaign financing makes them captives of concentrated economic interests (Ames, 1995).

The importance of state legislatures for environmental policy making varies with the stability of government majorities. Where partisan forces are balanced, legislatures can become quite important. With less competition, legislatures are marginalized, because gubernatorial bills automatically pass. Legislators' political roles, moreover, are not limited to what they do in the assembly itself. Because legislators mediate the distribution of "pork" among electoral bailiwicks, they help determine priority areas for locating public investment, and they may seek to weaken environmental measures that affect their constituencies.

Fragmented Corporatism

From the 1930s through the 1970s, most Brazilianists argued that parties and legislatures were not the major channel for making demands. Brazil was a classic case of state corporatism. Interest groups existed at the discretion of the government, channeling demands directly to the bureaucracy (Schmitter, 1971). In the 1990s, however, Brazilian corporatism is breaking down under the twin pressures of neoliberal, market-oriented ideology and the demands of society itself. But while state corporatism is declining, a transition to societal corporatism appears unlikely. Changes in global production and the state's reduced role in the economy are more likely to fragment interests than to promote their aggregation. This rule has sectoral exceptions, leading some analysts to refer to a rise in "sectoral corporatism" (Almeida, 1994; Tapia, 1994; Martin, 1997).

Claims by social movements or nongovernmental organizations to "represent" civil society are often tenuous, and many NGOs still rely on government financial support. Governments also have found the transition difficult. Current laws, for example, often require the inclusion of organized interest groups in policy-making councils, but governments frequently restrict these councils to purely consultative roles or fail to convene them at all. In environmental policy making, furthermore, it is sometimes difficult to determine what range of interests must be represented.

The field therefore is ripe for political entrepreneurship (Vogel, 1993). The most successful NGOs look for and even invent opportunities in the breakdown of the corporative order. In this sense, strategically minded actors engage in what Baumgartner and Jones (1991: 1050) call "venue shopping," finding ways to influence policy agendas without appealing to a mass public.

A wide variety of societal organizations participate in environmental politics. Most of Brazil's more than one thousand environmental organizations are small, local, single-issue, or educational (Viola,1992; Hochstetler,1997; Minc,1985). Those few that are capable of intervening effectively in policy making do so inconsistently. Environmental organizations participate in state-level environmental councils, as well as the federal CONAMA, but these essentially normative bodies have generally played a weak role on substantive questions (Acselrad,1996).

Few environmental NGOs are "professionalized," and most of these are concentrated in Rio de Janeiro, Sao Paulo, and Brasilia. Some of the more scientific organizations carry out research projects, either with international funding or under contracts with state agencies. Although some environmental organizations insist on total independence from the state, others assume formal roles monitoring government projects. As in the United States, grassroots groups often accuse the professional conservation organizations of elitism and overdependence on state or foreign donors. Political allegiances, furthermore, sometimes divide organizations purporting to have the same goals.

Until the early 1990s, most Brazilian environmentalists worked in relative isolation from each other. Communication and collaboration among NGOs increased with the preparations for the UN Conference on Environment and Development (UNCED) in 1992, when new networks developed among Brazilian environmental organizations and other social activist groups (Hochstetler, 1997; Keck and Sikkink, forthcoming; Viola, 1992, 1996).

Environmental organizations are obviously not the only social actors with an interest in being heard on environmental issues, and the private sector is far from homogeneous in its concerns. Most corporate enterprises try to shift the cost of environmental controls onto the public sector, but alliances change with the issues and over time. Cattle ranchers in the Center-West might support regulations on gold mining to decrease the mercury content of regional waters, but they oppose limits on deforestation caused by pasture expansion. Construction firms thrive on sanitation projects, and their large campaign contributions make them politically powerful. Sugar mills once had difficulty disposing of refining residues; now residues are made into fertilizer, but the mills have problems with other wastes.3 Private sector diversity is rarely visible, however, because the Federations of Agriculture and Industry strive to present a unified position on environmental questions.

Bureaucratic Competence and Economic Crisis

Analysts of Brazilian politics use the term political clientelism to refer to the pervasiveness in the public sector of individual exchanges between actors of unequal power; that is, between patrons and clients. The result of persistent clientelism is a politicized policy making that usually furthers personal gain rather than group or ideological interests. Appointments are "political" rather than technical, not merely at the level of minister or secretary general but five or more levels down. Political parties expect to nominate party faithful to apparently techni cal positions, and major disputes develop over fair division of the spoils. Many holders of high bureaucratic positions expect to run for elective office, so they use their positions to develop personal followings (Hagopian, 1996).

In general, presidents and state governors have tried to preserve islands of competence in the bureaucracy to administer areas they considered top priorities. Competence has always been quite unevenly distributed within the central government, with the economic policymaking apparatus normally receiving priority. Areas like education are at the opposite extreme, with incompetence and clientelism the rule rather than the exception. Environmental agencies have rarely been granted a high priority.

At both the federal and state levels, Brazil's environmental agencies suffer from a chronic inability to influence the early stages of policy planning in more powerful agencies whose projects affect the environment. Although a 1986 CONAMA resolution required environmental organs to approve impact reports for certain kinds of projects, these agencies have little control over such crucial areas as land use, sanitation and waste policy, or energy generation (Tauk-Tornisielo et al., 1991). Planning, agriculture, energy, education, and health departments all have significant environmental impacts. All are normally represented on state environmental councils (CONSEMAs) and on bodies coordinating particular projects, but coordination at the level of project design is rare. It follows that the environmental secretaries' personal and political connections in other areas of government are likely to be quite important.

The prolonged economic crisis of the 1980s and early 1990s also affected the environmental policy-making process by privileging shortterm economic gain over such long-term issues as sustainable development, and by starving government agencies of resources.4 Public sector salaries were often a fraction of those in the private sector; equipment, travel, and per diem budgets were scarce. Legally mandated federal transfer payments often arrived in states and municipalities so late that inflation had eroded their value. With fewer and poorer human resources trying to solve problems, burdened by more uncertainty and less time, ill-conceived policy should not have been surprising. Recent improvements in economic performance can be expected eventually to redress some of these ills, but the effects will not disappear quickly.

Oversight: The Judiciary and the Press

Policy oversight in democracies can be exercised by legislative committees, the courts, and the press, as well as by citizens' groups. The Brazilian congress and state assemblies often convene parliamentary inquiry commissions to investigate charges of environmental abuse, but few standing committees effectively exercise oversight. In principle, Brazilian administrative law offers access to the courts, but the weakness and extraordinary slowness of the judicial system has historically made litigation less attractive than in the United States. Although some signs of change are visible, at least in the more developed South, the courts are still much more effective at short-term remedies (such as obtaining temporary injunctions) than at resolving long-term issues.

The judicial system, however, is beginning to affect environmental policy making. In 1985, the federal Diffuse Interests Law (No. 7347/85) gave environmental organizations (along with government agencies and the district attorney's office) the right to sue individuals, enterprises, or government agencies damaging the environment. It is significant that these lawsuits do not have to show personal harm. Because few NGOs have enough legal expertise or confidence in the judiciary to take advantage of the new instrument, the governmental Ministerio Publico (similar to the district attorney's office) brings most lawsuits. Litigation is likely to increase, but progress will be slow until the judiciary is stronger and the rule of law gains greater acceptance (Fuks, 1996).

In recent years, environmental reporting has expanded in the major newspapers, and some smaller regional papers have followed suit. Many papers have an environment page. Activists report media receptiveness to information denouncing irregularities in environmental administration. Television coverage has also increased, and the largest network, TV Globo, broadcasts a weekly ecology program. Journalists are beginning to specialize in this area. Although attention may have peaked around the 1992 UN conference, it remains significantly above earlier levels.

