Public Policies for Sustainable Territorial Development in Brazil: Between Clientelism and Participation ^†

Abstract

The article discusses the permanence of clientelistic practices and their tensions with the participatory approach adopted within the framework of the Brazilian public policy of rural territorial development. It examines, in particular, the case of local implementation of the National Program of Territorial Development. The results come from the study of the functioning of the Territorial Collegiate and the projects implemented in the Águas Emendadas Territory in the Midwest of Brazil. It uses a socio-anthropological approach of patronage and political participation through the analysis of the social configuration and the relations of instrumentalisation in both the participatory spaces and the projects of this territory. The results show the existence of a not only social but also an affective dimension of clientele practice that can be analyzed as an asymmetrical reciprocity relationship based on the principle of anthropological reciprocity or as a process of unequal political exchange, considering a political science approach.

1. Introduction

This study examines the permanence or reconfiguration of clientelistic practices [1,2], in the context of the implementation of sustainable territorial rural development policies in Brazil [3]. The aim here is to discuss the social (socioeconomic and sociopolitical) dimension of sustainability in the context of rural territorial development policies.

For Carvalho (1997: 134) [2], the notion of clientelism ‘indicates a type of relationship between political actors that consists of granting public benefits, in the form of jobs, tax benefits, exemptions, in exchange for political support, mainly in the form of votes’.

The main studies on clientelism generally focus on elections, election campaigns and clientelistic vote-buying practices. There is very little work on clientelism in rural development policies and projects around the world, and even very few on the permanence of clientelism in political participation mechanisms. This article, based on a case study in Brazil, aims to contribute to reducing this specific gap.

Through successive research and action research projects, I have had the opportunity to verify such practices where I least expected it—during the implementation of territorial policy by Brazil’s Ministry of Agrarian Development (MDA)—precisely because these are programmes based on the participation of beneficiaries in the formulation and evaluation of actions. It cannot be said that this policy was specifically marked by clientelistic practices. On the contrary, the reduced budgets allocated to the various rural territories, combined with the participation and social control of the definition and implementation of territorial actions by the beneficiaries, have probably contributed to a reduction in clientelism compared with previous practices and policies. However, in this renewed political context, the specific object of my analysis was the permanence or reconfiguration of new forms or new expressions of clientelism associated with the participatory approach, or in tension with it.

This case study corresponds to the implementation of the National Territorial Development Program (PRONAT) between 2005 and 2015 in the Territory of Águas Emendadas (TAE), which includes the Federal District (DF) and seven municipalities in its periphery in the states of Goiás and Minas Gerais in central-west Brazil. PRONAT has existed since 2004 and was extended after 2008 by the Territories of Citizenship Program (PTC) that lasted until 2016. It was coordinated by the Civil House but executed by the Secretariat for Territorial Development (SDT) of the MDA. Both programmes, characterised by innovative designs and practices, including the implementation of an intermunicipal territorial scale and a participatory dimension, were dismantled in October 2016 following the congress impeachment of President Dilma Roussef and the transformation of the MDA into the Secretariat for Family Agriculture and Development (SEAD). Since 2019, under the Jair Bolsonaro administration, the SEAD has been closed, and its functions have been transferred to the Secretariat of Family Agriculture within the Ministry of Agriculture, Livestock, and Food (MAPA). The rural territorial policy has been permanently abandoned.

Following this introduction, the article is divided into four parts. The first part provides an international review of the literature on clientelism in rural development projects and policies. The second part presents the Brazilian sociopolitical context during the period of study, the scientific framework, and the study’s methods. The third part examines the social configuration and clientelistic relations in the local implementation of PRONAT in the TAE. Finally, the fourth discusses the results with a reading of the affective dimension of the clientelist relationship according to the principle of anthropological reciprocity.

2. Clientelism and Rural Development: Research Overview

2.1. The Permanence and Strength of Clientelist Networks

Since the 1950s, a social anthropology literature on patronage and clientelism has developed with authors such as Eric Wolf (1966) [4] on Latin America, Frederick G. Bailey (1960, 1963) [5] on India, David Hass (1978) [6] on Thailand, Sydel Silverman (1968, 1977) [7] and Jeremy Boissevain (1969, 1974) [8] on the Mediterranean. These authors make patronage networks the product of a particular situation of economic or political elite domination.

For example, Ọmọbọwale and Olutayọ (2010) [9] examine clientelistic relationships in south-western Nigeria and their possible impact on rural development. They find that rural clients attract the attention of the political class to promote the exchange of goods for loyalty through associations that afford clients a sort of cohesive power and a common front, the basis of their relevance in the political clientelistic chain. This clientelistic chain also serves as the channel through which development projects are conceived and implemented. However, since the projects provided only serve symbolic purposes, they easily collapse: clients may have the opportunity of changing patrons, but they remain subservient to the political/economic elite.

By this fact, a dominant centralised political authority enacts general norms that it does not have the real capacity to implement or impose on local societies. The intervention of the state is therefore affected by a significant margin of unpredictability.

For Bierschenk, et al. (2017: 10) [10] “It is true that state intervenes in local arenas, but it uses patronage relationships rather than its own bureaucratic-co-universalist rules, which reinforces the arbitrary aspect of its intervention. The creation of clientelistic links between representatives of the administrative apparatus and certain actors in local societies does, however, make it possible to reduce the unpredictability of state interventions.”

In spite of the universality of clientelism, it is often seen as a peculiar aspect of southern hemisphere countries’ politics, one which inevitably stifles development

For Williams (2017) [11], development projects such as schools and boreholes in Africa are very often abandoned midconstruction. He examines three plausible explanations through a case study in Ghana: unfinished projects are primarily an inefficient outcome of failed intertemporal bargaining among local political actors, but it is inconsistent to associate them solely with corruption and clientelism as major causes of noncompletion. Fiscal institutions can increase completion rates by mitigating the operational consequences of these distributive pressures.

