Mining shares a close relationship with Brazil’s economy. This relationship has existed at least since colonial times, when in the seventeenth century the Portuguese discovered gold in the state of Minas Gerais and quickly set up infrastructure to recover the resource. Indeed, mining even justifies the state’s name to this day––Minas Gerais translated to English means “general mines.” Today, mining in Brazil involves much more than gold, silver, and precious stones. It is equivalent to 5 percent of Brazil’s gross domestic product, creating 180,000 direct jobs and resulting in the extraction of a vast range of minerals. Vale, Sobre a Vale (2017), www.vale.com/brasil/pt/aboutvale/news/paginas/qual-a-importancia-da-mineracao-para-a-economia-do-pais.aspx.
Despite heavy reliance on and a long history with the mining industry, mining remains a very dangerous activity and involves significant environmental impacts. These risks came to a head on January 25, 2019, when a mining dam owned by the Vale company and located in Minas Gerais collapsed, releasing a mining tailings mudflow of approximately 12 million cubic meters. This massive human-caused mudflow immediately destroyed Vale’s local facilities and nearby communities in Brumadinho. Semad/MG, Clarification Note 1—Disaster Damage B1, Jan. 25, 2019, www.meioambiente.mg.gov.br/noticias/1/3734-nota-de-esclarecimento-brumadinho-.
This disaster, known as the “Brumadinho Disaster,” caused unprecedented environmental damage to the entire region by contaminating soil, impairing important water resources, and destroying millions of hectares of forests. Even more tragically, the disaster killed 203 people and left another 103 missing, making it the worst mining disaster in Brazil’s history.
While the Brumadinho Disaster caught global headlines and created a sense of concern around the world, the disaster was not the first of its kind in Brazil. For example, on November 5, 2015, the Fundão dam, also owned by Vale and located in the state of Minas Gerais, collapsed in what is known as the Mariana Disaster. That event released a mining tailings mudflow of 43 million cubic meters, destroying the city of Bento Rodrigues, killing 19 people and causing similar severe environmental damages. Ibama, Breaking of the Fundão Dam: Documents Related to the Samarco Disaster in Mariana/MG, Oct. 23, 2018, www.ibama.gov.br/cites-e-comercio-exterior/cites?id=117.
Thus, in the aftermath of the Brumadinho Disaster, Brazilians were forced to grapple with the question of whether their environmental laws will be effective in holding mining companies like Vale responsible for protecting their workers and the environment.
As to Brumadinho, the authorities are still assessing the best approach to redress damages and investigating possible punishments. And while some have worried that little has been done to punish the violators, actions were taken after the Mariana disaster. Specifically, in June 2018, the federal public prosecutor and the corporations associated with the Mariana Disaster (i.e., Fundão Dam) entered into a settlement agreement that created the Renova Foundation. This foundation was established with the purpose of managing and assisting in numerous restoration actions, and collectively this settlement identified 42 projects to be funded. See Fundação Renova, www.fundacaorenova.org.
However, neither Brumadinho nor Mariana is likely to be the last major collapse as the Barão de Cocais Dam, also located in Minas Gerais and owned by Vale, presents severe stability problems with the risk of collapsing at any time. Smaller mining disasters are also relatively common.
Even if Brazilian mining companies are not going unpunished for these major catastrophes, it is alarming that both massive mining disasters occurred within such a short period of time in Brazil. It is little wonder then that people are concerned about whether Brazil’s environmental mining regulations are fundamentally flawed or, alternatively, whether compliance and enforcement are ineffective such that the question is not “if” but “when” the next disaster will occur.
This article sets out to reflect on whether Brazil has a problematic regulatory framework or otherwise lacks capacity to enforce the law by giving an overview of the regulations most relevant to the Brumadinho case and identifying possible problems with that framework. With this assessment, suggestions will be made for mitigating the problem and insights provided regarding what to expect as a result of these unfortunate events.
Environmental Permitting in Brazil
Environmental law in Brazil began with the National Environmental Policy Act (Política Nacional do Meio Ambiente—PNMA Lei No. 6.938/1981) in 1981, which set forth a system known as the Sistema Nacional do Meio Ambiente (SISNAMA) to address environmental protection. One of the several purposes of the law was the creation of a federal environmental agency, Instituto Brasileiro do Meio Ambiente (IBAMA), and state environmental agencies. Each of these agencies have their own respective jurisdictions to enforce environmental regulations and issue permits. As in many systems, permitting was established as a primary instrument for environmental protection.
