Advocacy coalitions and negotiation strategies in the revision of the forest code

This study aims to analyze the advocacy coalitions (classified as ‘environment’ and ‘agriculture’) established during the revision of the Brazilian Forest Code and the main negotiation strategies used. Interviews, analysis of documents and newspaper’ reports allowed capturing how the managers of the Ministries of Environment (MMA) and Agriculture, Livestock and Supply (Mapa) led these coalitions. Coalition analysis used the Advocacy Coalition Framework (ACF). Results showed both ministries used negotiation as their main strategy. The agriculture coalition invested in scientific information, while environment coalition carried out social mobilization.


INTRODUCTION
The first Brazilian Forest Code was established by Decree 23793, on January 23, 1934. The legislation's purpose was to preserve part of the native vegetation inside private properties, considering that the government did not have enough structure to oversee all the public territories. According to the decree, the Ministry of Agriculture was responsible for classifying the remaining and protected areas and forests, establish the location of the national parks, and organize model forests, to recognize the totality of the forest area of the country. The group leading this code, according to Franco and Drummond (2012), sought to establish a connection between environmental protection and building national identity in Brazil.
According to Drummond and Barros-Platiau (2006), the public control over the use of forests proved much weaker than water and mining control, not due to the legislation but to poor management. In 1965In , the 1934 Forest Code was repealed, and a year before on November 30, 1964, the military government published the Law 4504, known as the Land Statute. For Drummond and Barros-Platiau (2006), the Land Statute temporarily united the themes of environmental conservation and land reform, to constitute a considerably progressive law. The statute provided, for example, that the social function of land is only fulfilled if combined with equitable distribution, satisfactory levels of productivity, and conservation of natural resources. In other words, land reform could be associated with environmental protection policies. However, the authors state that none of the set of policies advanced during the 1960s and 1970s.
The 1934 Forest Code was repealed by the 1965 Forest Code (Law 4771/1965). For Drummond and Barros-Platiau (2006), the National Congress debated around the 1965 law since 1948, i.e., it took 17 years for its approval and the amendments to the code were presented only during the military dictatorship. Cureau and Leuzinger (2013), say that the provisions on environmental protection of different territories during that time were a result of the generals' perception of forests as some form of guarantee of territorial integrity. On the other hand, the military government also established territories to extract the resources needed to promote development. In article 1, the 1965 Forest Code provided that forests and other forms of vegetation, recognized as useful to the lands they cover, are assets of common interest to the entire population, and, therefore, the property rights are secured observing the limitations established by law.
In 1996, the provisional measure (MP) 1511 expanded the area considered as legal reserve in the Amazon Rain Forest to 80%, due to the negative response of the international community to the region's increasing deforestation (Cureau & Leuzinger, 2013). As the provisional measures are legal instruments valid for 30 days (as provided in Article 62 of the Brazilian 1988 Federal Constitution -the rule was changed by the Constitutional Amendment 32, passed in 2001), the MP 1511 was reissued every month and, during this time, new changes in the forest law gradually passed. The legislation was continuously discussed with representatives of the various organizations and agencies involved, including the Ministry of Agriculture and the Ministry of the Environment. The Law 9605 of February 12, 1998, known as the 'Law of Environmental Crimes, ' brought important changes to the MP 1605-30, which provided on the conversion of forest areas to agriculture areas in the North and Central-West regions of Brazil (Cureau & Leuzinger, 2013 Law 11934/2009;Law 7511/1986;Law 7803/1989;Law 7875/1989;Law 9985/2000;Law 11284/2006;Law 5870/1973;and Law 5106/1966. The MP 2166-67/2001, as well as the 1965 Forest Code, were repealed by the 2012 Forest Code. As pointed out by Araújo (2010, p. 188), "the debate on the changes in the Forest Code and, in general, on the federal norms to control deforestation has been historically marked by polarization with environmentalists on one side and the productive sector on the other. " Therefore, this article seeks to identify the members of two coalitions, one led by the Ministry of the Environment (MMA) and the other by the Ministry of Agriculture, Livestock, and Supply (Mapa), which traditionally polarize discussions about the Forest Code (Araújo, 2010). Also, the study aims to point out the strategies adopted by the coalitions to achieve the goals they established during the revision of the 1965 Forest Code.
The 2012 Forest Code is the object of several studies since it was passed, considering the importance of the theme and also the great social mobilization created during the revision of the 1965 Forest Code. A survey on the main articles about the subject shows studies analyzing the impact of the application of the code (or part of it) in a particular region or biome. Also, some works observe the legislation's impact from the ethical and economic point of view, as well as studies with socio-political analyses of the legislation's approval process. This article, therefore, presents the backstage of the Forest Code revision process, based on the perspective of the coalitions connected to the areas of agriculture and environmental protection.
The next section introduces the Advocacy Coalition Framework (ACF), and the third section describes the methodology of the study. The following section presents the results and summarizes the debate around the 1965 Forest Code revision process, introducing the members of both coalitions, who were identified through interviews and documentary analysis. Also, the strategies the coalitions adopted are pointed out in this section. The fifth and final section discusses the findings of the study and concludes indicating elements to be explored in future research.