Koster and Eiró (2021) [12] have developed an approach that builds on theories of brokerage in anthropology and social network studies. Based on long-term ethnographic fieldwork in low-income neighbourhoods in Recife, Brazil, it shows how clientelism is based on informal exchanges both within and outside election periods. Through a study of community leaders, their projects and their search for resources, the article advances a more comprehensive understanding of how clientelism works as a social mechanism in the ordering of life in these neighbourhoods.

2.2. Social and Development Policies vs. Clientelism

Social policies, cash transfer programs and particularly participative local government or territorial development policies were supposed to reduce clientelist practices, especially in rural areas (Abbers, 1998; [13] Daieff, 2016 [14]; Gacitua M et al., 2009 [15]). The literature shows us that things are rather more complex, such as the study of Arun R. Swamy (2016) [16] in the Philippines. Flavio Eiro (2017) [17] shows how the implementation of the ”Bolsa Família” Programme in Brazil is based on informal rules, reflecting the dominant social representations of poverty in Brazil. In addition, the allocation of social benefits depends in part on personalised rapports between the programme’s beneficiaries and political candidates and elected representatives.

For Ansel and Mitchell (2011) [18] vertically integrated, corporatist clientelism in Mexico and more locally oriented, bossist clientelism in Brazil differentially shape the choices of governments to turn piecemeal, discretionary cash transfers into more expansive and secure benefits.

In Kenya, Ringera (2011) [19] examined the multiple layers of formal and informal political powers at the local level using selected Community Constitutional Fund projects in South Intend, Meru County. CDF policy is essentially about distribution of power and resources among different levels of the society and among different interests in their relationships to ruling elites. The study therefore suggests that the institution of the office of Prime Minister in Kenya, while vibrant, is conducive to the provision of goods and service mainly in clientelistic networks.

For Jonathan Fox (1994) [20], “the politics of social policy can tell us a great deal about nonelectoral dimensions of democratization. So many types of regimes are now experimenting with “demand-based” antipoverty funds aimed at making structural economic adjustment politically viable”. Fox examined Mexico’s Solidarity program, Bolivia’s social emergency fund from 1986, and similar programs carried out by Peru, Chile, Colombia, Zambia, Senegal, Ghana, Poland, El Salvador, and Honduras. Similar to Solidarity, some of these new targeted antipoverty programs created political openings for social movements and nongovernmental organisations, while others reinforced partisan clientelistic controls. Peru’s program largely perpetuated semiclientelism. Fox concludes “the degree of political conditionality required for access to these new social funds is a key indicator of the extent of the transition from clientelism to citizen-ship. This focus on the politics of social policy shows that the relationship between electoral competition and the erosion of authoritarian clientelism is not obvious. In other words, electoral competition can either strengthen or weaken coercive clientelism, which in turn can be either strengthened or weakened by electoral competition”.

2.3. The Capture of Democratic Political and Economic Reforms by Traditional Elites

As Fox (1994) conclusion, clientelistic machines around the world have also shown that the threat of electoral competition can also create incentives for elites to limit political choices sharply. For example, Anderson et al. (2015) [21] studied the operation of local governments (Panchayats) in rural Maharashtra, India. Elections are freely contested, fairly tallied, highly participatory, noncoerced, and lead to appointment of representative politicians. However, beneath this veneer of ideal democracy they find evidence of deeply ingrained clientelist vote-trading structures maintained through extrapolitical means. Elite minorities undermine policies that would redistribute income toward the majority poor. These elites use their dominance of land ownership and traditional social superiority to achieve political control in light of successful majoritarian institutional reforms.

In general, international norms of social, economic and political rights are presented as a means of transforming social relations in developing countries (Brinkerhoff and Goldsmith, 2004) [22]. Schneider and Zúniga-Hamlin (2005) [23] examine this kind of political process in rural Peru. “International norms of social, economic and political rights are presented as a means of transforming social relations in developing countries. Yet, when rights norms are introduced into domestic practice, they do not always produce liberal, democratic results. Instead, rights and local practices of clientelism mix”. For these authors, this calls for an explicit politics of advancing rights by any means necessary: “accepting hybrid forms when inevitable, incorporating excluded groups when possible, and striking alliances that displace traditional elites”.

If such studies have underlined the advantages of clientelism as an alternative to violence and class conflict, C. Escobar (2002) [24] shows that the case of Sucre in Colombia seems to suggest that clientelism does not preclude the possibility of violence. For Escobar, the combination of clientelism, which dominates the institutional political channels, and the escalating violence outside these channels has taken over the countryside. “Because clientelism rests on the exclusion of most of the population from the exercise of power, it undermines the value of laws, the basis of the democratic state. Institutional reforms, social programs, and grassroots organizations can do little to recover the political rights of the rural population if the rule of law and the state itself are eroded by the escalating warfare between guerrilla and paramilitary armies”. (Escobar, 41:2002) [24].

Because the culture of clientelism in Latin America is rooted in long traditions of economic dependency and political exclusion, only local community organisations providing both economic support and a locale for democratic participation can serve as the basis for developing a culture of democratic citizenship. But the obstacles are numerous, the paths often very long and tortuous, as shown by the following elements from this brief review of the world literature.

3. Institutional Context, Theoretical and Methodological Framework

3.1. Brazil: A Still Clientelist Country

As developed recently by Carvalho (2003) [25], Brazil is a country with a clientelist tradition [1,2]. The seminal work of Nunes Leal (1948) [26] analysed the Brazilian political system of the 19th century, which was based on the power of militias led by large landowners called “coroneis” (colonels). These colonels imposed the vote on peasants and rural workers over whom they had authority and a duty of protection, which was generally guaranteed by the central political power. Later, with the advent of the Republic (1889), President Campos Sales (1898–1902) created the “Governors’ Policy”, strengthening sociopolitical relations through a chain of favours from the President to the rural populations, via the governors of the federated states and, of course, the colonels as the masters of local politics [26,27].

