Practicing Radical Democracy
Lessons from Brazil

Introduction

Since the late seventies, a growing critique of the "statism" prevalent in both the socialist and social-democratic Left has led many progressive scholars and activists to declare "civil society" the privileged sphere of emancipatory social change.[1] Experience suggests that the redistributive state envisioned by the Left tends to be rigid, technocratic and ultimately, authoritarian. "New Left" democratic theory has therefore increasingly called for combining the pursuit of economic and social equality with increased citizen participation in public life and the celebration of social diversity. But more often then not, the literature has envisioned civil society rather than the state as the sphere where such an inclusive, egalitarian, participatory "radical democracy" is most likely to emerge. Models for a radically democratic state usually describe distant utopias in which widespread citizen participation is possible because social inequality and political alienation have been already eliminated.[2] But such models rarely make clear what would be necessary to "get there from here."

This paper seeks to show how actual attempts to "radically democratize" governance can shed light on what might kind of transformative practice is possible by state actors in the "here-and-now." When they do gain some control of the state apparatus, radical actors are forced to confront the contradictions between utopian aspirations and present day conditions. Obviously, successful attempts to overcome these dilemmas are few-and-far between. But those few examples can tell us many things about practicing radical democracy that abstract theorizing cannot.

Practicing Radical Democracy in Brazil

Perhaps surprisingly, some of the most innovative thinking on a new progressive model of state action is coming from a country typically considered to be on the political, economic, and theoretical periphery. One of the only growing "democratic socialist" parties in the world is Brazil's Worker's Party (PT) — a heterodox political coalition that has been expanding on Brazil's political scene since the early 1980s. Many sectors of the PT, in some regions the majority of the party bureaucracy, continue to preach "big state" doctrine, avidly opposing the "flexibilization" of public employment and the privatization of state-owned businesses. But where the PT has reached elected office in Brazil it has tended to experiment with creative alternatives for governance that challenge such bureaucratic traditions and promote a participatory, redistributive, and inclusive "radical democracy".

When, in 1980, the Brazilian military regime (1964-1985) first allowed the creation of new parties, a coalition of radical labor unions, urban and rural social movements, and formerly revolutionary Marxist political groups that now rejected violent action decided to found a new type of socialist party. From the beginning, the PT repudiated ties to the Soviet Union and sought a democratic road to socialism that would preserve diversity, civil liberties, and tolerance. A coalition of a variety of social groups, it also questioned traditional socialist doctrine, seeking to include a variety of actors other than the industrial working class into its project for change. During the first decade of its existence, the PT created a name for itself as an important locus of political opposition in Brazil, articulating the concerns of a wide spectrum of social movements and political groups that demanded social justice for the poor and criticized the corruption and clientelism that was the status quo in Brazilian politics.

In 1988, the party made its first major electoral inroad, winning 36 municipal governments, including three state capitals, Porto Alegre, Victoria, and most importantly, the huge megalopolis, São Paulo. In the following year, the party's national leader, Luis Inácio "Lula" da Silva, lost the country's first direct presidential elections in nearly thirty years by only a small margin, generating massive popular support for the PT, particularly in the more developed southern and southeastern regions of the country. Although it lost the presidency, the prefeituras (municipal governments) that the party now controlled became important proving grounds in which the PT could test out its ideals. Many would agree that the most successful of these "Popular Democratic" governments has been in Porto Alegre, a regional capital of 1.3 million inhabitants in the far south of the country where the PT is currently in its third consecutive term in office. My reflections on the PT's contribution to a politically viable "radical democratic" redefinition of the state come largely out of the Porto Alegre experience where I conducted research over the course of several years including extensive participant observation and over 100 interviews with government officials, neighborhood activists, and non-governmental organization staff.[3]

PT ideology is rather contradictory, which is not surprising given the multiplicity of groups making up the party. Over the first decade after its foundation, a consensus emerged around two "mottos" for governing: "participação popular" or grassroots participation and "invertendo prioridades", literally, "inverting priorities" of government policies away from the privileged classes that have traditionally benefited from them and towards the poor. But despite this consensus, a constant struggle has dominated party debates about what specific groups should be included as the participants and beneficiaries. More orthodox sectors of the party see organized workers as the principal targets of PT policy-making and generally reject the idea of making alliances with moderate political parties and middle class or business groups. More heterodox and moderate groups are more inclusive, focussing on not only organized working- class sectors, but also on the un-organized, informally employed poor and on the middle class. They are also much more prone to accept the political need to make alliances in order to survive electorally. In recent years, the former group has tended to capture the majority of posts within the party bureaucracy. But the more heterodox groups have been more successful in electoral terms and have thus dominated the way PT governments have evolved in practice. In part because they thus grow out of real experiences with governing, that is, out of an effort to reach a broad electorate and to survive in office, I believe that the experiences of these governments can provide much insight on how the Left can practically promote something that at least moves towards an ideal of radical democracy.

