Narrative Disputes over Family-Farming Public Policies in Brazil: Conservative Attacks and Restricted Countermovements

This article analyzes the conservative narratives underlying current changes in Brazilian rural development public policies. It first presents an overview of the recognition and institutionalization of family farming and the main policies created in support of this segment. Subsequently, it discusses how a conservative narrative started to question the ability of these policies to integrate family farmers into modern agricultural markets. Focusing on how this narrative tries to legitimate a new “referential” for public action, the article analyzes the discourse on newly occurring segregation between agricultural policies for productive farmers and social policies for unproductive ones. It demonstrates how this discourse excludes family farmers from the social pact that prevailed over the past three decades and depended on the state as an actor to mediate the contradictions in the unstable coexistence of agribusiness and family-farming logics. Finally, it analyzes how rural social movements are reacting to this process.

Brazil's experience developing public policies for family farming is globally acknowledged as a success story (FAO 2014(FAO , 2015World Food Programme 2013). During the past decade, these experiences became essential elements of important international programs, especially those focused on food security (Milhorance 2013). In the same period, however, some of these policies have been questioned because of their limits and misapplications. Criticism was leveled at the concentration of resources in rural credit programs, which financed technological modernization and subordinated integration of a relatively small number of family farmers to conventional agricultural markets (Aquino and Schneider 2015). In addition, difficulties have been acknowledged in promoting the productive inclusion of the poorest farmers, who, despite advances in food security programs, have remained dependent on social assistance (Niederle 2017). Other criticisms highlighted institutional hindrances to structural changes, whether toward broad agrarian reform (Sauer and Mészáros 2017;Wolford 2016;Medeiros 2015) or creating new dynamics of sustainable territorial development (Favareto 2010).
At the end of President Lula da Silva's (2007Silva's ( -2010 second term, and especially during President Dilma Rousseff's first term (2011)(2012)(2013)(2014), debates intensified over the limits and future of family-farming policies, and even over the fate of the Ministry of Agrarian Development (Ministério do Desenvolvimento Agrário, MDA), created in 1999 to respond to the demands of this social segment. Two narratives began to dispute the "referentials" of the policies (Radaelli 2006;Muller 2008). 1 On the one hand, the conservative narrative tried to build a story about the ineffectiveness of rural development policies, suggesting that the economic strength of Brazilian agriculture was confined to a small segment of highly productive modern farmers provided with technology, which, though including some family farms, would be better supported by sectoral agricultural policies through the Ministry of Agriculture (Alves and Rocha 2010). On the other hand, researchers and agrarian social movements began to question the effects of such policies, given their limitations on expanding the scope of action and countering the advance of a new type of agrarian predatory capitalism (Fernandes 2014).
This article analyzes the development of this dispute since the end of President Rousseff's first term, identifying the main actors and their stances. Despite ongoing developments, 2 the article also examines the main consequences of the institutional coup d'état that ousted Rousseff in 2016 and has contributed to consolidation of the conservative narrative, eliminating the MDA and eliminating or weakening various public policies. These moves put the main protagonists in the fight to recognize and legitimize family farming as a sociopolitical category in Brazil on the defense.
We briefly recount the history of the institutional recognition of family farming in Brazil. We then present the referentials underlying three generations of family-farming policies over the past twenty years. Another section focuses on narrative disputes around the outcomes of these policies, emphasizing how, given transformations in the "global referential" (Jobert and Muller 1987), the conservative narrative gained strength. We then examine effects of ongoing institutional changes, highlighting the widening gap between agricultural and social policies, the dismantling of food security policies, deregulation of the land market, the attack on social minorities, and changes to qualifying criteria for family farming. Finally, we discuss the dilemmas faced by family-farming organizations in organizing a countermovement to resist their exclusion from the social pact that has been partially effective since the restoration of democracy and the 1988 national constitution.

From Struggles for Recognition to Institutionalization
Since the colonial period in Brazil, the state and economic elites have relegated family farming to a position subservient to the interests of the exporting plantation. Family farming became responsible for functions deemed secondary, such as food production for the domestic market and the maintenance of a seasonal workforce reserve for large farms (Prado 1996;Martins 1981). Despite some isolated state actions to support small farms in specific localities and periods (especially in the southern states), family farming has historically been left to face economic precariousness and social vulnerability (Wanderley 2015). Moreover, family farmers have faced difficulties in constructing autonomous political forces to challenge large landowners and the mainstream agricultural model.
Although peasant movements were already present in some regions, it was not until the 1960s that they gained national labor union representation. The National Confederation of Agricultural Workers (Confederação Nacional dos Trabalhadores na Agricultura, CONTAG) brought small and landless farmers together with rural wage earners (Medeiros 1989). It was only in the 1980s, when redemocratization movements broke from their submission to military governments, that CONTAG began to play a relevant role in social criticism. Family farmers have since organized important political actions to demand better prices for agricultural products, differentiated credit programs, agrarian reform, technical assistance, and access to social security (Picolotto 2014). The increasing visibility and political importance of these actors allowed them to intervene in debates on the Agricultural Law of 1991 (Law No. 1771), guaranteeing special treatment for smallholders.