Environmental Policymaking in four states

n Brazil as in other countries, understanding a policy-making process requires examining not merely the formal institutional structures but the functioning of the political process as well: its competitiveness, its transparency, its alliances. A brief sketch of the central features of the policy process in each of our four states will highlight the differences among them, creating the context to examine particular cases of environmental policy making in each state. These cases generate some hypotheses that might help guide research both in Brazil and in other developing countries.

Parana

Because its occupation was comparatively recent, with major migratory flows in the 1930s and 1940s (Martine,1995: 62), Parana lacks the kind of traditional economic and social oligarchy common in much of Brazil. Immigrant groups have traditionally kept out of statewide politics. Many of Parana's migrants are of European or Japanese stock, with large groups also entering from Minas Gerais and Rio Grande do Sul. Income distribution is more equal than in most states, partly because migration out of the state has been substantial: responding to increased concentrations of landholdings, in less than ten years one million Paranaenses have left for the state of Amazonas.

Politics is not a high-status vocation in Parana. If the state can be said to have a political oligarchy, the growers of mate-the herbal tea popular in southern Brazil, Argentina, and Uruguay-have filled that role for much of this century. Even after economic predominance shifted north to the coffee regions, the mate producers retained a disproportionate share of political power.5

In 1972-73, under the military regime, Parana underwent an unusually successful administrative reform. As a result, the state's bureaucracy is more competent than those of its neighbors, and budget processes are taken more seriously. At the same time, the military's restrictions on open politics contributed to a "feudalization" of the state bureaucracy. Each secretariat became an entity unto itself, and interagency coordination suffered. Most high officials still strive to build their own political capital for future political campaigns or higher bureaucratic positions.

Parana has its share of the vices of Brazilian politics. In 1993, Governor Roberto Requiao was temporarily removed from office for campaign malfeasance, and his political troubles paralyzed the state government. The story of his reorganization of the environmental secretariat is a good illustration of how little it takes to destabilize even apparently solid institutions.

In July 1992, the state government replaced the previous Special Secretariat for the Environment, located in the Agriculture Secretariat, with a new, cabinet-level coordinating and policy-making organ, the Instituto Ambiental do Parana, or IAP. Removing environmental policy making from the control of agricultural interests was supposed to reflect the increased salience of environmental policy. The two most important elements in the IAP were the state Superintendency of Water Resources and Environment (SUREHMA) and the Institute of Land, Cartography, and Forests (ITCF).6 Created in 1976, SUREHMA already had a reputation for competence, and its programs had been credited with reducing overall pollution levels. Its weakness lay in a lack of inspection: it was about 25 percent understaffed, and low salaries led to high personnel turnover (Interview no. 8, June 1993).

ITCF, by contrast, dated from Parana's settlement, in the 1920s, when it was involved in land titling and mapmaking. In the mid-1980s, ITCF turned to forests and protection, initially with a focus on production, and gradually became more conservationist. After its merger with SUREHMA, its eight hundred employees were to remain responsible for cartography and land surveys and were also to monitor the state parks and the state environmental protection areas (APAs), which had grown from three thousand hectares in 1978 to almost one million in 1992 (Interview no. 11, June 1992).

The merger's rationale was to involve ITCF in monitoring pollution, but the disparate missions of the two organs made this problematic. Their training, personnel qualifications, and areas of local strength were very different. Thus, when the Requiao administration first created LAP and placed SUREHMA and ITCF inside it, many feared for the new marriage. Actually, the institution was undermined in other ways. Requiao named his brother, a psychoanalyst with no experience and apparently no ideas about environmental issues, to be agency head. The director of ITCF resigned in protest. Training, inspections, and data collection ground to a near-halt. According to one estimate, in 1993,10 percent of the technical staff would move to other agencies or to the private sector (Interviews no. 3, 6, 8, 11, June 1992). Governor Requiao's political troubles also encouraged a search for scapegoats: local newspapers, quoting the governor's brother, blamed the debacle on "foreign interests" like the U.S.-based Nature Conservancy, which was helping to purchase land in the protection area of Guaraqueraba.

Although federal law required each state to have an environmental council (CONSEMA) that included representatives of government, nongovernmental organizations, and the private sector, no agent exercised effective oversight in this instance. Parana's NGOs could not apply enough pressure, either through the media or through the legislative assembly, to compel Governor Requiao even to convene their CONSEMA.

Persons interviewed from both NGOs and the state government expected little improvement in the IAP until the next administration took over in 1995 (Interviews no. 3, 11, 12, June 1992). In terms of bureaucratic efficiency, they were correct: the next governor, Jaime Lerner, created a new environmental secretariat and persevered in the SUREHMA-ITCF merger until it became an accomplished reality. Indeed, by 1997, most observers felt that the two former agencies had disappeared as bureaucratic categories. In terms of the substantive output of the new agency, however, the activists were sorely disappointed: the Lerner administration emphasized attracting investment to Parana and slighted environmental demands. In addition, business interests learned to form "front" organizations claiming to be environmental in nature, thereby allowing the new administration to fill the places reserved for environmental interests with prodevelopment organizations.

Parana's state legislature has always been weak. Turnover among deputies is very high at both the state and federal levels, and Parana's political tradition is weaker than neighbors Santa Catarina and Rio Grande do Sul. Parana is also one of the most municipalized states in Brazil: almost all Parana's deputies have very strong geographic constituencies, which their party colleagues avoid. As a result, deputies focus on local problems, and in most of the state, environmental demands carry little weight with assembly members.7

In 1988, during the state's constitutional convention, about twenty deputies formed a "Green Front"; but it was thought that by 1993, only about five (of sixty deputies) would be consistently sympathetic to environmental issues (Interviews no. 2,7, and 12,July 1992). No deputy specialized in environmental questions, and the Assembly's Environmental Committee was an unpopular assignment. Committee Chair Luciano Pizzato, moreover, had ties to timber interests.8

In a legislature characterized by high turnover, deputies with seniority develop great power. A example of this power was played out in the 1991-94 legislature with respect to the state's overall environmental law (Lei Estadual do Meio Ambiente), which had been in the legislature since 1986. The Assembly president, Anibal Curi, had blocked the law. Curi had had a long career: he was president of the Assembly during the pre-1964 days before the military regime; he lost his political rights after 1964; but by 1995 he had spent many more years in the state assembly. His opposition doomed the environmental law in the 199094 term, and it became mostly a marketing tool for politicians.9

Lei Beraldin. If the state's failure to pass an overall statute signals the weakness of environmental forces, the passage of the "Lei Beraldin," formally titled the Law of Ecological Royalties (Brazil, Parana State Legislature,1991), demonstrates that unlikely coalitions can sometimes yield positive results. The Lei Beraldin mandates the return of part of a major consumption tax to municipalities that have established environmental protection areas (APAs). What's more, a single technocrat, an economist at Parana's now-defunct development bank, was the prime mover in the legislation.10

The bank had been studying ways to increase municipal revenue, so the economist suggested creating a fund for municipalities. As it happened, the state constitution granted municipalities a certain portion of the revenue from the Merchandise and Services Consumption Tax (ICMS). Of the part allocated to municipalities, industrial production determined the distribution of 75 percent, and the remaining 25 percent was apportioned by a variety of criteria. So the economist called a meeting of mayors in municipalities with APAs. Only one responded: the mayor of Piraquara, a municipality in which an APA accounted for 75 percent of the land area. Together, they wrote a piece of legislation. Because assembly proposals need sponsors, the mayor of Piraquara supplied one Neiva Beraldin, a state deputy beholden to the mayor.

Most government agencies opposed the proposal, which failed on its first try. So the law's supporters resorted to negotiation, and a compromise gave one-fifth of the 25 percent allocated by multiple criteria to areas with APAs and freshwater springs. As a result, the municipality of Guaraquecaba, which includes a large APA and a national park, saw its receipts increase 400 percent.11

The passage of the Lei Beraldin illustrates two aspects of environmental policy making. Opposition to environmental legislation will inevitably arise from bureaucratic actors who fear that their own programs will lose revenue. At the same time, creative legislative tactics potentially can yield positive results. Indeed, the failure to pass new initiatives sometimes stems simply from lack of effort.

