How Privatizations Created New State Companies in Brazil

In July 2014, Brazilian President Dilma Rousseff’s Facebook page celebrated the record production of iron ore by Vale. According to the page, the mining company “broke a record for iron ore production in the second quarter,” representing a “12.6% increase compared to the same time frame in 2013.”

Several pages quickly pointed out Rousseff’s “mistake,” because she had commemorated the performance of a privatized company — an anathema for the Worker’s Party (PT), to which Dilma is affiliated, always opposed to privatizations and particularly against the sale of Vale itself in 1997. Rodrigo Constantino, in his blog on Veja magazine’s website, did not skip a beat in pointing to Dilma’s inconsistency: “Could that be a late acknowledgment that the privatization of the former state-owned enterprise, so opposed by PT, was a good idea?”

However, there was no inconsistency on Dilma’s and the government’s part, because Vale is a state company. By that, I mean that Vale, fundamentally, has never been out of Brazilian state control.

It is worth repeating so that there is no doubt left: contrary to what Rodrigo Constantino (who, by the way, wrote a book called Privatize Now), Vale is literally controlled by the Brazil government. There was clear proof of that when the company’s president Roger Agnelli was fired due to pressures by PT’s government. The occurrence, widely commented at the time, was extremely educational. It showed not only the close connection between large businesses and the Brazilian state, but also demonstrated how we have a wholly inadequate conception about the privatization process that happened in Brazil.

“Privatizations” in Brazil did not involve any transference or pulverization of power and economic control; they were effectively corporate restructurings that changed very little the distribution of economic control and altered their legal regimes only as little as necessary to make them economically viable again.

There were, evidently, technical and productive improvements; it is also evident that those were the initial goals of the restructurings, which did not include any substantial change in stock control of the “sold” companies. Brazilian privatizations were not a way to cut the government off from the control of state companies, but a way the government found to keep their control.

The electoral campaigns in 2014 feature a few candidates who wish to reevaluate the merits of privatizations. Discussing privatizations is nothing new; every four years there is a new cycle of condemnations of them punctuated by a few unfounded praises. The reality is that both supporters and detractors of the privatizations talk about imaginary processes that bear no relation to reality. Few speak of the actual process Brazil underwent: it was not a “handover,” nor “privateering;” it also was not a new era of “efficiency” and “government downsizing.” It was a reorganization of the state apparatus and the inclusion of the corporate class in its ranks.

The Privatization of Vale

State-owned companies were a failed model by the 1990s and the Brazilian state was bankrupt after a decade of hyperinflation. The privatization of state enterprises was included as one of the factors for the success of the Real Plan, the implementation of a new currency that required the “zeroing of public deficit.” This cut on public deficit had to include revenues from state companies’s sales.

The sale of Vale was the largest privatization ever made in Brazil and was the most resisted one — and, yes, PT led the opposition to it, along with a large part of the left and social movements. To sidestep protestations, the government promoted a “coalition of support” that consisted in forming new investment groups spearheaded by state pension funds. The state development bank BNDES sponsored the formation of Valepar S.A., which controls the Vale’s Administrative Council, with 53.3% of the voting capital. Valepar is controlled by four state pension funds, led by Previ, which is the pension fund of state-owned bank Banco do Brasil’s employees and the largest pension fund in the country, with 58% of the stock. Besides pension funds, Valepar is also controlled by private bank Bradesco, multinational Mitsui and by BNDES itself, which owns 9.5% of the shares.

Through the actions of BNDES and the inclusion of state pension funds, government made privatizations “viable.” From then on, the new Vale, privatized with government money, would be owned by state pension funds and BNDES. Since the beginning of the 2000s, BNDES and pension funds compose the network of control that not only runs companies that have formally been cut off from the government, but also put nominally “private” businesses (even if they were not previously state-owned) in service of the government.

Pension Funds and the Control of the Unions by the State

Pension funds, created in the 1970s to boost savings, have become the largest investment tool in Brazil. Their investment potential was over $150 billion in 2010 (16% of the GDP), and was expected to grow. Pension funds’s investments, considered as a whole, are even more significant than BNDES — which is already the largest “development” bank in the world, far surpassing the World Bank (for instance, in 2009-10, the World Bank loaned about $40 billion, less than half of what BNDES invested).

From the late 1980s on, pension funds were increasingly managed by union leaders, especially because of some reforms that happened during Fernando Henrique Cardoso’s government, which opened up managing positions to workers. Union leaders became pension fund managers. Lula’s presidential campaign in 2002 specifically encouraged workers to form those funds, not only as a means to boost their consumption levels, but also to form controlling blocks in investment positions. Those pension funds could then be controlled by the government and used as an instrument of policy to “domesticate” capitalism.

The single union system and the union tax always helped the state in that regard, because they kept the unions under government tutelage — something that has never been challenged by the PT administration. It is no wonder that, from the 1980s on, the strongest Brazilian unions (connected to the car manufacturers in the interior of São Paulo) started to adopt a stance of “propositive unionism” or “citizen unionism,” which opposes conflicts between workers and managing classes, and urges the inclusion of workers themselves in managing positions. CUT and Força Sindical, the largest union federations in the country, represent that paradigm perfectly and act as managers’s mouthpieces (see, for instance, their urges to workers in the Amazon infrastructure projects to resume working after their strikes for better work conditions).

