Since 1989, the local government of Porto Alegre, a city of 1.3 million people in southern Brazil, has implemented what it calls the "participatory budget." One of the central goals of this policy is to hand over decisions about the distribution of municipal funds for basic capital improvements - paved streets, drainage and sewer investments, school construction, and so on - to neighborhood-based forums. The policy has fostered a dramatic increase in neighborhood activism in the poorest neighborhoods of the city, with over 14,000 people participating each year in budget assemblies. Innumerable new neighborhood organizations have appeared in response to the policy, often in areas that were previously dominated by closed, ineffective associations that served as little more than tools of clientelist party politics. The Porto Alegre policy has combined a substantial amount of government investment in social programs with a successful state-sponsored effort at capacitating civic groups to control that investment and, in doing so, to dramatically improve their quality of life.

This article will examine how this process of civic empowerment occurred[1]. The next section will consider the role that state actors can play in helping those with little previous experience to begin to organize collectively. I then go on to briefly examine the history of neighborhood associationalism in Porto Alegre and to describe the budget policy. The main body of the article looks at one district of the city that, prior to 1989, had virtually no experience with broad-based, participatory civic organizing, showing how the budget policy mobilized neighborhood groups, discouraged clientelist forms of neighborhood action, and promoted the emergence of participatory groups that not only struggled collectively to bring benefits to their neighborhoods but also learned to work in collaboration with other neighborhood groups in the pursuit of broader goals.

Clientelism, Associationalism, and the State
It has often been suggested that strong democratic institutions will only emerge once strong civic groups have formed that can, essentially, force the state to represent them. That is, the organization of civil society is a prior condition for democratization. In an influential 1993 study, for example, Putnam argues that the primary explanation for the rise of democratic institutions in northern Italy is that region's long history of civic organizing, a legacy that goes back nearly a thousand years. The result is a rather deterministic view of democratization: only those regions that have an ingrained culture of associationalism are likely to develop a responsive, accountable, participatory state. Others have suggested, however, that a culture of civic organization can be created in places where communities do not benefit from such legacies. In their recent work on "associational democracy," for example, Cohen and Rogers argue that associative environments are "artifactual." Networks of civic organizations can be created and transformed as political circumstances, institutional environments, and balances of power change. With this possibility as a starting point, this article will explore how, under certain circumstances, state actors can be the transformative force that helps promote the growth of associationalism.

Lacking the long civic history of northern Italy, Brazil much more closely resembles the Italian south, where Putnam argues that clientelist traditions have almost inexorably suppressed the emergence of organized civil society. In political systems dominated by clientelism, those in power use their access to state resources to provide personal favors to a broad-based clientele who, in turn, mobilize votes for their patrons. In Brazil, clientelist traditions have led to the development of vast political machines that link local bosses to state-level and federal politicians through such troca defavores, or favor exchange. In the cities, neighborhood associations play a critical role in these linkages. Cabos eleitorais, or ward bosses, head up a large percentage if not most neighborhood associations and use them to mobilize votes for their party candidates by conveying promises of favors to local residents.

Yet a number of recent authors have shown that certain types of state institutions, usually implanted "from above" by reformers, can build up civic activism among groups that have little previous experience with community-based cooperation. In a recent issue of World Development, Peter Evans brings together a series of studies that demonstrate how such state-fostered civic empowerment can occur. Ostrom describes a participatory sewer construction policy in Recife, Brazil, where participation was directly facilitated by government officials. Lam argues that the success of a decentralized irrigation policy in Taiwan grew out of the close and flexible relations between government workers and farmer participants. Fox argues that even in the context of authoritarianism, reformers within the Mexican government were able to promote policies that gave power to and strengthened peasant organizations. In related work, Tendler writes of how government practice in the Brazilian state of Ceara rapidly moved "from bad to good" with the implementation of policies that included efforts by the state government to assist directly in the organization of civil associations in the municipalities. In all of these cases, both the close ties that developed between government officials and citizens and the fact that the state policies explicitly encouraged civic organizing led not only to strengthening civic groups but also to a growing political capacity of those organizations to pressure for government accountability and effective policy. The importance of what Evans calls "state-society synergy" suggests that the relationship between civil society and strong democratic institutions is far from a simple, one-way, bottom-up process.

To understand how state actors can promote associationalism, it is useful to look at how the literature on social movements has explored what causes people to mobilize. Many authors have noted that purely economic explanations are not satisfactory: associationalism does not seem to be related to certain levels of misery or well-being. Instead, political factors seem to best explain why civic organizing booms at certain times and places and not others. General changes in political conditions provide "windows of opportunity" in which enabling environments emerge that make people believe that collective action is likely to be fruitful. "Social movements form when ordinary citizens, sometimes encouraged by leaders, respond to changes in opportunities that lower the costs of collective action, reveal potential allies and show where elites and authorities are vulnerable."

At the same time, studies have shown that poor people's movements often only mobilize with the direct help of outsiders. External agents can help people acquire financial resources and political skills for taking advantage of windows of opportunity. Those who have more resources--disaffected elites who, sympathize with a cause or organized groups seeking to build up potential allies--can provide the funding and the technical skills that make it possible for those with few resources or previous experience to mobilize. In Latin America, the Catholic church, student activists, left, wing militants, non-governmental organizations, and liberal professionals have historically played a critical role in helping popular organizations get off the ground.