Pernambuco

Pernambuco is a highly politicized state in which long-established oligarchies coexist with strong populist traditions. Traditionally, Pernambuco's major crop has been sugar, and despite its decline, planters from the sugar region, the zona da mata, still wield tremendous influence. Farther from the coast, the semi-arid agreste and the arid sertao have supported cattle and cotton, although a plague has recently hit the latter. From the 1940s on, the decline in sugar and the frequent droughts in the sertao stimulated a steady stream of migrants to move to the state capital, Recife, and from there to the urban centers of the Southeast. The city's growth rate began to slow only in the 1970s.12 Greater Recife's population now constitutes more than 50 percent of Pernambuco's total, and the city's poverty and inadequate services have long provided fertile ground for a wide variety of political organizations.

During the early 1990s, three political leaders competed for control of Pernambucan politics. Miguel Arraes was a charismatic politician and former governor whose career began before the coup of 1964. Jarbas Vasconcellos, mayor of Recife, was on the center-left, with a career rooted in opposition to the military dictatorship. Joaquim Francisco, former mayor and governor from 1990 to 1994, was a conservative from the Liberal Front Party (PFL). All three had links to popular groups as well as to specific elites. Arraes's financial support came mostly from the ranchers of the sertao and agreste; Vasconcellos got his backing from the modern construction sector (the empreiteiras); and Joaquim Francisco depended on the sugar planters. Each also had supporters in the others' camps.

Politics is more important as a career and business in Pernambuco than in Parana. In the early 1990s, about half of Pernambuco's deputies, at both state and federal levels, came from families in which other members held elective office. (In Parana, by contrast, less than 10 percent were similarly linked.) For these political businesses, central government resources, especially drought relief, are a fundamental currency of influence.

Pernambuco's politicians tend to have much weaker links with specific communities than their counterparts in Parana. The demo graphic importance of greater Recife means that virtually all legislative candidates, whatever their communities of origin, seek votes in the state capital. In the absence of a diverse immigrant population, state politics is more centralized and less community based.

Without question, bureaucratic competence is lower in Pernambuco than in Parana. Politicization increases bureaucratic turnover. New governors and mayors feel obligated to cleanse their administrations of holdovers from previous governments in order to minimize the risk of losing the support of the "bases."

Under Arraes's administration (1986-90), Pernambuco had established a cabinet-rank secretariat of the environment (SEMA). In 1991, Governor Francisco eliminated the SEMA, mainly to save money but also to reduce the interference of environmental criteria in policy making (Interviews no. 24, 27, 29, June 1993). In its place he established an Environmental Directorate within the Secretariat of Planning, Science, Technology, and Environment (SEPLAN). Environmentalists regarded this change as a mistake; in the absence of a high-level office, major projects received no environmental vetting. In 1993, the Legislative Assembly approved a proposal, which the governor accepted, creating a new Secretariat of Environment and Science and Technology.

Like Governor Requiao, Joaquim Francisco never convened his state's environmental council (CONSEMA). When the Environment Committee of the legislative assembly threatened to convene the council by itself (and when newspapers took up the issue), the environmental directorate in SEPLAN proposed to reactivate the council with a new structure. The proposal envisioned equal representation between the government and "civil society."'3 Formally, Pernambuco's new CONSEMA would have the power to discuss and resolve major environmental questions. Still, this reorganizing of the CONSEMA was more rhetorical than real. Given that the NGOs had only four votes in a council of twenty-four members, the government would rarely lose disputes to environmental interests.

Despite these limitations, Pernambuco's competitive political life, along with an active civil society, forms the basis for an independent capacity to critique and propose alternatives to government plans. The federal university is strong enough to serve as the center of intellectual and scientific life in the state. Pernambuco hosts some of the bestestablished social-development NGOs in Brazil, along with autonomous research institutes. In addition, Recife's two newspapers provide lively coverage of political issues and political malfeasance.

The Port of Suape and the Costa Dourada. Two economic development projects illustrate central aspects of environmental policy making in Pernambuco. These cases show how durable environmental NGOs often evolve from discrete, short-term conflicts. They also show how local NGOs tend to mirror the state's broader political context. In Pernambuco, the broader context was the high level of left-right polarization.

Suape was a port development project fifty kilometers north of Recife. Initially intended as a dry-dock during the closure of the Suez Canal (for boats coming around the Cape of Good Hope toward the United States), it had a short life as a project for an aluminum factory designed to process ore mined elsewhere. Eventually, even without any persuasive economic justification, the project took on a life of its own.

Building the port required destroying part of the inshore reef protecting the neighboring mangrove forest. The government contracted university consultants to study the ecological aspects of this move, but it ignored the consultants' criticisms of the project and their recommendations against cutting the mangroves (Interviews no.18,21, June 1993). The movement that formed to oppose the port project included former government officials and consultants; it later became the Pernambucan Association for the Defense of Nature (ASPAN), the state's leading environmental NGO. Other participants in the antiport movement included Elias Gomes, then a municipal council member from an area affected by the project, who became chair of the Environment Committee in the legislative assembly and served from 1991 until 1994.

The finished port created few jobs and caused considerable environmental damage. Perhaps its only benefit was the shift of a refinery from Recife to Suape, reducing pollution in the Recife area. A squatter settlement developed around the port, populated by seekers of jobs that never materialized. The port destroyed about one-fifth of the mangrove forest, and its pier illegally decimated part of the reef.

In early 1992, a hotel chain proposed to build a hotel on the site to serve tourists and port workers. The state water resource company undertook the required environmental impact study. The study recommended imposing a series of conditions and limitations on the hotel, including protection for the mangrove swamp and development of a plan for collecting garbage and sewage. The governor, who supported the project, fired the technical team, but the conditions themselves had the force of law. The project was never submitted either to the state secretariat of planning or to the federal environmental agency, IBAMA (Interview no. 26, June 1993). Indeed, IBAMA's superintendent remarked that even though the law required an IBAMA opinion, his only knowledge of the project came from the newspapers. The criticism, however, had an effect: as of June 1995 (42 months later), the hotel chain had not submitted a new proposal.

Costa Dourada, by comparison, was an ambitious attempt to develop "ecological" tourism in an undeveloped area south of Recife. The first phase of the project, the Tourist Center of Guadalupe, was on paper by 1994, while a second phase, in a nearby area known as Carneiros, remained just an idea. When complete, the entire project would add more hotel beds than the state then possessed and would include a casino on a nearby island. In an area of 8,900 hectares, the state would invest in roads, communications, and sanitation.

The Guadalupe area at the time contained little but a few large farms (mostly coconut plantations), which, through death and inheritance, would inevitably break up. The typical "development" sequence in this region would begin with the initial conversion of the farms into small lots for beach homes, followed by cutting of the mangrove forest, contamination of the swamp with garbage, installation of piers (which would compete with local fishermen), intrusion of jet skis in the estuary, and so on.

The government asserted that the project would preserve the estuary and its fishing villages. It proposed to divide the region into seven districts, three for hotels and four for biological reserves and small farms. A governing council, including representatives of the state, the two communities, and the hotels, would be responsible for an annual environmental quality report. A "Center for Environmental Training" would actually produce the report, guide tourists, and coordinate activities with nearby communities.

Pernambuco's environmental regulations required the state government to prepare an environmental impact statement, so once again the state contracted university consultants. Their report recommended that only one pier be built in the area and that this pier be constructed outside the mangrove zone, with no motorized boats allowed near it. Immediate sanitation projects should be initiated in a couple of nearby towns as a way of preventing pollution (which would discourage tourists) and aiding the local communities. The report also suggested training locals as guides for programs of ecological tourism.

