In the history of politics, instances of direct people’s participation in local government are rare, but participation on higher levels of government is even rarer. Two such unusual incidents of citizen involvement on a large scale occurred in the Brazilian state Rio Grande do Sul from 1999-2002 and 2011-2014, when the government gave away some of its powers over the budget to its inhabitants. How did it work? Did people participate? Did it make any difference? And what does this story tell us about the possibilities and limitations of establishing a participatory democracy at the state level?

Since the late 1980s, Rio Grande do Sul has been a stronghold of the Brazilian Workers' Party (PT). All through the 1990s and up to the election of Luiz Ignacio Lula da Silva in 2002, the electorate always came out in favor of the presidential candidate from the Worker’s Party in the national elections.

In the state capital, Porto Alegre, it held power continuously from 1989 to 2004. It was in this city that something called participatory budgeting was first created. What this meant was that the mayor’s administration invited all citizens above age 16 to take part in a process to make a people’s investment budget for the municipality. It soon became a success: every year new people came to the meetings, poor neighborhoods got more money than ever before (they got connected to the water network, muddy roads were paved, their sewer system was improved, schools renovated, health clinics established and computer labs and kindergartens came to the unprivileged part of the city) and the party got reelected over and over again.

Video: This is part one of a good documentary on participatory budgeting at the local level in Porto Alegre and some other cities. It offers a view of many of the features that was shared on the local and the larger level: like the public assembly, election of delegates etc.

The movements enter the state

One important reason to why both the Worker’s Party and the participatory budget became entrenched in Porto Alegre, was the development of strong ties between the party and local neighborhood associations that supported its policies and were able to get people with lower incomes and short education to participate in the process (Bootstrapping Democracy is one of many books).

Such ties also existed at the state level, where a range of organizations for rural and industrial workers, landless peasants, public employees and coalitions of neighborhood associations actively supported the Workers' Party, and the party in turn included the demands of these movements in its election programs. Many party activists were also activists in these organizations.

In 1998 the Workers' Party headed a coalition of left-wing parties called the Popular Front in the gubernatorial elections. They won and took over the state government in 1999. One of the main promises of the coalition was to open up for social movements to influence the government’s policies, and to create spaces where citizens could plan and decide on the budget. A central reason for doing so was to redistribute public funds to the poorest sectors in the population.

The newly elected governor, Olívio Dutra, had been mayor in Porto Alegre during the years that participatory budgeting was established, and the model for participation was taken from the state capital.

Image: Olívio Dutra (here seen in the middle with his arms wide open) – the governor who introduced participatory budgeting in Rio Grande do Sul in 1999 – was one of the founders of the Workers' Party. He was perceived as a politician who cherished close ties with social movements and trade unions.

Modelled on Porto Alegre

As in Porto Alegre, the popular assembly became the basis of the state’s participatory budget. The government invited citizens to participate in both municipal and regional assemblies to debate and vote on social policies, infrastructure and economic development projects. Municipal assemblies mainly discussed local demands, but also voted on regional and state-wide issues. Regional assemblies had a greater emphasis on services and public works that affected two or more municipalities, and they also provided a list of policies and projects that the local assemblies could choose from.

Both set of assemblies elected delegates to the regional level, and these delegates in turn elected councilors for the whole state. These representatives had an important task in compiling and prioritizing all the local and regional budget proposals. They negotiated with both the government and other delegates in order to make up a list of priorities. In addition they discussed and modified the rules of the process each year.

The role of the state government was then to make an investment and service plan based on the delegate’s work. This plan included a distribution formula which directed investment towards the poorest regions in Rio Grande do Sul. Finally, the governor sent the plan to the state’s legislative assembly, where the elected politicians debated it and took a vote. Once it had passed there it was sent back to the delegates and councilors again, who used the document to monitor the execution of the budget.

Limited participation and indirect representation

Although the state government modeled its version on the local participatory budgeting in Porto Alegre, it was notably different in some aspects. In Porto Alegre the local government invited the citizens to two rounds of open assemblies. The first round was used to discuss the previous year’s budget, come up with new proposal and elect delegates. The second round was used to discuss the delegates’ work and to vote on priorities.

Video: This is the only footage of the participatory budget from 1999-2002 that I have found online. It is only in Portuguese, but if you don’t speak the language it nevertheless gives an idea of how the assemblies in the process looked like and who participated.

At the state level, all of this was done in just one round. As a consequence the assemblies had very limited space for discussion or exchange of information among participants. It has been pointed out that they were more a place for participants to get to know each other and each other’s demands, than arenas for public discussion.