For the most part, the social movement literature envisions the state's impact on civic organizing as an indirect one. The state is seen as highly unlikely actually to encourage autonomous civic groups to form, and any attempt on the part of the state to act as an "external agent," directly intervening in local communities to help associations gain organizing capacity, is often assumed to be no more than an attempt to co-opt them. Porto Alegre provides us with an unusual example in which state reformers took on a more proactive role, not only providing an enabling environment, in which the formation of civic groups was explicitly promoted, but also working directly and closely with local communities to help them organize[2]. Examining how this occurred can help us rethink the potential relationship between state and civil society.

Neighborhood Organisations and the Participatory Buget Policy in Porto Alegre

In 1964, the military took power in Brazil and the PTB was outlawed. FRACAB was taken over by ARENA, the military-aligned party. Those SABs that continued to be active did so by supporting the regime and by generally avoiding making demands. As the regime began to weaken after 1974, however, renewed possibilities for clientelist politics began to emerge as ARENA sought to generate backing for the regime by distributing patronage to supporters. But in the late 1970s, dissatisfaction with the military spread throughout Brazil, and protest movements of all kinds began to gain strength. During the military period, a massive migration from the countryside had occurred, and Brazil's cities were swelling with impoverished settlements totally lacking in basic infrastructure. For the vast majority of neighborhoods, fragile promises and petty exchanges with local bosses were nowhere near capable of satisfying the need for public transportation, paved streets, running water, sewers, basic health care, and other services. At the same time, the weakening regime was becoming more responsive to popular demands. Within this window of opportunity, a new kind of neighborhood organization emerged in Brazilian cities that refused to play according to clientelist rules.

Nevertheless, such "combative" neighborhood groups never represented more than about one-third of the city's neighborhood associations. Although by the end of the 1980s, several other districts of the city had formed similar neighborhood coalitions, most city neighborhoods were still dominated by clientelist associations that had also been expanding, especially under the aegis of the Partido Democratico Trabalhista (PDt, the successor to the pre-coup PTB. The PDT had developed a powerful network of neighborhood leaders that helped elect it to the mayor's office in 1985. Vast impoverished districts of the periphery continued largely unmobilized, with their only associations controlled by ward bosses working for political parties, who discouraged more participatory forms of organizing.

Elsewhere, I have explored, at length how the participatory budget policy evolved over time and why it became a central policy of the PT administration while other participatory programs fell by the wayside. Briefly, as the PT re-organized the administration, generated revenues by passing new tax laws and streamlining spending, and worked to build broader political support, it found that the budget policy was more viable than the other participatory efforts. On one hand, it clearly corresponded to the two PT party mottoes by promoting both popular participation and investment of public resources in poor neighborhoods. In doing so, it attended the demands that the "combative" neighborhood coalitions had long been making: the systematic allocation of basic infrastructure in poor neighborhoods and the control over infrastructure decisions by neighborhood residents themselves.

Perhaps most important, the budget policy proved to have a tremendous capacity to mobilize. As people realized that, through mobilization, they could bring tangible benefits to their areas, participation in neighborhood associations boomed. Although the majority of participants in the budget policy were not PT militants, as the government showed its commitment toward attending their demands, they naturally grew to sympathize with the party. This legion of neighborhood activists was essential for maintaining the budget policy: on numerous occasions, budget participants flooded the city assembly chambers to demand that the deputies approve the budget they had designed and vote for tax increases that financed investments. They also helped mobilize support for the PT at election time, although not in the traditional clientelist fashion. Rather than providing political support in exchange for favors, the budget participants promoted the re-election of a government that had effectively intended their needs in an open, transparent way. In part because of the support generated around the budget policy, the PT has been re-elected two times and continues with high levels of approval to this day.


From Clientelism to Cooperation: The Extremo Sul District

The remainder of this article will examine how the budget policy succeeded in mobilizing neighborhoods that had been previously dominated by clientelism and how this led to a more general strengthening of civil society in Porto Alegre. The discussion will focus on one district of the city where I conducted extensive interviews and participant observation over the course of a two-year period: the Extremo Sul district. In earlier times, the "Extreme South" of Porto Alegre, a beautiful area of green hills bordering the Guaiba River, was known best for the riverside settlement of Belem Novo, the city's most accessible waterside resort until the mid-twentieth century. When paved highways made travel easy to the ocean beaches of Rio Grande do Sul State, the area lost its importance as a vacation spot. Eventually bathing became altogether impossible there, as Porto Alegre's rapid growth during the 1960s and 1970s led to the contamination of the river as far as the next municipality, The district was largely ignored by city officials for many years as development efforts were directed toward the northern "industrial zone." But by the early eighties, the largely rural southern periphery had become a district of expansion, not of economic investment, but of people seeking affordable housing. In the Extremo Sul, lower middle-class people who could not afford to purchase land elsewhere could buy a nice-sized lot, albeit in a subdivision totally lacking in basic infrastructure. Along the district's dirt roads, dozens of impoverished settlements and tiny roadside land invasions also appeared, where people collected who were too poor to make it even in the in town squatter settlements. By the 1990s, about 24,000 people lived in the district. Figure 1 is a map of the district identifying some of the larger settlements.
  
As in many other districts of the city, wide-scale community organizing in the Extremo Sul district was virtually nonexistent prior to the PT administration. Official neighborhood associations were largely dominated by a few of the better-off residents with clientelist ties to political parties, especially the PDT Most of the neighborhoods and settlements in the district either had no association at all, or associations were closed to the participation of residents in general. Except for older parts of Belem Novo, none of the district's settlements had basic sanitation, storm drainage, or pavement infrastructure, and transportation to the city on the unpaved roads was extremely slow and expensive. Occasional candidates would travel through the district promising to bring the much-needed infrastructure if elected, but the promises were rarely fulfilled. In the years after the PT came to office in Porto Alegre, however, a dramatic transformation of this picture occurred, as a result of the budget policy.

