The World Bank’s ‘Market Assisted Land Reform’ in Colombia and Brazil (1994-2002)

ABSTRACT

The article analyzes the implementation of the World Bank's 'market-assisted land reform' in Colombia and Brazil. It shows that MALR was designed and implemented as a model opposed to redistributive agrarian reform, based on the disappropriation of private lands by the state. It also analyzes the results of MALR in the two countries, arguing that its principal function was not economic, but rather political, to adjust agrarian policy to the neoliberal agenda and serve as an instrument to undermine the popular struggle for the democratization of the agrarian structure in highly unequal societies.

Keywords:
World Bank; agrarian reform; land markets

RESUMO

O artigo analisa a implantação da "reforma agrária assistida pelo mercado" (RAAM) do Banco Mundial na Colômbia e no Brasil. Mostra que a RAAM foi concebida e implementada como um modelo contrário às reformas agrárias redistributivas, baseadas na desapropriação de terras pelo Estado. Também analisa os resultados da RAAM nos dois países, argumentando que a sua função principal não foi econômica, mas sim política, ao ajustar a política agrária à agenda neoliberal e servir como instrumento para esvaziar a luta popular pela democratização da estrutura agrária em sociedades altamente desiguais.

Palavras-chave:
Banco Mundial; reforma agrária; mercados de terras

This article analyzes the implementation of market-assisted land reform, MALR) in Colombia and in Brazil. Propelled by the World Bank (WB) in various countries from 1994 onwards, MALR was one of the central components of the neoliberalization agenda of agrarian policies and social relations in the countryside undertaken by the institution. Aimed at countries with high rates of landholding concentration and significant social tensions in the countryside, it had the objective of replacing redistributive land reform, based on either expropriation (without indemnification) or disappropriation (with indemnification but below the market price), with land purchase and purchase/sale relations between private agents. MALR was the subject of debates and intense conflicts waged on an international scale around its use. Adopting it as a subject of investigation makes it possible to understand many of the arguments which have molded the discussion over the last twenty years about the scope, the instruments, and the means of financing agrarian reform in Latin America.

Colombia was the first country to institutionalize MALR as a model of access to land attuned to neoliberalism. This pioneering case was shortly afterwards followed by South Africa in 1995 and Brazil in 1997. A decade after it began, MALR had been implemented in different forms in countries such as Guatemala, the Philippines, Malawi, Honduras, Mexico, and El Salvador. In a few years a wave of a new type of agrarian policies had been established internationally.

The experiences discussed here were the most relevant in Latin America, since they served as shop windows for the WB to spread MALR to other countries. It should be highlighted that since the end of the 1940s, Brazil and Colombia have figured among the five biggest clients of the WB in the region and among the principal ones on a global scale.

The text does not propose to analyze or compare the complex configuration of the agrarian question in these two countries, nor the set of changing relations between national states and the WB. Rather, its objective is to analyse MALR in the two cases with the greatest weight in the region and to establish some relevant comparisons. The interpretation emphasizes the political dimension of social processes.

Initially this paper discusses the reasons which led the WB to promote an agrarian agenda during that period, as well as what its components were. Afterwards, it analyzes the assumptions and the political rationality of its principal item: MALR. Next, the WB's arguments to legitimize this are summarized and the international politics which guided its formulation and implementation in Colombia and Brazil are examined. Finally, it evaluates the performance of this public policy in the two countries. The paper is based on official documents and empirical research, some financed by the WB itself.

THE OFFENSIVE OF THE NEOLIBERAL AGRARIAN AGENDA

At the beginning of the 1990s, the diagnosis of top-ranking WB employees was that the liberalization of labor markets, land, and credit had advanced little in the previous decade in comparison with the macroeconomic and fiscal adjustment. It was then that the WB began its re-engagement with agrarian subjects, after a decade of practically abandoning questions linked to land ownership and rural development. Various reasons can explain this change of direction.

The first was that for the WB the end of the Cold War opened the possibility of dealing with problems associated with the concentration of landholding in a purely technical form (Deininger; Binswanger, 1999, p. 248). According to this perspective, different from the 1960s and 1970s, when the dispute for agrarian reform was associated with 'revolutionary' ideologies, the collapse of the USSR opened a new phase in which the agrarian question could be dealt with in a pragmatic manner and 'without ideology.'