Furthermore, two levels of representation were created to tie all the regions in the state together, as opposed to only one level in Porto Alegre. As a consequence, the bonds between participating citizens in the municipal assemblies and elected councilors at the top level were very weak. Since the regional delegates elected the state councilors, those who participated in the assemblies had no opportunity to recall councilors and limited opportunities to control and affect their actions.

The fight with the opposition

A third important difference between participatory budgeting in the capital and the state, was that there existed organizations in all the regions of Rio Grande do Sul who already directed a similar participatory process. These were Regional Development Councils (COREDEs) who had been allocated a certain amount of money to use in a state-wide referendum called the Popular Consultation (’Consulta Popular’). Unlike the participatory budget that was initiated by the new Popular Front government, these councils didn’t hold open assemblies but rather decided the proposals that the public could vote on themselves.

The Left in Rio Grande do Sul perceived these councils to be a political instrument for the right-wing, as they typically consisted of state and federal deputies, mayors, city council presidents, representatives from universities, unions and organized civil society with ties to the Right. The new government therefore ignored these councils almost completely, and instead created forums where their supporters, social movements as well as unorganized citizens could participate. In the words of Goldfrank and Schneider, the participatory budget “provided a direct link in between the executive and the population, bypassing both the legislative branch, where the PT again was a minority, and those preexisting organizations favored by prior administrations.”

This provoked an intense reaction from both the parliamentary opposition and the development councils. The latter didn’t want to have anything to do with the participatory budget and effectively broke with the government.

To counter the legitimacy of the participatory budget, the right-wing in the legislative assembly established a series of public hearings called the Democratic Forum where the participatory budget proposal was criticized by various opposition groups and institutions. A former state governor sued the Popular Front government for improper use of funds, which resulted in a court order that halted the government’s ability to run the participatory process for almost one year (it was then organized by a coalition of social movements instead until the court order was reversed). And it also led to attacks in the media, which presented the participatory budget as an illegal, parallel power to the democratic state, which the Workers' Party used to undermine and eventually supplant legal representative institutions.

Image: When the Workers' Party government initiated participatory budgeting in Rio Grande do Sul in 1999, there was already a participatory structure in place called Regional Development Councils (COREDEs). As the image above shows, the state was divided into 28 such councils. The government’s attempt to bypass these councils led to a fierce conflict, and in the end the Workers' Party was forced to compromise their version of participatory budgeting by giving concessions to the opposition.

It wasn’t possible to continue with the participatory budget amidst the reactions from the opposition. In the end, the government made several concessions to its opponents to be able to continue the process. The development councils, for example, got a quota of about 25 per cent of the representatives in the state budget council. Also, the budget process was organized in such a way that the councils could restrict the scope of the municipal assemblies’ decision-making powers by giving the regional assemblies a role where it provided guidelines for local discussions. Finally, the government accepted to execute projects from previous years when the government was controlled by the right-wing.

Increasing attendance and more money to the poorest regions

Surprisingly, the conflict did not discourage people from participating in the budget process. The number of people attending the assemblies rose through the government period, to a peak of 378,340 persons in the third year of government. This amounted to six per cent of the electorate.

Research has shown that this form of mass participation at the local level has contributed to more investments in areas that are important to poor persons such as education, health and basic sanitation (see for example here, here and here). This was also the case in Rio Grande do Sul that spent more on these areas than on any other state in Brazil in the same period.

The rules of the system dictated that more money automatically was allocated to poorer regions in the state. This meant that municipalities who lacked basic services in health and education were awarded more investments than others. But funds were also allocated according to the number of participants in the municipal assemblies. Since social movements connected to the Workers' Party were best organized and most positive to the process, municipalities with a high number of members from the government party got more resources than those who hadn’t.

Unlike Porto Alegre were the participatory budget secured the reelection of the Workers' Party for four consecutive periods, the party’s governor failed to get reelected in 2002. It is a saying in Rio Grande do Sul that nobody wins the gubernatorial elections twice, but researchers have also attributed the loss of the Workers' Party to the expectation that were not met in the participatory budget process. The election in 2002 was very tight and it might be that the lack of sufficient revenues to deliver on its investment promises swung the vote in the sitting governor’s disfavor.

Want to read more?: Budgets and Ballots is the only good quality study in English of participatory budgeting at the state level that is freely available online. The authors of this report have also written an academic journal article that can be accessed through many university libraries. For those who read Portuguese, Cláudia Feres Faria have written a doctoral dissertation about the process and also an article on the clash with the opposition.