The leading environmental NGOs were quite hostile to the project. While their representatives occasionally argued that Pernambuco's heavy May-to-July rains would discourage tourists, the opposition seemed motivated more by a fear that the government simply could not be trusted to carry out any plan without degrading the environment. Although few participants denied the need for more tourist facilities in the state, the NGOs saw no middle ground that would enable any degree of controlled development (Interviews no. 17, 21, 25, June 1993).

The state government did little to mollify its opponents. It never sent the Costa Dourada project to IBAMA for evaluation. It built an onsite heliport to attract investors but never bothered to obtain the required environmental impact statement for it. The Secretariat of Industry, Commerce, and Tourism refused to send representatives to discuss the project with the Assembly's Environment Committee. At the same time, the Costa Dourada project could not proceed without financing from the Inter-American Development Bank, and it seemed widely understood that the environmental impact statement would play an important role in the IDB's acceptance or rejection of the project. The state government may have had no choice but to swallow strict controls on the project. The government's disorganized bureaucracy and indifference to the law had striking effects on policy conflicts in Pernambuco. Still, actors in civil society turned out to have considerable power at times, even enough power to stop major development projects.

Mato Grosso

Mato Grosso, in Brazil's Center-West, includes important areas of Amazonian forest, savannah, and pantanal (lowland swamp). It also contains key parts of the Amazon, Paraguay, and Araguaia-Tocantins river basins. The state is rich in mineral resources and timber. Cattle raising has long been important, and in recent years the savannah has proved well suited to soybeans.

During the 1970s, colonization along the main roads, coupled with large-scale investment in mechanized soybean production, fueled an expansion of agribusiness in Mato Grosso and throughout the CenterWest. By 1980, properties exceeding ten thousand hectares (representing 1 percent of all productive units) occupied 51 percent of the state's territory. Traditional activities like cattle ranching continued to expand. Colonization companies from southern Brazil made huge profits selling land acquired cheaply in the 1960s, and settlement increased in the north of the state.

In politics, after 1945, a long history of violence, political instability, and central government intervention gave way to a politics of elite accommodation (Neves, 1988). Between 1945 and 1965, the PSD and UDN systematically alternated in power. In the post-1964 military period, ARENA-the pro-regime party-dominated the state. Under the military regime, access to resources depended increasingly on the central government, and Mato Grosso was chronically dependent. In effect, the military period maintained clientelism as the glue of the intraelite bargain.

Although politics is more competitive since democracy's return, Mato Grosso still reflects this accommodative tendency. The administration of Jaime Campos (1991-94) enjoyed a stable majority in the state assembly. Although Campos's brother Julio, who governed the state from 1983 to 1987, and Cuiaba mayor Dante de Oliveira (of the PDT) were both potentially strong successor candidates, most observers expected them to reach an accommodation (Interviews no. 35, 36, 42, July 1993). As expected, Oliveira was elected governor in 1994.

Administrative capacity in Mato Grosso developed primarily in response to federal regional development programs. Until the administration of Carlos Bezerra (1987-90) began a cleanup of Cuiaba through construction of public housing and sewer improvements, social issues took a back seat to roads and telecommunications. By 1990, the region's agricultural development had generated a population explosion, but revenue shortages undermined the state administration's ability to increase services.

Mato Grosso's state government established its environmental secretariat (SEMA) in 1987. The agency got off to a rocky start, with five secretaries during the four years of the Bezerra administration. The SEMA's main problem was its lack of trained personnel. Its technical team was ill equipped to handle large, World Bank-funded projects, such as the National Environmental Program and Prodeagro.l4 Federal environmental organs in the state were equally weak, and cooperation with state officials was poor.

Mato Grosso's environmental advisory council (CONSEMA) allocated its seats in thirds among government representatives, environmental organizations, and the private sector. The council, which met regularly, approved environmental impact statements and proposed an environmental code to the state assembly. The NGO members of the CONSEMA, arguing that governmental and industrial members were exercising a permanent veto over environmental measures, periodically walked out in protest (Interview no. 35, June 1993).

Most environmental NGOs in Mato Grosso were small and poorly institutionalized, but a number of NGO leaders were experienced in government and had spent time in environmental organizations abroad. NGO leaders in the early 1990s were usually quite skeptical about the state government's desire and capacity to implement sound environmental policy (Interviews no. 35, 36, 37, June 1993). The Environment and Development Forum, an NGO coordinating body formed in preparation for UNCED, attempted to negotiate a formal role in deliberations over Prodeagro, a World Bank-financed initiative promoting environmentally sound rural development with protection for ecological and indigenous areas. The forum argued that the bank's efforts to promote sustainable agriculture were at cross-purposes with other ongoing development projects. In particular, the proposed hidrovia, a waterway intended to bring Mato Grosso 's soybean production to the markets of MERCOSUL, would encourage the expansion of monoculture relying heavily on chemical inputs.ls

The Federal University of Mato Grosso had enough strength in the sciences to deepen the capacities of state agencies and societal organizations alike. In 1993 about 80 percent of its faculty was full-time; 6 percent had doctorates, and about 40 percent had master's degrees. The university offered a master's in agronomy and had begun master's programs in public health, biology, environment, and biodiversity. Its environmental education course served all of Latin America, and it maintained a number of collaborative environmental research programs with universities in Canada and Bolivia and institutions in Europe.

During the early 1990s, the Federation of Industries of Mato Grosso (FIEMAT) was openly hostile to environmentalists both in civil society and in government agencies. The former it regarded as leftist extremists; the latter, it contended, only wanted to shut enterprises down. In the timber sector, FIEMAT was beginning to sponsor agroforestry projects and was collaborating with the university in projects researching native and exotic species.16

The Pocone Gold Mine. The conflict over the Pocone gold mine illustrates how jurisdictional disputes and differing administrative philosophies can reduce environmental effectiveness. Pocone is located at the northern end of the pantanal. Improper burning of amalgam and the resulting mercury residue in the air and waterways made placer mining an environmental problem in the town and in ecologically sensitive areas of the pantanal. Most people believed, however, that mining, by its very nature, could not be regulated. Policymakers took miners' reputation for radical independence and violent behavior as a given, a "natural" element in the cultural framework within which they worked.17

After a decision by the state environmental agency in 1987, IBAMA had attempted to close the Pocone mine, but it was unable to enforce the interdiction. Simultaneously, both private and public efforts sought to rationalize production. A new cooperative tried to stimulate the use of environmentally safer methods to extract gold. Although only a minority of miners belonged to the cooperative, its organizers asserted that it was undertaking an educational effort and that other miners would join eventually (Interviews no. 43, 51, June 1993).

The State Environmental Foundation (FEMA), in whose primary jurisdiction the problem fell, worked both with the cooperative and with the more broadly representative miners' union. In an accord negotiated in 1992, FEMA agreed to monitor the area, conduct environmental education programs, support cooperative projects, and undertake technical studies. In early 1993, Pocone established a municipal environmental commission, which also became involved with the mine. At a meeting of the relevant parties, it was agreed that all the urban mines, along with all mines in areas of permanent preservation, would close. The state then assigned a technical team from FEMA to educate the miners.

In March 1993, a university chemist found high concentrations of mercury in waters near Pocone. IBAMA responded immediately, closing the mine.'8 In spite of vigorous protest from the governor, senators, deputies, and Pocone miners, the director of IBAMA held firm, claiming that the six years since the initial CONSEMA resolution were sufficient time to rationalize operations. Once their equipment was up to code, the director argued, the miners could resume their operations. By law, mining activities must be licensed by FEMA to operate, and the director estimated that only one mine in Pocone could be licensed (Interview no. 38, June 1993).