A second reason is related to the disruptive potential which the accumulation of agrarian conflicts represented in countries in the global south (Binswanger; Deininger, 1993). Published in 2003, the WB's principal report systematized a decade of reflection on the issue and prescribed the creation of mechanisms aimed at avoiding or reducing the impact of disputes over the possession and ownership of rural land, arguing not only that many of the most important political conflicts experienced by various twentieth century societies had roots in struggles for land (e.g., Guatemala, Colombia, El Salvador), but also that the resolution of agrarian conflicts had been crucial in making feasible peace accords which brought to an end long civil wars (e.g., Mozambique Ethiopia, Cambodia, Nicaragua). This concern was at the root of the WB's return to the issue of agrarian reform (2003a, p. 151).

The third reason was related to the institutionalization of the fight against extreme poverty in the international arena. Due to the function of the socially regressive impact of adjustment policies, the (WB, 1997a) came to prescribe in some cases the distribution of land as a cheap means to relieve rural poverty.

Finally, the (WB, 1997a) was one of the protagonists of the transition of Eastern European societies and the old USSR to neoliberal capitalism, prioritizing the privatization of the agrarian structure in its technical and financial assistance during the 1990s.

The agrarian policy of the (WB, 1997a; 2003a) was aimed at the complete conversion of rural land into merchandise, promoting three lines of action: stimulating commercial relations of leasing, purchase and sale; accelerating the privatization of property rights in public, communal, and collective lands; and overcoming the informality of property rights, ensuring private ownership.

In order to make this agenda feasible, the WB, through technical and/or financial assistance, induced the promotion of new administrative mechanisms for landholdings, seeking to introduce the broader state reform agenda into the repertoire of public policies for rural development. In Latin America other central themes of this agenda included the de-federalization of land policy, the creation of public-private partnerships for the administration of landholdings, and private ownership. Furthermore, the WB also advocated the revision of legislation, in particular in those countries which had carried out agrarian reforms, in order to remove any legal restrictions on leasing relations and land purchases.

MALR ON PAPER

MALR was a central item on WB's agrarian agenda in the 1990s. The book organized by (Van Zyl et al., 1996) condenses the theoretical discussion on which the MALR proposal is based. The core of the debate focused on the centrality of 'land markets.' Based on neoclassical neo-institutionalism, the concentration of landholdings, differentiated land uses, the exploitation of natural resources, the exploitation of natural resources, and inequality of income and rent in the countryside - in other words questions fundamentally about what is produced, how, and for whom - stopped being seen from the perspective of relations of power existing in society, and came to be read as problems linked to the functioning of land markets and institutions. According to this focus, the 'imperfections' of markets and the 'distortions' provoked by 'erroneous' macroeconomic and sectorial policies (protectionism, directed subsidised credit, tax exemptions, administered exchange rates, etc.) inhibited the allocation of the land of the least efficient producers to the most efficient.⁸ In light of this, the state had to abandon the roles of provider and executor, limiting itself to guaranteeing the 'rules of the game' - the institutions - which would facilitate market relations. Moreover, the new model also combined methodological individualism with the idea of homo economicus, assuming a maximizing individual rationality which imposed itself in accordance with the 'correct' incentives.

Not by chance, the assumption of MALR was the historic fallacy which WB economists called state-led land reform. Considered to be a discretionary, confiscatory, and conflictual model, as well as politically unfeasible for democratic regimes - as it was based on an instrument of disappropriation, paying compensation (when existent) to landholders through land bonds -, this model was also centralizing, bureaucratized, paternalistic, economically inefficient, and ineffective in the reduction of rural poverty.

In tune with the new times, MALR was a 'voluntary' and 'negotiated' model (Burki; Perry, 1997, p. 95), based on the principle that anyone who wants to buy and sell land could do so. For this reason, strictly speaking, it consisted of a commercial transaction among private agents financed by the state, which was responsible for subsidizing investments in productive infrastructure and expenditure with technical assistance. The lower the price paid for the land (loan), the greater the subsidy available for investment and vice-versa. Sellers would be paid in advance in money at the market price, while purchasers, organized in community associations, would be responsible for the costs of acquiring the land and other transactions costs.

What advantages did this model offer? First, it stimulated mercantile bargaining between purchasers (interested in paying less) and sellers (interested in earning more) over the price of land, reducing the total cost of reform. Second, it did not economically 'penalize' landholders, the reason why legal disputes over the value of indemnifications were eliminated. Third, since it had a voluntary, decentralized, and non-bureaucratic nature, it favored the participation and autonomy of beneficiaries. Fourth, it encouraged the economic development of peasants, by requiring the planning of productive activities before purchasing the land, providing subsidies for this purpose, and stimulating associationalism. Fifth, it made land markets more dynamic, contributing to formalize the property rights, as well as rural financial markets, as the new landowners, with secure title as guarantees, would be eligible for bank loans (see Van Zyl et al., 1996).