THEORETICAL FRAMEWORK
The Advocacy Coalition Framework (ACF) is an approach to policy processes developed by Sabatier and Jenkins-Smith (1993) to deal with wicked problems, i.e., those problems where conflicts of goals stand out, where there are significant technical disputes and multiple actors at various levels of government. The increasing use of the ACF led to the need to improve the framework, as Sabatier and Weible (2007) pointed out. The concepts presented in this study consider the extended version of the ACF.
According to Sabatier and Weible (2007), the ACF considers the complexity of policy-making in current times, both substantively and legally, and that participants should specialize in order to have influence. Such specialization occurs within political subsystems, composed of participants who regularly seek to influence politics within the political subsystem. A subsystem is characterized by functional/substantive dimensions, for example, water policy, and territorial policy (state or municipal). Policy participants have strong beliefs and seek to translate them into policies. Because technical and scientific information plays an important role in changing the participants' beliefs, researchers (such as political analysts and consultants), play a central role in the political process. However, beliefs are stable over a period, which restricts political changes.
The vast majority of policy formulations occur within political subsystems and involve negotiations between experts. The behavior of the political participants within the subsystems is, however, affected by two sets of exogenous factors, one relatively stable and the other quite dynamic (Sabatier, 1988). Stable parameters include the problem's basic attributes, distribution of natural resources, structures, and fundamental sociocultural values, as well as basic constitutional structure. These factors rarely change over a decade, although they are important in establishing the resources and constraints in which actors must operate. External dynamic factors include changes in socioeconomic conditions, changes in government coalitions, and political decisions made in other subsystems. They also affect the behavior of actors, but their ability to change them makes them critical factors that influence most policy changes. One of the assumptions of the ACF, according to Sabatier and Weible (2007), is that changes in one of these dynamic factors are a necessary condition for significant political changes. Another form of policy change is through learning (experience or new information) gained through achieving or revising policy objectives. The strong pressure of the agriculture coalition, affected by the implementation of the Forest Policy, as well as the learning in the implementation of the 1965 Forest Code, pointed out the need for revision of the law.
For Sabatier and Weible (2007), the ACF conceptualizes a hierarchical structure of three levels. At the broadest level are the deeper beliefs, which reach most political subsystems. As these beliefs are acquired in childhood, there is great difficulty in changing them. At the next level are political beliefs, which achieve an entire political subsystem. As the political participants are great connoisseurs of relations within the political subsystem, they wish to invest in the application of certain deep beliefs to the development of political beliefs in that subsystem. However, there is no one-to-one relationship between these beliefs. Political beliefs are also difficult to change because they deal with fundamental political choices. Sabatier and Weible (2007, p. 195) use the term "policy core policy preferences, " i.e., normative beliefs that project an image of how the political subsystem should be, helping to unite allies and divide opponents. Policy core policy references can hold coalitions together. The last level consists of secondary beliefs, with a narrower scope in comparison to political beliefs. Changing secondary beliefs is easier since they require less evidence and agreement among subsystem actors.
The ACF predicts that the beliefs and behaviors of the various stakeholders are embedded in informal networks and that the policy-making process is structured in part by networks among important political participants (Sabatier & Jenkins-Smith, 1993). The framework assumes that these participants struggle to translate the components of their belief systems into policies before their opponents do. In order to have any chance of success, they must seek allies, share resources, and develop complementary strategies. The prospect of loss motivates the actors to align and cooperate with the allies.
According to Sabatier and Weible (2007), allies are people who have similar political beliefs, and by engaging in a higher degree of coordination, they form an advocacy coalition. Coordination involves some degree of joint work to achieve common policy objectives. For the authors, this framework provides the most useful tool for aggregating the behavior of hundreds of organizations and individuals involved in a political subsystem. In each subsystem, there are between two and five advocacy coalitions. Therefore, the ACF will contribute to the identification of the main values (relatively stable parameters) of the two advocacy coalitions involved in the forest policy in Brazil -agriculture and environmental. Also, it will allow analyzing how the external events and the short and long term factors led to strategies during the negotiation of the new Forest Code. The following research question guided the study: Which negotiation strategies were most effective during the process of discussing the new Brazilian Forest Code?