Since it first came to office in 1989, the Porto Alegre prefeitura has been one of the PT administrations most successful at implementing policies adhering to the two party mottos. In the first place, the government dramatically opened up government decision-making to citizen participation. Most big cities in Brazil have in recent years created citizen councils to monitor or approve major planning efforts. But few of these councils have real decision-making power and they are rarely formed out of an open, widely participatory mobilizing process. In many cases, "community" members are nominated by government authorities or, at best, specific civic groups — such as certain non-government organizations, unions, universities, and businesses — are bequeathed the right to send their representatives.

The Porto Alegre administration has gone much further. It has developed an entirely "bottom-up" system of participatory governing in which openly elected forums have a great deal of deliberative power over an expanding number of policy arenas. The central participatory policy is the "participatory budget": a system of decision-making that gives power over public resource allocation to forums elected at neighborhood-level assemblies and at open "thematic" meetings. This system has grown in scope and power over time, at first largely addressing decisions about community level capital expenditures and eventually gaining decision-making power over major capital investments, service and maintenance programs and personnel issues. At the top of the participatory pyramid is a Municipal Budget Council, elected in open assemblies, which has final say over the entire budget proposal that the mayor sends to the city council each year for approval. Alongside the budget forums, most city departments have also formed special councils in areas such as transportation policy, housing, culture and health care. These councils are all open to the participation of ordinary citizens, requiring neither special technical qualifications or official nomination. Over the nine years that the PT has held office in Porto Alegre, citizen participation has been in constant expansion, as more and more city agencies have created citizen forums and more and more local residents have joined them. Each year, over 15,000 people attend Participatory Budget assemblies alone, with about 1,000 of them taking part in year round forums that meet regularly.

In the second place, the Porto Alegre administration has presided over a dramatic redirectioning of government priorities, focussing a large part of policy-making towards the needs of the poor. This process began with a massive effort to raise revenues, which occurred in Porto Alegre by raising rates and reassessing values for the two taxes under municipal control: the tax on revenues in the service sector and the urban property tax. Over the first four years of the administration, city revenues almost doubled. A government that began with nearly 100% of its revenues earmarked to cover payroll now had 20% of revenues for "discretionary" spending.[4] With these funds, the government could give special attention to capital investments in poor neighborhood that historically in Brazil have been abandoned by the state.

Like other Brazilian cities, Porto Alegre has a large urban periphery and many hillside slums that prior to the PT's arrival were desperately lacking in basic infrastructure and services such as public transportation, running water and sanitation, paved streets, garbage collection and so on. In the last nine years, however, water and sewer services have been extended to all but a tiny percentage of neighborhoods. Bus services have reached neighborhoods that were previously unattended. Extensive efforts have been made to legalize squatter areas and to transform disorganized settlements into neighborhoods. The result has been to dramatically transform the quality of life of some areas of the city. With paved streets, bus lines, trash collection, flood control and sanitation, places previously perceived as "dangerous" and "inaccessible", where residents had to trudge long distances through the mud to reach public transportation, became integrated parts of the city. The administration has dramatically improved in municipal schools — widely perceived as much better than those in the parallel state-government school system. It has funded community-run day care centers and adult literacy programs and has installed health clinics in distant neighborhoods. Virtually all decisions about which neighborhoods and regions of the city would receive what types of benefits has occurred through the highly representative participatory budget process, or through other sectoral councils, with citizen groups closely monitoring the details of project design and implementation.

In a country where a hierarchical political machine has largely benefited elite groups and where poor people's access to government officials is typically mediated by a highly personalistic and arbitrary clientelist system, the participatory, redistributive government of the Porto Alegre administration is indeed impressive. Yet, I would argue, the ideas of participação popular and of inversão de prioridades are not enough to present a useful model of radical democracy: it is in the process of implementing these ideas that the Porto Alegre administration provides us with some clues for what a radically democratizing project would be in today's world.