Mobilizing Neighborhood Residents
In an effort to bring better service provision to the south of the city, in its first year in office, the PT administration created a Centro Administrativo Regional (District Administrative Center; CAR) in Restinga, a more densely populated district adjacent to the Extremo Sul. The CAR staff was faced with the difficult task of making contact with neighborhood leaders in the Extremo Sul and of promoting the participatory budget policy. The task was daunting in such a large, sparsely populated district, and the administration's initial financial difficulties made things more complicated. Without even an automobile in the first year to travel through the district, initially the only way to call people to the budget assemblies was a mailing list of formally established neighborhood associations. The vast majority of associations on that list were of the traditional, non-participatory clientelist type. Association leaders were invited to the first participatory budget assembly in Belem Novo in 1990. Elections for the Extremo Sul representative in the Municipal Budget Council held at the very poorly attended meeting gave the job to two wealthy men, one who owned most of the land around Lami and dominated the neighborhood association there and another who was an aspiring politician who a few years later would be elected to the City Assembly by the PDT. Making little effort to meet with the Extremo Sul residents, these councillors formulated a very long and unordered list of demands, most of which would be left unattended in a year that the PT administration still had few funds for capital improvements.

The following year, the annual districtwide assembly drew eighty people, still largely from the Belem Novo area. The new budget councillors elected were a PT militant active in the progressive Catholic church and a small-time politician who worked for a conservative state congressman. To collect information on the district's needs, they traveled through the district talking to residents. Still only one general assembly to discuss district priorities took place. The first major investment of the new administration in the Extremo Sul district--the pavement of Chapeu do Sol Road--resulted out of this process. "If there had been any real participation," one neighborhood leader I interviewed remarked, "that road would never have been voted on, because very few people live there" (Pedrosa).

Yet the Chapeu do Sol project had an important demonstration effect. As the CAR director noted, "even though it wasn't an important project from a technical perspective, it was still very important because it showed that through the participatory budget, things actually happen." Many Extremo Sul residents found out about the participatory budget policy after contacting the CAR (or some other municipal agency) to find out why the city was paving Chapeu do Sol Road and to request similar investment in their neighborhoods. The response would invariably be an explanation of the participatory budget process: "If you want such investment in your neighborhood, you need to organize and take part in next year's budget assemblies."

In addition to thus advising those who came to them with questions about district investments, the CAR staff also went out into the district-- especially after 1991, when they obtained a car - and sought out potential neighborhood organizers. One major "discovery" in 1992 was Senhor Azevedo, an extremely energetic retired rail worker living in Ponta Grossa who had been a militant of the underground Communist Party during the dictatorship. Identifying him as a "potential leader" the CAR director encouraged him to found a neighborhood association. Another neighborhood, Lageado - which was suffering from lack of potable water--also began to mobilize after CAR officials encouraged residents to join the budget assemblies. In response to these efforts, in 1992, mobilizing for the big district budget assemblies in the Extreme Sul intensified: 569 people attended the second-round assembly, mostly from Lageado and Ponta Grossa neighborhoods. Sr. Azevedo and a Lageado resident were elected to the Municipal Budget Council, and regular assemblies of delegates were scheduled for the first time. Not surprisingly, the district's official priorities for that year were the paving of Ponta Grossa Road and the provision of potable water for Lageado. Large numbers had begun to participate for the first time, but still mostly from these two neighborhoods.

In the following years, however, as the first investments started to be implemented, other neighborhoods began to mobilize. In my interviews with participants from Belem Novo, Lami, Silo Caetano, Jardim Veronese, and Chapeu do Sol, I heard story after story of how they saw the administration investing elsewhere and how the demonstration effect of those capital improvements first implemented brought them into contact with administration officials who then encouraged them to take part in the budget assemblies. Mobilization and revitalized neighborhood organizations meant that by 1993, the dispute for investments in the Extreme Sul was no longer limited to Ponta Grossa and Lageado. The larger neighborhoods of Belem Nero, Lami, Sao Caetano, Santa Monica, Chapeu do Sol, and many other tiny settlements began to participate effectively. Over the years, different groups were able to mobilize a majority and to obtain major investments in their parts of the district: a health post in Lami; pavement in Ponta Grossa, Chapeu do Sol, Sao Caetano, Belem Novo, and Lageado; a major resettlement project for squatters on the river's edge in Belem Novo; storm drainage and sewer projects dispersed throughout the district; and new and more frequent bus lines all the way to Lami and Sao Caetano. For the first time ever, major city investments were being made in this distant periphery of the city. Each one of those investments corresponded to a neighborhood organizing effort.

It is clear that such mobilization only occurred because the administration was able to convince skeptical residents that participation would actually bring them fruits.

The constant attention of high-level administration staff who regularly appeared at meetings and who saw to it that smaller demands outside of the participatory budget (such as road maintenance) were attended gave further credibility to the participatory process. The fact that upper-echelon administration personnel would go all the way to the Extremo Sul district and meet with residents carried a lot of weight. In my interviews, statements like the following were common.

Opening Closed Neighborhood Associations
The organizing that occurred throughout the Extremo Sul district was only possible because emerging neighborhood leaders were able to break with traditional associations that prohibited the broader based participation of residents and that made little effort to bring government services to the neighborhoods, especially when that government was in the hands of an opposition party. This usually took some time, since even after they knew of the opportunities presented by budget policy, residents often expected the "official leadership" of their associations to organize discussions about capital improvement priorities. In many cases, however, some residents eventually tired of waiting and began to work together on their own The first neighborhood to thus topple a clientelist neighborhood association was Lageado, which for years had a "lifelong" president who, according to residents, never did anything except knock on doors at election time. By 1992, with the scarcity of clean running water in the neighborhood reaching intolerable proportions, residents were able to force an election in the association and to begin holding regular meetings in which many residents participated. It was this new organization that had the capacity to mobilize the neighborhood in sufficient numbers to put "water for Lageado" at the top of the district's priority list for that year.