The consolidation of family farmers as political actors occurred in the 1990s through a coalition of three groups of social actors. The first comprised researchers who had shown the historical place of family faming in the development of advanced capitalist economies. Therefore, the adoption of the notion of family farming in social theory not only replaced that of smallholders (until then predominant) but also promoted a theoretical and normative displacement of what these farmers represent. Researchers began to challenge the supposed inferiority of this social segment (as well as the signs of its imminent disappearance) to show its socioeconomic importance, diversity, productive capacity, and, above all, major contributions to reducing inflation in food prices and generating rural employment (Veiga 1991;Abramovay 1992;Lamarche 1993).
The second group comprised rural workers' unions, as well as nongovernmental organizations and agrarian social movements such as the Landless Workers' Movement (Movimento sem Terra, MST). Although in the early 1990s the main demands of this group involved rural workers' rights and agrarian reform, from the mid-1990s onward, the promotion of family farming as the backbone of an alternative rural development project gained more relevance. This era is also marked by national mass mobilization that, combined with other events, was marked by nationwide mass demonstrations that led the government to create, in 1996, the National Program for Strengthening Family Farming (Programa Nacional de Fortalecimento da Agricultura Familiar, PRONAF), which remains the main policy for family farmers (Medeiros 2015).
In the late 1990s, political dissent began to break out in the broad coalition around family farming. The emergence of new labor unions and nonunion organizations revealed dissatisfactions with political choices and an increasing diversity of actors in rural areas (e.g., indigenous, quilombolas, fishers, extractivist farmers, women), each demanding attention to their circumstances (Futemma, Munari, and Adams 2015). In the labor union sphere, the Federation of Workers in Family Agriculture of the Southern Region (Federação dos Trabalhadores na Agricultura Familiar da Região Sul, FETRAF-Sul) was created in 2001, soon acquiring national scope as FETRAF-Brasil. Apart from the formal structures of unionism, other movements were created, such as the Small Farmers Movement (Movimento dos Pequenos Agricultores, MPA). Created in 1997, MPA disputed the political representation of family farmers by the unions. However, while FETRAF and CONTAG support the identity of family farmers, MPA is linked to Via Campesina's political platform. Similar to MST, this movement seeks to strengthen the political identity of the peasantry, which implies a critical narrative about the instrumentality of family farming to the capitalist logic (Fernandes 2014;Wolford 2016).
The third group of actors is policy makers. The Brazilian state's liberalizing agenda of the 1990s (Bresser-Pereira 2015), confronted with the new possibilities for social participation granted by the 1988 constitution, placed rural labor unions in an ambiguous position vis-à-vis the state: in some situations, they directly opposed trade liberalization and deregulation; in others, they found opportunities to propose and collaborate in the design of public policies. For instance, the creation of Mercosur sought to turn Brazil into an exporter of industrial products and an importer of agricultural products; this was interpreted as a huge threat to family farming and triggered several discussions with the government on the need for differentiated agricultural policies. The objective was to reduce the impact of trade liberalization on specific value chains (e.g., milk, wine, wheat) and regions (mainly the south).
To define this new social category of family farming and guide implementation of related policies, a broad project was launched. Results were released in 1994 and contributed to elaborate guidelines for a new strategy for rural development (FAO and INCRA 1994). The project created two categories of Brazilian agricultural establishment, "employers" (patronal) and "family"; within the latter it distinguished three subcategories: "consolidated," "in transition," and "peripheral." The project suggested that policies in support of family farming should address the farmers in transition, who, despite their weaknesses, had the potential to become economically viable. This was the strategy adopted by the government when it created the first public policies for this social segment.
In short, the creation of PRONAF in 1996 was the beginning of the trajectory of public policies for family farming in Brazil. The MDA's creation in 1999 affirmed the struggles for family farming's recognition, increasing visibility and legitimacy vis-à-vis the society and the state. As of 2003, the political reconfiguration engendered by the election of Lula gave a fundamental boost to this trajectory. Legitimized by their contribution to the political process, scholars, policy makers, and social leaders linked to family farming not only began to demand new policies and greater institutionalization but also became protagonists of those processes, occupying important state posts (Leite 2016).

Three Referentials and Generations of Family-Farming Policies
The public policy "referential" (référentiel) is a concept proposed by Jobert and Muller (1987), who define it as the substantiation of ideas into policy instruments that make public action effective. In other words, a referential is a representation of the reality social actors (mainly policy makers) use to interpret public problems and to create solutions for them. The concept can be separated into two levels. The global (or general) referential refers to a framework for interpreting the reality beyond the limits of a sector, field, or policy. It is a wide representation of a social paradigm that prevails in a specific context (e.g., developmentalism, liberalism). In turn, the sectoral referential is a representation of a specific field or sector (Muller 2008). In general, alterations in the global referential result in changes in the sectoral referential.