To explain the patrimonial and bureaucratic character of the Brazilian political system, which is based on individualistic, personalised, and very unequal power relations, Faoro (1958/2001) [28], mobilised colonial history and the country’s slave heritage, indicating the late permanence of the colonels or their heirs that persists even today in the federated states of the northeast. Lanna [29] (1995) analysed the reciprocal favour relationship between local politicians and the rural population in the Northeast Region, defending the theory of the sacred gift in terms of a “divine debt”. Also in the Northeast Region, Bursztyn (1984) [27], examined clientelistic relations in the context of land use planning in the 1970s and 1980s, which centred around large integrated development projects and access to rural credit. Thirty years later, in the afterword to the third edition of his book in 2008, he noted that this type of relationship continues, but through new mediators: municipal councillors, agricultural technicians, and nongovernmental organisations (NGOs) have replaced the “colonels”. I made the same observation in the irrigated agriculture projects of the State of Bahia, where technical services and NGOs compete with the Catholic Church and regionally elected officials for the power of mediation and supervision (Sabourin et al., 1997) [30].

José de Souza Martins (1994) [31] included the Catholic Church among these trusteeships; today, the evangelist churches should be added [32]. In his book O poder do atraso (The Power of Backwardness, 1994 [31]), Martins, one of the pioneers of rural sociology in Brazil, defined clientelism as “a relationship of exchange of political favours for economic benefits”. For Martins, it is essentially “a relationship between the powerful and the rich and not only between politicians and their voters”. In fact, according to him, long before the vote of the poor, the State established a relationship involving the exchange of favours with the rich. It is in this context that the “culture of debt” develops (Lanna, 1995) [29] and that access to political, social or constitutional rights is negotiated or reduced on the basis of personalised favours (Martins, 1994). The clientelistic politics of voting has replaced old bonds of loyalty with the offer of material benefits. This intermediation takes place through the political currency of favours, which implies a debtor status [1,2,27]. According to Dagnino (1994) [33], the political culture of clientelism is implanted in various spheres of social life within Brazilian society: family, state, work, educational institutions, medical and social assistance, and culture. In this tradition, social relations are based on clientelism and patronage, which are factors of inequality and violence [33].

Yasbek (1993:50) [34] referred to the welfare culture as a “matrix of favours”, in place of social development, which should, a priori, be in tension or contradiction with active social movements that claim full citizenship [34]. This is what could be expected from the participatory approach to public policies, especially social or rural development, as claimed by rural social and trade union movements [34].

3.2. New Participatory Policies

Since the 1990s, Brazil has also become a real laboratory of participatory public action [35,36]. For Pelegrini and Rovere (2011: 4) [37], “social and political participation has an intrinsic value as a right and duty of the population to participate in decisions that affect them, and at the same time, an instrumental value, guaranteeing the necessary political support to make viable the redistribution of power and resources that allow the reduction of inequities”.

Moreover, according to Guilhardes and Costa Conssenza (2015: 1) [38], “social participation in the management of public action must be used as a means of influencing and contributing to the construction of local public policies, through relations between various social actors and the State”. These authors recalled that the principle of popular participation in the management of public action is enshrined in Article 1 of the Federal Constitution of 1988, which states that “all power emanates from the people, who exercise it through the representation of elected officials or directly” (Brazil 1988 and 2008) [39,40]. This direct popular participation is also seen as a form of approximation between society and the State, taking into account a diversity of interests, and, above all, the emergence of spaces for debating collective interests, particularly around sectoral or territorial councils, management committees, chambers, and forums (Sabourin, 2015) [35], (Gohn, 2011) [41].

In fact, alongside citizen participation via elections, referendums, or plebiscites—that is, so-called representative democracy (citizens elect members of the executive or legislative branches)—there has been a gradual development of participation by organised civil society in the formulation or implementation of programmes, projects, or public policy instruments in discussion or negotiation forums. This second form of social and citizen participation is also referred to as “technical democracy” (Callon et al., 2001) [42]: the elected or appointed representatives on the management boards or committees have a technical skill, experience, expertise, or practise in a given field of activity.

Decentralisation has led to the creation of sectoral municipal councils (health, education, rural development, food security, Agenda 21), participatory budgets (Porto 2017) [43], and, finally, to the territorialisation of development [43,44,45,46,47]. In the rural sector, the dual dimension of territorialisation and participation began to take hold from 1996 with the creation of the Municipal Councils for Sustainable Rural Development (CMDRS) in the wake of the initiation of the National Programme for the Strengthening of Family Agriculture (PRONAF) [37]. One component of the programme did not deal with individual credit but, rather, with the allocation of grants to finance collective municipal infrastructure, which were decided after a diagnostic process within the CMDRS [44]. From 2003, the process was expanded with the allocation of funds from the National Infrastructure Program (PROINF) for intermunicipal actions discussed and decided in the Territorial Council, which brings together representatives of family farmers’ organisations and the public sector.

Whatever the successes and limitations of the multiplication of councils and forums for territorial public action, various authors have noted the maintenance of clientelistic relations in this new participatory framework [47,48,49]. For Dagnino and Tatagiba (2010: 179) [50], this persistence is contrary to the expectations that presided over the creation of participatory bodies: “It was hoped that the creation of these spaces intended to channel demands presented as rights to debate and negotiation would put an end to the traditional mechanisms of favourable relations”.

This observation is not entirely paradoxical, however, insofar as the introduction of participatory approaches cannot claim to eliminate several centuries of political culture; although poverty has recently been reduced, social and political inequalities persist. This is the case in terms of access to information, health, education, and, of course, decision-making capacities.