In order for MALR to be able to function, it was necessary to increase the mercantile offer of land and reduce its price. Six complementary actions were considered vital for this: a) the elimination of subsidies, fiscal exemptions, and protective tariffs which privileged large landowners and, together with inflation, contributed to increasing the price of land above its agricultural profitability; b) the end of any legal restrictions on the purchase and leasing of land; c) any type of land taxation, in order to discourage underuse and speculation; d) the legal security of the land ownership rights, through systematic private ownership; e) creation or improvement of market information systems, in order to orientate economic agents about property prices and characteristics; f) reduction of transaction costs, through administrative and judicial simplification (see Van Zyl et al., 1996).

MALR IN COLOMBIA: CONTEXT, INTERESTS, AND MOTIVATIONS

In the 1960s, there was a fear on the part of the US establishment in relation to ongoing political processes in the Latin American countryside. Given this panorama, the US government designed the Alliance for Progress, combining in this the promotion of capitalist development with the doctrine of counterinsurgency. The civil war in Colombia had been dragging on, with highs and lows, for decades, and one of its ingredients was "the greatest armed mobilization of peasants in the recent history of the western hemisphere, with the possible exception of determined periods of the Mexican Revolution" (Hobsbawm, 1968, p. 226).

Government strategy, in this context, consisted of maintaining severe restrictions on political participation and trade unionism along with fiscal and commercial overprotection in favor of industries and latifundiários (large landowners). Furthermore, in accordance with Law 135, from 1961, the government launched a preventative land reform, which was the shop window of the Alliance for Progress (García, 1973). However, the dominant national class felt threatened by the initial steps of the law and after twelve years, the timid agrarian reform was replaced by a policy of forest colonization.

In relation to counter-insurgency, American political and military circles designed a strategy of preventative war, the LASO (Latin American Security Operation) Plan, to be applied to resisting peasants under Law 135. As a result armed conflicts intensified around the country, giving rise to organizations such as Ejército de Liberación Nacional (ELN), the Fuerzas Armadas Revolucionarias de Colombia (FARCs) and Ejército Popular de Liberación (EPL), which existed in various parts of the country over the following years, as well as counterinsurgent paramilitaries (see Restrepo, 2009; Peña, 2014; González, 2014).

During the 1980s, the neoliberal agenda was consolidated in Latin America, resulting in a cascade of constitutional reforms, one of which took place in Colombia. It was in this context that Law 160 was passed in 1994, drafted to implement MALR mechanisms. It was proposed to redistribute land in this manner. Furthermore, the agrarian question had been impacted by drug trafficking, beginning in the 1970s. Estimated a few years afterwards at between four and six billion dollars (Chernick, 2008), the funds from this activity were being legalized through fiscal reforms and channeled into civil construction and the purchase of extensive rural properties, worsening an already high concentration of landholding, increasing the costs of agricultural production and restricting its ability to reach international markets. Meanwhile peasant demands were meagerly responded to by rural development programs aimed at agricultural production and precarious social services in 'colonization' areas.

With the neoliberal reforms, it became necessary to face the agrarian problem, due to its impact on the costs of production and agricultural competitiveness. Law 160 intended to respond to these factors, as its central purpose was to make the rural sector suitable to commercial opening and to globalization. The transformation of the property system was to be propelled by the 'dynamization of land markets,' according to a broad agenda of privatization and reduction of the role of the state in the social area and in sectorial policies.

Law 160 was drafted with the guidance of a WB team and preceded by an important study sponsored by the (FAO, 1994), which warned that there did not exist an integrated national land market, but rather highly 'imperfect' markets whose characteristics were a high level of informality, the asymmetry of information among agents, the segmentation of transactions (of small properties among small landholders and of large properties among large landholders) and the existence of vast areas of territory subject to distinct forms of control (economic, political, or military, linked to leftwing guerrillas, or rightwing paramilitaries). On the other hand, the structural problems of the landholding system in the country had been looked at by a team of experts from the WB (Binswanger et al., 1995), whose work showed that, far from having reduced the concentration of landholding, the monopolization tendency was continuing (WB, 2004). This dynamic resulted both from capitalist development and the particularities of Colombian society, amongst which were highlighted relations between the latifúndio and political power, as well as the occurrence of drug trafficking and the violence in these relations (Kalmanovitz; López, 2006, pp. 338-341).