Starting again – as a compromise

It was not until 2011 that the Workers' Party regained the government in Rio Grande do Sul. With Tarso Genro, twice the mayor of Porto Alegre, as its gubernatorial candidate, the Workers' Party headed an alliance of left-wing parties called Popular Unity for Rio Grande.

The alliance won in the first election round, and soon formed a government with two other center-right parties which gave it a majority in the legislative assembly. Although this was an advantage for the new government — as compared to the paralyzing opposition experienced by the first Worker’s Party government — it also meant that it began its term as a compromise with less radical parties.

Image: Tarso Genro (here talking to a crowd in an election campaign) was the governor in the new Workers' Party government that came to power in 2011. Genro is generally seen as a less radical politician than Olívio Dutra, and who also emphasize the importance of bonds with other sectors in the population than social movements and trade unions.

Another early compromise was made with the Regional Development Councils (COREDEs). While the Workers' Party had been out of government, the competitor to its participatory budget — the Popular Consultation — had increased in popularity. The number of people voting in its referendum had increased from 462,299 in 2003 to more than the double in 2009.

As a consequence the Genro-government gave the Development Councils a prominent role in its new participatory scheme. The responsibility for organizing the budgeting process was now shared between the government and the councils. In each of the 28 regions the regional and municipal assemblies were convoked by a coordination consisting of government and council representatives. Together they also facilitated the work of the elected delegates in the budget process, and oversaw the referendum that decided the budget priorities and the subsequent implementation of chosen works and services.

This lead to a mostly harmonious relationship between the Development Councils and the government. At one point, however, the councils threatened to boycott the process, but their reason to do so was not to undermine it, but rather to force the government to give it higher importance. Their main complaint was the slow execution of the investments and the government’s unwillingness to increase its funding.

Even the right-wing opposition in the legislative assembly had a more pacific stance toward the government’s participatory ambitions. Their main objections were that the government on one hand promised more than they could deliver and thereby risked to undermine the Popular Consultation, and that the government exploited the participatory process to mobilize support for the Workers' Party.

Just one out of many

But the most crucial difference between the participatory budget under the new Workers' Party government and the one which preceded it ten years earlier was that it was no longer the main channel of participation. Under the Dutra-administration it was by far the most significant channel, but in 2011 a System of Popular and Citizen Participation was established where the budget became only one of several branches of participation. A range of other opportunities emerged for organized social actors to influence the government’s work. Even budgetary questions were handled in other fora such as sectoral councils and thematic conferences.

In addition many new spaces of participation and dialogue had opened up with the Workers' Party presidency at the federal level. The social movements now went to the places they could affect their causes most effectively. For the large movements that previously were so important in mobilizing people to come to the assemblies, this was apparently no longer the state participatory budget. This didn’t mean that they stopped participating, but rather that they gave it less importance and basically involved themselves if their local branches found it suitable.

Hence, a survey that I did in 2014 showed that only 18 per cent of the delegates in the budget process said to represent a civil society association, NGO, social movements, popular council or trade union. In comparison, a survey among delegates in 2001showed that 54 per cent of delegates in the metropolitan region participated in a neighborhood association, 14,5 per cent in a trade union, 14,7 per cent in a popular council, and 9 per cent in an NGO.

A regionalized process

Several important investment areas were also excluded from the new government’s scheme. As opposed to the participatory budget in early 2000s, participants could not discuss or vote on issues concerned more than one region or the whole state. This meant that infrastructure projects such as electric transmission lines, ports, airports and highways were left to other forums. This was also the case for investments in statewide institutions such as universities, central sports or cultural venues, or specialized hospitals.

Image: In the new Popular and Citizen Participation scheme, participants could decide over investments on a regional and local scale. One very common demand in 2011-2014 was to buy new fire trucks to the fire brigades.

Correspondingly, the share of the budget that was decided through the participatory process had also diminished. From 1999-2002 all the investments in the state budget were under discussion, but according to the government’s own transparency website this was only the case for around 10 per cent during the Genro-administration.

This didn’t mean, however, that those who attended the meetings had nothing to talk about. Since most public works and services are delivered locally or regionally, they could propose investments within most government areas such as education, health, rural and economic development, sanitation, public security, social services, poverty reduction, irrigation, sports, culture, tourism, human rights and environmental sustainability. Typical demands raised at the assemblies were buying medical equipment to health clinics and hospitals, renovating schools or providing them with computer and lab gear, updating the vehicle fleet of fire brigades and police departments, or supporting peasant-run dairies and other agricultural cooperatives financially.