One of the interpretations of current clientelism, including the clientelism found within social movements, is that of public demands channelled from private meetings, “that is, outside the participatory channels within which these same movements act” [50] (Dagnino and Tatagiba, 2010: 182) or in parallel spaces (Massardier, 2008) [51]. Dagnino and Tatagiba (2010: 185) [50] proposed, on the one hand, the consideration of “the coexistence of distinct cultural matrices that place side by side the discourse of rights and the mobilization of personal networks, and, on the other, the insistence on autonomy and the practice of clientelism. The principles of participation, citizenship, and democracy coexist with the use of personalized and clientelistic relationships as a way of accessing the State, in a situation (...) defined as a constitutive tension”. However, it is not a situation of opposition but a contradictory and ambivalent combination that persists and guides the action of social movements.

3.3. A Framework: Clientelism as a Reciprocity Relationship

In this sense of the coexistence of cultural matrices, I propose to deepen the conception of J.-F. Médard (1976: 107–108 [52], who defined clientelism as “a relationship of reciprocity or reciprocal exchange”. Médard [52] notes that this reciprocity, even if it is mutually beneficial to both parties, is ambiguous because it is unequal. However, he concludes “the relationship of dependence in the client relationship is in reality based on reciprocity” (1976: 109), which implies not only the expectation of a return of a service or favour through political membership or voting but also through the reproduction of the bond between the two parties. The conjunction between bilaterality and inequality determines the client relationship, but more complex structures can be built on this basis’ (Médard, 1976: 114) [52].

According to the theory of reciprocity in anthropology (Temple [53], Martinez-Gutierrez [54] Sabourin [55]), the clientelistic relationship corresponds to a structure of centralised and generally unequal reciprocity, which is comparable to certain forms of redistribution in the sense proposed by Karl Polanyi (1944) [56]. This asymmetrical relationship of mutual services and favours generates social ties and is, therefore, irreducible to market or money exchange [54]. The difference between the unequal exchange relationship and the reciprocal relationship (even if unequal) lies in the social bond produced and the associated human values, (Polanyi, 1994) [56] as we have seen in the case of rural Brazil, (Sabourin, 2077) [55]. Thus, the client relationship differs from reciprocal exchange because the instrumental relationship is associated with an affective relationship that redoubles the production of social ties.

Political clientelism in northeast Brazil reproduces the legacy of rural patronage, that is, the “morada” system of sharecroppers (Garcia, 1980, [57]. This bilateral and asymmetrical reciprocal relationship still functions due to dependence but also because of the respect for the human values attached to reciprocity relationships. The peasants’ loyalty to the obligatory or “captive” vote [57], even in today’s less-violent context, is based on respect for their voice: “The Silva family helped me, and I will vote for their candidate. I cannot deny the last good that is mine, my voice”. The return to democracy in 1984 in a context of illiteracy, compulsory voting, and socioeconomic dependence revived the clientelistic practices of the Brazilian political apparatus (see Bursztyn) [27].

The novelty of my research was to associate an anthropological approach of reciprocity to the policy science approach of clientelism (Avelino, Carvalho, Briquet, Medard, etc.). Regarding anthropological reciprocity, clientelism is an asymmetric mutual relationship of reciprocity (Sabourin, 2012) [58], (Landel, 2009) [59].

3.4. Methods

The study’s methods mobilised a sociological and anthropological analysis of the functioning of the Council and Commissions of the TAE and of the projects developed in this arena [60,61,62]. The empirical approach combined the study of the social configuration of actors (Elias) [63], interest groups (Grossman, E.; Saurugger [64], 2006), and projects formulated by the territorial Council of Aguas Emendadas (COTAE) and comprised participant observation during meetings, archival analysis, and interviews between 2005 and 2015 during three action-research projects. (Project ‘Contribuição dos dispositivos coletivos dos agricultores a renovação dos instrumentos de políticas públicas de desenvolvimento rural’ (CNPq, UnB-SOL, Cirad, 2005–2009); Project ‘Renovação da Ação Pública Territorializada de desenvolvimento rural’, RAPT (Cirad, UnB-CDS, Capes, 2009–2014); Project ‘Território, Pobreza e Políticas Públicas: uma abordagem pela territorialização’ (Capes-Cofecub, CPDA-UFRRJ, Cirad-Art-Dev, UFSC-CAA, UnB-CDS, 2014–2016).) Meanwhile, the sociological approach examined the recomposition of social relations, power games, alliances, and co-optations between actors within the framework of the new spaces of coordination and discussion that constitute the plenary sessions and the commissions of the territorial collegiality.

Based on this anthropological approach and following Marcel Mauss ([65] 1931), the study proposes a consideration of the importance of structuring relations of reciprocity between political actors and analyses clientelism as a form of asymmetrical reciprocity [53]. According to Temple’s (1998) [53] classification, there are five elementary structuring relationships of reciprocity. Binary reciprocity has two subtypes: (i) Face-to-face relationships, which correspond to practices of cooperation, alliance, and mutual aid between two individuals, two families, two groups, or two organisations. This type of reciprocity generates values of respect and friendship. (ii) The sharing of resources, the second subtype, develops trust among the members of a group. (iii) Ternary reciprocity (involving at least three parties) can be unilateral (between generations) or bilateral (reversible between several subjects) and produces values of responsibility between subjects. (iv) Generalised ternary reciprocity (symmetrical relationships between all parties) produces values of justice and equality. (v) Finally, centralised reciprocity exists when a centre of power and redistribution has the possibility of collecting contributions from subjects over whom it exercises both a power of domination and a responsibility for distribution. This corresponds to the figure of the colonel, the farm owner, or the local boss.

This asymmetrical relationship generates prestige for the power centre and submission for the subjects. It may correspond to many cases of clientelistic relations in rural Brazil, as defined by Souza Martins (1994) [32] and Médard (1976) [52].