There are three generations of family-farming public policies, each associated with a specific referential (Grisa and Schneider 2015). The first generation, from the mid-1990s, encompasses instruments that are underpinned by an agricultural referential for rural development. In line with the mainstream global (neoliberal) referential, these policies aimed to correct market failures and, at the same time, to structure family farming for economic integration into value chains (Guanziroli 2007;World Bank 1994). In addition to PRONAF, this generation also includes policies addressing rural settlements and production and price insurance, including the Family Farming Insurance program of 2004 and the Family Farming Assured Prices Program of 2005. These are differentiated agricultural policies directed at a specific public, although they are still similar to those historically designed for the larger agricultural sector. Such policies remain in force today and account for the majority of resources, especially rural credit-a mechanism that perpetuates itself as the driving force behind the modernization of family farming and whose effects, according to some authors, are not quite different from those of the 1970s: it is concentrated on more capitalized farmers, commodity production, and especially the southern region (Aquino and Schneider 2015).
The second generation of policies began shortly after the creation of PRONAF, when the idea that the state should act more effectively to fight poverty (considered a market failure) gained space on the international agenda (Grisa and Schneider 2015). Thus, greater proximity was demanded between family-farming policies and the Solidarity Community Program, created in 1995 by the government of Fernando Henrique Cardoso. Accordingly, rural labor unions and social movements also began to demand adequate public policies for the most vulnerable rural population. This was accomplished by creating new modalities of rural credit (PRONAF) aimed at the groups, regions, and productive activities less served by that program (Grisa 2012).
The emergence of a new hegemonic political coalition following Lula's election further strengthened efforts to combat poverty and hunger (Bohn 2011), spurring the creation of policy instruments under the same social assistance referential: Crop Insurance Program (2002) (2011). In all these cases, the focus was on improving rural living conditions (e.g., access to food, other basic rights) and rural infrastructure.
This referential soon became the subject of a dispute between two narratives (Grisa and Schneider 2015). Representatives of family-farming social movements, public officials, and the researchers who identified with the movements started to advocate for these policies in view of the need for the state to pay greater attention to the most vulnerable rural groups. In their view, social assistance policies should not only ensure social reproduction of these groups but also facilitate their chances of productive inclusion so they could gradually overcome the structural barriers to agricultural and nonagricultural markets (Niederle 2017). At the same time, representatives of the coalition associated with agribusiness, the National Confederation of Agriculture and Livestock (Confederação Nacional da Agricultura e Pecuária, CNA), began to claim this same referential, as did some academics and public officials (FGV and IBRE 2010;Navarro 2016). From this perspective, noting the high concentration of the national agricultural production, Alves and Rocha (2010) suggest that, for most family farmers, fostering agricultural production would not solve the problem of their poverty. Therefore, "a good deal of social assistance policy is necessary for farmer families in the countryside" (Alves and Rocha 2010, 288). This leads directly to the claim that agricultural policies like PRONAF should target more established farmers who have access to technologies and markets.
However, a new generation of public policies began to challenge this dualistic view of agricultural and social policies (Grisa and Schneider 2015). Lula's election also created room for other actors advocating the construction of new sustainable markets to promote food security as an alternative to the inclusion of family farming. Advocated mainly by public officials, scholars, social movements, and nongovernmental organizations working under the umbrella of the National Council of Food and Nutritional Security (Conselho Nacional de Segurança Alimentar e Nutricional, CONSEA), this referential guided the actions of the Ministry of Social Development (Ministério do Desenvolvimento Social). In this view, the Zero Hunger Program is remarkable, as are its several initiatives to promote access to food, to foster diversified and agroecological production, and to promote food quality: Food Acquisition Program (2003), National School Feeding Program (2009), and National Policy on Agroecology and Organic Production (2012).
Though emerging at different times, these three generations of public policies, associated with three distinct referentials (agricultural, social, and food security), coexist in the public arena (Grisa and Schneider 2015) but not harmoniously. There are tensions, sometimes escalating or sometimes waning, depending on the power at stake, such as elections, governmental negotiations, political and economic context, familyfarming unions' and movements' ability to pressure, and the very influence of the agribusiness coalition.
The conditions for the prevalence of one or another referential are also associated with disputes over the place and role of family farming in rural development strategies.
Since the end of Dilma Rousseff's first term, many discussions have dealt with this subject, and several questions about the results achieved by these policies have arisen. Some actors began to construct a narrative of the exhaustion of such policies, suggesting the need for a full redesign of public action. Despite being directed by the agribusiness coalition, this narrative also engaged social activists (e.g., Via Campesina, agroecological movements) and scholars dissatisfied with the limitations of existing policies. However, as we discuss shortly, there were already considerable differences in the way each group devised alternatives for a new referential, for example, in their positions in support of or against the dissolution of the MDA.
Another narrative mainly supported by MDA officials, though also by researchers and rural labor unions (CONTAG and FETRAF), emphasized the need to review existing policies and especially improving coordination between programs that seemed to have contradictory rural development strategies (i.e., social policies to alleviate rural poverty, which was partly produced by indebtedness resulting from credit policies). Furthermore, this narrative also sought to strengthen the idea that a new generation of policies should be designed to include improvements in infrastructure for agricultural production, processing, and transportation and technical assistance to promote new economic alternatives.