By 1994, state agencies and IBAMA authorities in Brasilia had worked out a compromise allowing the Pocone mine to reopen. How could the mine reopen with ongoing statutory violations? Mato Grosso 's attorney general provided an answer. The question of the Pocone mine, he said, was a question that could be resolved only by personal relations, not by the courts (Interview no. 44, July 1993). Thus, this case demonstrates that while negotiation is a desirable strategy for dispute resolution, law in Brazil remains quite fragile (Interviews no. 38, 39, 43, 51, 44, June 1993).

Rondonia

Rondonia is a frontier state in the western Amazon. Rapid colonization began with the completion in 1960 of the federal highway from Cuiaba to Porto Velho. Between 1970 and 1980, the state's population grew at a yearly rate of 16 percent, continuing at 13 percent annually between 1980 and 1985. By 1992, the population had reached one million, divided equally between urban and rural areas.

Rondonia's economy has centered on timber extraction, mining (gold and cassiterite), and agriculture. Because of resource exhaustion and falling prices, all three have declined. In the early years, the state's social and development programs depended heavily on federal transfers of funds and personnel. When these transfers declined radically, in the late 1980s, Rondonia lost population, especially among the more educated and skilled sectors.

Patronage-based grupos of politicians are central actors in Rondonia's politics. These alliances include members of federal and state congressional delegations, state government agencies, and local governments. A group's influence depends on its federal and state patronage resources and on its control of major local governments. In 1992, for example, the "ruling group" in Rondonia was composed primarily of Governor Osvaldo Piana, the mayors of Porto Velho and JiParana, and Senator Odacir Soares. Inside the state administration, the selection of secretaries and deputy secretaries was allocated among the various group leaders (Interview no. 60, August 1992). Intense bargaining resulted, with political entrepreneurs reaping high rewards. A good illustration is the career of William Curi, planning secretary in the latter part of the Piana government. While head of the state assembly's Agriculture and Ecology Committee, Curi parlayed his opposition to the state's agroecological and zoning plan into influence over a number of key government posts. He ultimately became secretary of planningcharged with implementing the zoning plan.

As in Parana, the municipalization of the legislature produces a strong interest in local pork-barrel issues; but these are mediated by Rondonia's shifting political alliances. The executive's ability to mobilize federal and international resources is a key element in maintaining legislative support. Although assembly members might participate in the political grupos, the governor is equally likely to deal directly with local mayors.

Rondonia's bureaucracy includes relatively autonomous state agencies and state secretariats. During periods of economic downturn, agencies like the water and sanitation company (CAERD) are more successful than ordinary secretariats in maintaining their salaries and their technical competence. But even relatively autonomous autarchies like CAERD are subject to political influence in the location of priority service areas and in project bidding (Interview no. 56, August 1992).

In the early 1990s, the most important state environmental organs either lacked qualified personnel or borrowed personnel from other agencies. The environmental secretariat (SEDAM) was chronically understaffed, partly as a result of delays in the disbursement of funds from the state and federal governments. The secretariat also lacked infrastructure, particularly transport and technical equipment, and frequently it could not find funds for ordinary operating expenses. Although Governor Piana's well-qualified environmental secretary appointed able assistants, he was largely ignored by the secretaries of more powerful government agencies (Interview no. 54, August 1992). High turnover among skilled employees retarded the development of institutional memory.

Most of these government agencies expected their salvation-at least over the short term-to come from a $229 million project, 75 percent funded by the World Bank, called the Planafloro.19 The project was intended to assist in the implantation of the zoning plan, which called for intensification of agriculture in already settled areas and intensification of conservation activities in parks and reserves, both indigenous and extractive. It supported roads, infrastructure, and capacity building for state institutions, but it provided no support for salaries. It also supported NGO and local participation in a variety of activities, but this aspect of the project generated conflict as early as the drafting stage of the program.20

Rondonia's newspapers were weak, and its federal university was especially poor. Lacking both infrastructure and scientific personnelin 1993 it had only two or three Ph.D.'s on the faculty-the university undertook little research and rarely collaborated either with state organs or with NGOs. Although local groups organizing around the preservation of some small natural area at times received good publicity, newspapers in Rondonia were normally hostile to environmental issues.

Rural unions, however, played a potentially important role. Rondonia's rural unions were composed primarily of small proprietors living mainly along the federal highway running from Vilhena to Porto Velho. Until 1995, most of the unions were affiliated with CUT-Rural, a militant union federation linked to the Workers' Party. In recent years, several of these unions have increased their contact with environmentalists. Some rural leaders have recognized the importance of environmental ideas in imagining alternative enterprises that could make staying on the land economically viable. The union sector has been relatively receptive to new ideas, even though its rhetorical positions have appeared more ideological than pragmatic.

The centrality of World Bank projects in Rondonia's development meant that savvy governmental and nongovernmental actors had to speak to both domestic and foreign audiences. Because Rondonia's environmental problems threatened the pillars of the state's economy, ecology and ecologists were bad words in local political discourse. They became scapegoats to explain the state's problems to the local population. More sophisticated politicians, however, recognized that the Amazon forest was unique in its ability to generate external support. The result was a radical bifurcation between externally and internally directed political discourse. Externally directed discourse stressed sustainability; internally directed discourse stressed the need to develop the area's potential wealth to support farmers, timber companies, and miners. It demanded that those who cared about environmental issues (at the federal and international levels) pay for them.

The Planafloro. In the opinion of many external observers, Rondonia's NGOs gained considerable influence in the implementation of the Planafloro. A closer look shows that this influence was often more formal than real. Although formally the World Bank and the state government recognized the desirability of NGO input, the actual format may have been established merely to satisfy external critics. The participating agencies initially were not set up to mediate among the diverse interests at stake. The Planafloro Deliberative Council, for example, was supposed to be the overall policy-setting body advising the project secretariat. While negotiating its formation, in 1991, environmental and indigenista NGOs had demanded parity between government and nongovernment members. To their surprise, the state actually agreed, and several of these organizations became actively involved in the council's debates on the annual plans presented by organs seeking Planafloro funds.21

In practice, however, the Deliberative Council did not act as a planning organ. Instead, it was simply a mechanism for approving annual plans. This was not unexpected, given that several of the state secretariats involved, including SEDAM, believed that most NGOs had little capacity to participate effectively (Interview no. 54, August 1992). Moreover, the council initially did not include large and well-organized interest groups crucial to the implantation of sustainable projects.22

The most important omission was FIERO, the association of industrialists. FIERO was one of the most active regional supporters of the proposed road to the Pacific, a project that could have a major environmental impact. It was also a major actor on agroforestry issues (Interview no. 57, August 1992). Excluding FIERO meant that the state government had not organized the Deliberative Council as a venue where politically and socially significant negotiation could occur.

This decision-making structure created a dilemma for the NGOs. They were nominally included in deliberative bodies dealing with a project whose aims they supported, but in practice these organs did not function. The independent commission convened to evaluate the first year of the project presented an initial draft of its report in February 1994. This draft identified a number of serious deficits in project planning, monitoring, and implementation. Lacking support to finish the final draft, and under considerable pressure from the state government to soften its findings, the commission essentially disbanded. In mid-June, the Rondonia NGO Forum wrote an long letter to the World Bank, detailing its concerns, calling on the bank to suspend disbursements of the loan, and announcing its withdrawal from the Deliberative Council (Rondonia NGO Forum, 1994). Although an agreement between NGOs and the bank followed in August, at the end of 1994 Friends of the Earth encouraged the forum to bring the Planafloro case before the bank's new independent review panel. Under pressure to avoid a full-scale review by the panel, the state government finally demarcated indigenous and ecological areas and negotiated a longdelayed agreement committing the federal land institute to respect the zoning plan.

Overall, the Planafloro case illuminates the opportunities and constraints of venue shopping. Rond6nia's environmental and indigenista NGOs were politically weak in the state itself, but they had considerable capacity to move information both nationally and internationally. Armed with computer and fax technology, international contacts, and the ability (through resident foreigners) to produce materials in English, these groups reached external audiences more easily than local ones.