A chain of events ending in a referendum

It is worth noting that the participants in the municipal assemblies could no choose these investments freely. Their demands came at the end of a chain of discussions which provided a set of guidelines for all subsequent demands. The main event in this chain was the multi-annual budget plan which was made in the first year of government. Around six thousand persons from public institutions, social movements and community organizations met in nine big preparatory meetings in the state. Here they discussed the main content of the budget for the next four years and elected representatives to formulate the thematic areas and government programs in the plan. All demands in the subsequent participatory budget process had to come from one of the programs that were identified in this plan.

Before the annual municipal assemblies could take place, regional assemblies were held that gathered local leaders, provided information about the status of the previous year’s budget and the present year’s funding, established rules on how to distribute these resources within the region, and appointed those who were to coordinate the process. Next, local assemblies took place in nearly all of the state’s 497 municipalities. In each of these meetings, up to ten local demands and five regional demands were formulated. Delegates were also elected to represent the municipality at the regional level.

In contrast to the state participatory budget in the early 2000s, the delegates under the Genro-government didn’t play an equally important role. Their job was not to make decisions but mainly to comprise the demands from the municipal assemblies into 10-20 items that would be presented in a referendum where the whole electorate could participate. It was the referendum that finally decided on which proposals that would be prioritized.

Video: The government put a lot of effort in trying to get people to vote in the referendum that ended the participatory budget process. This is one of many videos that were made to get people to vote.

In this sense, participants in the municipal assemblies could directly influence the outcome of the process, since their delegates could not alter the proposals which were supported in the municipal assemblies. The cost of this ability was that it was not possible to propose projects for the whole state, and only very limited opportunities to vote through regional investments. Even though the ballot contained a field where voters could choose between a few large regional projects – such as establishing a new hospital or a new educational program in the schools – this was only suggestive and not binding for the government.

Redistribution, but poor performance

As in the early 2000s the redistributive dimension was also central in the participatory process of the Genro-administration. The bylaws of the process dictated that regions with higher poverty rates and lower social development scores would get a larger chunk of the funds than richer regions. The poorest municipalities within these regions also got more people to vote in the referendum, which in turn increased their share of the resources (these facts are taken from an unpublished manuscript where only the abstract is available online).

Whether this money ended up being spent in the municipalities or not, is a different question. On average, the performance rate — meaning how much of the budgeted money eventually was used on projects and services — was only 49 per cent for the whole period. In 2011 it was as low as 30 per cent, but it had risen to 80 per cent in 2014. The reasons the government gave for the low performance rate, was poor planning capacities and technical expertise in small municipalities, as well as a general difficulty for the whole government to raise as much revenues as expected.

Less popular participation

This didn’t discourage people from coming to the meetings or voting in the referendum. The number of people directly involved in regional and assemblies increased from 60,000 in the first year to 87,000 in the last year of government municipal. The number of voters increased from 1,1 million to 1,3 million in the same period. In 2014 as many as 16 per cent of the electorate “participated” in making the budget, as opposed to 6 per cent in the early 2000s. But this conceals the fact that the type of “participation” was very different from the one in the early 2000s. Under the Genro-administration it was enough to vote to be counted as a participant, but under the Dutra-government you had to come to a meeting and at least listen to the discussions that took place. In 2014 the share of the electorate that came to the budget meetings was just a little above one per cent.

There are reasons to believe that the people who came to the meetings and who were elected as delegates were quite different from the first Workers' Party government to the last. Two researchers who studied the participatory budget under both governments, wrote in 2014 that "while in the original experiences participating citizens came mainly from what me might call civil society: people who came on their own, who became mobilized by neighborhood initiatives, or were participants in social movements, trade unions or other civil organisations - more than half of the participants in public meetings of the participatory budgeting process are public servants." My own survey confirms this picture. Most delegates in 2014 were employed by the government. 67 per cent told that they represented a municipality, a state organ or a hospital, school or university - not a movement or community.

Whether this would have changed in a second period of government is unknown. In 2014 Tarso Genro lost clearly to his adversary, and a right-wing government was formed. It continued with the Popular Consultation in a similar format to that of the Worker’s Party government, but the funding was significantly lower from the previous period.

Want to read more? There hasn’t really been written a lot about the state participatory budget process in 2011-2014. In English, one article has been printed that compare the two periods (only available through certain university libraries), and another on electronic voting in the process can be read online. If you read Portuguese you may consult the chapter “Sistema Estadual de Participação Cidadã in a recent book on public policies in Rio Grande do Sul. Other data for this section is taken from my own field-work in the state.