Narrative Battles: The Conservative Turn
In public policy analysis, there is a growing interest in the performativity of discourses that create a common cognitive referential guiding public action (Schmidt 2008). The concept of narrative is at the core of these studies. Policy narratives are discursive resources used by groups of actors to persuade other actors into creating or changing a policy (Jones and McBeth 2010;Radaelli 2000Radaelli , 2006. They can be understood as arguments, or truths, that try to influence a policy process. According to Radaelli (2000), narratives are stories that restore a temporal and causal order to social problems, thereby suggesting measures to solve them.
The argumentative sequence of a narrative establishes a plot that comprises beginning, middle, and end. The denouement, then, is the moral of the story, whose aim is to suggest solutions and bring about changes in public policy (Jones and McBeth 2010). Consequently, the appropriation of the discursive rhetoric guides the action of policy makers and other social actors involved in public action. The instability resulting from distinct concurrent narratives is inherent in the conflict between different actors and coalitions. Therefore, it is common for a situation to be interpreted differently in each narrative, the essential function of which is to object to the current referential or to argue in favor of it.
Besides being challenged by agribusiness, the referentials guiding family-farming policies are rife with inherent contradictions, ambiguities, and conflicts in the coalition that supports this social segment. The debate has intensified, especially after the 2009 release of the Agricultural Census 2006 (IBGE 2009). For the first time, the census distinguished family from nonfamily farms. A cooperative project between MDA and IBGE compared data to show the economic and social relevance of family farming. The first published data revealed that the country's 4.3 million family farms (84 percent of total farms), despite occupying less than 25 percent of agricultural area in use, accounted for 38 percent of gross production value. Specific statistics also stated that family farming accounted for 70 percent of food consumed in Brazil (França, Grossi, and Marques 2009;Marques, Grossi, and França 2012). Embedded in the discourse of government officials and social activists, and in many scholarly studies, this information has rapidly spread, giving rise to the narrative that family farming is the main responsible actor for domestic supply. This narrative about the importance of family farming to food production gained more legitimacy after debates on the global crisis in food prices in 2008, which also affected Brazil (Maluf 2008).
The backlash to this argument included several argumentative and statistical strategies. The agribusiness coalition soon realized that another narrative was needed, one more favorable to its interests, on food prices and possible solutions. Thus, an often presented data point showed that only 423,000 rural establishments of the country's nearly 5.2 million establishments (or 8.19 percent) accounted for 85 percent of gross production value (Alves and Rocha 2010). This point began to be used as evidence of the inefficacy of public policies in promoting the economic integration of family farming, as well as to play down the importance of family farming in food security.
Another research project was established, this time between CNA and the Getúlio Vargas Foundation (Fundação Getúlio Vargas, FGV), one of the most important Brazilian think tanks, to present "a reliable answer to the question about who produces what in the field: how much and where" (FGV and IBRE 2010, 1). In this case, the adopted classification of rural establishments contrasted farmers who were qualified and not qualified to access the PRONAF. Considering that the criteria to access credit are more restrictive than those described in the Family Farming Law (used by the Agricultural Census), the results indicated that Family Farming (qualified under PRONAF) would cover 64.4 percent (not 84.4 percent) of the total number of establishments and would account for 22.9 percent (not 38 percent) of gross production value. The document also suggested that family farmers, especially milk, pork, and poultry producers, were largely dependent on the production of non-PRONAF farmers, given the acquisition of agricultural commodities for animal feed (intermediary consumption). Hence, it also argued that the categories of family and nonfamily farms were inappropriate. The document articulated a narrative of complementarity regarding commodity chains, which sought to sustain the emergence of a new type of referential for policies that would be concentrated in a single ministry and focus on productive sectors rather than social groups (FGV and IBRE 2010).
Besides challenging the number of rural establishments and share of gross value, the methodology also raised a new debate involving a large contingent of medium-sized producers who were identified as "rural middle class." The differences between the IBGE and CNA-FGV classifications (20 percentage points for number of establishments and 15 for gross production value) supported the existence of one million establishments that were identified neither as family farms, according to the MDA's methodology (IBGE 2009), or as agribusiness, according to the CNA's criteria (FGV and IBRE 2010). In line with the latter, this group of establishments would be better represented by the logic of entrepreneurial agriculture, which is technologically modernized and integrated into the market and whose labor force is not exclusively family based. The logic conceptually related to family farming does not apply to this stratum and, therefore, it should be served by the agricultural policies of MAPA (Ministério da Agricultura, Pecuária e Abastecimento).
The distinction between efficient and market-integrated farmers and those whose social reproduction logic is based on subsistence is nothing but the dressed-up old distinction between modern and backward. By emphasizing the importance of medium-scale rural producers and questioning the productive capacity of most family farms, the conservative narrative created a new dualism, not between agribusiness and family farming but between productive and unproductive farmers. Productive ones would intensify market integration, expand and facilitate access to agricultural credit, boost agricultural income generation through technology transfer, increase productivity, correct market imperfections, and improve productive infrastructure, especially for food distribution.