The data collection process combined three techniques: (i) participatory observation of numerous CMDRS meetings in the municipalities of Unaí and Cabeceiras (Minas Gerais), the councils of three DF administrative regions (Brazlãndia, Planaltina, and São Sebastião), and the TAE Council between 2005 and 2013; (ii) interviews with COTAE leaders, PRONAF beneficiary farmers, PRONAF and PTC officials between 2008 and 2014; and (iii), the analysis of COTAE archives and project documents.

4. Instrumental Relations of Clientelism in the TAE

4.1. The Case of PRONAT’s Territorial Public Action

In Brazil, the participatory and territorialised approach to rural development began in 1996 with the implementation of the National Program for the Strengthening of Family Agriculture (PRONAF) and the creation of the Municipal Councils for Sustainable Rural Development (CMDRS), which had a responsibility to decide upon municipal investments in collective infrastructure (MDA, 2008) [66].

The mayors who presided over the CMDRS often took advantage of their power to co-opt representatives of the family farmers’ organisations of their choice and, thus, capture new federal resources while favouring their clientele. As these practices were criticised by social movements, the Rural Territorial Development Policy created by the Lula da Silva government in 2003 sought to reintroduce equity and democracy in favour of family farming, while going beyond the limits of project funding at the municipal level. The principle of PRONAT was to plan and implement collective projects on the multi-municipality scale through consultation between organised civil society and the public authorities within the framework of territorial rural development collegial bodies, the CODETERs [35]. The aim was to promote intermunicipal dynamic economies of scale and bypass the powers of local oligarchies or municipal executives dominated by traditional political parties.

Within this framework, in 2004, about 100 priority territories were identified and approved by the SDT, which offered methodological support for the preparation of diagnoses and Territorial Development Plans. PRONAT thus introduced a new level of territorial management between the federal State and the municipality, creating groups of 10 to 20 municipalities, as in the case of the TAE (Figure 1) around Brasilia.

The MDA sought to bring about a territorial and political rebalancing in favour of marginalised rural areas with a high density of family farmers. To give coherence “to the projects of the territories as well as to the project territories” [46,66], the aim was to build development territories around a geographical, cultural, or productive identity (typical product). However, sometimes, as in the case of the TAE, a clientelist political logic based on the interests of the Federal District has drawn the contours of a territory where the rural component is not clear-cut or is even artificial. In fact, the TAE is comprised of the DF’s supply basin, which is essentially urban or periurban since it includes the federal capital, Brasília. This territory also has the particularity of bringing together municipalities from several federal units: the DF, the state of Goiás (GO), and the state of Minas Gerais (MG).

Territorial management is collegial and participatory; however, the CODETERs are essentially consultative. This new level of governance, without its own legal status or bureaucracy, remains fragile. Even though family farmers initially had a representative majority, they generally had difficulty implementing the projects they had formulated.

The case of the TAE illustrates the difficulties faced by the COTAE in developing quality projects with a territorial dimension, although, above all, it reveals its dependence on the technical and administrative services of the municipal, state, and federal levels of government.

4.2. Functioning of the Águas Emendadas Territorial Council

The composition of the COTAE brings together different colleges (civil society, family farmers, elected representatives or local officials, and technical services). Initially, it was favourable to family farmers and civil society organisations. From 2008, a directive from the Presidency of the Republic imposed parity between government (public sector) and civil society representatives, eliminating the possibility of establishing a majority in favour of rural social movements. The case of the COTAE, thus, well illustrates the effects of the conjunction between the heritage of clientelist routines (Briquet, 1998) [67] and the participatory approach (Schneider et al. 2003) [68].

The first problematic aspect concerns the inclusion or exclusion of certain sectors of the programme’s target population, family farmers, represented by their organisations. The power to select projects is no longer in the hands of mayors and municipal executives, as in the case of the CMDRS and the former PRONAF “Municipal Infrastructures”. Rather, this selection is orchestrated by the services of the federal government (the MDA) or the federated state with the passive or active complicity of social movements. In the COTAE, this was the case for the minority sectors of the family farming segment (native and traditional peoples, quilombolas, and landless people), who offered no electoral interest (and/or had no political or administrative intermediaries), and for the beneficiaries of agrarian reform linked to the Landless Workers Movement (MST). While the Águas Emendadas territorial development plan indicated the regularisation of the land titles of the landless as a priority, this meant the exclusion of a large segment of the target population from the programme.

The second mechanism observed within the COTAE is the permanence of old clientelistic relationships between the agents of the Technical Assistance and Rural Extension Services (ATER), especially the Technical Assistance and Rural Extension Enterprises (EMATER), and the representatives of family farmers. The three farmer coordinators of the COTAE were, at the time of its creation, employed by the Secretariat of Agriculture of the Federal District, the Rural Agency of Goiás, and the Municipality of Unaí, Minas Gerais. This status gave them the advantage of being able to easily attend numerous meetings and negotiations; however, in practice, they had no autonomy from their employers who formatted their position (Avila et al., 2011) [35,61].

The third modality is the instrumentalisation of the collegial participation mechanism by a supervisory body, which, in the case of the TAE, is the EMATER. At the time of the first COTAE assembly, there was significant representation of EMATERs in all the collegial bodies: the federal service through its agents seconded to the MDA; the State service through the Secretariat of Agriculture; the guardianship of EMATER; the municipal level through the local EMATER offices; civil society through its representatives at the CMDRS; and even a seat allocated to the Association of EMATER servants. This omnipresence quickly translated into the formulation and approval of projects that benefited their institutions more directly than family farmers (usually equipment, computers, and vehicles for the technical agricultural services).

The fourth mechanism observed is the conjunction of clientelism and activist participation. The SDT of the MDA is a new structure within a newly created ministry with no permanent civil servants; therefore, it needed to hire many consultants in a short time. This process, which was conducted without selection or competition during the first few years, gave free rein to the expression of the forms of favouritism and nepotism that are common in Brazil—this time militant, as it was marked by the ideological adherence to the Workers’ Party (PT) current that dominated within the MDA.