One of the central problems with socialist ideology is that it often assumed that redistribution could not be arrived at democratically since it would require a violent crushing of the elite groups that would most certainly have more resources to dominate any electoral process. Many theorist of participatory democracy have suggested that genuine participation is only possible once social inequalities have been eliminated. But if violent, revolutionary or authoritarianism methods of achieving social equality are rejected, then "how to get there from here" becomes less clear. If radicalism is also democratic, if redistribution must be combined with open participation, then the road to radical democracy must pass through existing power inequalities, dealing with them democratically. This requires jumping some difficult hurdles. If those to whom one wishes to distribute are not well organized, how to prevent an open participatory process from being controlled by those who are already more powerful? If redistributing government resources means taking benefits away from those with access to money and media, how to "invert government priorities" and still win elections? The remainder of this paper will look at how the Porto Alegre administration responded to such problems, in the attempt to develop some criteria for how to overcome these dilemmas.

Getting There from Here

1) Mobilizing the Poor

Participation is unlikely to bring about redistribution if the poor are not mobilized. This fact has been widely noted in the empirical literature: in case after case, middle class and elite groups dominate participatory forums because it is they who are best organized. Even where the poor do participate, it is often a select group of organized associations who, through the very process of participation tend to gain even more capacity to dominate citizen forums while those groups that are not organized initially find it increasingly difficult to make it in the door. In Porto Alegre, however, the central participatory forum, the Participatory Budget, drew principally poor people. In fact, the income levels of participants were significantly lower than the population as a whole.[5] As a result, the policy disproportionately benefited poor neighborhoods rather than wealthy ones because they were the ones that sent the largest numbers to the assemblies and that elected the largest numbers of delegates to the district forums and to the municipal budget council. What is more, within a few years after the policy was initiated, most of those neighborhoods that participated were ones that prior to 1989 had little or no experience with civic mobilizing. All this contrasts most experiences with participation, where the wealthy and most organized dominated citizen councils.

This mobilization of the poor and of the previously unorganized occurred for a variety of reasons. In the first place, since participation initially focussed on basic community infrastructure, the process began by directly responding to the demands of poor neighborhood organizations. Demands for neighborhood infrastructure had long been the subject of demonstrations and protest and several regions of the city had developed strong popular organizations to defend these demands. The issues involved in the Participatory Budget were, therefore, immediately meaningful and understandable to poor neighborhood residents. At the same time, they were not particularly interesting to middle class neighborhoods, which for the most part already enjoyed access to such basic infrastructure, or to business groups that would seek government investments in other areas. This initial focus on an arena immediately attractive to the poor contrasts many other experiments with participation that opened up public discussion on subjects about which poor people have little understanding or that put the poor in direct competition with better off groups who had more resources to articulate their demands. In Porto Alegre, participation only expanded to such arenas after a strong mobilization had occurred around basic neighborhood infrastructure. For example, four years after the budget policy was initiated, a series of "thematic" forums were incorporated into the budget process to discuss issues such as transportation policy and economic development. These forums attracted substantially more middle class residents and business groups, but also drew poor residents, who were able to be effective participants in part because of the skills they had acquired through experience with the neighborhood-based process.

While on the one hand, the policy attracted the poor, it was also able to work against the tendency of citizen forums to be dominated by those groups that happen to be organized at the time participation is initiated. At first, the main participants in the budget forums were indeed associations representing those poor regions of the city that had organized over the course of the decade before the PT came to office. But the Participatory Budget system also worked. As word spread that the Porto Alegre administration was investing in projects prioritized by neighborhood assemblies, innumerable new neighborhood groups formed and joined the process. Once again the thematic focus of the policy spurred on this mobilization. For the most part, the type of investments made — such as paved streets and new schools — were highly visible. Many participants I interviewed declared to have joined the budget assemblies only after they saw such projects being implemented in adjacent areas and went to local authorities or community groups to inquire about how they could receive such benefits for their own neighborhoods. Upon hearing that only those groups that participated in the budget assemblies had a chance at receiving such projects, they began to organize their neighborhoods. Many skeptical residents of poor neighborhoods — whose ingrained distrust of politician's promises might have kept them from "wasting" precious time on participation — came to believe that the policy had credibility and that civic action might be worthwhile because of this "demonstration effect."

