Consultor Jurídico


Negotiated punishment

"Agreements can't be turned into the cornerstone for conviction in corruption cases"


If you are the executive of a large corporation who happens to be accused of a crime under the Anti-Corruption Law or antitrust laws, you will need to prepare yourself to face a few years of wading through government bureaucracy. But if you are the authority in charge of the investigation and prosecution, you better be prepared to use the agreement programs for the benefit of the criminals themselves, or for the overlap of powers of government entities to serve as discouragement for the informants to keep from collaborating with the investigations.

According to lawyer Victor Rufino, an expert in Competition Law and antitrust laws, such are inevitable consequences of investing in agreement programs. The best remedy, in his view, is not to choose only one strategy to trust for any of the parties involved in a negotiation, whether it is a plea bargain, or a leniency agreement, in the case of companies.

Rufino knows what he's talking about. Former federal prosecutor, Rufino was an attorney for the Brazilian Antitrust watchdog (Cade) for 11 years, and also became the general prosecutor of the entity. He was one of the people in charge of the council's leniency program, the acclaimed Brazilian antitrust entity.

He is currently pursuing a doctorate in Competition Law from São Paulo University (USP) and is an associate at the law firm Mudrovitsch Advogados. He offers the following recommendations to investigators: "Do not make the agreement, which is a means of investigation, the sole and only means of proof, the cornerstone for conviction."

To the lawyers involved, he explains that the tendency is for institutions to become stronger and more articulate. The solution is to articulate the defense as well, and structure it in several different areas. Despite the differences that exist in investigative techniques in the criminal, competition, and misconduct areas, "the defenses must be in harmony, they must speak the same language." "The attack is articulate, the defense has to be also."

Read the interview:

ConJur — The most important matter when discussing the leniency program is effectiveness. The law does not speak of the role of the Public Prosecution Office, but how can an agreement be reached without a guarantee that the prosecutor will not later use that information against the employee?
Victor Rufino —
 It's not explicit in the law, but there is already a project to make it explicit. Now, the fact that it's not explicit in the law does not mean that it's not legally required. It's a discussion. But today at Cade all agreements are merged together with the help of the Public Prosecutor's Office. It's also apparent today that the Public Prosecutor's Office enters into some leniency agreements and the challenge comes from other entities, like the Federal Attorney General's Office (AGU) and the Federal Comptroller General's Office (CGU), claiming that "there will be no agreement if we're not together." Not knowing what the sphere of protection is ends up being a problem in an agreement. Back in the day, Cade and the Public Prosecutor's Office realized that it was better to join forces in cartel agreements so the collaborator could know that the promised protection would be delivered. It's important for him to know this before making the decision to make a deal.

ConJur — There has already been an example of a leniency agreement that involved all the players, which was that of the advertising agency. It took three years to get signed though. Has this problem already been solved at Cade?
Victor Rufino —
 Cade needed not only the law, but also the time. Today we are reaching the five-year milestone of the "lava jato" (car wash) operation and there is a consensus now that the prosecution of financial and white collar crimes has now reached another level in Brazil. There are two factors. The law needs to clearly define the roles of each and what the procedures are, and the institute needs time to mature. But it's safe to say that the anti-corruption system today is no longer the same as it was five years ago. We now have a considerable critical mass of laws, regulations and decrees. Not every definition needs to come from law, but it's better if it does, to make it more stable.

ConJur — In view of all this experience, and based on your own experience, what needs to be considered for a good leniency system?
Victor Rufino —
 In my master's dissertation I analyzed the program of Brazilian antitrust laws. There are three elements that are essential: predictability regarding requirements and procedures; existence of degrees of secrecy and stability. It's important for these three things to be well defined in the abstract, with little room for subjectivity.

ConJur — An issue we commented on before starting the interview was the interdisciplinarity of crimes. Can you explain this better?
Victor Rufino —
 As much as we talk about corporate crimes, the same act is subject to several spheres. If you engage in a cartel with a competitor to defraud a bid and, along the way, you paid a government agent, you are subject to criminal prosecution, as an individual as well as a legal entity of your company. An act of corruption, impropriety, fraud to competition has been committed and you may also suffer civil penalties. The Federal Attorney General's Office may demand the money back, for example. So it's an act that places you within five different spheres. The main problem today is how to impute to an enterprise the actions of an executive.