Another practice designed to compensate for the lack of personnel was the SDT’s direct financing of partner NGOs to provide support to the CODETERs in the form of training advisers, technical studies, or the hiring of a territorial animator. When the COTAE was created, this facilitator was a technician from a local NGO linked to the National Confederation of Agricultural Workers (CONTAG). His union alliance, and especially his nonalignment with the EMATER technicians, led to his rapid replacement: first by a technician from an agronomist NGO favourable to EMATER but not aligned with the PT and, thereafter, by a consultant directly hired by the SDT-MDA and a militant of the “good” PT current. The role of the territorial facilitator financed by the SDT’s NGO partners is at once ambiguous, militant, and precarious because they are poorly and very irregularly paid. However, these agents benefit from a form of professional remuneration for their activism, as Nonjon (2005) [69] observed in French territorial mechanisms. On the one hand, they are subject to precarious employment, lack of recognition, and occasional and irregular remuneration for their activism, while on the other, they work on a daily basis with well-paid career civil servants who are comfortable performing the same function. Thus, they will likely be tempted to compensate for this lack of resources by directing projects or study contracts in the territory towards the NGO by which they are employed, the trade union that promoted them, or the politician who sponsors them. In the case of the TAE, the facilitator commandeered the vehicle intended to support the marketing of farmers’ products and ensured that the management of the training centre was entrusted to his NGO.

Similarly, at the federal level, the neocorporatist relationship between the MDA and the social movements of family farming (as in other ministries with other sectors) has given way to clientelistic relationships. We can observe a form of “distributed neo-corporatism” (Bobbio et al. [70]) rather than a social or territorial management [68] according to the French model between the FNSEA and the Ministry of Agriculture, cf. Jobert and Muller [71], because each Secretariat or sector of the MDA was entrusted to one of the social movements of family farming (CONTAG, FETRAF, MST, Family Farming Cooperatives) cf. Sabourin, 2015, [35].

In the case of the COTAE, while also leading to resistance or breakdowns, as in the case of the Landless Workers Movement (MST) and Via Campesina Brazil organisations, or even from public service agents hostile to participation. Among the three levels of government, the most virulent corporatist resistance rarely comes from federal state agents, who are comforted by their privileged status as senior officials. Opposition and boycotts of territorial dynamics are mostly observed when the state or municipal government are in opposition to the federal governing coalition, which was the case in the DF and the other two federated member states of the TAE during this study. Strong corporatist or clientelistic reactions and relations also oppose family farmers’ unions and NGOs among themselves, based on both ideological and financial considerations in the competition for the establishment and control of social bases (Jobert, 1983) [72].

4.3. Territorial Projects and Client Relationships in the COTAE

Despite the PRONAT proposal of “territorial and participatory social management” [68], the analysis of the elaboration, selection, and financing of “collective” TAE projects shows several types of dysfunction and decision making in parallel or “ad hoc” spaces outside the COTAE framework (Massardier et al., 2012) [73], (Egret, 2013) [74]. The content and funding modalities of territorial projects depended largely on direct relations between MDA officials and beneficiaries through various intermediaries or mediators: the territorial animator themselves, SDT consultants, partner NGOs, and, of course, EMATER agents [61]. These mediations stem from local clientelism or neo-corporatist administrative practices [69]. From the moment, the State proposed a new territorial level to trade union movements and civil society (NGOs) without administrative capacity and, above all, without validated operating rules, these practices entered into competition to represent the same social bases as family farming and to collect funds, thus contributing to the reproduction of existing clientelistic patterns. Elsewhere in Brazil, the findings are similar: in the state of Acre, the rural territorial process has been placed under the supervision of NGOs, in the region of Marabá (state of Pará), the trade union linked to the Workers’ Party (PT) and the MDA has taken control of the Territorial Council, going against the historical trade union majority and linked to the struggle for land and agrarian reform; in the territory of Portal da Amazônia, Mato Grosso, despite the revival of family farming movements that are members of Via Campesina, the young leaders remained under the control of NGOs linked to the PT and the Catholic Church.

In the TAE, the case of the DF’s organic market construction and management project is an example of this type of mechanism. A minority of small organic producers in the Federal District managed to monopolise PRONAT resources, which were theoretically reserved for the family farming segment, through various political resources. This group of organic producers obtained funding for the market and for equipment with the support of an agronomist and businessman from the sector who was the president of EMATER-DF, the Union of Organic Producers, and Secretary of State for Science and Technology of DF, and who has since become a member of parliament in the DF.

This group of 10 producers justified the project on the basis that it would benefit 80 organic family farmers in the DF. They managed to get three of them involved in their cooperative and quickly established barriers to the entry of other family farmers, requiring a high social rating and certification by external audit [60]. A small group of fishing entrepreneurs in the DF, directly sponsored by the Minister of Fisheries, had an even easier time securing funding for a fish market and refrigerated vehicles by forming an association with 70% of artisanal fishers (with family farmer status), although none of them managed to sell even a single fish in this market [73].

In the DF and the municipalities of Goiás, this minority of multioccupational entrepreneurs, either self-employed or retired from public service, shared important PRONAT resources with extension technicians. They were able to influence the management and assembly of the COTAE through the powerful network of agronomists in the region. They had both the means to influence decisions at different levels of PRONAT governance and the legitimacy or technical competence to guide project content [61].

Through this professional network, they were able to build alliances with other agronomists or technicians from NGOs, the private sector, and municipal agriculture secretariats and form a majority bloc in the COTAE. This was facilitated by the absence of other categories of actors, including family farmers (due to their lack of time and resources) and less motivated representatives of federal or state nonagricultural services [61]. They were able to finance their institutions with computer equipment, vehicles and, most importantly, study or training contracts to support family farming. In the three cases indicated above, most of PRONAT’s target population, that is, family farmers and agrarian reform beneficiaries, were only able to obtain small projects or benefit from the indirect effects of structuring projects.

