In June 2005 a scandal shook the administration of Brazilian President Luiz Inácio Lula da Silva. His Workers’ Party (PT) had been charged with making monthly payments to members of Congress to support its favored legislation. Allegations of vote buying have never been proved, and the accused were never linked directly to Lula, who managed to avoid the fallout. But there is no denying what the scandal revealed about the deep structural problems in Brazil’s democracy and the formation of majorities in the Brazilian Congress.

In 1988 Brazil enacted a new constitution. Its approval concluded the transition back to democracy. More than two decades of authoritarian rule were over. But it remained to be seen how far the country had come in changing long-standing traditions of political exclusion. While the constitution included fourteen articles enhancing popular participation, the organization of the political system was largely unchanged. Presidents would be elected by popular vote, but Congress remained as weak and fragmented as it had been under the authoritarian regime. Even a president elected with majority support would have trouble gaining Congress’s backing. In the last five elections, the president’s party has never claimed more than 20 percent of the seats in Congress.

Fernando Henrique Cardoso, who served as president from 1995 until 2003, created an ingenious way of governing within this framework. Despite lacking majority control, he was able to get his legislation approved in Congress. He established broad alliances between his party and an array of other parties. This “presidential coalition” system involved the strategic distribution of ministry positions in order to form a coalition. The result was a 70.7 percent success rate in Congress for executive legal initiatives, rivaling outcomes in the British Parliament.

When Lula became president in 2003, pork-barrel politics were not an option. The PT had run with grassroots support, emphasized public participation, and criticized politics as usual. Moreover, Lula could not dole out ministries as Cardoso had, particularly in social policy areas (housing, income support, education), because the PT had strong links to social movements focused on these concerns. As a result, Lula’s efforts to get PT policies through Congress foundered. Hence the scandal: his supporters believed buying political support in Congress was the easier way to win over the legislature. Lula’s response to the uproar that followed opened a path toward a deeper, more direct form of democracy in Brazil.

Lula reacted to the scandal in two ways. First, he dismantled the PT machine headed by his chief of staff, José Dirceu, and filled empty spots with PT members largely from outside the party’s base in São Paulo. He brought Dilma Rousseff, Tarso Genro, Fernando Haddad, and Jaques Wagner into key positions. These officials reconfigured Lula’s government and have maintained power: Rousseff is now president, the first woman to hold that office; Genro and Wagner are state governors; and Haddad was recently elected mayor of São Paulo.

But Lula didn’t just shake up his leadership. In an effort to pursue his agenda outside the boundaries of Congress, he sought to bring further democratic control directly to the people through national conferences.

National conferences are joint state and civil society initiatives aimed at forging consensus in a specific policy area. The conferences have existed in Brazil since the 1940s, when President Getúlio Vargas sponsored the first national conference on health. Lula’s government made them a basic feature of his administration. Of the 115 national conferences that have taken place in Brazil since 1941, 74 were held during Lula’s eight years in office.

Lula’s government standardized national conferences. They are initiated by one of the ministries through an administrative act, and after deliberations among citizens and officials at the city, state, and national levels, the conferences make recommendations to the federal government. Conference decisions can become legal decrees signed by the president. In many cases, conference outcomes lead to government-sponsored changes, particularly to social policies.

In the last ten years, close to six million adults (6.5 percent of Brazil’s population) have participated in national conferences on topics ranging from gender equality to domestic abuse to curbing rural violence in Brazil. Human rights are a major theme. Participation at the national conferences strongly resembles local political participation in terms of the incomes, levels of education, and genders of those who take part. Just over half of participants at national conferences are women with four years of education and an income between one and four times the minimum wage. Turnout at conferences also resembles that at the PT’s previous local experiment with direct democracy, its successful participatory budgeting project in Porto Alegre.

Through national conferences, veteran activists are finally gaining a voice in Brazil’s public policy.

The PT’s investment in direct participation is a hopeful sign for Brazil’s democracy. National conferences bring forward the voices of people who historically have been excluded by the Brazilian political system. For instance, less than 10 percent of representatives in Congress are women, and low-income groups typically find their access to centers of power massively constrained. The presence of women and poorer and less-educated citizens at the conferences represents a departure from traditional Brazilian ways.

Historically, social movements were the only form of action for those excluded. Participation in national conferences bears an important relationship to social movements. São Paulo’s health movement grew as Brazil was democratizing, and its members became “co-designers” of the national conferences. Organizing surrounding social assistance, urban reform, and landless peasants also took off during the same period.

While the conferences have inspired new mobilization, issues surrounding health and social assistance continue to rank high on the list of conference-goers’ concerns. The continued interest in traditional areas of social policy suggests that activists who have long been participating in local politics are gaining an outlet at the national level through the conference system. They are now able to directly challenge the government and the political system.

But how effective are the conferences? How are results reflected in policy?

I conducted a survey of conference participants, which found that only 5.6 percent thought all conference decisions were implemented by government; 16.1 percent thought that most of them were; and 19.6 percent thought that an “average number” of decisions were implemented. These numbers no doubt correspond to the reality: many decisions undergo considerable revision before they are implemented.

The Institute for Applied Economic Research, the government’s think tank in Brasília, asked managers of government programs whether they account for conference decisions when establishing program guidelines. Of the 140 largest programs, about 14 percent draw on conference outcomes. However, close to 30 percent of social programs—including the National Program on Social Assistance, the National Program on Food Security, and most of the programs on urban infrastructure and housing—take national conference results into account. In this crucial area, widespread participation seems to be successful.

Thus, we still have mixed results on participation in the federal government. It is successful in the areas that traditionally have had strong social movements and in which the PT’s policies have been effective. But we still do not see participation in decisions about national infrastructure, such as large mineral and oil projects and the recently approved Belo Monte Dam. These engineering plans are growing in importance as the Brazilian economy rebounds.

Brazil is trying hard to create what the government calls a “national system of participation.” In the future this system may well be able to connect citizen participation to government decision making at the local, state, and federal levels. National conferences are already having an impact in the production of new legislation in Congress. Conference decisions on social policy are more easily approved than government-sponsored legislation. They are also improving the management of large programs at every level of government. The national system of participation may offer a way of offsetting the problems of a fragmented Congress.

While there is more to be done, Brazil’s experience is encouraging. The conferences demonstrate that large-scale democratic engagement beyond the voting booth is feasible. All democracies could learn from Brazil’s political creativity.