A political competition for the rural middle class was established (Picolotto 2012). The strengthening of this narrative led political coalitions to readjust their actions. Regarding rural unionism, the CNA (entrepreneurial farmers) launched an effort to expand its labor union base to the productive segment of family farming, thus amplifying its dispute with CONTAG. As to the state actors, MDA and MAPA sought to readjust their policies to cover this segment. MDA expanded and added flexibility to conditions for accessing PRONAF for more capitalized farmers. In addition, it created a specific program (More Food Program, in 2008) to finance productive infrastructure (machinery and equipment) and invested in the establishment of the National Agency for Technical Assistance and Rural Extension (Agência Nacional de Assistência Técnica e Extensão Rural, ANATER), created in 2014 to coordinate technical assistance to farmers at the national level. In general, these policies were mainly focused on farmers already integrated into markets, which would help consolidate the divide between agricultural policies for productive farmers and social policies for poor farms (Niederle 2017).
In turn, in 2010, MAPA reformulated and expanded the Program for Rural Employment and Income Generation (PROGER), changing it to the National Program in Support of Medium-Scale Farmers (PRONAMP), which would become one of the ministry's main agricultural credit programs. In 2015, MAPA launched the program Middle Class in the Countryside with the objective to promote, within four years, four hundred thousand low-income farmers to the rural middle class. This program envisioned partnerships with state governments, municipalities, labor unions, and the parastatal Micro and Small Enterprises Support System, which provides technical assistance and training to improve efficiency, profitability, and management of rural enterprises. In general, MAPA's policies focused on increasing food supply through technological increments; type of labor, whether family or hired, was less relevant.
In line with this narrative, several authors began to support the emergence of a new agricultural and agrarian model, one in which the notion of family farming and the policies especially designed for this segment, particularly agrarian reform, would lose their relevance (Buainain et al. 2013). However, even within this coalition disinclined to family farming, the proposed solutions diverged, taking two non-mutuallyexclusive stances. On the one hand, the focus was on social policies to ensure social reproduction of poor family farmers (about three million farms). Such policies would mitigate these farmers' situation until most farmers could complete the "natural" exodus to cities or see their farms disappear because of a lack of successors (Alves and Rocha 2010). On the other hand was the argument for a new wave of technological modernization, which, though highly selective, would make viable a (small) part of these farms (Navarro 2016). Both ideas presume a "discursive selectivity that imposes a hierarchy on who is able or not to be productively integrated" (Bruno 2016, 1).

Immediate Effects of the Conservative Turn
In large measure, the support of Lula's and Rousseff's governments for family-farming public policies was associated with Brazilian economic growth, partially driven by the commodity price boom of the 2000s (Sauer, Balestro, and Schneider 2017;Bresser-Pereira 2015). According to Gudynas (2012), it reinforced the role of state as a compensator that promoted a neo-extractivist (or neo-developmentalist) mode of accumulation in agriculture yet also used its dividends to support social policies.
Following the financial crisis of 2008 and the associated reduction in commodity prices, the government was obliged to maintain high levels of public spending to sustain an anticyclical strategy of development and attend to its allies' demands (Singer 2015). Although this assured the political base in Congress for Rousseff's election in 2010 and reelection in 2014, it also created unsustainable public debt, reducing the creditworthiness of the government. This problem was among the most important discursive resources (together with corruption scandals) in the conservative narrative to support the massive protests in July 2013 and July 2016. The economic and political crisis worsened, and the government completely lost support in Congress and the streets. The window of opportunity for a neoliberal return was again opened.
Soon after Rousseff's impeachment in August 2016, major changes began to take place regarding familyfarming policies. Backed by a broader narrative on the need to reduce state spending and control the government's budget, the MDA was shuttered and its attributions reallocated to other ministries. First, given the understanding that family farming should be a focus of social policies, most MDA policies were reallocated to the Ministry of Social Development (MDS), later the Ministry of Social and Agrarian Development (MDSA). To this end, the Special Department for Family Agriculture and Agrarian Development (Secretaria Especial de Agricultura Familiar e do Desenvolvimento Agrário, SEAD) was created in this ministry. However, as Mattei (2016, 9) argues, "due to political differences between the various forces that comprise the interim government, this structure did not even work under the MDSA," with SEAD soon transferred to the Cabinet. 3 These institutional changes are an outcome of new ideas regarding the place and importance of family farming and sectoral understandings of rural development in Brazil. Through this process, a relatively autonomous ministerial structure changes into a department subordinate to the president, thus reducing the political and institutional influence of family-farming actors. As highlighted by an interviewed public officer, these changes indicate that the family-farming "agenda has no place in the government." Furthermore, the "institutional downgrading would imply loss of political capacity for coordination and would directly impact the operation of some public policies" (quoted in Soldera 2017, 134).