What does this story prove?

Is this a good or a bad story? Does it tell us whether participatory democracy can be scaled up or not? I am not sure this is the right question to pose, since many kind of “proof” can be read into the story depending on what perspective you have.

Those who think that participatory democracy is impossible on the larger scale would probably argue:

“Didn’t we tell you this would happen? That a democracy on this scale would no longer be direct, but representative? This is exactly what happened under the Dutra-administration when delegates at the state level had decision-making powers not different from in a parliamentary democracy. Yes, the system was more direct under the Genro-administration, but was only because participants then could only decide over very local investments. And didn’t we tell you that the participation would be elitist? Only a tiny minority in the population participated in both instances. Most of the delegates under the Genro-administration were not ordinary people, but professional public employees. And under the Dutra-administration they were probably professional activists that had a far better understanding of regional and state-wide issues than the vast majority.”

Those who favor participatory democracy on all scales – not just the local – may reason:

“Both of these experiences had their limits, but at least they show that it is possible to coordinate the participation of hundreds of thousands individuals in the same decision-making process. Together they managed to make something as difficult and complex as an investment budget. Their model of combining direct attendance in local and regional assemblies, the election of delegates and a referendum, could be enlarged to include millions of people. Yes, only a minority participated and this was probably some kind of elite in both instances. But this is not a problem of scale as such, but rather a problem of a highly unequal society where most citizens lack the basic education, time and economic resources that are necessary to participate meaningfully on a state-wide scale. If people were equal and if the whole economic, cultural and social infrastructure supported participatory democracy, these two instances would look quite different.”

Correct as some of these arguments may be, it is not possible to reject the feasibility of participatory democracy on a large scale by pointing to only these two examples. They are, as mentioned above, restricted most notably by economic, social and cultural inequality. On the other hand, it is not possible to prove the feasibility of such a large scale direct democracy by referring to these two instances. After all they were experiences where only a minor share of an already austere budget was decided. If the rest of the budget or more long-term issues were included in the process, it probably would have been way different.

Limits and possibilities

I think it is more interesting to ask what these two instances tell us about the limits and possibilities of trying to establish participatory democracy at the state level.

In the case of Rio Grande do Sul, one of the advantages of going through the state was that popular participation got an official stamp of recognition. It has sent a message that participatory budgeting is a legitimate way of allocating public investments, and this has made it very difficult to reverse those arrangements that have been installed. It has also opened up for demands for expanding popular participation to other areas: “If parts of the budget can be decided this way, why can’t it apply to the rest of the budget or other areas of the state?”

Furthermore, implementing it through the government meant that the advocates of participatory democracy got access to resources they wouldn’t otherwise have had. They could dedicate public revenues to the outcome of the process, employ personnel to facilitate meetings and mobilize participants, and not at least hold the state administration accountable to executing the decisions that the participants made. The flip-side of this was that the bureaucratic agencies to a large extent defined the scope of what could be demanded in the assemblies. Participation was not “free” in the sense that those who showed up themselves could ask exactly what they wanted from the bureaucracy. Rather the professionals employed in the state administration defended their “expertise” and autonomy, which meant distinguishing themselves from the lay people in the assemblies.

Image: Supporters of participatory democracy may argue that the model of combing direct attendance in local and regional assemblies, the election of delegates and a referendum, could be expanded to include millions of people. But they mustn’t forget that this process only dealt with a tiny share of the budget.

Going through the state means a top-down approach to establishing participatory democracy, which has its peculiar challenges. Since the implementation of the process in Rio Grande do Sul wasn’t supported by a majority of the voters in the election the resulting institutions were a product of compromises with other parties and the parliamentary opposition. Also almost none of the municipalities had any previous experience with participatory budget. Most mayors were even against the system, and wanted to reap the political benefits from distributing state money in their own way. The participatory process was unleashed from a small group in the center of the state administration and sent into to a partly hostile, partly indifferent and largely inexperienced territory, which gave a participatory budget that in many places was very different from the ideals of its creators.

The state is not a blank canvas on whichever vision can be painted. It consists of multiple centers of authority, which all have a say in the formulation of implementation of public policies. Representatives from large organizations and sectors in the population form subsystems with government agencies and political parties, with considerable autonomy in commanding public resources. The state’s powers – especially its ability to collect taxes – depend on economic growth, and it plays a significant role in ensuring market stability and fiscal responsibility.

Participatory spaces that want to utilize the resources that are available through the state, also have to fit in with this public machinery. The story of the scaling up of participatory budgeting illustrates some of these restrictions.