Additionally, Resolution No. 237/1997 of Conselho Nacional do Meio Ambiente (CONAMA) details the general environmental permitting instruments that apply in the country. Specifically, environmental permitting in Brazil consists of three consecutive phases, each phase corresponding to the issuance of a different permit. These three permits are issued according to the stage of development of the activity proposed. First, the Preliminary Permit, which is granted during the planning phase, serves as approval for the activity or project. This stage also includes consideration of environmental feasibility through an Environmental Impact Assessment, which is similar to the assessment produced in the United States under the National Environmental Policy Act. Second, the Installation/Construction Permit authorizes the activity’s construction. Third, the Operation Permit authorizes the activity’s operation.
In Brazil, an agency’s authority to issue permits, enforce regulations, and monitor activities arises from its basic, plenary police power, which means that all levels of government can determine whether it is necessary to regulate an activity. Constituição Federal, art. 23, Sc. VII (Braz.). However, the CONAMA Resolution and Federal Law No. 140/2011, which were enacted with the goal of better organizing and clarifying such common authority, established that only one level of government can issue particular permits. In most cases, these permits are issued by Brazilian state agencies, though for certain matters, such as when two or more states are involved, or when the permit relates to nuclear technology, military facilities, or national borders, IBAMA has jurisdiction.
On the other hand, when it comes to the authority to legislate on issues of environmental protection, Brazil recognizes the concept of concurrent authority. Constituição Federal, art. 24, Sc. VI (BRAZ.). Therefore, similar to American federalism, the federal government has the authority to establish and enact a minimum standard of protection and procedure, and the states can supplement those standards established in federal statutes by enacting even stricter standards. The states may also freely legislate in the event of federal inaction. Constituição Federal, art. 24, §§ 2, 3, 4 (BRAZ.).
The authority responsible for permitting Vale’s operations, including the dam that collapsed in the Brumadinho mining disaster, was the state of Minas Gerais through its state environmental agency (Secretaries do Estado do Meio Ambiente e Desenvolvimento Sustentável (SEMAD)). But, as is often the case in the United States, the state was acting under the guidance of various federal standards. If one simply looked at the paperwork of inspections required by law before the disaster, one would not necessarily suspect something was wrong.
According to SEMAD, the dam’s Operational Permit had had its last renewal performed in 2011 with no irregularities and the permit was valid until 2017. Additionally, inspections indicated that the dam was not receiving any more waste or discharge. Therefore, in 2018 Vale filed a request for a permit to discontinue its operations at the dam and to reuse those materials. Since decommissioning can itself lead to pollution, the activity was classified as a low environmental impact and, under state regulation, did not have to go through the full three-phased permitting programs discussed above. Instead, the discontinuance permit was subject only to a single phase of “simplified permitting.” Thus, at the time of disaster, the mining dam was permitted to be decommissioned. After the accident, SEMAD claimed that this procedure did not affect the quality of the permitting process.
National Dam Safety Policy Act (Política Nacional de Segurança de Barragens)
In addition to the permitting process established by Brazil’s National Environmental Policy Act, Federal Law No. 12.334/2010 enacted a national policy to ensure the safety of dams. In short, it sets forth a system of dam risk classification, listing assessment, and reporting instruments, and authorizes the National Mining Agency, Agência Nacional de Mineração (ANM), as a competent authority to monitor mine dam risks.
Thus, the ANM is responsible for regulating private companies to ensure they are properly following safety measures, including the implementation of various required safety reports and inspections. To allow ANM to accomplish its mission, there exists a system of shared responsibility between the federal agency and the private sector, in which the private corporations generate the monitoring data upon which the ANM evaluates compliance.