METHODOLOGY
This case study used multiple evidence sources (interviews, document analysis, and news in major newspapers) to identify existing advocacy coalitions and their key strategies. The interviews followed a semi-structured script based on categories related to the ACF model (Sabatier & Weible, 2007). They were recorded, transcribed and validated for content analysis. The script was tested with the first interviewee and subsequently improved. Twelve interviews were carried out with people that participated in the process of revising the Forest Code (Mapa, MMA, Ministry of Agrarian Development -MDA, Office of the Chief of the Staff -directly connected with the president, Chamber of Deputies, representatives of the productive sector and environmental nonprofits) from April 2013 to April 2014, totaling 672 minutes of interviews (average of 56 minutes per interview). There was an agreement of non-identification of the interviewees, indicated in this article by the organizational link. One of the interviews was held in Florianópolis, SC and the others in Brasília, DF, always in the interviewees' workplaces. Despite many attempts, no representative of the Federal Senate was interviewed.
The interviewees were selected through researching for their participation in the process of news reports, and their participation was confirmed in each interview, using the snowball technique (Malhotra, 2006). The largest number of interviews was conducted in the Executive Branch, given the research focus. The cycle of data collection ended when the researchers identified saturation.
According to Severino (2007), the word 'documents' has a broad sense in the approach of documentary research, meaning printed documents, newspapers, photographs, films, recordings, and legal documents. In this study, the documentary source was mainly legal documents, reports and digitized and printed documents, such as reports, and strategic plans (Brazil, 2009).
The interviews and documentary research (documents and reports) were analyzed using categorical content analysis (Bardin, 2011). Four categories were analyzed: influence strategies; external events (system); opportunity for coalitions; and actors' restrictions and resources. The first category emerged from the content analysis of interviews and documentary research. The others were organized using the ACF model (Sabatier & Weible, 2007). The categories were subdivided into nodes, as shown in Box 1.
The documents underwent an initial analysis and the most relevant ones (such as the interviews), were included in the software NVivo 10. Although the phrase inserted in the software was initially selected as a coding unit, it was often necessary to expand the records for a better understanding of the meaning. The elements were coded into nodes representing the categories, as shown in Box 1, and then analyzed again for the understanding of the dynamics of the two coalitions studied. The word frequency cloud was one of the tools used to analyze the main consensus in each coalition.

RESULTS AND DISCUSSION
According to Araújo (2010), in January 2010, 36 bills were being discussed in the Brazilian Chamber of Deputies aimed at some form of amendment of the Forest Code. The author divides the bills into two main blocks. The first block of propositions was bound to Bill 6424/2005 and its appendices, in a process analyzed by the Committee on the Environment and Sustainable Development (CMADS) and the Committee on Agriculture, Livestock, Supply, and Rural Development (CAPADR). The focus was on adjustments to Law 4771/1965. This bill was archived on July 18, 2013, due to the approval of the 2012 Forest Code, under the Law 12651 of May 25, 2012. The second block grouped the propositions analyzed by the special commission created ad hoc to study the Bill 1876/1999 and its appendices, which aimed to significantly restructure the normative basis of the theme, especially about consolidated settlements in disagreement with the 1965 Forest Code. This was the process that led to the approval of the 2012 Forest Code. The negotiation intended to adapt the reality of rural properties to the current environmental legislation. Among the principles approved in the 2012 Forest Code, the environmental protection continues to be the obligation of the farmer that owns the land, through the implementation of permanent preservation areas (APPs) and areas of legal reserves.
Currently, the Environmental Regularization Program and the Rural Environmental Registry are instruments used by both coalitions to environmental regularization of properties and rural possessions, implementing the provisions of the 2012 Forest Code. Initially created by the Decree 7830 of October 17, 2012, these instruments had their general complementary norms approved by the Decree 8235, of May 5, 2014. As of November 30, 2017, according to information from the Rural Environmental Registry system, more than 3.5 million properties had already been registered, as shown in Graph 1. The next section presents the results of the efforts to identify the two coalitions involved in the negotiations around the Forest Code, and their beliefs and main strategies. As reported in the interviews, coalition members are primarily those involved in the implementation of their respective public policies (related to environmental protection or agriculture). As observed by Sabatier and Weible (2007), the coalitions were formed based on their members' identity and beliefs.