ConJur — …who often acts without the company even knowing.
Victor Rufino —
 That's one of the key issues the world is trying to solve. Brazil is not an island in that. Anti-corruption and antitrust laws come from international commitments undertaken by Brazil, because if you committed an illegal act and the money resulting from such act goes through the international financial system, you need to have a dialog with authorities in different countries. Operation "lava jato" itself brings to mind dozens of examples of this.

ConJur — Now the inevitable question is: how to defend oneself from such an integrated and internationalized system? There is no point in hiring a criminal lawyer anymore - you have to defend yourself on several fronts.
Victor Rufino —
 A solution that is being adopted by all is to create stricter compliance systems in order to navigate through this jumble of laws that are based on different criteria. There is an interest in preparing and understanding in advance the problems that may arise.

ConJur — And how does one equate this with the company's target, which is to make a profit, and not obey the compliance programs?
Victor Rufino —
 It's hard. All these wrongdoings go in the direction of the legitimate interests of the company. Adultery, for example, goes against the legitimate purpose of a monogamous relationship. A white-collar crime, however, is perpetrated with the same goal of legitimate company business in mind. It's an identification problem. Particularly, in competition cases.

ConJur — Why?
Victor Rufino —
 Because the law says that it's okay to try to destroy your competitor, as long as you do not use unfair mechanisms. But it's difficult to set the boundaries, it requires talking constantly to Compliance, with preventive legal representation in place, to keep from making a slip or being unjustly accused of willful misconduct. So it's necessary to speak these different languages. Today, most of Cade's search and seizure operations in the fight against cartels is a result of leniency agreements signed with the Federal Prosecutor's Office. And this kind of order can only be issued by a judge. Therefore, information to support a criminal, as well as a competitive proceeding, is now being requested in the same operation. Fairly or unfairly, you are being accused in two different spheres and the defenses, although the techniques are different, have to be in harmony, they must speak the same language, because the world requires it today. The attack is articulate, the defense has to be also.

ConJur — We recently interviewed a former attorney of the United States Department of Justice and his reccomendations were very clear: "If anything happens in your company, be prepared to collaborate because you have no defense".
Victor Rufino —
 The US has a system in which almost everything is settled by agreement. At the federal level, more than 90%. Brazil is still introducing systems of agreement - today more than yesterday, but we're still a long way from the US, which is not necessarily the most desirable. There is much questioning in American doctrine today about if this system is going too far. The fact now is that in Brazil today you can consider making a deal, whether it's convenient for the defense or not. It's not the ultimate goal of a defense, but it's a strategy that has to be considered. If you go for an agreement, you have to know exactly what you're doing because of those five fronts I mentioned. It's impossible to have a problem in Brazil today that starts and ends with a single authority.

ConJur — That's what I was trying to get at: with so many fronts and with such articulate and rigid authorities, defense has become a rather cruel job, has it not?
Victor Rufino —
 I don't think that the act of being investigated automatically means that you need to raise the white flag and go for collaboration. It depends on each case. But of course, knowing that there are several areas that you have to turn to, it's best to be well prepared. If we look at the control laws in place in Brazil since the 90s, the problem started becoming less purely criminal with the attributions of the Federal Audit Court, the Federal Attorney General's Office (AGU), the antitrust authority (Cade) ... The arsenal today is heavy and will tend to get more so because, as the authorities talk more, it will become more robust.

ConJur — So this so-called classic legal path, of those who refuse to make a plea bargain, for example, is losing ground?
Victor Rufino —
 The question is this: if you're going only for the criminal argument, you might even get a good outcome for the individual. But will it be a good solution for the company? The company wants to continue. It has large and expensive operations, thousands of employees, so solving the issue of an executive's freedom is only a part of the problem.