Besides the extinction of the MDA, the resources intended for family farming were strongly reduced under the argument of "necessary" fiscal adjustment, and several public policies were paralyzed, reformulated, or put under review. Family-farming unions estimate that, "with the end of the MDA, many policies and programs that were accomplished through struggle by the rural working class over the last few decades ended up eliminated or lost strength due to reduced budget" (CONTAG 2017, 17). Table 1 summarizes the budgetary situation of the most important programs aimed at family farming and agrarian reform over the past four years. Even though resources for most policies were already being reduced during Rousseff's second term, when the government started to control public indebtedness, this tendency has continued since 2016.
The significance of the budgetary cuts is illustrated by the case of the Food Acquisition Program, which in January 2017 was to undergo a 30 percent reduction over the previous year's spending. However, the 2018 Federal Budget Project suggests a 98.7 percent reduction, meaning that one of the most innovative Brazilian food security programs receives only R$4 million (US$1.2 million) (FAO 2015;World Food Programme 2013). In practice, this would make the program defunct. However, political reactions to this, including those from allied government parties, forced a partial recomposing of the budget by means of parliamentary amendments to the project. Whereas it can raise the food program's budget to R$375 million (US$100 million), the program then is dependent on resources controlled by deputies, and the government can still block them according to political disputes and concessions.
It is important to emphasize that reduced resources are not only circumstantial; the reduction also echoes Constitutional Amendment No. 95 of December 2016, known as PEC do Teto dos Gastos Públicos (Constitutional Amendment Bill on Public Spending Ceiling). This rule established an upper limit to the state's primary expenditures for the following twenty years that corresponded to the spending of the immediately previous year, adjusted for inflation. It imposes tight constraints on spending over the long term, not only for family-farming policies but also for those focused on health, education, infrastructure, and social policies that affect the conditions for social reproduction of this particular segment as well as all of Brazil (Paiva et al. 2016). It also reflects the articulation between the sectoral referential and the new neoliberal global referential (Jobert and Muller 1987).
Territorial development was also strongly affected by political and institutional changes. The main evidence here is the elimination of the Department of Territorial Development (SDT) and the National Program of Sustainable Development of Rural Territories (PRONAT), which was responsible for various development projects in 243 territories. The latter had faced problems since Rousseff's first term because the territories were never institutionalized as governance units (Delgado and Leite 2015). Even so, PRONAT had strengthened not only a territorial logic of development but also a learning-by-doing process of participatory public policy, which involved state and nonstate actors (Cavalcanti, Wanderley, and Niederle 2014). Since 2016, however, this territorial, multifunctional, and participatory logic of state action has given way to sectoral and top-down public policies.
Another important change relates to policies of agrarian reform and land regularization. In line with the conservative narrative, and given the increase in land grabbing (Flexor and Leite 2017;Sauer and Leite 2012), institutional changes have sought to make new areas available to the market. These changes were exacerbated by Provisional Act No. 759, proposed in December 2016, andenacted as Law No. 13,465/2017 (Land Regularization Law). The law does the following: • Confers the title of the land to those who obtained land under agrarian reform programs, allowing the sale of the land. • Permits monetary payments for lands acquired for the implementation of agrarian reform projects (previously payment was made only in state agricultural debt securities), thus making the sale of land more attractive for settlers. • Establishes new criteria for beneficiaries, which is now carried out in the municipal sphere and conducted by the Instituto Nacional de Colonização e Reforma Agrária (INCRA), thus reducing the participation of social movements in the process. • Regularizes areas "directly and peacefully cultivated" that were illegally occupied before 2008 (2004 in previous legislation), or after 2008 if the land has been demonstrably occupied over at least five years and is at most 2,500 hectares (maximum was previously 1,500 hectares).
As Sauer and Leite (2017, 32) suggest, such provisions, "besides placing huge extensions of land on the market, point to a drain in the agrarian reform policy in Brazil, prioritizing allocations to public lands that weaken the creation of settlement projects and the regularization of areas owned by traditional communities." These changes in agrarian reform policies have been followed for other initiatives such as deregulation of land acquisition by foreigners and stricter rules for demarcating indigenous and traditional community territories. Some of these proposals have long been debated in Congress, 4 and the political changes under way since 2015 may offer a "window of opportunity" (Kingdon 1995) for their institutionalization. Also noteworthy is the recent decision (No. 001/2017/GAB/CGU/AGU, July 19, 2017) that only the indigenous communities that were in possession of their territory when the 1988 constitution was promulgated are entitled to their land. Such change was influenced by the Parliamentarian Agricultural and Livestock Front (known as the ruralist legislators) and was celebrated by the agribusiness coalition. This offensive against indigenous and traditional communities evidenced a conservative narrative associated with a neoliberal referential that, to promote financial and industrial investments, questions why the state should support "unproductive" groups that, in their view, should not even be recognized as farmers.