On the day of the Brumadinho Disaster, ANM through a press release stated that there were no pending documents regarding the dam monitoring, and it was classified as “low risk.” Also, it stated that according to the reports presented by Vale, no irregularities with the dam stability and safety were detected. See Agência Nacional de Mineração, www.anm.gov.br/noticias/nota-a-imprensa-1, last modified Jan. 28, 2019. However, recent investigations performed by SEMAD and a Congress Inquiry Committee have suggested that Vale might have ignored data demonstrating potential stability risks of the dam and then possibly passed false information on to ANM through the monitoring report. Congresso Nactional, CPI Relatório, https://legis.senado.leg.br/sdleg-getter/documento?dm=7974085&ts=1562172217942&disposition=inline.
Hence, despite having a system in place to identify and address risks such as failing tailings dams, the system clearly did not work in the case of Brumadinho. The reliance of regulators on the private sector is one of the reasons many Brazilians have begun to question the system in the aftermath of the disaster. Questions have also arisen because the dam was permitted by SEMAD while the monitoring inspections must be attested by ANM, which is a federal agency. This gap between state level for permitting and federal level for monitoring might also be problematic.
Once environmental damages occur, Brazil holds violators liable through three independent and different spheres: civil, administrative, and criminal. The strict, joint civil liability applies to all persons responsible, directly or indirectly, for the activity that causes the environmental deterioration irrespective of fault, resulting in the obligation to indemnify or remedy the damage caused to the environment and to third parties. In addition, the obligation to redress damages has a retroactive effect for all potentially responsible parties (Federal Law No 6.938/1981, art. 14§ 1º). In other words, the law applies somewhat similarly to the United States’ Comprehensive Environmental Response, Compensation, and Liability Act.
Administrative liability applies to parties that engaged in building, installing, or operating potentially polluting establishments, projects, or services, without permit or authorization from the proper environmental bodies, and subjects the violators to punishments such as fines, permit suspensions, and embargoes. Criminal liability applies to a company or person that commits infractions characterized as crimes under the Environmental Crimes Law (Federal Law No. 9.605/98), and subjects the infringing parties to criminal sanctions such as prohibition from doing business with the government and detention, among other penalties.
It must be stressed that unlike civil liability, the applicable criminal and administrative laws do not apply strict liability; there must be demonstrated fault, negligence, or recklessness by the infringing parties. Thus, the system is designed relatively similar, though not exactly like, the United States’ general system of enforcement in statutes such as the Clean Water Act.
Possible Outcomes of Brumadinho Disaster and Redressability
A range of legal outcomes has started to play out in the wake of the Brumadinho Disaster, and more can be expected to play out in the coming years.
With respect to administrative liability, on January 29, 2019, IBAMA issued four fines against Vale with a total of approximately $60 million USD for (1) causing pollution to the environment and human health, (2) making urban or rural areas unfit for human habitation, (3) causing water pollution and interruption of water supply services, and (4) polluting effluents and causing biodiversity loss. See Ibama, Ibama Fines Vale in R $250 Million for Catastrophe in Brumadinho, Jan. 26, 2019, http://www.ibama.gov.br/noticias/730-2019/1879-ibama-multa-vale-em-r-250-milhoes-por-catastrofe-em-brumadinho-mg. Additionally, SEMAD also issued five fines against Vale for similar reasons amounting to approximately $25 million USD. Recently, in May 2019, a new fine of $80,000 USD was issued for possible false information on the monitoring report given to ANM. See Semad/MG, Infractions—Disaster Damage B1, Jan. 25, 2019, www.meioambiente.mg.gov.br/component/content/article/13-informativo/3750-autos-de-infracao-desastre-barragem-b1.
As to criminal liability, at present it is hard to predict whether criminal sanctions will be pursued in the Brumadinho case. As to strict civil liability, in Brazil the public prosecutor has standing to bring public legal claims, with the option of seeking collective moral damages (i.e., the Brazilian version of punitive damages). Due to the extent of the destruction that resulted from the Brumadinho Disaster, the state public prosecutor recently filed a civil action against Vale to redress the environmental damages seeking about $12.5 billion USD for emergency measures to mitigate and compensate the environment. ACP No 5000056-68.2019.8.13.0090.
Because Vale is responsible for two severe accidents in similar circumstances (Brumadinho and Mariana), there is a strong case for bringing a claim for more punitive damages and new civil actions in the future. It is important to point out that since the National Dam Safety Policy sets a shared responsibility system of monitoring between the entrepreneur and the National Mining Agency, according to the strict, joint civil liability, the agency can also be held liable.