A similar process occurred in a number of other neighborhoods as the years went by. One particularly impressive tale of a rejuvenated neighborhood association took place in Lami, where a very wealthy landowner, Lira, had been president for thirteen years, never holding open meetings. In the first years of the participatory budget, Lira participated in the district budget assemblies and presented demands for Lami without holding any discussions in his neighborhood or informing residents about the policy. Lami residents only began to meet after administration officials, concerned that few were participating there, called an open assembly in the neighborhood to explain the budget process. At the first such meeting, only six residents showed up, and it was canceled for lack of quorum. But a month later, those six brought with them another forty, and from that time on participation grew in the neighborhood. As one of the original six noted,

The result of this parallel mobilization was the appearance of new leaders in the neighborhood who began to push Lira to hold elections for the association executive committee.

The revitalized neighborhood association went on to become one of the most effective associations in the Extremo Sul district, holding large and regular meetings, bringing in many participants from the impoverished settlements near Lami village, and promoting a number of activities that went beyond the budget discussion.

At the same time, in another part of the district, a very different process occurred: one neighborhood leader managed to combine the mobilization required by the budget policy with what might be called "quasi-clientelist" techniques. Nelson, a small-time politician affiliated with the PDT, mobilized residents in one of the poorest squatter settlements in the Ponta Grossa area to attend the big district assemblies with promises that if they voted him into the Municipal Budget Council, he would get basic improvements for the settlement. According to residents, he even promised investments that they later discovered were not within the purview of the participatory budget, such as electricity, which was a state government service. Providing buses for transportation and, according to some rumors, other fringe benefits (ranging from sandwiches to children's shoes, depending on the source of the rumors), he brought dozens of residents to the big district assemblies and was elected by them to the Municipal Budget Council. Having mobilized so many people, he also clued the fight to appoint his friends and neighbors to a large number of seats in the District Budget Forum. But rather than using the influence he thus gained in the District Budget Forum to propose improvements to the settlement he had originally mobilized, he rallied to pass a project that would pave the road passing in front of his house in a different part of Ponta Grossa.

This process, which strikingly resembled the vote-getting techniques of clientelist neighborhood leaders, is what Porto Alegrenses call inchaco, or the one-time "swelling" of assemblies by people who do not participate regularly, who do not have reliable information about the process, and who simply vote according to the dictates of their leader. It was a relatively common problem in a participatory process that privileged those neighborhoods that could bring large numbers to the assemblies. But, as I will discuss further below, such practices became increasingly difficult to carry out in the Extremo Sul as the rest of the district organized. In any case, whatever the method of mobilizing--through collective discussion and participation, or through quasi-clientelist techniques--many more people were joining neighborhood associations and intending public assemblies in the district than ever before.

Building Cooperative Alliances
While the budget policy stimulated neighborhood mobilization in the competitive effort to win investments, the policy also tended to promote cooperation among neighborhoods. This is because the dynamics of decision making in the forums not only encouraged neighborhood associations to mobilize residents but also to make alliances with other neighborhoods. The District Budget Forums incorporated representatives from too many neighborhoods to allow a single one t fully dominate decision-making sessions. In 1994, for example, the Extremo Sul elected representatives from about ten settlements. Other, more densely occupied districts included up to thirty neighborhoods and settlements. This meant that even well-mobilized neighborhood groups usually sought alliances with others to ensure that their demands were included in the district's top priorities for a particular year.

In the Extremo Sul, Nelson's above-described attempt to monopolize district priority-setting assemblies had the unexpected effect of encouraging neighborhoods in the rest of the district to unify. Since it had the majority at the big district assembly in 1994, Ponta Grossa was in the unusual position of being able to name over half of the district's delegates, for the 1995 budget year. Most of these delegates did not regularly participate in the District Budget Forum. But on the day that the forum discussed how the district's "pavement" quota would be divided up among subdistricts--a meeting that took place only after several months of discussion within the forum–Nelson successfully mobilized the full body of delegates from his area. The result was that he pushed through an agreement in which Ponta Grossa would receive twice as much investment in street pavement as any other neighborhood. This lopsided distribution infuriated residents from the rest of the district. At the next forum meeting, they argued that the inchaco of Ponta Grossa delegates the month before had been unfair and demanded a new vote. Ponta Grossa still had a majority of delegates at this meeting, but the other neighborhoods had also ensured that all non-Ponta Grossa delegates were present at the assembly. In addition, they lured a number of Ponta Grossa delegates to their side, taking advantage of a dispute occurring between Nelson and others within the neighborhood who feared that the pavement would only go to his part of the neighborhood. This new majority voted in a much more equal distribution of pavement. It also voted to hold a recall assembly to revoke Nelson's election to the Municipal Budget Council. A few weeks later, he was removed from office. In the end, his attempt at inchaco did little more than bring about his own demise because it led participants from other neighborhoods to unite against him.

In the following years, the group that led the organization against inchaco made a special effort to help the residents of the settlement Nelson had used to get elected to organize on their own. This group became a regular presence within the District Budget Forum. What is more, this first attempt at cooperating among districts led to the development of a long-term alliance of emerging neighborhood leaders in the Extremo Sul. In the following year, those who organized the resistance against Nelson's inchaco attempt made speeches throughout the district on the importance of unity, of defending "the needs of the district" rather than the "needs of specific neighborhoods," and of ensuring that all the neighborhoods that participated had at least some of their demands attended each year. This discourse proved highly popular. At the end of that year's budget cycle, the "cooperative group" put together a coalition for the Municipal Budget Council elections, selecting representatives from four different neighborhoods and waging a campaign that emphasized the importance of districtwide unity. This coalition won the elections several years in a row and thus came to be the major force within the District Budget Forum.