Last but not least is the enactment of Decree No. 9,064 of May 31, 2017, which defines the family unit of agricultural production and establishes the Family Farming Registry. On the one hand, this decree emphasizes a new concept-the family agrarian production unit-to the detriment of family farming, which might weaken the political coalition that has coalesced around this social category since the 1990s. This is a conceptual change that exposes disagreement over the term and points to new interpretations of the category and identity of family farming. On the other hand, the decree creates a new mechanism to identify family farming that may overcome the limits of PRONAF's Declaration of Entitlement (depending on the criteria established in the conceptualization).
The ongoing changes in public polices around family farming suggest that rural development is again giving way to agricultural development driven by so-called productive and economically viable agriculture. State action involves deregulation and sanctions to make agricultural activity even more attractive to large national and international investors. In this scenario, there is no room for family farming, including traditional groups and indigenous communities. The conservative narrative shows its ability to guide a new referential, as the self-fulfilling prophecy of family farming's lack of capacity is met not by virtue of conditions but by the change in the state's modus operandi and the breakdown in the social pact that sustained the coexistence of family farming and agribusiness over the past decades.

Restricted Countermovements
Despite the acknowledged advancements of family-farming policies in Brazil, insofar as their limits and deficiencies became more evident, labor unions and social movements increased their criticism of government action. Such criticism varied in intensity and reflected different understandings of the role of public policies and rural development strategies. In general, while family-farming unions (e.g., CONTAG, FETRAF) favored restructuring and expanding existing policies, 5 representatives of Via Campesina (MPA and MST) and a more dispersed group of agroecological movements adopted a discourse attacking some policies, particularly PRONAF, which was to reproduce economic, social, and environmental effects similar to the old programs of agricultural modernization and promotion of agribusiness. According to Marcelo Leal and Sérgio Görgen (2015), two national leaders of Via Campesina, "PRONAF is experiencing its exhaustion stage as a massive public policy for upward mobility, change of technological standard towards agroecology and universalization of the access. In recent years, PRONAF has been seized by the banking logic, was transformed into forms of profit transfer for the industries of agricultural machinery, agrochemicals, and transgenic seeds. It moved away from the peasant logic and now strengthens the rural middle-class agribusiness and the pesticides industry." Inadvertently, these criticisms and the conservative narrative have created a new kind of "perverse confluence." 6 Perverse because both narratives were used to question the results of family-farming public policies but promote different changes in their referentials. While the conservative narrative blamed the policies' inefficiency for their inadequacy in integrating most family farmers into markets and promoting the transfer of modern technologies, the criticism from those social movements denounced the perverse effects of the process, responsible for the socioeconomic vulnerability of family farmers and their dependence on industrial and financial capital.
The intensification of criticism from both sides of the political spectrum has led state actors to a defensive political stance, hindering a broader dialogue on adjustments and innovations that could have been made to existing programs. Only belatedly, on the eve of the impeachment of President Rousseff, did some initiatives involving MDA representatives, social leaders, and researchers indicate a thorough evaluation of PRONAF, but this was not carried out given the subsequent political changes.
When Michel Temer's interim government started the administrative restructuring that led to the end of MDA, the backlash was less significant than most public officials, social leaders, and scholars had expected. Protests were mostly dispersed and disjointed. Internal disputes in the labor union movements regarding appraisal of the new political context contributed to that weak reaction. The period between promises of re-creating the MDA and consolidating SEAD in the Civil House involved disputes between coalitions that wanted a stronger position in the new government. This interval was just enough to break up stronger reactions and cause splits between social movements and labor unions. Today, as the new institutional design becomes clearer, the political coalitions are entrenched in state structures (albeit precariously), and the impacts of the new conservative (and liberalizing) fiscal adjustment begin to drastically affect the implementation of familyfarming policies, the reactions to this process are beginning to acquire clearer contours.
Four examples are particularly relevant to an understanding of the countermovements. First, regarding the key issues that mobilized the agenda of movements and unions during Temer's government, the restructuring of social security achieved the greatest political unity not only among movements and rural unions but also with urban workers' unions. The reactions generated on the eve of a new presidential election (October 2018) obliged the government to withdraw its initial proposal to increase the retirement age (Constitutional Amendment Bill No. 287/2016) and to wait for a new political opportunity to take up the subject again.
If social movements were able to stop the institutional changes around social security policies, that has not been the case with other reforms, and some movements have not been affected in the same ways. For instance, changes in policies of agrarian reform became the main concern of MST. The granting of land titles to settlers (and hence the possibility of selling that land), as well as the transfer of obligations to municipalities, may result in a general disruption of MST's social base and in its decreasing participation in managing these policies. CONTAG's agenda, in turn, remained focused on adjustments to different agricultural policies, particularly rural credit (reduced interest rates and access criteria), demands that are not as important to MST. Because of these distinct political agendas and strategies, until May 2018 it was difficult to find broad unity in the face of the restructuring of these public policies.