The salience of the Planafloro's predecessor in Rondonia, the Polonoroeste, in the roster of past critiques of the World Bank's environmental performance gave the NGOs leverage they might not otherwise have had. But it remained extremely difficult to translate this international leverage into anything more than a purely formal voice in the process itself. The groups themselves failed to recognize early enough that they had gained important technical input into the project but had been outflanked politically.23

DISCUSSION AND CONCLUSION

No simple measure of capacity or effectiveness captures the diversity of political interactions uncovered in the four states surveyed. In every case, bureaucratic capacity proved to be as much a function of political decisions as of economic and human resources. But although political will is a crucial variable, the costs of exercising leadership in a given policy area are a function of political-institutional histories and cultures. The institutional grid against which policy advocates act varies accordingly. In some instances, well-established policy networks, akin to the "iron triangles" that have become entrenched in U.S. politics, crowd out even the most entrepreneurial of challengers. In others, the amorphous nature of existing jurisdictions and structures deprives challengers of stable targets. Nevertheless, this interaction of policy activists and technocrats in and outside of government is clearly central to the formulation and implementation of environmental policies. Under what conditions does their interaction produce effective strategies?

Effective environmental strategies are especially problematic in Brazil, where, historically, patterns of political interaction deemphasized diffuse interests. State corporatism recognized and regulated wellorganized interests and rendered the rest invisible. With the breakdown of state corporatism, group politics in Brazilian civil society became a scramble for recognition. In some areas, such as trade unionism, the existence of bounded constituencies made evaluating rights to representation relatively easy. Other areas, however, were murkier.

In the absence of political parties that can make credible claims to represent diffuse interests, the system rewards entrepreneurship, but only up to a point. Because many environmental goods are indivisible, they can rarely serve as currency in a milieu of clientelistic political bargaining. In addition, some environmental measures offend powerful private interests. Park set-asides, for example, may represent a victory for environmentalists but a defeat for private developers. Securing consensus thus requires clearly linking environmental policy to other benefits.

We highlighted three dimensions of the political context of environmental policy: the competitiveness of state politics, clientelism and municipalization, and capacity-both bureaucratic capacity and the independent capacity of nongovernmental policy activists. Although competitiveness has historically depended on the existence of multiple groups caring about controlling public goods, neither its form nor its outcome is uniform. In Parana, where private business activities are more rewarding than politics, clientelism is less rampant. In Pernambuco, a long history of competing elites has produced a highly competitive and politically organized society, in which clientelism is held somewhat in check by a mass politics that requires substantive results to maintain popular support. In Mato Grosso, a history of extreme competitiveness (to the point of continuous violence) has led to a Colombialike accommodation among the state's older elites.24 Finally, Rondonia is a model of highly competitive politics among loosely defined alliances of politicians. In these alliances, clientelistic goods purchase allegiance, and parties are not the main vehicles of competition.

Efficient and stable bureaucratic organization facilitates the optimal utilization of whatever resources states do have. But even where bureaucratic structures were stable, we found that environmental organs were often left out of the policy-making loop, consulted only after governors had already made policy choices. Jurisdictional confusion or uncertainty as to which agency should undertake which task was rampant in each of the states. At times, environmental organs struggled to find bureaucratic homes, and interagency coordination and collaboration were sporadic.

Under these conditions, individual power and connections gain importance. Even in Parana, environmentally the most successful state, policy making remains highly dependent on individuals: the installation of the governor's brother as the head of the state's new environmental organ halted new initiatives and slowed implementation of old programs. A new, growth-oriented governor resolved efficiency problems to the detriment of substantive environmental outputs.

The stability of political support-that is, whether support for government policy initiatives must be reconquered for each new initiative-is the second component of the strength of state governments. In a political environment characterized by incoherent and undisciplined parties, personalized politics, and weak ties between legislators and constituents, governing coalitions are likely to be both unstable and distributively focused. Governors reward their allies with power over high-level appointments, even though they often see these appointees as political rivals. Where coalitions are unstable, control of "pork" is the key to power; and particularly in light of Brazil's prolonged economic crisis, any externally funded project becomes a potential source of pork. In Rondonia, access to funding from Brasilia or from international sources became a means to leverage power. Mato Grosso, by contrast, had a tradition of accommodative elite politics, a politics reconciling elites in mutually beneficial pacts. Elite accommodation typically gave governors enough control over legislatures to implement programs if they chose to do so.

The instability of political support for environmental issues also results from a scarcity of politicians who care. Committed environmentalists were rare among deputies, mayors, and other players in state politics. Except for a handful in Parana, few state-level politicians were well informed on environmental questions. Knowledgeable staff were equally scarce.

Monitoring the environmental policy process is costly. At times, government agencies deliberately and illegally withheld information from other agencies and from state environmental councils. They often elaborated projects with little or no consultation with each other, with legislatures, or with directly affected communities. Information was often accessible, but finding it required time, a framework for assessment, and technical expertise.

The main processors of policy information in industrial societieslegislatures, courts, political parties, unions, and the media-are ill equipped to play this role in Brazil. The four state assemblies played secondary roles in environmental policy making, although the legislatures of Parana and Pernambuco took some positive initiatives. In some states, the Assembly was so highly municipalized that deputies cared mostly about local-level pork barrel items; in others, deputies' bailiwicks were so scattered that citizen-legislator ties were inevitably tenuous.

Brazil's political parties rarely took environmental issues seriously. Although the mass media did devote space to environmental questions, they served up a hodgepodge of items, seldom following an environmental story from beginning to end. Unions-particularly rural unionsat times became involved in environmental questions, but they lacked staff resources to maintain a steady involvement in the area, and their primary foci were employment and land. Although a legal framework for judicial review of administrative acts did exist, its use was cumbersome and slow.

The fragility of the rule of law clearly hampered transparency in environmental policy making. Often, neither private nor public actors really expected laws to be upheld. When the judiciary cannot uphold the law, the outcome of policy struggles is purely a function of short-term power relations. The creation of environmental sections of public prosecutors' offices was a positive step, but these offices are only as strong as their resources-normally paltry-and the judiciary's overall clout.

The politicization of environmental policy making meant that technical criteria yielded to broad ideological programs or to the career considerations of individual politicians. New administrations felt compelled to terminate their predecessors' projects and begin new ones. Politically generated turnover among bureaucrats hindered environmental project development and weakened institutional memory.25

At the same time, political competition at the state level can increase efficiency. In Pernambuco, each major competitor for state power had ties with activist "bases." Although activists take more extreme positions on policy questions, sharp political competition often generates more active news media and a popular desire for visible results. Parana's antipolitical tradition, by contrast, was partly responsible for the weakness of the legislative assembly, the absence of strong social science programs in the university, and the thinness of the media.

Civic intervention in the environmental area in Brazil currently takes place almost exclusively through nongovernmental organizations. These organizations are sometimes professionalized, but more often they are very small and dominated by volunteers. Frequently, their claim to expertise in particular issue areas wins them the right to a kind of de facto recognition as participants in environmental policy-making councils.

This representation poses two problems. The smaller problem is that although social networks of personal ties may exist between NGO activists and some state agency personnel, ties that often look like the basis for an issue network around environmental matters, the job prospects for agency workers are too exposed to shifting coalitional winds for such networks to function effectively. The larger problem lies in the false sense of security that a guarantee of participatory rights can bring. The recent experience of Parana (although our investigation largely ended before these events took place) should raise a warning flag to environmentalists. Developers quickly learned to form their own NGOs, adopt the word environmental in their titles, and accept official invitations to represent environmental interests on policy-making councils. Only when the political leadership is truly supportive will a guaranteed place for environmental groups on decision-making bodies lead to environmentally friendly representation. Environmental groups ignore electoral politics at their own peril.