Under the leadership of this group, investments in the district were from then on distributed much more fairly, a result of annual accords in which each participating neighborhood would be benefited. The group also proceeded to promote a series of larger districtwide projects that depended on the allied support of a number of neighborhoods. One such effort was a project to resettle and rehouse a group of squatter settlements to a well-localized area supplied with infrastructure. Throughout the Extremo Sul, there were dozens of tiny settlements on public lands, most in areas unfit for habitation because they were subject to flooding and erosion, because the soil was too porous for pipelines and sewers, because they were ecological preservation areas, or because they occupied the shoulders of roads. In 1994, the new leadership in the District Budget Forum went to all the neighborhood assemblies in the Extremo Sul to convince budget participants that the district's first priority should be resettling residents of such areas. They succeeded in passing a project to purchase a new area along the district's main road to be used for resettlement. The following year, they mobilized even more intensely to give districtwide priority to the construction of houses for the settlers on the new subdivision. The growing concern for the district's general development was also reflected in the intense participation of Extremo Sul budget participants in a series of public meetings in 1995 and 1996, organized by leaders from the budget forums of the southern part of the city and by local small farmers living in the rural area within the municipality. This discussion led to the formulation of a "sustainable development plan" for the district that was presented to the municipal government.

Transforming Participants' Perspectives
All this shows that within a few years after the budget policy was initiated, people who had initially joined up to obtain localized benefits for their neighborhoods were now thinking more broadly about the potential of the district as a whole and about how neighborhood groups could work together to realize that potential. This reflected changes not only in the way people voted, but in the way they perceived the process of deliberation. There were several aspects to these changes.

In the first place, participants with little or no previous experience in collective action had to learn the elementary rules of holding a meeting. Few of the participants in the Extremo Sul forum had previously been activists--for the most part, they were just ordinary people hoping to improve their neighborhoods. But in order to do so, they had to learn some basic democratic practices--simple ones, such as how to hold a meeting. In the period that I attended assemblies in the Extremo Sul--which coincided with the period of consolidation of the District Budget Council there--a slow but clear transition occurred. Initially, the meetings were chaotic, everyone interrupted everyone, people yelled and cursed, and offended participants walked out before decisions were made. Numerous "scenes" occurred, in which residents angry with decisions made or with perceived betrayals of loyalty disrupted meetings. Such conflicts alienated many participants. As one remarked, "We don't want to be insulted with obscenities. People should use their good sense. Many delegates stopped participating because of all that mess" (Francisco). Many of my informants interpreted the chaos of the meetings in the initial phase as a result of self-centered attitudes:

Over time, the leaders in the forum, many of whom initially engaged in such personal disputes, name calling, and temper tantrums, began to see that those tactics were not working. As they were criticized by their peers for inappropriate behavior and as assembly attendance notably declined after such scenes took place, the more fiery leaders began to temper their ways. At the same time, calmer neighborhood residents noticeably rose in favor among participants. The cooperative group mentioned above, which came to coordinate the meetings, slowly learned to control interruptions, to keep discussion on the agenda, to hold careful and well-counted votes, and so on.

This transformation was much more complex than simply getting people to hold their tongues. People also had to learn to create rules guiding speech that would ensure all voices being heard. By the end of my observation period, it had become commonplace to hold lengthy discussions prior to substantive debates about whether all who wanted to speak should be able to, whether each neighborhood should have one representative present its case, or whether all individuals should have their say, how much time would be given to a certain issue, and how much time each participant could speak. Participants also learned the necessity of carefully defining voting rules. The budget prioritizing process often involved very complex voting procedures as the group placed seven budget categories in order of priority and then placed numerous specific demands in order within each category. The simple yes or no vote did not work for these intricate orderings. But it required some experience before meeting coordinators learned how to explain the purpose of the vote and how to make the rules clear. Developing skills in "rule setting" was the first step in working together collectively.

As they learned to work together, participants' reasons for participating also tended to change. Most people joined the assemblies with very specific, narrow goals in mind: I want the street paved in front of my house, I want to stop the seasonal flooding in my neighborhood. The confrontation between these localized concerns and the similar concerns of others often forced them to change perspectives. Participants from better-off areas began to recognize that some neighborhoods had far greater needs and began to give preference to them.

The districtwide support for relocating squatter settlements, mentioned above, was an example of such emerging solidarity. That process began with the pleas of one riverside squatter settlement resident, Dona Carolina. She was the only delegate of her weakly organized settlement and was therefore unable to pass the demand for resettlement to a better area without the support of others. But with the help of the cooperative leadership, she was able to convince the vast majority of delegates in the district that the needs of her area were so great that they should give it preference, The project, which eventually expanded to include other squatter settlements, was voted "top district priority" largely because delegates came to believe that in addition to dramatically improving the quality of life of squatter residents, it would also increase the general economic potential of the erstwhile tourist region by beautifying the riverside areas and reducing the pollution of swimming areas caused by untreated sewage. Although this rationale suggests that delegates voted for the project at least in pan because they themselves would indirectly benefit, this decision still involved putting put their original, neighborhood-based demands in second place. The fact that this occurred showed that, through participation, people, not only began, to feel solidarity for others but also began to see their own interests more broadly. Many of the people who began to participate for the sake of their street were, a year or two later, champions of "districtwide" issues, defending projects that would promote the economic revitalization of the southern pan of the city or protect ecological reserves in the rural area of the municipality.