ENVIRONMENTAL COALITION
The environmental coalition observed in the interviews and documents was led by MMA and is presented in Figure 1. The content analysis of the interviews showed strong identification between MMA and environmental nonprofits, especially during the periods Ministers Marina Silva and Carlos Minc were in office. This relationship changed when Minister Izabella Teixeira took office. She was more technical and conducted the ministry's activity independent from the position of the nonprofits. Minister Teixeira was a technician, civil servant working at the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA) (Brazilian institute of the environment and renewable natural resources) since 1984. The relationship of the MMA with family farming started after Marina Silva left the office. The coalition also counts on other institutions and people related to environmental protection such as the Public Ministry and the academia.
There were no major differences between the positions of the Ministry and nonprofits. The appointed executive secretary was an activist in a nonprofit. People used to make jokes saying it was a Ministry-Nonprofit. At that moment, the ministry was not an actor that stood out, that was worthwhile. The environmental movement included the environmental deputies, the Ministry of the Environment, and the environmental organizations. They all felt very well represented by the Minister Marina (Silva) (Interviewee 3 -MMA).
The first act was to weaken the rural movement. For the Minc, the allies were the environmentalists and family farming. So, groups were created in the Ministry, with representatives of these two actors. Moreover, then there was a difference in there. Those who do not want to change anything; those who accept change, as long as it is by means of infralegal instruments, such as decree, resolutions of the Conama, normative instructions and everything that could be created without taking the discussion to the Congress; and those who thought, like CONTAG, that in the end the discussion would have to go to Congress, there is no way to solve it, these are our demands and they have to go there (Congress) (Interviewee 3 -MMA).
Because with (Minister) Minc, you had a polarization such as the one I mentioned. He did not care about it. With Izabella, the tendency was always avoiding polarization, because there was this recommendation in the Office of the Chief of Staff. The fact that she was not a person with a history in politics, in general, she did not take the lead in any position (Interviewee 4 -MMA).
Despite shared political beliefs about the importance of protecting natural resources, the coalition's secondary beliefs, i.e., the way of instrumentalizing forest policy changed over time. At first, with the strongest influence of environmental nonprofits, the agreement was to maintain the Forest Code as it stood (Law 4771/1965 andMP 2166-67). With the expansion of the coalition, which meant a greater diversity of participants with different visions about how the process should occur, based on their original beliefs, other strategies were considered: 1) no change; 2) alteration using infra-legal instruments; and 3) passing a new law. While the members from family farming are originally part of the subsystem 'family farming'; Environmentalists and other members of the coalition, such as the Public Ministry, share the original beliefs of the subsystem 'environmental protection. ' Thus, it is possible to observe that the level of coordination in the environmental coalition is weaker than the coordination in the agriculture coalition, which will impact the outcome of the new law's negotiation.

Environmental coalition beliefs system
Political beliefs • Protection of natural resources; • Agribusiness as responsible for environmental degradation; • Need to harmonize social (family farming) and environmental.

Secondary beliefs/instruments
• No changes in the Forest Code; • Changes in the infra-legal instruments; • Publication of the new law.
Source: Elaborated by the authors.

AGRICULTURE COALITION
The stakeholders of the coalition led by the Mapa, identified in the interviews and the documents analyzed, are presented in Figure 2. The productive sector is formed by several organizations, some of them stood out for their participation in the the The analysis of interviews and the documents show common values among the members of this coalition. The first one refers to the importance given to agribusiness in the context of the national economy. Regarding the Forest Code, specifically, there is a general understanding that the small farmer would be the most harmed by the full implementation of the 1965 Forest Code and that the priority is to unite production and environmental sustainability. The environmental coalition shares this particular understanding.
The legislation of the (Forest) Code is much more difficult to comply in the case of the small farmers than it is for the large farmers. A person who has 5,000 hectares makes 20-30 hectares of APP. A person who has 10 hectares does not make 1 hectare of APP. For one of them, this is 10% of the property, for the other, it is 0.0 something (sic). So I explained that this affected the small farmer, who could not comply with the legislation the way it was (Interviewee 3 -Mapa).
The discussion with the Ruralist Parliamentary Front was that all parts were engaged in defense of small farmers. The small farmer is the one that suffers, who has no credit (Interviewee 4 -MMA).
The MDA also participated in this discussion because it was observed that the Forest Code poses a restriction on the use of property, both to the small and large farmer, but that especially affected the small farmer. This situation moved the decision makers. This issue was heartbreaking and before that changing the Forest Code was something widely disapproved (Interviewee 2 -Office of the Chief of Staff). The interviews showed an alignment between Mapa, productive sector and the FPA, which is a result of shared values and beliefs. The close relationship between Mapa, the FPA, and the productive sector was weakened, especially during the time Minister Jorge Alberto Mendes Ribeiro Filho was in office. Minister Mendes Ribeiro Filho was not familiar with the sector and was appointed by President Rousseff as retribution to his work as leader of the government in the Chamber of Deputies in 2011.
In the initial working group to discuss the Forest Code, the CNA was participating representing the agriculture, as well as some well-known deputies from the Ruralist Parliamentary Front, such as Marcos Montes, Homero Pereira, Heinz, and Moreira Mendes (Interviewee 3 -Mapa).