At the same time, the government made specific efforts to draw new neighborhoods into the process when they did not find out about the policy on their own. The prefeitura employed a league of civic organizers — usually experienced neighborhood activists themselves — to mobilize residents and to provide information on how to participate. Often, these organizers would make visits to neighborhoods that had not sent representatives to the budget assemblies and would help incipient groups to rally support and bring together residents. One could say that these state-employed organizers took the place of the "external agents" (priests, students, non-government organizations) that had historically played a critical role in helping poor people's movements in Brazil and elsewhere gain strength: they provided information and skills that helped people overcome the barriers to participation.

Many authors have argued that participation is "developmental": through the very experience of participation, people learn the skills that they need to participate effectively.[6] In Porto Alegre, the government capitalized on this potential for "social learning" to insure that participation would have a distributional effect by "starting small," by focussing on policy arenas that were particularly attractive to the poor, and by investing both financial and human resources in ways that drew the poor and unorganized into the process.

2) Transforming the Bureaucracy

A second dilemma of implementing radical democracy for which the Porto Alegre case provides some insight has to do with the bureaucracy. When a new party comes to office, only the high and some of the middle echelon posts are taken over by party nominees. The remainder are typically permanent administrative posts, made up of people who cannot be easily laid off. Participatory policy-making challenged technocratic traditions that are particularly strong in Brazil. Experts responsible for designing and implementing infrastructure projects and other services were now asked to take commands from ordinary citizens. What is worse, they needed to work with people, most of whom had little formal education, explaining the technical details of each project demanded, negotiating time limits and technical criteria. Often this meant meeting with community groups in distant neighborhoods and outside of normal working hours and listening to often scathing critiques and challenges. Transforming a rigid, hierarchical bureaucracy into a flexible, responsive organization capable of attending the often idiosyncratic demands of citizens would not be easy.

In Porto Alegre, there were several reasons that the administration was able to adapt the city bureaucracy to new ways. In the first place, the administration always paid comparatively high wages, even when this came at tremendous cost. In the first year of the government, for example, the government was faced with a payroll that accounted for nearly 100% of revenues, a result of the fact that the previous mayor had dramatically raised wages in his last days in office,. Rather than reducing wages and risking the wrath of the public employees union, the government maintained the rates inherited from the previous administration, even though that this meant that there would be no funds available for investments or service expansion until revenue levels could be raised, a process that took over a year. The result was low approval rates in public opinion surveys, but a generally satisfied work force and a labor union that backed the administration.

This fairly peaceful relationship did not mean, however, that the city's civil servants went along easily with the changes in decision-making that the participatory budget policy involved. Certain agencies, most importantly the Secretariat of Planning, were characterized by an entrenched technocratic bureaucracy that largely disapproved of the idea of giving decision-making power to ordinary people. The agency heads I interviewed generally noted that although outright sabotage of the decisions made by budget participants was very rare, at times employees resisted designing or implementing projects that challenged technical norms. Despite these problems, a substantial sector of the bureaucracy generally supported the PT administration. The vision expressed by a number of employees I interviewed was one of appreciation of the administration's invigorated governing capacity. Since civil servants were the ones who would have to live with the government's mistakes long after the elected government hierarchy would leave office, they typically preferred a competent government that invested, without corruption, in properly-designed and well-maintained projects.

Even with this relatively widespread support within the bureaucracy, the workload of organizing participatory policy was largely born not by the civil servant bureaucracy, but by a fleet of appointed employees. According to the agency heads I interviewed, most public employees did not object to receiving participants and responding to their questions and problems during the ordinary hours of the workday and in their offices. But, with a few exceptions, they largely refused to work nights and weekends, attending assemblies and meetings all over town. This outreach work, which was essential for making the participatory process function, was largely carried out by the about 300 appointed employees named by the Mayor and agency heads. The community relations personnel in most departments were either appointed or were PT militants who happened to have passed the civil service exam. It was they who attended community meetings, articulating the complex exchange of information between city agencies and participants. Agency heads and their closest advisors also were extremely active at nights and weekends, attending meetings to answer questions at the budget forums. Much of the success of the participatory budget policy was a result of the administration's investment in this constant "pounding the pavement" as administration personnel dedicated enormous amounts of time to communicating directly with participants about the details of government actions. The burden of this "on-the-street" activity was born, however, by a small minority of the municipal employees, for the most part PT militants in appointed positions.