What is perhaps most impressive about the learning process that occurred through the budget debates is that participants slowly began to develop the capacity to systematize their changing attitudes into general rules for how resources ought to be distributed. As they gained experience, they began to discuss the criteria that should, determine which neighborhoods should receive their demands before actually discussing the demands themselves, thereby abstracting from specific interests to think more generally. At least up to the end of my research period, this process of developing rules was only incipient in the Extremo Sul district, seeming to result from a sort of exhaustion as after much fighting budget delegates sought a more peaceful way to resolve their differences. At the suggestion of the city-employed organizer, delegates agreed to use a system, in which each neighborhood would list general priorities (sanitation, health care, pavement, etc.). First priority for the district in each category would be given to the neighborhoods that gave first priority to that category. When more than one neighborhood put the same category first, "technical" considerations would be taken into account: how many people would benefit, did the road go by a school or allow for an extension of bus lines, and so on. In this way, final distributions would not simply reflect which neighborhoods could mobilize more votes. Those projects that were located in more needy areas or that would benefit larger numbers would prevail.

Ultimately, in the Extremo Sul, the rule system ended up being combined with a certain amount of direct negotiation in which those neighborhoods that had elected more delegates, pressured to get their projects even if they could not be justified "technically." Even so, with the growing awareness that decisions should correspond to such rules made it more difficult to argue, that the group with more delegates on the day of the vote should simply have its way. In other districts of the city that I studied, the use of distributional rules had developed even further by the period of my research, becoming the central focus of debate as participants designed complex "point" systems to decide which neighborhoods should have priority. This sort of discussion represented a whole new level of deliberation: people began to think about what they thought "distributional justice" should practically mean.

These social learning processes were complex, slow, and often frustrating. One week participants might agree on a set of general roles that should guide decision making, and the next week those same rules might be utterly ignored. But it is certain that over a period of months and years, the rules slowly began to stick as participants became conscious of their broader interests, beginning to consider how their personal concerns could be made compatible with the interests of others and, ultimately, even campaigning for others' causes. This process of developing what Benhabib calls "enlarged thinking" started, however, only because people were drawn to the participatory process in the pursuit of more narrow, individualistic concerns and, through that process, were forced to confront their own concerns with those of others. That is, the fact that the participatory budget provided the incentives to participate discussed earlier ultimately created the conditions in which people ginned experience with participation that not only gave them a chance to build ties of trust and reciprocity with others but also to build new, enlarged worldviews.


Conclusions: Local Policy and Civic Organising

In 1989, when the participatory budget was initiated, the capacity for neighborhood mobilization was distributed very unevenly throughout Porto Alegre. But the result was not that those more organized neighborhoods and districts dominated the process over time as some might have expected[3]. Instead, certain characteristics of the budget policy and of the way it was implemented led to mobilization in poor neighborhoods where previously residents had little or no experience with collective action. Although I have focused here on the Extremo Sul, that district was not particularly unusual. In other districts of the city where neighborhood organizations had been similarly weak before the PT came to office, the budget policy also created an enabling environment in which associations gained membership and force and districtwide alliances consolidated. In those districts where civic organizing had been strong before the budget policy, important transformations also occurred, with new neighborhoods gaining strength and older leadership losing influence[4]. These processes did not occur without conflict, in the Extremo Sul and elsewhere, it took four or five years after the policy was initiated before strong civic groups emerged. But within six or seven years after the PT came to office, civic organizing was no longer the purview' of a few "historically combative" districts of the city' there had appeared innumerable new neighborhood and districtwide organizations that mobilized large numbers, met regularly with administration staff to discuss government policy, worked together to articulate demands, and engaged in broader and more general discussions on how to improve their communities.

Quantitative data on the budget policy confirm these observations. In the first place, assembly attendance changed dramatically over the years that the policy was in place. In 1989, 60% of budget participants at the district assemblies lived in the six districts of the city that had some history of protest-based neighborhood activism, while only 40% came from the ten districts that had little activist history. By 1995, this picture had reversed dramatically: 62% of participants came from the latter ten districts.

In the second place, questions on participation in neighborhood associations asked during a 1995 survey of participants in the district assemblies showed that associationalism had increased as a result of the budget policy. Of those interviewed, 67% were members of some kind of civic association. Of those, 83% participated in associations located in their neighborhoods, including not only official neighborhood associations but also "community centers," "mothers clubs," and informal "street commissions." When asked how their participation in civic groups changed since the policy began, 33% of those who were members responded that prior to the budget policy, they did not participate in civic groups at all. Another 25% said that they had previously participated in civic groups, but since the budget policy, that participation had increased. This means that nearly 60% of those who were active in associations had become more active in the context of the budget policy.

Finally, confirming my qualitative evidence that the residents of poor neighborhoods have been the prime mobilizers in the budget policy, a comparison of our assembly survey data with the 1991 population census shows that the budget participants had significantly lower income levels than the municipal population as a whole. While 29% of the Porto Alegre's households earned three times the minimum salary or less, 45% of participants in the district assemblies declared a household income in this range. Likewise, while 53% of Porto Alegre's households had incomes of more than five times the minimum salary, only 34% of those participating in district assemblies had household incomes at this level. This corroborates the conclusion that the processes of civic organizing described here have been particularly intense for social groups that have been historically dis empowered in Brazil: poor and working-class people living in an urban periphery grossly lacking in basic services and infrastructure.