Thus, the Brazilian legislation works to transform monopoly unions into political and economic tools. The largest pension funds in Brazil (Previ, Petros, and Funcef) are still under direct control of the government, being the funds of employees of state-owned Banco do Brasil, state oil company Petrobras, and state bank Caixa. With the conversion of union leaders (mostly belonging to the Articulation, the dominant caucus in PT) in pension fund managers, becoming a new managing class, the government secured direct access to those funds.

In 2011, Exame magazine reported on how Roger Agnelli was fired from Vale’s presidency. “Roger, wait! This is a shareholders’s matter. And it’s being discussed by us, shareholders.” It was Ricardo Flores, then president of Previ, the main pension fund in Vale’s Administrative Council, who said that before discussing whether Agnelli would be let go, giving in to pressures by the Dilma administration. Ironically, later on he was removed from Previ’s presidency due to power disputes.

BNDES: State Privatizations, Private Nationalizations

BNDES is the largest development bank in the world. It was instrumental in the privatization process and transferred formal control of about 30% of the GDP. During this same process, BNDES positioned itself as a key partner of the new companies, like Vale and others, such as the twelve telecommunication companies that were created when Telebrás was privatized. Later, those companies were unified under the name Oi and BNDES had control over 25% of its capital. To facilitate the buying of Brasil Telecom — which was another company that was created from the “privatization” of Telebrás —, BNDES made new loans. After the acquisition of Brasil Telecom by Oi, the company was 50% owned by the government, through BNDES and the three largest pension funds (Previ, Petros, and Funcef). Over 20% of the company’s shares are owned by Andrade Gutierrez, a construction company that is extremely dependent on, and sympathetic to, the government.

It is actually difficult to find former state-owned companies that underwent different processes. In fact, the stock control by BNDES and unions does not tell the whole story. In the 1990s, Brazil went through a process of regulatory capture by design. Following privatizations, regulatory agencies were established for the new sectors the state had “left.” It was the first great moment of transit between the government and large corporations. With the subsidies to privatization processes, the new business and shareholders classes secured not only access to productive capital, but also to the state in its regulatory instances. It was an almost instantaneous process in the case of telecommunications.

Hence, “privatizations,” far from cutting the state off productive resources, were in fact only an organizational reconfiguration of capital. Capital formally left the scope of the state, but remained under its effective control, changing its legal regime without greater economic consequences. It is not a matter of ascertaining whether the capital that was “sold off” during the 1990s had taken a crony role; rather, the capital remains a part of the state, as it is controlled directly (through BNDES and pension funds) or indirectly (through the regulatory apparatus of joint control by the government and the corporations) by it.

The opposite process also took place very quickly during the PT administration (especially during the aftermath of the 2008 financial crisis) and is still underway. BNDES started to inject capital in private corporations and to elect its political-economic branches. This included the Sadia-Perdigão merger, the acquisition of Aracruz Celulose by VCP, the Friboi-Bertin merger, the Ambev buyouts, among many others. Construction companies are also policy branches of the Brazilian government. Odebrecht, in particularly, is a PT ally since 1992 and has repositioned itself in countless infrastructure and military endeavors. Other companies, such as Andrade Gutierrez and Camargo Corrêa, that have had their growth inextricably linked to national infrastructure projects, are nowadays nothing if not policy enacting arms of the state. The government has a toolbox with contracts as well as direct stock control to influence the “private” sector in Brazil.

In reality, it is a mistake to consider that large conglomerates in Brazil are “private” or “state-owned.” It is a meaningless distinction in this context; privatizations created mixed conglomerates controlled both by private and state interests, whereas companies that were already private have a level of government influence large enough that their interests and the government’s interests are interconnected. There is no opposition between “particular” and “public,” between “private” and “state,” because there is a convergence of the interests and desires of large corporations and the state that unite them.

The Vocabulary of Privatizations

Both supporters and detractors of the privatizations tend to defend their stances for the wrong reasons.

The technical and service improvements that took place after privatizations were not due to fundamental changes in capital control. They were reforms that modified organizational structure and the incentives of “public” companies, improving their capitalization and rationalizing their actions. The improvement in performance of the former state-owned companies did not take place due to denationalization, which did not occur, but due to restructuring. (In the same way, there was an improvement in performance and capitalization of Petrobras, which was not privatized. Privatizations, and the public offering of Petrobras’s stock, can be seen as new capitalization strategies rather than a dial back in state power.)

Our language reflects a duality between “private” and “state,” between “privatizing” and “nationalizing,” that is simply untrue. These dichotomies have no explanatory power because the state is not limited by its formal range of action, and because the state is not an insurmountable barrier where private businesses cannot venture. We can take as an example the trajectory of the former Minister of Development, Industry, and International Commerce Luiz Fernando Furlan, who pressed for the merger of food companies Sadia and Perdigão. Furlan left Sadia to join the government. After the merger, he left the government and resumed his presidency of the Administrative Council of Sadia.