Agriculture coalition beliefs system
Political beliefs • Contribution of agriculture for food security and the country's development; • Property rights; • Harmony between production and sustainability; • All farmers in Brazil (small, medium and large) suffered from environmental rules (1965 Forest Code and Decree 6514 of July 22, 2008, which poses the regulations of the Law of Environmental Crimes).

Secondary beliefs/instruments • The law needs to change.
Source: Elaborated by the authors.

STRATEGIES
The analysis showed that negotiation was the main strategy adopted by both coalitions. Scientific information was also an important strategy for the agriculture coalition. The strategies of each of the coalitions during the negotiations of the new Forest Code are presented below.

AGRICULTURE COALITION
The main beliefs of this coalition were based on the importance of agriculture for the development of the country, the priority to be attributed to family farming and the compatibility of production with environmental sustainability, according to the analysis of interviews and documents. When analyzing the frequency of words of the interviewed coalition members, it is possible to see some differences between the results with and without the MDA. In the agriculture coalition without MDA (Figure 4), the environmental issue was more prominent, including the concern for APPs, and there is greater influence from the relationship between the ministry (represented by the minister) and the FPA (deputies committed to the issue). In the coalition considering the MDA as a member (Figure 3), agriculture gains greater prominence than environmental issues, probably due to the 'pendulum' relationship that the ministry established between the two coalitions.   According to one of the interviewees, another strategy used by the Mapa was to remove the radical people from the process. This has contributed to the reduction of conflicts during intragovernmental negotiations. Also, the coalition sought scientific information that could strengthen the arguments in favor of modifications in the Forest Code.

ENVIRONMENTAL COALITION
The environmental coalition was larger than the agriculture coalition. Its performance, however, was not uniform over time. As occurred in the Mapa, the profile of the ministers of the MMA influenced the form of negotiation. When Minister Marina Silva was in office, there was no difference between the position of the MMA and that of the environmental and socio-environmental nonprofits. During Minister Minc's administration, other organizations were included in the coalition, such as the MDA and family farming movements, as mentioned before when the coalition members were identified. However, with the MMA's change of attitude when Minister Izabella Teixeira took office, the Ministry took positions coordinated with the Executive. The nonprofits and family farming movements joined the Brazilian Committee for the Defense of Forest Law and Sustainable Development and started to operate independently. The committee was created on June 7, 2011, to mobilize the population against the changes in the Forest Code, being discussed in Congress.
The main strategies used initially by the environmental coalition and, later, by the partner nonprofits to obtain greater influence in the process, according to the interviewee of the environmental sector, were: 1) Denounce. It does not participate initially. 2) Negotiation when there was a perception that the process would advance.
3) Bringing in academia, because the nonprofits have less credibility when isolated. 4) Partnership with part of the business sector that supported the not so radical change (Brazilian Pulp and Paper Association -Bracelpa). 5) Attempt to consolidate understandings and seek movements related to family farming (Via Campesina and Landless Workers Movement -MST) and CNBB. 6) Hand-to-hand negotiation with government and members of the Congress. 7) Mobilization of communication -media. Critical coverage -proportionately more voice for environmentalists. The result was a change in the media, after discussion in the Senate. 8) Mobilization in the electoral campaign of the candidates. Nonprofits pressed candidates to commit to the agenda during electoral campaign.
According to the interviewees, an unanimity was reached regarding the importance of changing the Forest Code. This influenced the change of the ministry's position in the negotiations and contributed to its distancing from the position of non-governmental members of the environmental coalition, who did not want to change the code. Despite the concentration in negotiation strategies led by the MMA, the environmental coalition also invested in social mobilization and other forms of influence, such as the "Veta, Dilma" campaign, requesting President Rousseff to veto the law, which was observed in the analysis of countless news items.

And in the federal government, during the first (President) Lula's administration, Minister Marina
(Silva) had an attitude based on "better not to let anyone touch it". It was a kind of autistic attitude, or not, sometimes she was aware that if she allowed it to move forward she would lose so much that it was better not to let anyone touch it. She sought President Lula's support to keep the process from moving (Interviewee 2 -MMA).
The Ministry of the Environment stopped pressing because there was a point where they could no longer sustain the pressure for two reasons: the MDA target public, with whom they had always been pari passu, needed these changes. And the MDA's target public, as I told you, led them (the MMA) to submit a proposal for a provisional measure. So that public voiced discussion between Ministers Stephanes and Minc stopped. When the discussion went to Congress, it cooled down inside the Executive, and the fight was carried out in parliament (Interviewee 3 -Mapa).
The frequency of words of the environmental coalition ( Figure 5) shows, in addition to the focus on the environmental issue, as expected, also the words 'governo' (government), 'ministra' (minister), and 'Izabella, ' which were emphasized. 'Agricultura' (agriculture) appears more often than the term 'ambientalistas' (environmentalists), which can be explained by the distance between Ministry of the Environment and the nonprofits throughout the negotiation and also by the focus on family farming, which started to be part of the environmental coalition. The MMA became open for negotiation regarding the Forest Code when Minister Izabella Teixeira took office.