As noted earlier, theorists of social movements have pointed out that people mobilize when there are windows of opportunities that lead them to believe that action will more likely bring results. Often, such enabling environments have to do with changes in the state power structure, such as the weakening of a powerful elite or the strengthening of reformist policy makers. Obviously, the creation of a responsive, participatory policy represents a particularly radical change in the "opportunity structure for collective action. Some people are born activists, struggling for improvements even when government is not responsive to them. Others may be interested in their communities and in improving their conditions but are unwilling to go through the long, tortuous, and usually frustrating process of organization and protest. Many of those who began participating in the participatory budget policy fit into this second category. For them, the participatory budget policy provided an environment in which, for the first time, it was easy and rewarding to participate in public life. As one participant I interviewed noted,

The first step in a state-sponsored process of promoting civic organizing is that the state must be genuinely open to participation and genuinely responsive to participant demands. But that is not enough: potential participants also have to become aware of that responsiveness. In Porto Alegre, this did not occur immediately. Indeed, after such a long history of having the promises of government officials go unfulfilled, it took some time before a large numbers began to participate in earnest. The administration had to prove its credibility by consistently providing publicly perceivable responses to the demands that participants prioritized in a fairly short period of time. A demonstration effect helped provide previously unmobilized groups with the incentives to organize their neighborhoods.

At the same time, participants also responded to the policy because the issues it targeted were easily comprehensible and clearly relevant to their lives. The policy focused on basic needs that all residents of poor neighborhoods would agree were important to resolve. Where people do not fully understand the purpose of a policy issue, it is much less likely that they will mobilize. One example of such difficulties was the attempt by the same Porto Alegre administration to carry out a participatory debate on reformulating the city's Master Plan within the District Budget Forums. Although Master Plan regulations--determining the minimum width and maximum incline of roads, the size of lots, and so on--have a dramatic effect on the neighborhood infrastructure issues with which budget participants were preoccupied, the municipal agency responsible for the process was never able to clearly communicate to the residents how the Master Plan affected their lives. The use of planning jargon dominated, and educational efforts to translate that jargon into "popular" terms were never carried out on a large scale. In consequence, initial interest by neighborhood delegates soon petered out, and the participatory process was dominated by more educated representatives of sectoral organizations, nongovernment organizations, and business groups.

While both its "demonstration effect" and the fact that the budget policy targeted issues that were meaningful to poor neighborhood residents helped people recognize that participation could bring benefits, other factors helped reduce the costs of participation. In the first place, the fact that the budget forums were open to new participants and to providing benefits to them reduced the costs for newcomers to join up. The institutional structure created by the administration was critical for ensuring this openness: all adult residents could participate in the assemblies, and proportional representation roles ensured that minority groups would hold seats in the District Budget Forums and the Municipal Council. Another effort at reducing the costs of participation was m send government officials to far-away neighborhoods rather than obliging participants to travel to the city center. High-level agency officials were required to spend a significant amount of their time traveling throughout the city to meet with residents to discuss capital improvements and other city services. The creation of a number of District Administrative Centers also brought city services and officials closer to the residents of the impoverished city periphery. The result was that, as suggested by some of the comments cited earlier, participants were often surprised by the regular appearance of high-echelon city officials, including the mayor himself, at neighborhood and district assemblies, This proximity made the participation of poor residents less costly, since they no longer had to travel long distances to engage in negotiations with city officials about capital improvements.

But perhaps the most important effort by the administration to reduce the costs of participation and to increase awareness of potential benefits was the work of community organizers who acted as external agents, visiting unmobilized neighborhoods, seeking out new leaders, helping people organize, and disseminating information about what could be ginned Sough collective action. In the Extremo Sul district, direct contact with government organizers was absolutely essential for drawing new neighborhoods into the process. One government-employed community organizer referred to this integration as "politicizing the pothole." The administration had a coherent policy of encouraging all people who came to government offices with small demands to begin to think about the larger needs they may have and to organize their communities in the pursuit of those demands. Often, this required innumerable trips to unorganized neighborhoods to help residents call assemblies and to present information to neighborhood groups on what they needed to do in order to participate effectively in the budget process. Indeed, such "popular education efforts" on the part of administration officials could have been used even more intensely in other areas of decision making. For example, had more effective information seminars been held around the Master Plan discussion, that policy might have been better able to mobilize residents[5].

While all these aspects of the policy helped provide incentives for collective action by increasing the expectation of benefits and reducing the costs of participation, the policy also provided an environment in which people could gain experience with cooperative action and, in that context, could become aware of broader needs going beyond their own neighborhoods. The specific institutional structure of the policy helped encourage associations to become more inclusive and cooperative. On one hand, neighborhoods had to mobilize in order to obtain benefits. In all the neighborhoods I examined, associations membership increased and open discussions of priorities occurred, although in some cases, quasi-clientelist techniques were used to attract people to the assemblies. On the other hand, the dynamics of decision making within the District Budget Forums, through which neighborhood groups had to negotiate with one another to come up with lists of priorities, encouraged the formation of alliances among neighborhood groups. The experience of building alliances from year to year contributed to the creation of strong formal and informal districtwide networks of neighborhoods throughout the city. Individuals within the networks changed the way they understood their own interests, beginning to see themselves as members of larger groups and beginning to take the interests of others into account. In these ways, the budget process discouraged longstanding traditions of clientelism and promoted what some authors would call "social capital" or networks of reciprocity and trust.