Although the statist model of development has yielded to neoliberalism and privatization, Brazilians hold on to an ethos of heroic appropriation by human enterprise of the country's vast potential. This ethos creates a permanent bias against policies aimed at regulating development and makes it difficult to be optimistic about the future of sustainable development policy. In the end, good policy depends on the will of political leaders: they have to care about adopting environmental protection policies, even when such policies may impede short-term growth.

Still, environmental policy making in Brazilian states is not without its bright spots. Brazil's universities are rapidly increasing their capacity to deal with sustainable development issues. Although the quality of the universities varies, broad improvement is general and likely to continue. The universities supply consultants capable of undertaking environmental impact statements; they advise governments on specific technical problems; and they furnish members and support to NGOs. A second hopeful sign is citizens' expanded ability to use the courts as a way of forcing governments to obey environmental regulations. Often these suits come from semi-independent branches of government itself, but they help close the gap between the law and its implementation.

A third cause for optimism is the increasing interest in environmental issues in the print and electronic media. The media will report on such questions if the public wants to hear about them, and the visibility of environmental issues in the media encourages politicians to take an interest. Finally, environmental NGOs have improved their capability enormously. Although most NGOs have yet to resolve their organizational problems, and although many still expect policymakers to accept their claims to "represent" civil society, NGOs are still a strong and growing force for the adoption and maintenance of sustainable development policies.

Notes

1. Studies of environmental policy making have often focused on developing an environmental economics, leaving out the political dynamics of the process. On Brazil, see Cavalcanti 1995.

2. Because our study focuses on state-level policy making, we did not consider the innovations begun in Curitiba under Mayor Jaime Lerner (now governor of Parana).

3. The old problem was a residue called vinhoto; the current residue problem is "rinse water" (agua de lavagem).

4. The interviews on which this article is based were conducted during Brazil's years of very high inflation. Prices have since stabilized, but resources for government agencies have not increased; indeed, squeezing public sector expenditures was still a major tool for price stability in 1997.

5. Before 1964 the Workers' Party (PTB) had dominated Parana's north, but the military regime crushed the PTB and left the old mate oligarchs with no challengers.

6. SUREHMA was an outgrowth of SANEPAR, the state sanitation company. It comprised hydrologists, administrators of small dams, and water quality experts. Chemical engineers joined later, and in recent years a third group has arrived, mostly biologists with an essentially protectionist outlook.

7. One of the few exceptions was Pedro Tonelli, a deputy of the PT from the southwest part of the state. Tonelli began with a municipal orientation, then gradually took positions more sympathetic to environmentalists. He moved in 1990 to the national Chamber of Deputies. Tonelli's type, however, is a rarity; a more common figure is Luciano Pizzato, former head of IBAMA and a former state deputy now also in the Chamber of Deputies. Pizzato portrays himself as an environmentalist, but his personal economic interests include sawmills in protected areas.

8. It is interesting that the Environmental Committee was more desirable in 1992 because deputies thought the Rio de Janeiro conference ECO-92 would bring them the possibility of profit.

9. From one perspective, the attempt to pass a single, overarching environmental regulation law may be typical of the Brazilian mentality of profetacos, or megaprojects. Environmentally sympathetic officials in Parana have begun to realize that sweeping proposals make too many enemies; a majority always opposes such broad legislation. Officials in other states, including Pernambuco, did not absorb this lesson as quickly, and some found it surprising that the United States does not have a single law regulating the environment in detail, although the National Environmental Policy Act comes close (Interviews no. 17 and 21, June, 1993). From another perspective, however, what really matters is the size of one's supporting coalition. The Lerner administration, which took office in 1995 with a broad base of legislative support, quickly passed a "Forest Law" that did nothing to reduce the rapid cutting of Parana's remaining forest reserves.

10. This economist had offered to help one of the grassroots environmental interest groups in Parana. But after discovering that the group had only one member, he formed his own group!

11. The legislation's sponsors hoped that the money would be put into a fund, with the income used for projects, and would not be spent purely on

salaries for public officials. So far, that has not happened. Furthermore, two current legislative projects would modify the Lei Beraldin. The environmental groups have taken no position on these proposals; indeed, many did not even know about them when they were introduced in the Assembly, because these groups reportedly had no contacts in the legislature.

12. Between 1940 and 1970, the growth rate for metropolitan Recife was 3.99 percent; from 1970 to 1980 it was down to 2.74 percent, declining further to 1.85 percent between 1980 and 1991 (Martine, 1995: 275).

13. The government would select twelve members, including eight from the state, one from IBAMA, two from municipalities, and one from the state public ministry. The Environment Committee of the state legislature would pick one representative, as would the scientific community, the union of urban workers, the business community, rural workers, lawyers, fishermen, and finally engineers, architects, and agronomists. NGOs would pick four representatives, divided according to region.

14. The former involved increased monitoring of protected areas and institutional capacity building; the latter was intended to rationalize agricultural development according to an agroecological and economic zoning plan and to safeguard protected ecological and indigenous areas.

15. Beginning in 1993, the hidrovia became the focus of a transnational campaign involving NGOs in Bolivia, Uruguay, Paraguay, and Argentina, as well as in Mato Grosso and Mato Grosso do Sul. Internationally it involved the International Rivers Network, the Bank Information Center, and the Environmental Defense Fund "Both Ends" (The Netherlands), among others. The "Rios Vivos" coalition received funding from HIVOS (Holland) and from the W. Alton Jones and Charles Stewart Mott Foundations in the United States. The campaign focused on a demand for a serious environmental impact assessment and sought to persuade the United Nations Development Program and the Inter-American Development Bank to withdraw their support from the project.

16. FIEMAT blames the state for the absence of reforestation, claiming that timber companies have neither a seed bank nor information about what and how to plant.

17. We frequently heard the phrase ninguem segura garimpeironobody controls gold miners.

18. IBAMA has authority in mining when the state organ is deemed omissive.

19. The acronym Planafloro comes from the project's Portuguese name, Plano Agropecuario e Florestal de Rondonia (Agriculture, Livestock, and Forest Plan of Rondonia). The World Bank's name for the project, interestingly enough, is quite different: the Rondonia Natural Resources Management Project.

20. Disbursements for the program began in June 1993.

21. NGO members on the council at the outset were the National Council of Rubber Tappers, Organization of Rubber Tappers of Rondonia, Environmental Protection of Cacoal, Ecological Action of the Valley of Guapore-Ecopore, Missionary Indigenous Council, Religious Land Committee, Department of Rural Workers, Research Institute for the Defense of Amazonian Identity, Association for Environmental Preservation and Recovery of Indigenous Areas, Union of Indigenous Nations, the Institute of Amazonian Studies, Articulation of Indigenous Peoples, and Institute of Pre-history, Anthropology and Ecology. The presidency of the council was exercised by the state secretary of planning, and the council secretary was the secretary general of the Planafloro.

22. Other groups left out were the somewhat weaker agricultural and commercial associations, along with significant corporate entities like the Council of Architects, Engineers, and Agronomists.

23. A more detailed look at this case is found in Keck (forthcoming). The "boomerang pattern" by which politically weak local actors may leverage their influence by working with stronger external ones is described in Keck and Sikkink (forthcoming).

24. In Colombia, however, the accommodation resulted from formal pacts, including the Pact of Sitges and Pact of San Carlos. 25. Political clientelism was not limited to the poorest states. In 1993, the federal government replaced most of its IBAMA superintendents after a deal between the Party of the Democratic Movement (PMDB) and President Itamar Franco, a deal that had nothing to do with policy and everything to do with rewarding PMDB supporters with federal jobs.