FIGURE 5 FREQUENCY OF WORDS OF THE ENVIRONMENTAL COALITION (IN PORTUGUESE)
Source: Elaborated by the authors.

POLICY BROKER
In the Advocacy Coalition Framework, public policy subsystems count, in addition to coalitions, with 'policy brokers. ' They are mediating actors whose main concern, according to Sabatier (1988), is to keep political conflict within acceptable limits, reaching some reasonable solution to the problem. The result of this mediation is often some governmental action, such as a program. However, according to Bratt (2013), many brokers may be politically biased, while coalition members may be concerned about maintaining the system. Therefore, policy brokers are not disinterested actors. Their interest can be both material and institutional, which leads them to act strategically in mediation. This characteristic, pointed out by Bratt (2013), was confirmed during the analysis of the negotiation carried out for the revision of the Forest Code.
In the case of the Forest Code, important policy brokers representing the federal government were Presidents Lula and Rousseff. During the negotiation period, from 2008 to 2012 (the focus of this study), the presidents played a significant role in mediating the coalitions, but with different positions. The federal government under President Lula issued Decree 6514, of July 22, 2008, which regulates environmental administrative infractions and sanctions, establishing the administrative process for the determination of such infractions. According to several interviewees, this decree caused a great impact in the FPA and was considered by all the main factor that accelerated the discussions for the changes in the Forest Code.
Minister Minc, at the time, issued a decree regulating the law of environmental crimes, and this decree imposed huge fines for those who did not have a legal reserve. Then, there was a reaction from the Minister of Agriculture and here also, from the deputies, in a very strong way (Interviewee 2 -Chamber of Deputies).
The reaction of the agriculture coalition led to a move from both ministers toward a reconciliation. Some of the interviewees reported a meeting with representatives of the MMA, Mapa, MDA, environmental nonprofits and the FPA to discuss Decree 6514/2008. As interests were conflicting, Minister of Agriculture, Livestock and Supply Reinhold Stephanes told members of the parliamentary front that the best debate arena would be the Congress, given the resistance of the MMA.
This was the result of that meeting because of a very simple situation: Minister Stephanes realized that the decree was issued, and Minister Minc would not take a very different position from what was in the Decree. Also, the Executive was not going to propose a change in the Forest Code since the elections were approaching and there was a whole context. Moreover, at the table with the members of Congress, Stephanes told them that he did not know why they were bringing this matter to the Executive because this issue would be decided in the Legislative Branch instead. He told the parliamentary front they had the strength to set the agenda in Congress. At the meeting were the Mapa, the MDA, the MMA, and the parliamentary front. Stephanes said, "If I were in your place, I would make a proposition. " And that is what happened (Interviewee 1 -MDA).
Even with the rupture that occurred in the initial group, the government continued to seek an internal solution to the issue. Following guidance from President Lula, the group was restricted to the ministries, intermediated by the Office of the Chief of Staff. The president also ordered Ministers Reinhold Stephanes (Mapa) and Carlos Minc (MMA) not to take the discussions to the press. The unification of the government's discourse would only occur during the administration of President Rousseff, when the bill was already at an advanced phase of discussion in the Congress. During this period, the MMA, through Minister Izabella Teixeira, was the official negotiator on the part of the federal government. According to the interviewees, the internal discussions in the government were held in three levels. The first was a discussion of the technicians, mediated by managers of Office of the Chief of Staff. The second level was a decision among ministers, with the intermediation of the Chief of Staff. The last level was the decision by President Dilma Rousseff. According to the interviews, the president participated actively in the agreement around the federal government position on the issue, a presence that sometimes made it hard to technically finalize controversial points.