“Privatizations” are a smokescreen, because they were only a revolution inside power, facilitating the maintenance of state control over vital sectors of the economy. It is impossible to complain about government interference in private companies; the large business owners in Brazil are part of the state. Vale is a political and economic branch of the state; its privatization and capitalization were structured with this goal in mind. When the head of Vale became an obstacle to the ends of the Brazilian government, he was fired.

Our language is unprepared to reflect this lack of differentiation between the public and the private. It is also difficult for most people to think about government and large conglomerates as part of the same system.

Moreover, we tend to treat what is “state-owned” as something public, and “private” resources as something out of the reach of the public. It is perfectly plausible, and indeed it is what happens in most cases, that a private resource can be “private;” that is, it is perfectly possible (and, I would argue, inevitable) that state resources serve the interests of only a small caste. The expressions we use are so completely inadequate that we speak of “nationalizations” when talking about state companies and “handovers” when speaking of privatizations. The political and economic experience of Brazil proves that these are all inadequate terms and that we should develop a vocabulary that represents reality such as it is: where state or “nationalized” companies serve the interest of a few state-connected groups and where private companies and the government possess converging, rather than antagonistic, interests — and both are contrary to general welfare.

Our political ideas are only prepared to deal with broad generalizations that stipulate that the government and the private sector are categorically distinct and that their influences over each other are a few small deviations — we tend to think that, in the majority of cases, government and businesses stick to their ideal roles. Privatizations, according to this mode of thinking, served to take companies out of the control of the government and put resources in a sphere over which government would have no influence. Even though people generally recognize forces that act in the relationship between government and businesses, most are likely to adopt this naive and a-historical view when analyzing processes and defending their political ideas.

The fact remains: privatizations were not a reduction, but a way of extending and rearranging state power. The pro-privatization discourse, thus, defends them in these terms, not under ideal conditions. The opposite is true as well: opponents and critics of the privatizations think of them as a decrease in state power. But if the privatized companies are still under government control, what could be the problem?


Any pro-privatization campaign in Brazil, such as some that have surfaced during the electoral period, must take into account the following fact: the Brazilian state and large corporations are one single entity.

This means that the privatizing effort has to take into account the presence and influence of the state as a fundamental fact. “Privatizing,” thus, does not mean a radical change in the state’s power structure, but slight adjustments in their legal regimes to provide for the capitalization of companies that, ultimately, remain under the state umbrella. Thus, both the idea of privatizing as well as the thought of nationalizing are founding ideologies of the state.

It should be obvious that privatizing, in itself, is not a passport for the dismantling of the state; in Russia, for instance, the same Soviet elite took control over the “privatized” resources during the transition to capitalism.

In Brazil, government control over large national corporate “privatized” conglomerates and even over companies that have always been nominally “private” did not happen by sheer chance nor did it suffer internal resistance; the business class was always a willing partner in this relationship. There has been, in the last decade and half, especially, an alignment of the views of the PT elite and national bourgeoisie. This alignment also included an incorporation of the old nationalism supported by the military elite, who is comfortably fortified and represented inside the government (despite the claims of several conservatives, who think the military is ignored and humiliated by the current regime).

Brazil has developed a quasi-fascistic system of systematic subsidies to capitalists, and direct and indirect control by the government of businesses and unions (which have given rise, through pension funds, to a new capitalist class themselves).

Conventional right and left-wing criticisms to this system are misguided because they end up defending a different aspect of the same system while they attack it. A defense of privatizations, for instance, can serve as a criticism of government power, but if executed as it was in Brazil, it extends the power over businesses and capital the government possesses.

Allies and Enemies

Privatizing is not enough. The corporate sector and the government are one single class. The deregulation that took place in the country were not able to break state influence over the economy, but have simply changed their character. Our political vocabulary does not reflect well real political issues, because it pits as polar opposites categories that are not fundamentally distinct: private and state-owned, corporations and government. Real opposition exists between those who have power and those who do not.

As I mentioned in two previous articles on the union actions in Brazil, there is an articulation taking place currently in Brazil between the business sector, the state elite and union leaders. They have formed a new managing class that represents the individual’s aspirations and decides on the sharing of the economic pie. The only way to resist this reality — that, yes, has been created by privatizations — is with the perception that the ruling class is not limited to a categorical sector of “businesspeople” or “bureaucrats.” It is a mixed class, with free transit inside government, unions and administrative councils.

After the newest billionaire corruption scandal in Petrobras, some have already spoken of the need of privatizing it and taking it away from the political sphere. But what we should remember is that Brazilian privatizations were never intended to strip the state of its power.

Public and private, capital and labor are now not opposites, but allies. Thus it is not surprising that Dilma should celebrate the 12.6% increase in iron ore production.

The one who paid for that record was you.

Anarchy and Democracy
Fighting Fascism
Markets Not Capitalism
The Anatomy of Escape
Organization Theory