REFERENCES

ACSELRAD, HENRI (1996). "Politica ambiental e discurso democratico-o caso do Conselho Nacional do Meio Ambiente." Paper presented to the Working Group Ecologia e Sociedade, 20th Annual Meeting of ANPOCS, Caxambu, Minas Gerais, 22-26 October.

ALMEIDA, MARIA HERMEINIA TAVARES DE (1994). "O corporativismo em declinio?" pp. 51-58 in Evelina Dagnino (ed.), Anos 90: Politica e Sociedade no Brasil. Sao Paulo: Brasiliense. AMES, BARRY (1995). "Electoral Strategy Under Open-List Proportional Representation." American Journal of Political Science 39, 2 (May): 406433.

BAUMGARTNER, FRANK R., and BRYAN D. JONES (1991). "Agenda Dynamics

and Policy Subsystems." Journal of Politics 53, 4: 1044-74. Brazil. Parana State Legislature (1991). Law of Ecological Royalties ("Lei Beraldin"). Complementary law no. 59.

CAVALCANTI, CLEOVIS (1995). Desenvolvimento e natureza: estudos para uma sociedade sustentavel. Sio Paulo: Cortez. COLLIER, DAVID and RICHARD MESSICK (1975). "Prerequisites Versus Diffusion: Testing Alternative Explanations of Social Security Adoption." American Political Science Review 69, 4: 1299-1315. FUKS, MARIO (1996). "Do discurso ao recurso: uma analise da protecao judicial ao meio ambiente do Rio de Janeiro," pp. 189-216 in Leila da Costa Ferreira and Eduardo Viola (eds.), Incertezas de sustentabilidade na globalizarao. Campinas: Editora Unicamp.

GUIMARAES, ROBERTO PEREIRA (1991). The Ecopolitics of Development in the Third World: Politics and Environment in Brazil. Boulder Lynne Rienner.

GUIMARAES NETO, LEONARDO (1995). "Desigualdades regionais e federalismo," pp.13-59 in Rui de Britto Alvares Affonso and Pedro Linz Barros Silva (eds.), Desigualdades regionais e desenvolvimento. Sao Paulo: FUNDAP/Editora da Universidade Estadual Paulista. HAGOPIAN, FRANCES (1996). Traditional Politics and Regime Change in Brazil. Cambridge: Cambridge University Press. HOCHSTETLER, KATHRYN (1997). "The Evolution of the Brazilian Environmental Movement and Its Political Roles," pp.192-216 in Douglas Chalmers, Carlos M. Vilas, Katherine R. Hite, Scott B. Martin, Kerianne Piester, and Monique Segarra (eds.), The New Politics of Inequality in Latin America: Rethinking Participation and Representation. Oxford: Oxford University Press.

KECK, MARGARET E. (forthcoming). "Brazil's Planafloro: The Limits of Leverage," in Jonathan Fox and L. David Brown (eds.), The Struggle for Accountability: The World Bank, NGOs, and Grassroots Movements. Cambridge: MT Press, 1998.

KECK, MARGARET E. and KATHRYN SIKKINK (forthcoming). Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca: Cornell University Press, 1998.

LAMOUNIER, BOLIVAR and DIETER NOHLEN (1993). Presidencialismo ou

parlamentarismo. Sao Paulo: IDESP.

MAINWARING, SCOTT (1995). "Brazil: Weak Parties, Feckless Democracy," pp. 354-398 in Mainwaring and Timothy Scully (eds.), Building Democratic Institutions: Party Systems in Latin America. Stanford: Stanford University Press.

MARTIN, SCOTT(1997). "Beyond Corporatism: New Patterns of Representation in the Brazilian Auto Industry," pp. 45-71 in Douglas Chalmers et al. (eds.), The New Politics of Inequality in Latin America: Rethinking Participation and Representation. Oxford: Oxford University Press. MARTINE, GEORGE (1995). UA evolucao espacial da populacao brasileira," pp. 61-91 in Rui de Britto Alvares Affonso and Pedro Linz Barros Silva (eds.), Desigualdades regionais e desenvolvimento. Sao Paulo: FUNDAP/Editora da Universidade Estadual Paulista.

MINC, CARLOS (1985). Como fazer movimento ecologico e defender a natureza e as liberdades. Petropolis: Vozes.

NEVES, MARIA MANUELA RENHA DE NOVIS (1988). Elites politicas: competigao e dinamica partidario-eleitoral: o caso de Mato Grosso. Rio de Janeiro: IUPERJ/Vertice.

Rondonia NGO Forum (1994). "Letter from the Forum of Non-Governmental Organizations of Rondonia to the President and Executive Directors of the World Bank." Porto Velho, Rondonia. 15 June. SCHMITTER, PHILIPPE (1971). Interest Conflict and Political Change in Brazil. Stanford: Stanford University Press.

SIKKINK, KATHRYN (1991). Ideas and Institutions: Developmentalism in

Brazil and Argentina. Ithaca: Cornell University Press. TAPIA, JORGE RUBEM BITON (1994) "Corporativismo societal no Brasil: uma transicao incompleta?" pp. 65-80 in Evelina Dagnino (ed.), Anos 90: politica e sociedade no Brasil. Sio Paulo: Brasiliense. TAUK-TORNISIELO, SAMIA MARIA, NIVAR GOBBI, and HAROLD GORDON FOWLER (eds.) (1991). Analise ambiental: uma visao multidisciplinar. Sao Paulo: UNESP.

VIOLA, EDUARDO (1996). "A multidimensionalidade da globalizacao, as novas forcas sociais transnacionais e seu impacto na politica ambiental do Brasil, 1989-1995," pp. 15-66 in Leila da Costa Ferreira and Eduardo Viola (eds.), Incertezas de sustentabilidade na globalizacao. Campinas: Editora Unicamp.

(1992). aO movimento ambientalista no Brasil(1971-1991): da denuncia e conscientizacao publica para ainstitucionalizacao eo desenvolvimento sustentavel," pp. 49-76 in Miriam Goldemberg (ed.), Ecologia, ciencia e politica. Rio de Janeiro: Revan.

VOGEL, DAVID (1993). "Representing Diffuse Interests in Environmental Policy Making," pp. 237-271 in Bert Rockman and R. Kent Weaver (eds.), Do Institutions Matter: Government Capabilities in the U.S. and Abroad. Washington, DC: Brookings Institution.

Barry Ames is Andrew Mellon Professor of Comparative Politics at the University of Pittsburgh. His recent publications include "Electoral Strategy Under Open-List Proportional Representation," American Journal of Political Science, 1995; "Electoral Rules, Constituency Pressures, and Pork Barrel: Bases of Voting in the Brazilian Congress," Journal of Politics, 1995; and "Approaches to the Study of Institutions in Latin American Politics," Latin American Research Review, forthcoming, 1998.

Margaret E. Keck is associate professor of political science at Johns Hopkins University. She is the author of THE WORKERS' PARTY AND DEMOCRATIZATION IN BRAZIL (Yale University Press, 1992) and co-author, with Kathryn Sikkink, of ACTIVISTS BEYOND BORDERS: ADVOCACY NETWORKS IN INTERNATIONAL POLITICS (Cornell University Press, forthcoming, 1998). Her current project is a book on environmental politics in Brazil.

For research contributing to this article, Margaret E. Keck received financial assistance from the Howard Heinz Endowment (Center for Latin American Studies, University of Pittsburgh, Research Grant on Current Latin American Issues), the Joint Committee on Latin American Studies of the Social Science Research Council, the American Council of Learned Societies (with funds provided by the Ford Foundation), and the John D. and Catherine T. MacArthur Foundation. Barry Ames received financial assistance from Washington University, St. Louis; and the University of Pittsburgh. This research began as a consulting project for the World Bank, which naturally bears no responsibility for any statements in this article.

Copyright Journal of Interamerican Studies Winter 1997-1998
Provided by ProQuest Information and Learning Company. All rights Reserved