This use of "ideologically mobilized" personnel has to a large extent been effective at insuring government responsiveness to participants. But certainly this solution to the problem of transforming bureaucracy has its limitations. Over time, those militants who joined the administration with a strong sense of urgency and purpose are likely to wear out, either to seek less demanding employment or to insist that their jobs conform to something closer to the forty hour work week. Furthermore, there are practical limitations to the amount of outreach work that a few hundred employees can do. As participation in Porto Alegre has gone from the "small phase" of discussing primarily neighborhood level investment to the "big phase" of debating long-term policy goals and city-wide issues, the need for popular education and close regular, contact between technical personnel and participants has only grown.

To be able to grow with the policy, the administration will have to institutionalize bureaucratic flexibility to a greater extent than it has yet done. Tendler (1996) has suggested some of the ways that government bureaucracies may make such changes, arguing that state agencies need to follow the lead of certain cutting-edge industries. In a study on the state of Ceará in Brazil, she has shown how increasing worker autonomy, customizing each job to specific situations, involving workers in multiple and varying tasks, and promoting a culture of respect for public employees can help bureaucracies become more responsive and flexible, and thus more capable of working with citizen participation. Porto Alegre would do well to experiment with some of these ideas.

3) Building Political Support

The new priorities of the PT administration in Porto Alegre obviously had a great deal of impact on groups outside as well as inside the government bureaucracy. To implement the kinds of grassroots, participatory, redistributive policies the PT called for would require raising revenues and using them to benefit groups that had previously seen only a tiny minority of public funds: poor neighborhoods requiring basic infrastructure. This meant that other groups lost priority for benefits or paid higher taxes and fees to the prefeitura. Property owners were faced with significant tax increases and the elimination of traditional amnesties and exceptions that had historically been handed out as patronage by political leaders seeking to secure elite political support.[7] The city assembly, accustomed to adding in innumerable amendments to pay back supporters, would now both be asked not only to vote for controversial tax increases and but also to approve in its entirety a budget that has been designed by highly representative citizen forums. Finally, middle class neighborhoods and central commercial areas that historically received the bulk of city services and projects would lose priority. Yet property owners, the middle and business class, the city council were all extremely influential in city politics. They had control over the media and provided critical financing to political candidates. How was a city government to take benefits away from them and still expect to get critical legislation through city council, no less to win the next election?

How the Porto Alegre administration acquired the political backing to implement policies that benefited the poor is a very complex story that cannot be easily summarized in a few paragraphs. In essence, I would argue, the administration was able to build an alternative coalition of political support that included not only poor and working class sectors, but also what could be referred to as the "progressive" middle class. In addition, while its policies damaged some sectors of the economic elite, other business sectors benefited. This mixed group of supporters was sufficiently influential to be able to pressure the city council to pass critical tax and budget legislation. It also meant that media, while not exactly friendly to the administration, did not engage in the kind of blitzkrieg seen in other places where Left parties have come to office.

Three important components of this alternative coalition should be emphasized. In the first place, as already noted, the budget policy had a dramatic mobilizing effect. While this effect helped insure that the policy reached large numbers, it also indirectly fostered a mobilized support network on the city's periphery. This helped the administration politically in two ways. On the one hand, those who received benefits for their neighborhoods through the budget process (as well as those who did not receive benefits but had come to believe through participation that the administration's mode of distributing resources was legitimate and fair) would mobilize public opinion in favor of the administration at election time. On the other hand, participants also pressured the city council to pass legislation critical for the implementation of their budget proposals. Many council members had come to office with the help of community leaders on the periphery who, according to clientelist traditions, had long mobilized support for particular politicians in exchange for promises that their neighborhoods would benefit should their candidates be elected.[8] All too often, these same leaders now participated in the budget assemblies, which proved to be much more reliable forms of negotiating with those in power. Even if they did not change their party sympathies (which many did), they now exerted pressure on the council members they helped elect to insure that the budget they had designed was approved.

In the second place, the policy was able to generate the support of certain sectors of the business class, most notably construction agencies. This group, often seen as extremely conservative and complicit in corruption schemes in Brazil, initially boycotted the administration. But, as it became clear that the government was prepared to spend significant sums on civil construction, the companies began to break the cartel to take advantage of the new contracts that the administration was offering. Ultimately, these companies provided essential, albeit veiled, elite support for tax increases that would increase revenues which would be used to build infrastructure. They also counterbalanced the influence of the city's speculative land owning elite over the media, which avoided coverage of the administration, but for the most part did not engage in active campaigns against it.