In the case of Minas Gerais, the bulk of the TAE’s funding was captured by the municipality of Unaí, which mobilised a broader coalition of municipal departments, EMATER-MG, and the CONTAG Municipal Rural Workers’ Union (STR) on a local clientelist basis (the exchange of positions in the municipal government). The COTAE, which has no legal status to receive or administer federal funds, has seen its infrastructure and collective facilities for family farmers (warehouses, farmers’ markets, tractors, etc.) handed over to the Unaí city council, which has redistributed access to them according to a local clientelist logic. The mayor did not hesitate to replace the COTAE logo with that of his municipal government on tractors financed by the MDA through the TAE and intended to provide services to family farmers in the municipality.

Finally, the refinement of the clientelist logic by small entrepreneurs and extension service agronomists has proceeded to take advantage of their position as territorial councillors to “enter politics” by running in municipal, district, or federal elections [73]. A technician from the rural agency of Goiás, who was also a member of the COTAE and a candidate for deputy in the Legislative Assembly, campaigned at the Territorial Council meetings. Quite naturally, he called for the massive vote of family farmers to obtain improved wages and working conditions for EMATER agents [73]. I asked him about the narrow and corporatist focus of his speech: why he did not base his campaign on improving services to family farmers, which was precisely the mission of his institution, who represent far more voters than the extension workers in his constituency? He earnestly assured me that, in any case, farmers would vote for the candidate nominated by the rural agency of Goiás. His priority was really to secure the endorsement and support of his peers and colleagues who might also be candidates.

5. Discussion: Clientelism and Reciprocity

In the previous section, the empirical reading of clientelist practices at the level of the TAE and its projects primarily examined the relationships and instrumental games applied for the benefit of material or political resources (positions of power). However, an in-depth examination of the practices described as clientelistic through the prism of reciprocity theory also allows for the identification of the affective, social, and ethical values associated with these relationships, thus configuring a less mechanical and more complex social framework.

5.1. Reciprocal Relationships and the Affective Dimension of the Clientele

The quick association between clientelism and corruption may in fact mask forms of solidarity, resistance (Scott, 1985) [75], and redistribution [56], (Temple 2003) [76]. These are relationships of reciprocity, often centralised and even asymmetrical, which generate mutual services, as well as feelings and ethical values between the two parties. Many of the practices of clientelism observed in the context of the TAE do not correspond to relations of instrumentation, exploitation, or unequal exchange but, rather, to redistributive services associated with the existence of differentiated powers and status.

The difference between unequal exchange and redistribution (or centralised reciprocity) lies in the human values engendered by these modes of relationship. The social or economic obligations that the associative or trade union leaders of the TAE are forced to fulfil according to the rules of reciprocity are part of these redistribution practices, which are often described as clientelistic or even as misappropriation or corruption. This is the case when the use of funds or infrastructures does not correspond to the schemes established by the technical or administrative imperatives of the projects.

Leaders of farmers’ organisations that receive aid or equipment are subject to two contradictory pressures: that of their community or social base, which demands redistribution, and that of the development services, which demand productive investment for accumulation. In order not to be condemned by their own people, the peasant leader must redistribute. However, as Temple (2003) [76] reminded us, “farmers who want to perpetuate their system of reciprocity consider this redistribution, which puts an end to the investment of a system that destroys them and their values, as a just act”. A farmer leader is, therefore, often the object of this double criticism: that which comes from his community, his family, and his peers, and that which comes from the outside, from development, and administrative and financial services. Community pressure leads to the privileging of proximity networks and reciprocal/redistributive relationships; for example, giving relatives and neighbour access to the project’s tractors and to the training offered by the COTAE. These practices are often interpreted by local elected officials and technicians as an abuse of power, as if it were different from the policy of the State and the municipal administration in Brazil. The aim is not to justify these behaviours, nor to take sides between the logic of redistribution or reciprocity and that of exchange, but to show how they differ in nature and correspond to equally differentiated social and economic projects.

Following the same logic of redistribution, the family farming communities of the TAE have sometimes accepted productive projects that at first sight seem ‘incoherent’ in the perspective of widening the social link or participating in new social relations, in other words, strengthening the structures of reciprocity. Productive infrastructures (cassava mills, tractor garages, storage warehouses) have been requested by some agrarian reform areas in the TAE communes in order to have a village hall, a chapel, a meeting centre, and even a school building, that is, for social, spiritual, and cultural relations. The leaders of the associations emphasise that even if their priorities and demands are of a different nature, the only way PRONAT can support them is by financing collective equipment and infrastructure.

Moreover, the feelings and ethical values produced by the reciprocal relations in the client relationship are identified by the protagonists in terms of affectivity: recognition, friendship, and trust, but also respect and obedience. The owner/farmer or boss/employee relationship, for example, includes mutual obligations that extend into affective relationships, which can sometimes lead to a form of asymmetric competition [57,58]. In the TAE, the most recent beneficiaries of the agrarian reform in Unaí-MG or São Sébastião-DF depend on the help of tractors from the TAE via the municipality and on external wage labour, while waiting for their plots or herds to become productive. Some of these farmers preferred to prolong the affective relationship as clientele with their former bosses rather than enter into a clientelistic relationship with the mayor and his councillors for access to tractors. They worked occasionally as day labourers for their former bosses, who come to plough or sow their plots.

Finally, in a vast territory such as the TAE, which is divided between three federated states and 19 municipalities, identity and local allegiances are superimposed upon the logics of party coalitions and social movements, generating phenomena of inclusion and exclusion that escape the rules of political participation [71,72]. Thus, the Municipality of Unaí-MG is the only one that was able to compete with the networks of rural entrepreneurs in the DF by obtaining funding from the COTAE for eight projects between 2004 and 2010 [73]. Solidarity and local identity, and affective and proximity relations, played as much of a role as the clientelist mechanism led by the Unaí City Council, along with CONTAG and EMATER-MG. In such cases, the affective and ethical dimensions are, therefore, difficult to dissociate from the instrumental relations.