And in (President) Lula's government each
Sometimes the president was obliged to hold meetings. Great progress was made with the participation of President Dilma, such as the "escadinha" (stepladder) for farmers that had in their properties rivers of different width, and for properties of different sizes -everyone should have APPs [permanent preservation areas], but with different dimensions (Interviewee 1 -Mapa).
In the final stage of the bill's analysis in the Congress, the president in person met with the ministers; she was personally involved. This closeness even hindered some details of the negotiation, because when the one at the top of the command chain says that they want something, who dares to contradict even though they are wrong? (Interviewee 2 -Office of the Chief of Staff). President Rousseff also resorted to her veto power to interfere with the outcome of the new Forest Code. She vetoed 10 points of the Law 12651/2012 and issued the Provisional Measure MP 571 to provide over the issues she vetoed in the legislation. The MP became Law 12727 / 2012, which was also vetoed by the president in 9 points.
It is possible to see, therefore, the difference in the position of the two main policy brokers of the negotiation, both representing the federal government. It is noteworthy the role played by President Rousseff in the definition of some important points of the new Forest Code. While President Lula sought internal conciliation but was not involved in the issue, President Dilma Rousseff acted directly in the final moments of the negotiation. Some of the interviewees highlighted the innovation of such an active president's participation in an environmental negotiation.
I think a big difference was that the president got personally involved. Regarding environmental issues, she was more dedicated than any other president. The president sat several times during the peak of the negotiation. In two weeks, she would devote part time to calling the ministers. And she directed the guidelines (Interviewee 1 -Office of the Chief of Staff).
It is important to note that in June 2012, during the discussion of the Forest Code in Congress, Brazil hosted the United Nations Conference on Sustainable Development (Rio+20). President Rousseff was heavily pressured to keep the environmental commitments made during the election campaign, which may have contributed to her involvement in the closure of the government's positions.
From the Legislative Branch, the policy broker was Deputy Aldo Rebelo (PCdoB party, from São Paulo), who was appointed as rapporteur for the special commission created to analyze the proposed changes to the 1965 Forest Code. The importance of Rebelo for the process was a consensus among the interviewees. The fact that he was from a communist party helped that both environmentalists and ruralists regarded him as a neutral person. His report, however, showed a clear alignment to the rural issues. Unusually, Deputy Aldo Rebelo (2010, p. 2) dedicated his report "to Brazilian farmers. " According to one of the interviewees, Deputy Rebelo has a strong nationalist tendency and rejected international nonprofits.
Deputy Aldo (Rebelo) cannot stand nonprofits with names in English, and he does not hide this. That is why environmentalists were not treated in a friendly way (Interviewee 1 -Chamber of Deputies).
The (Special Committee's) chairman shall be elected, and the rapporteur shall be appointed. Deputy Micheletto was elected and tried to appoint another Deputy that declined the position. Deputy Aldo Rebelo held numerous public hearings, including meeting requests from other members of Congress. At these hearings, representatives of the various stakeholders, not just members of the FPA and the Environmentalist Parliamentary Front, were heard. The list of participants in the hearings included people from the areas of transportation, industry, mines, and energy.
However, Santos (2012) points out that the participation of the FPA (83.5%) was much higher than that of the Environmentalist Parliamentary Front (16.5%) in the 26 hearings analyzed. The work of the special committee had to be extended so the hearings could be held. The opinion of the rapporteur was presented to the special committee on June 8, 2010. In his report, Deputy Aldo Rebelo highlighted the difficulties in applying the 1965 Forest Code because of the numerous legal changes and recognized that the state was the first to deny the application of the law, promoting its inefficiency. According to Rebelo, the legal instruments in place would allow one to consider an environmental crime even "the very act of living" (Rebelo, 2010, p. 4).
The data gathered in the interviews pointed to the direct participation of the Mapa in the discussions for the drafting of rapporteur Aldo Rebelo's opinion. This demonstrates a clear action strategy to change the rules (Freeman, 1984). At the same time, there was the internal debate in the government, mediated by the Office of the Chief of Staff. The MMA, initially against the changes in the Forest Code, ended up having to give in to the members of the family farming of its coalition.