Additionally, compensation for the deaths involved and socioeconomic impacts for local communities can be asserted in specific actions. To date, there is one ongoing civil action filed in April 2019 by the state public prosecutor addressing the socioeconomic damages and requesting collective moral damages up to $12.5 billion USD (i.e., ACP 0001827-69.2019.8.13.0090). See Ministério Público de Minas Gerais, www.mpmg.mp.br/comunicacao/noticias/caso-brumadinho-procedimentos-de-investigacao-instaurados-e-acoes-ajuizadas.htm.
Historically many mining disasters have been handled through settlement agreements between violators and the public prosecutor, as was the case with Vale’s earlier Mariana disaster. While these agreements have the advantage of allowing faster avenues to redress violations and help the families of victims, they also generally fall short of holding companies fully liable. Thus, these settlements might be viewed as falling short, in part because they have not had the desired deterrent effect.
Regulatory Framework on Debate
Because the Brumadinho Disaster happened in large part due to the inability to detect safety stability problems during dam inspections, the National Dam Safety Policy (Federal Law No. 12.334/2010) has become a center of attention and debate. Its application to upstream tailings dams, which was the type of dam that failed in both the Brumadinho and Mariana accidents, has faced scrutiny. An upstream tailings dam involves a type of construction in which the dam body (i.e., wall) is built using the deposited tailings. Water is then drained away, so that the waste sludge hardens and makes a tailings shell that does not need to be reinforced by concrete or stone. As such, this approach represents the cheapest and fastest option. Problems arise, however, if the sludge gets wet, leading to liquefaction that makes the dam susceptible to cracks and eventually collapse. Given this problem, and in light of recent accidents, in February 2019, ANM issued Resolution No. 4/2019 banning the use of new upstream tailings dams and requiring that all remaining 84 must be deactivated by 2021. See Agência Nacional de Mineração, www.anm.gov.br/assuntos/barragens/pasta-classificacao-de-barragens-de-mineracao/plano-de-seguranca-de-barragens.
Despite being phased out in Brazil, other companies in other countries still use this method. However, due to the risks involved the dams need constant monitoring, which in the Brumadinho case clearly was not effective. This failure highlights problems with the National Dam Safety Policy Act, which relies on private companies to perform all inspections and generate all monitoring data. While it is not ideal to have companies self-regulate, in most cases Brazilian agencies do not have the resources to perform official in loco inspections and thus conduct independent inspections to check the data. Given recent events, it appears that there are some flaws and irregularities occurring with this reporting procedure.
Ever since the Mariana accident, Brazilians technicians, scholars, and legislators have been discussing whether this policy should be improved or amended (e.g., Bills 18, 20, 109, 110, 184, 188, and 336/2019). For example, some have asked whether there should be additional defined punishment for corporations and prohibitions against new upstream tailings dams, a step already taken by ANM. The issue of how to craft an effective regulation has long been discussed in Brazil, with the main question being when to punish versus when to incentivize? In other words, which would be the best approach––harsh punishments to act as a deterrent to bad behavior, or persuasive benefits to obtain beneficial compliance? According to some scholars a combination of both is appropriate. John Braithwaite, To Punish or Persuade (1985).
A responsive regulation is a regulatory framework that escalates the intensity of government intervention in order to enable enforcement while adapting to the situation. This framework consists of a pyramid structure combining persuasion with command and control instruments, with the base of the pyramid consisting of instruments that encourage self-regulation and, where there is no cooperation from private actors, an escalation to the top of the pyramid through harsher punishments.
A preference for persuasion over command and control occurs because even though the latter approach often appears to be more effective, it is more expensive than self-regulation instruments. The additional cost stems from the need for more personnel and administrative infrastructure. Also, for a system strongly based on command and control to be productive, the regulator must have the capacity to escalate the punishments, creating a deterrence effect by making it cost more to break the law than respect it. This effect can be problematic and difficult to achieve where there are institutions and agencies with fewer resources and little history of full or aggressive enforcement.
Meanwhile, a system based only on self-regulation instruments rests on the belief of nurturing the virtuous citizen, but this approach does not always represent reality. It can be difficult to obtain desired outcomes using such a self-regulatory system even when the rewards for compliance (e.g., tax benefits) appear significant on the surface.