5.2. Research Perspectives on Brazilian Paternalistic Clientelism

Clientelism remains an unequal political relationship. However, the intersection between clientelism and participation must also be related to another characteristic of Brazilian society, namely the paternalistic oppression inherited from the slave tradition (Geffray, 1996) [77]. The study of rural development policies at the national level in Brazil has revealed a mechanism of double alienation resulting from the conjunction of paternalist oppression and capitalist exploitation [55,58]. Within this clientelistic and paternalistic framework, capitalist exploitation resists social or revolutionary critique, tending to perpetuate a pernicious cycle that combines these two forms of alienation.

One may wonder why, apart from the actions of the MST, and especially until 2003, peasant revolts have become so rare and the recuperation of social movements so easy? A first element of response is that if social, economic, and political relations in Brazil functioned solely according to the rules of the principle of capitalist exchange, the Marxist critique would have already motivated revolts, or even ruptures or alternations of power, during the 20th century. Second, beyond the limits of Marxist critique, in the absence of a critique of the alienation proper to asymmetrical reciprocity (paternalistic oppression), it continues to triumph over peasant demands—whether under the guise of republican progress [26,28] or the maintenance of the previous economic order: the military dictatorship against the peasant leagues of the northeast region; the foreign exchange inflow of agroexporting agribusiness, against the laws of agrarian reform and land regularisation.

Would clientelism not constitute, in certain cases, a form of immune self-defence against the “worst to come”, namely the generalisation of capitalist exchange and dehumanisation through social exclusion? As Temple (2003) [76] wrote, the worst of this could be “reciprocity in the service of exchange that would replace exchange in the service of reciprocity”. One may indeed wonder whether the practices of exchange and accumulation in the service of reciprocity (clientelism) are not healthier, less hypocritical, and more human than the systematic subjection of reciprocity to the development of capital in a perfectly conscious manner. The critique must first highlight the effects of two distinct notions: what reciprocity is, and what inequality is. However, it would also be necessary to examine the articulation of unequal reciprocity (tribute) with capitalist accumulation, namely the model of the large landed estate or plantation, the so-called “fazendas”. The system of sharecropping in the fazendas, called “morada” has almost disappeared, except in the northeast of Brazil. However, it was sometimes able to ward off the social exclusion that was linked to the extension of capitalist exchange. For example, during the military dictatorship in the 1970s and 1980s, the government imposed the Rural Worker’s Statute (designed for agricultural workers on large plantations, especially sugar cane plantations) on landowners who had sharecroppers on their land. The owners expelled many of these sharecroppers, whom they were unable or unwilling to register and pay as full-time employees. Most of these peasants, having lost land, work, food, housing, and protection, held the federal government and its regional expression, the Superintendence of Development of the Northeast, responsible for their fate, rather than their landlords [58].

6. Conclusions

The PRONAT model of sustainable territorial public action proposed a complex combination of participatory planning and vertical control of the financing of collective infrastructure projects. The implementation of bottom-up territorial planning is supposed to favour the emergence and formalisation of new social demands and new alliances between actors. It could, in fact, make possible more sustainable public actions that take regional, ecological, and ethnic diversity better into account. The elaboration of projects developed through consultation should open up possibilities for conquering spaces and negotiation between public and private actors or between local collective action and territorial public action. Beyond the difficulties of implementation in such vast territories, often lacking a real identity, the main limitation of PRONAT is undoubtedly that it has not been able to escape the path dependence of the federal administrative and political system set by the constitution. Since the existence of the new territorial level is limited to a council without legal personality, the rules for financing or implementing infrastructures of federal origin render actions and projects dependent on the power of municipal or state governments. Despite the participatory innovation, the passage through the old clientelist framework seems inevitable as it was pointed in other countries [12,13,14,15,16,17].

Territorial projects also offer examples of tensions between participatory technical democracy and representative electoral democracy [42]. What is the advantage, especially in the medium and long term, of having sought to bypass municipal representative democracy, whatever its limits, by opposing it with a fragile technical democracy that is more or less participatory? Fifteen years after the transition from PRONAF Municipal Infrastructures to PRONAT, one may wonder whether it would not have been easier and more critical to provide the MDA with the means to improve, evaluate, and control projects at the municipal level. To go beyond the limits of the municipal level while improving local democracy, the option could have been to democratise and develop intermunicipal development consortia. This form of intermunicipal structure already recognised by the federal constitution has a legal status that allows it to manage federal funds and allocations (Sabourin et al. 2011) [78].

The permanence, mutations, and adaptations of political clientelism in Brazil are linked to the fact that it combines both the structuring relations of capitalist mercantile exchange and the structuring relations of reciprocity, even if they are asymmetric and unequal. Thus, it combines two forms of alienation respective to the two systems: capitalist exploitation for private accumulation on the one hand, and paternalistic oppression through the domination of large rural, entrepreneurial, or capitalist estates on the other. Today, humanist criticism based on the rules of democracy or human rights is inoperative (as is revolutionary Marxist criticism) in the face of this double alienation. The specific critical analysis of this mixed system of asymmetric reciprocity and unequal exchange has yet to be developed, as shown by the great difficulty of the once-in-power Workers’ Party to do politics in a different way.

This study has shown the importance of including the anthropological category as reciprocity in the analysis of clientelistic practices in Brazil as in other contexts [19,20,21,22,23,24]. It allows and helps us to better understand the permanence of such dependence relationships in a modern Brazil and even in the context of renewed public development policies based on popular and citizen participation. Such categories and policies constitute the bases of the social or sociopolitical pillar of sustainable development.