CONCLUSION
This study analyzed the agriculture and environmental coalitions during the process of revision of the 1965 Forest Code, identifying their members and the main strategies used to achieve their goals.
The 2012 Forest Code (Law 12651, of May 25, 2012) has been studied since it was published, considering not only the importance of the theme but also the great social mobilization created during the process of revision of the 1965 Forest Code.
When studying the main articles on the topic, it was possible to find analyzes of the impact of the code's application (or application of part of it) in a given region or biome. Also, the research found studies about the impacts from an ethical and economic point of view, as well as a socio-political analysis of the legislation approval process. This study presented the backstage of this process, from both the agriculture and environmental coalitions point of view.
According to Sabatier and Weible (2007), the ACF considers the complexity of policy-making in current times, both substantively and legally, and that participants should specialize in order to have influence. Such specialization occurs within political subsystems, composed of participants who regularly seek to influence politics within the political subsystem. Policy participants have strong beliefs and seek to translate them into policies. However, beliefs are stable over a period of time, which restricts political changes.
The vast majority of policy formulations occur within political subsystems and involve negotiations between experts. However, the behavior of the political participants within the subsystems is influenced by two sets of exogenous factors, one relatively stable and the other quite dynamic (Sabatier, 1988). Stable parameters include basic characteristics of the problem, distribution of natural resources, structures and fundamental sociocultural values, as well as basic constitutional structure. External dynamic factors include changes in socioeconomic conditions, changes in government coalitions, and political decisions made in other subsystems. These factors also affect the behavior of actors, which makes them critical factors that influence most policy changes. Changes in government coalitions played an important role during the revision process of the 1965 Forest Code.
This case study used multiple evidence sources (interviews, document analysis, and news in major newspapers). The interviews followed a semi-structured script based on categories related to the ACF model (Sabatier & Weible, 2007). They were recorded, transcribed and validated for content analysis. The documentary research used mainly legal documents, reports and digitized and printed documents, such as reports.
The interviews and documentary research (documents and reports) were analyzed using categorical content analysis (Bardin, 2011). Four categories were analyzed: influence strategies; external events (system); opportunity for coalitions; and actors' restrictions and resources.
The strategies adopted by the agriculture coalition led to accomplish two objectives: changes in the 1965 Forest Code and reduction of environmental regulations without discussion with other coalitions. The Mapa identified the demands of its main public of interest -the productive sector -and organized to meet the challenge of reconciling production and sustainability. The necessary coordination with the other members of the coalition was made throughout the process, weakening in the last part of the negotiations, after the direct intervention of the president. Another crucial coordination was one held with the Deputy rapporteur of the Special Committee of the Chamber of Deputies, Aldo Rebelo (PCdoB party, from São Paulo). Most of the changes in the 1965 Forest Code that met the demands of the agriculture coalition were made from his work. The strategies of the agriculture coalition (coordination, use of scientific information, and removing radical activists from the process) demonstrated greater effectiveness in reaching the results planned.
The environmental coalition, on the other hand, defensively advocated against changes in the 1965 Forest Code. The expansion of coalition members, from the entry of family farming organizations, weakened the initial union of environmentalists and pressured for new strategies. The fact that President Rousseff, as a policy broker on the part of the federal government, favored the environmental coalition, minimized its initial losses. This offset, to a certain extent, the lack of identification of other stakeholders' demands that were not recognized by the MMA.
Currently, the Environmental Regularization Program and the Rural Environmental Registry are instruments used by both coalitions to environmental regularization of properties and rural possessions, implementing the provisions of the 2012 Forest Code. Initially created by the Decree 7830 of October 17, 2012, these instruments had their general complementary norms approved by the Decree 8235, of May 5, 2014. As of November 30, 2017, according to information from the Rural Environmental Registry system, more than 3.5 million properties had already been registered, as shown in Figure 1.
The analysis of the influence strategies, together with the other elements of the Advocacy Coalition Framework, was relevant in the study of public policy changes and can be replicated in other case studies. A limitation of this research was the failure to conduct interviews with people that were determinant in the negotiations -ministers and rapporteurs (Chamber of Deputies and Senate) -due to the complexity of their agendas. However, interviews were conducted with advisors and technicians who participated directly in the negotiations. Also, this limitation was reduced with the analysis of interviews granted to official media vehicles (Agência Brasil, Agência Camara, and Agência Senado). The information collected in these interviews corroborated the information provided by advisors. The interviews and the documents analyzed did not indicate the systematic participation of Brazilian councils in the negotiation, such as the Conselho Nacional do Meio Ambiente (Conama) (national council of the environment), and the Conselho Nacional de Florestas (Conaflor) (national council of forests). However, with the recent institution of the National Social Participation Policy (Decree 8243, of May 23, 2014), future studies may analyze the integration of these councils as stakeholders in the process of formulating and implementing public policies.
An interesting piece of evidence that emerged during the interviews concerns the role of political parties as negotiating stakeholders. In the opinion of the interviewees from the Chamber of Deputies, the PMDB was the party that made the difference in the votes regarding the Forest Code, while deputies from the Workers Party (PT), for the first time in its history, did not follow the direction of its leadership. As the research focused on the role of agencies of the Executive Branch, it was not possible to deepen the study of party differences in the Legislative. This is a topic to be analyzed in future research, especially adopting political science theories.
Finally, given the high turnover of top management positions in governmental bodies, a question to be investigated in future research is the impact of the managers' values on the organization's strategy. This study pointed out that there were changes in coalitions due to the profile of ministers in both the Mapa and the MMA, showing that this may be a topic to be further explored.