ConJur — In a certain section of your dissertation you say that the prisoner's dilemma, something that is in vogue at the moment especially among those involved with operation "lava jato," is not enough in competition cases. Is that right?
Victor Rufino —
 Every theory of collaboration speaks of the prisoner's dilemma, which is a consequence of the game theory and its rationale is mathematical. Professor John Nash, who developed it, needed an example to explain it to the students and created this dilemma. Basically he says that in a circumstance where there is an authority with the ability to punish and there are several people arrested, accused of the same crimes, he will help those who collaborate first with the investigations. And the collaborative programs were designed to follow this logic.

ConJur — And why is this model not efficient?
Victor Rufino —
 What I say in the dissertation is that there are other factors that weigh on the choice of those who will collaborate, not only the prisoner's dilemma. The choice based on the dilemma is made in ideal conditions, it's a hypothetical, didactic example. But if a suitable environment is not created for ideal conditions to develop, the natural state of those who have committed an antitrust crime is not to collaborate, not to give in.

ConJur — Why?
Victor Rufino —
 In practice, we see that the bonds that unite people are very strong. Sometimes they are religious bonds, or are people who have attended the same college, or know each other for many years, etc. They are human beings.

Secondly, they are people who are gaining from that situation, and to commit the same crime together they had to overcome many of the barriers that sometimes alienate people. So they trust each other a lot, they trust that the other guy won't talk, so he won't talk either. To break this bond is a very difficult decision.

Then we have the following situation: the bonds are strong, the trust is great, the gains from the crimes are high. The prisoner's dilemma then becomes "I won't say anything because I know he won't say anything either and we'll continue to gain together."

ConJur — What should be done then?
Victor Rufino —
 In order to have a program that is stable, other variables must be considered because the bonds will not be broken if the gains continue for everyone. What should be put on the table is an advantage that is so great that the guy considers it worthwhile to collaborate instead of continuing to commit the crime and the authority shows that if nobody cooperates and everything is found out anyway, it will be worse for everyone. In order to have a solid leniency program, the authority needs to conduct efficient investigations, he needs to show the accused that he will get the information on his own, but he can be helped and the prisoner can benefit from this.

ConJur — The moral issue enters the equation too, doesn't it?
Victor Rufino —
 The moral dilemma is one of the things that most favors the program because there are many people who want to get out of the scheme, but can't. What do they do? Confess and be arrested? Now if there's a way for people to say what's happening with some degree of protection, if the state shows that there's a way for the executive to speak, "the internal investigation has encountered this problem and there is a solution that is not to hide it," it's a different matter. If the agreement leaves the table, the director will go to the board and say "let's confess and pay a fine that may reach 20% of our sales revenue" or should they try to hide the problem?

But there's still the personal cost, of those who decide that they prefer sleeping at night rather than earning more and they confess. There is the pragmatic decision, of the executive who feels he will be excluded from the cartel and decides to confess before he is harmed. In short, there are several situations that can lead to collaboration, and for the authorities it doesn't much matter which one is the reason.

ConJur — And what should be done when the “concert of versions,” as coined by justice Gilmar Mendes, to harm or benefit someone, comes into play? We always talk about this in theory, "when it happens we'll see," but this has happened several times during operation "lava jato," and with proven lies.
Victor Rufino —
 This is a problem that is inherent in agreement programs, but it's not advisable to abandon them because of problems. The legal system will come up with a solution. What I recommend are two approaches: document as much as possible what's being said and revealed, always guaranteeing some degree of secrecy, and always have a moderate degree of skepticism in relation to what is being reported. It’s human to have a confirmation bias. That's why it's important not to make the agreement, which is a means of investigation, the sole and only means of proof, the cornerstone for conviction.

ConJur — Which means it's more efficient to work with the possibility of an error than to make believe it doesn't exist.
Victor Rufino —
 Be aware that although it's a fundamental mechanism, it's not the magic solution for all problems. The literature describes several risks and the evidence of these problems is overwhelming. In the game theory, as well as in any other. But it's not because our nails will grow back that we’re going to stop cutting them.

 é chefe de redação da revista Consultor Jurídico.

Revista Consultor Jurídico, 28 de novembro de 2018, 15h13

Comentários de leitores

0 comentários

Comentários encerrados em 06/12/2018.
A seção de comentários de cada texto é encerrada 7 dias após a data da sua publicação.