The role of the government as an external agent was also important in promoting those networks. Government officials sought to mobilize residents in areas where clientelist associations predominated, calling open meetings in neighborhoods, such as Lami, where the association president refused to disseminate information to residents about the budget policy. In the case of Ponta Grossa, where a neighborhood leader engaged in inchaco, the government-employed co--unity organizers worked closely with other budget participants to bring better information to residents and to help them organize on their own. The organizers working with the District Budget Forums also helped reduce discord and to promote agreements among delegates. In the Extremo Sul, although budget delegates coordinated the meetings, an administration organizer provided a constant source of information, advising the leadership on how they could better coordinate discussion and promoting "cooperative values" and the use of "distributive rules." This work much resembled the guidance that church activists, student groups, liberal professionals, and other groups always have provided in helping neighborhood groups organize.

The picture presented here of state-sponsored civic organizing sharply contrasts prevailing perceptions that government spending and proactive policy making are only likely to discourage collective action. Much to the contrary, the mobilization and organization I have described in the Porto Alegre case was the direct result of a policy that involved a great deal of government investment in two fundamental ways. First, countering international policy trends toward promoting self-financing and "cost recovery" in infrastructure investments, the Porto Alegre government gave top priority to raising revenues through taxation and administrative streamlining in order to pay for capital improvement expenditures. It spent between 12% and 20% of city revenues on that rubric each year - amounting in 1995 to about sixty million dollars - most of which went to neighborhood-level projects. In the context of the participatory budget policy, this intensive government spending set off, rather than discouraged, a boom in civic action throughout the city.

Second, the policy also prioritized a different type of investment: an investment in building a new relationship between government personnel and local citizens. The efforts of community organizers hired by the administration, and of many other government workers who regularly attended neighborhood meetings, reflected a new type of governing that privileged accessibility, flexibility, and negotiation. It was successful largely because of the willingness and passion of several hundred politically appointed employees who worked outside conventional hours going to meetings at night and on weekends. It was also successful because of extensive efforts within the administration to promote an "esprit de corps" of respecting ordinary citizens and avoiding bureaucratic rigidity. The result was that through a "case-by-case" process of organizing, popular education, and negotiation, the Porto Alegre administration helped neighborhood groups organize, providing inexperienced neighborhood leaders with practical assistance in mobilizing their communities, helping them learn the basic skills of organizing, and promoting values of cooperation and reciprocity.

None of this occurred, of course, outside the realm of politics. As always with state-sponsored participatory processes, the actions of the government were influenced by the need to generate political support for the PT But since the administration sought to do so in part by gaining allies in poor neighborhoods and in part, as mentioned earlier, by building a more general reputation for "democratic" governing, the administration's interests largely coincided with those of neighborhood organizations both sought a transparent process through which neighborhoods could obtain access to government infrastructure and services[6]. This coincidence of interests was perhaps the key that led to the cycle of civic organizing I have described, since while, on one hand, the government found that by providing infrastructure and services to those neighborhoods that effectively participated, it could build credibility and support, on the other hand, neighborhood residents tended to mobilize only when they believed that their participation would bring fruits. The end result went well beyond mere electoral strategy, leading to a fundamental transformation of political life in Porto Alegre as neighborhoods residents that had earlier served as powerless cogs in clientelist machines were now active participants in public life, organized into representative civic associations and engaging in an open, transparent debate about government policy.

Notes

1 In: Politics and Society, Vol. 26, nr. 4, dec. 1998, pp 511-537 See Rebecca Abers, Inventing Local Democracy: Neighborhood Organizing and Participatory Policy-Making in Porto Alegre, Brazil (ph.D. diss., University of California, Los Angeles, 1997) for a more detailed exami-
nation of these issues, as weil as a more general discussion of the politics of implementing the budget policy in Porto Alegre. Both the dissertation and this article are based on over a year and a half of field research in Porto Alegre between 1993 and 1997, funded by the Inter-American Foundation, the Fulbright U.S. Student Program, and the UCLA Latin American Center. The research involved extensive interviews with more than 50 local govemrnent officials and more than 40 budget participants. I also collaborated in the application of a
sample survey of 622 participants throughout the city (see note 9) and conducted participant observation at over 100 public assemblies and neighborhood meetings. Participant observation and interviews focused on two regions of the city--one with a long history of civic organizing and another where neighborhood organizations were historically weak. This article focuses on the latter region.

2 Although unusual, the Porto Alegre case is not unprecedented. For an interesting study of a sirnilar policy in a U.S. city, see Haeberle's study of the Citizen Participation Program in Birrningham, Alabama. As in Porto Alegre, Haeberle finds that local govemment investment in directly promoting citizen participation drarnatically strengthened the associational environment of that city. Steven H. Haeberle, Planting the Grassroots:Structuring Citizen Participation (New York: Praeger, 1989).

3 For a discussion of theories of participation that assume that those who are best organized early on win continue to dominate over time, see Rebecca Abers, Learning Democratic Practice: Distributing Govemment Resources through Popular Participation in Porto Alegre, Brazil in Michael Douglass and John Friedmann, eds., Cities for Citizens: Planning and the Rise of Civil Society in a Global Age (Chichester, UK:Wiley, 1998),39-65.

4 See Abers Inventing Local Democracy, for an examination of how the budget policy affected organizing patterns in one historically active region.

5 The work of an extension group from the local university's Urban Planning Program showed how it was possible to translate the Master Plan issues into terms more comprehensible to ordinary participants. They held a seminar with members of the budget forum in the Gloria region to help them respond to the Master Plan process. With same guidance about how physical planning and legal issues affected their region, a vigorous debate taak place as participants mapped the region and located areas that represented particular problems or that were sites ofpotential improvements.

6 See Rebecca Abers, Practicing Radical Democracy: Lessons from Brazil (in: Plurimondi, Vol. 1, nr. 2, juli/­december 1999 pp. 67-­82)