In the third place, the PT in Porto Alegre enjoyed wide-spread support from the middle class. Indeed, opinion surveys showed that the administration's approval rate increased according to income. My research suggested that there were two major reasons for this approval. On the one hand, the PT prefeitura insured that a substantial portion of increased revenues and administrative energy went towards maintaining high quality services in the center city: public transportation, trash collection, street repair and the like dramatically improved after the PT came to office. Opinion survey's showed that middle class voters saw the administration as a "competent" and "efficient," administration that effectively provided essential services. On the other hand, the participatory efforts of the administration, although largely attracting lower income residents, had much support among the middle class who identified the PT with "democracy", "transparency" and "social justice". At a period in history when corruption scandals were rampant in the media, leading in 1992 to the impeachment of the nation's president, these characteristics were seen as extremely positive by a growing progressive middle class that was tired of government mismanagement and chicanery.

All this suggests that it is possible for radical governments to gain sufficient political support when their policies are designed, implemented and advertised in ways that are broadly inclusive. I am not suggesting that it is necessary to "pay off" the rich and the middle class, making concessions to them even if they detract from radical objectives. To the contrary, the Porto Alegre administration creatively implemented its ideals in ways that incorporated the interests of a "critical mass" of poor, middle class and business groups. By being responsive to participant demands, the government generated support among participants. By engaging certain business sectors in the implementation of participatory projects, it built support among some elite groups. And by building a reputation as a competent, uncorrupt, transparent and socially responsible administration, the PT found support among an enlightened middle class that was particularly frustrated with traditional forms of governing at that historical moment. In this way, instead of being a political burden that brought on opposition, participatory, redistributive policies were actually an asset that helped generate political support.

Conclusions

Many theorists have argued for an inclusive, participatory, egalitarian state, but have said little about "how to get there from here". The in-depth examination of one "real-life" attempt to "radically democratize" the state can help us elaborate more realistic models of transformative action. This paper has suggested that under certain conditions, policy-making can promote radical democracy even if existing social and political conditions are far from ideal. But this can only occur if radical policy itself is used not only to implement an "ideal model" but to help transform conditions so that the ideal model becomes more feasible. Here I have suggested that making the state both participatory and redistributive at the same time requires that radical policy have at least three transformative objectives:

1. Participatory policy should not only seek to give civic organizations decision-making power, but also to mobilize the unorganized and the poor, so that they to will benefit from the policies. Doing so is not something that has to be done prior to implementing participatory policy, but can occur through "social learning" process of participation itself.

2. In order for participatory policy to be effective, the bureaucracy must become more flexible and responsive, capable of custom tailoring projects to participant demands and of disseminating information and skills to ordinary citizens.

3. Attempts to redistribute decision-making power and state resources to traditionally excluded groups must have widespread political support if they are to be long lasting. Progressive political groups must discover how to creatively use policy-making to attend the interests of a wide spectrum of groups without "selling out" their radical programs.

These reflections on one successful experience thus suggest that a successful radical project should not only seek to promote participation and redistribution (or participação popular and inversão de prioridades in the PT lexicon) but also must find ways to gain the support of the unorganized, of the middle class, of government employees and of potentially-progressive business sectors. In many political situations, gaining the backing of such a broad spectrum of actors may be impossible to achieve without renouncing essential principals. Learning the art of maintaining this fragile balance should be the primary goal of those who seek to put radical democracy in practice.

Bibliography

[1] See for example Cohen and Arato (1992: 417) and Laclau and Mouffe (1985)

[2] See for example, Barber (1984), Cohen and Rogers (1983), Dahl (1985), Green (1985), and Held (1987 (last chapter)).

[3] For more on PT history and ideology, see Gadotti & Pereira (1989), Keck (1992), and Meneguello (1989). On the experiences of other PT administrations, see Abers (1996), Alvarez (1993), and Nylen (1996, 1997).

[4] See Cassell and Verle (1994) for a detailed description of revenue raising in the Porto Alegre administration.

[5] As confirmed by the comparing the results of a sample survey I conducted in 1995 (in collaboration with two NGOs and the Porto Alegre Administration) to 1991 census data for the municipality.

[6] See especially Pateman's (1970) classic work.

[7] The tax on services was also an important source of revenue increases in Porto Alegre, but were much less controversial since they were simultaneously raised in all state capitals (as part of an agreement of capital mayors) and since they did not affect traditional local elites. The largest service tax payers were major companies such as the multinational, IBM.

[8] See Gay (1990), Banck (1986) and Diniz (1982) for descriptions of the critical role neighborhood-based clientelism plays in Brazilian politics.