A third example involves the different stances adopted by actors in negotiating the Family Farming Harvest Plan of 2017-2018. Since the MDA's launch of this policy management planning tool at the beginning of Lula's administration in 2002, for the first time the CNA (representing the entrepreneurial sector) participated in discussions and the launch (which took place in a formal meeting, without broad social participation). Formerly having participated only in the debates over MAPA's Agricultural and Livestock Plan, in 2017 the CNA proposed extending limits established by PRONAF on credit and annual gross income per farmer, what would imply increasing access to benefits for more capitalized family farmers. With regard to family farming and peasant social and labor union movements, CONTAG took a stance of dialogue, in order to preserve achievements. FETRAF and Via Campesina, in turn, refused to take part in discussions, explicitly manifesting their nonconformity with the "coup" that had brought down the government.
Finally, it is worth highlighting the new government's policy of social participation, especially the reconfiguration of the councils, forums, and committees established since the mid-1990s as arenas of dialogue with civil society to share management of public policies. Following an initial halt to these arenas, some reorganized around new bases. This is the case of the National Council for Sustainable Rural Development (CONDRAF). Soon after Rousseff's impeachment, civil society representatives elected a tripartite commission to run this council. The government had a different perspective: it reduced civil society's participation; mandated the nomination (without election) of CONDRAF's president; and in June 2017, created a working group to submit a proposal for restructuring the council. In this case, while Via Campesina maintained its opposition to the government, CONTAG, FETRAF, and some agroecological movements sent representatives to the working group. Among these latter movements, an understanding prevailed that nonparticipation would mean the ultimate dismantling of various policies as well as SEAD.
The new SEAD officials have announced the maintenance of family-farming policies. The political alliance that controls this space indeed has an interest in strengthening SEAD. Following the MDA's dissolution and the reallocation of its public servants, this secretariat became an institutional structure with little capacity for policy implementation. As the actors who control this office have a peripheral position in the new government, to some extent they depend on maintaining dialogue with social movements and labor unions, albeit with new bases of participation, to broaden their political capital. What makes this context even more complex is that these actors kept up political ties with another labor union confederation (Força Sindical); that confederation, with little influence in rural areas, sees in SEAD's policies an unprecedented chance to engage in dialogue with farmers. Although such an arrangement suggests a logic of political exchanges, the budgetary cuts and SEAD's institutional fragility make difficult any more substantial strategy.

Final Remarks
In Brazil, claims for more or less substantial changes to family-farming public policies have been the focus of debates among social movements, researchers, and public officials in recent years. Despite advancements in terms of institutionalization (e.g., Family Farming Law) and innovations, mainly in territorial and food security policies, the coexistence of different referentials guiding the design and implementation of these policies has exacerbated the contradictions of rural development (Niederle 2018). To some extent, family farming is victim of its own success. Increasing recognition by society, the upward mobility of some family farmers to the rural middle class, and intensified disputes with agribusiness organizations strained the social pact that, since the late 1990s, had allowed the segmenting of social, economic, and political spaces (including the coexistence of two ministries).
Of course, the breakdown of that pact goes beyond conflicts particular to the rural space. To a great extent, it reflects a change in the global referential (Jobert and Muller 1987) that was established following the impeachment of Dilma Rousseff. However, the contradictions that underlie this breakdown were already present in the conservative narrative. Reviving the idea of a new wave of technological modernization for the Brazilian countryside, this narrative proposes a new dualism, no longer between family farming and agribusiness but instead between productive and unproductive farmers. While transferring to the former the burden of sustaining the "modern" dependent agro-export model, which characterizes Brazilian participation in global value chains, it reserves for the latter a marginalized social condition, similar to that once occupied by small producers, a notion against which the identity of family farming was conceived in Brazil.
Although Brazil's political instability does not allow for the suggestion of any trends, this article shows that, in line with the conservative narrative, the main changes in public policies have so far focused on rather structural changes (agrarian reform), which were more innovative or legitimized the ways of life of social segments considered unproductive (traditional and indigenous communities). Conventional policies of rural credit (PRONAF), which support productive family farmers, have not seen substantial changes. However contentious these changes are, opposition to this process is restricted by the difficulty for social movements and labor unions in producing political mobilization and finding common agendas and strategies for action. We wrote this article in 2017, during Michel Temer's government and before the election of Jair Bolsonaro as Brazil's president, so we have not analyzed here new forms of resistance to the advancing conservative narrative and the dismantling of public policies around family farming. These processes are the focus of our ongoing research.

Author Information
Paulo Niederle is an adjunct professor in the Graduate Programs on Sociology and Rural Development in the Federal University of Rio Grande do Sul, Porto Alegre, Brazil. His research interests focus on food markets, family farming, and social theory. Catia Grisa is an adjunct professor in the Graduate Program on Rural Development in the Federal University of Rio Grande do Sul, Porto Alegre, Brazil. Her research focuses on rural development policies, public action, and food and nutritional security.
Everton Lazaretti Picolotto is an adjunct professor in the Graduate Program on Social Science in the Federal University of Santa Maria, Santa Maria, Brazil. His research interests focus on social movements, family farming, and sociological theory. Denis Soldera is a master's student in the Graduate Program on Rural Development in the Federal University of Rio Grande do Sul, Porto Alegre, Brazil. His master's thesis analyzes the evolution of public policies for family farming in Brazil.