A responsive regulation takes from both and can create a more balanced and efficient system––“it comes up with a way of reconciling the clear empirical evidence that sometimes punishment works and sometimes it backfires, and likewise with persuasion.” John Braithwaite, Responsive Regulation and Developing Economies, 34 World Development 884, 887 (Elsevier 2006) (doi: 10.1016/j.worlddev.2005.04.021); see also Ian Ayres & John Braithwaite, Responsive Regulation (1992). It must be noted, however, that responsive regulations can be more challenging for economies in Brazil’s current position to implement effectively; as scholars have pointed out, such economies often have more difficulty implementing this type of system because they have less regulatory capacity and resources, and are thus often more susceptible to corruption. Braithwaite, Responsive Regulation and Developing Economies, supra at 896.
One proposed solution to this problem that may work in Brazil is a networked governance strategy aimed at minimizing these deficits. This strategy relies on weak actors or regulators in a system becoming stronger by joining forces. In other words, the system must enable an environment of pluralism in which several other players can assist public authorities in an affirmative way, not simply by reducing corruption, but also by addressing the costs associated with infrastructure and monitoring. Such an approach would make the most of the limited resources available in Brazil. Third parties such as nongovernmental organizations (NGOs) can have a fundamental role in this strategy. Another front is to encourage whistleblowing among private actors to help identify problems in sectors that have capacity deficits. Id. at 890–893.
Regarding the Brumadinho Disaster and the National Dam Safety Policy Act, ANM’s prohibition on the use of new upstream tailings dams due to their high risks is a positive step. But the initiatives to amend the existing law that only impose higher punishments will not necessarily bring results. Strengthening command and control instruments without increasing the regulators’ capacity to actually enforce them is ineffective. Because Brazil has limited capacity and resources for monitoring and enforcement, as well as a high rate of litigation in courts and countless types of appeals, this approach alone likely will not prevent future accidents simply because enforcement is not robust in the first instance.
Brazil is a developing economy, and a networked governance approach could be helpful. In particular, regulators could benefit from the assistance of third parties for inspections and monitoring. In this sense, state agencies such as SEMAD in Minas Gerais and NGOs could help enforce the policy by assisting inspections and verifying the data. In the case of NGOs, this could include the use of name shaming for corporate activities that indicate procedural irregularities, and in the case of state agencies, this could include reducing the information gap between state-level permitting and federal-level monitoring. This networked governance could be positive during the period of deactivation of the other upstream tailings dams until 2021 and perhaps even prevent disasters by overcoming ANM’s lack of resources, personnel, and capacity to monitor. This approach could be applied similarly to other environmental regulatory challenges faced by Brazil.
Ultimately, Brazil has a wide range of environmental legislation, including for permitting, which is not only established in the law but also at the constitutional level. Unfortunately, as the Brumadinho Disaster illustrates, these systems are not always effective at protecting the environment or preventing disasters. Indeed, because Brumadinho is the second big mining accident in the country in recent years, it is likely that monitoring standards have been lacking or not properly executed. This directly leads to an inquiry into whether the National Safety Dam Policy can in fact ensure safety, especially as it relates to the method of dam construction.
The ANM decision to deactivate all upstream tailings dams is a positive step; however, it seems more of an aggressive response to try to cover and mitigate a deeper problem in Brazil such as the lack of resources and capacity to enforce the law. This policy has generic wording and relies upon a shared monitoring procedure between the entrepreneur and the competent authority, thus ANM is only informed about the monitoring data assessed by the entrepreneur, and not by the agency’s own inspections.
In this sense, reforming the existing policy toward a more responsive regulation may be appropriate by mixing self-regulation with command and control to make it more efficient. Such an approach could use a networked governance model that utilizes third parties such as NGOs and state agencies. Ultimately, this type of system would enable a smarter regulatory framework and also minimize capacity deficits and corruption.
Finally, concerning redressability of the damages already caused in Brumadinho, the possibility of another settlement agreement as happened in Mariana, must be carefully considered. The criminal, administrative, and strict civil spheres of environmental liability must be properly and fully enforced, and the regulated community must in fact be subject to an escalation of punishments. All environmental damages must be fully, not partially, redressed, to discourage possible future negligent or reckless behavior by mining companies.