The official discourse raised the expectation that a new type of agrarian reform, carried out with reduced state intervention and dynamized by land markets propelled by the new law, would reduce the concentration of property and land costs, making Colombian agricultural exports more competitive. At the same time, the possibility of access to land via market transactions subsidized by the state attracted the interest of landless rural workers and poor peasants, reducing the social gravitation of guerillas, trade union organizations, and social movements in the countryside, who demanded a more profound agrarian reform in the country. MALR therefore fulfilled economic and political objectives.

MALR IN BRAZIL: CONTEXT, INTERESTS, AND MOTIVATIONS

In 1994, the federal government implemented the Real Plan, oriented by a broader model of economic stabilization and re-structuration which combined the use of the exchange rate to combat inflation with a commercial opening, financial deregulation, the de-indexing of the economic, fiscal adjustment, and the privatization of public companies.

According to the (WB, 1995), the Real Plan led to a deterioration of living conditions in the countryside, which as a counterpart demanded the creation of compensatory social programs. At the same time, it opened a historic opportunity to establish the land market as a central mechanism for the reduction of rural poverty.

The introduction of MALR in Brazil occurred through the 'Solidarity Agrarian Reform Project,' a small experience initiated in the state of Ceará, based on a WB loan, with the aim of financing the purchase of land to alleviate the impacts of the macroeconomic adjustment in the rural environment. Beginning in April 1997, it financed the purchase of 44 properties by 694 families in a year. There was an increase in the price of land where the project was implemented (Brandão, 2000). From this experience a few months later the Cédula da Terra Pilot Project (PCT) was created, covering five states (Maranhão, Ceará, Bahia, Pernambuco, and the north of Minas Gerais), based on a new WB loan.

MALR was thus introduced in Brazil as a compensatory action to relieve the socially regressive impact of economic liberalization. But not just this. Politically, the aim of the federal government was to reduce the pressure provoked by the increase of land occupations, organized by the Landless Rural Workers' Movement (Movimento dos Trabalhadores Rurais Sem Terra - MST):

Increasingly, land occupations have generated significant pressure on the Brazilian Government to act rapidly on the existing land tenure problems ... For the Brazilian Government, the market-based land reform approach provides the opportunity to respond to landless farmer pressures in a way that is less conflictive than administrative approaches to land reform. (WB, 1997b, p. 7)

It should be noted that at the beginning of the Cardoso government (1995-2002), the agrarian reform program had an welfarist profile and was subordinated to Solidarity Community.⁹ Nevertheless, three factors radically altered the agrarian scenario.

First, the repercussions of police violence in Corumbiara, in August 1995, and principally in Eldorado dos Carajás, in April 1996. In the former ten landless peasants were killed and nineteen in the latter. Both episodes caused protests against violence in the countryside and in favor of land reform.

The second factor, already mentioned, was the increase in land occupations almost everywhere in the country. In particular, the occupations in Pontal do Paranapanema gained visibility - a region marked by the grilagem (illegal occupation) of public lands located in the state of São Paulo, the richest and most industrialized state in the country.

A third factor was the holding of the successful National March for Agrarian Reform, Employment, and Justice. Organized by MST, the march lasted for three months, departing from three different parts of the country, reaching the capital on 17 April 1997, a year after the 'massacre' in Eldorado dos Carajás. Despite the neglect of public authorities and the campaign aimed at belittling it by the media, the march succeeded in bringing together popular dissatisfaction with neoliberal policies, bringing together around one hundred thousand people in Brasília.

The repercussion of agrarian conflicts and violence in the countryside, combined with the increase in land occupations, led the Brazilian government to implement the PCT (WB, 1997b, p. 7). According to its operators, the introduction of MALR would break the connection between occupations and disappropriations, and ensure that the state only funded market relations between landless and landholders (WB, 2003b, p. 127). Therefore, the region with the greatest concentration of rural poverty in the country was selected, a place where the WB had operated for more than twenty years, so that the implementation of the new program would be rapid. The aim was to finance the purchase of land by 15,000 families in four years, and based on this to legitimate the extension of MALR on a national scale (WB, 1997c, p. 7).

PCT was opposed by MST and the National Confederation of Agricultural Workers (Confederação Nacional dos Trabalhadores na Agricultura - CONTAG), being identified as part of the withdrawal of the state from social questions and as incapable of meeting the demand for rural land, estimated at the time as 4.8 million potential families.

MALR IN PRACTICE: THE CASE OF COLOMBIA

Starting in 1994, MALR had the target of financing the purchase of land by 75,000 families in four years. However, by the end of 1997, only 17,058 had been financed (Höllinger, 1999, p. 162), leading the WB to describe its implementation as "disappointingly slow" (Deininger, 2000, p. 218). For the WB the case of Colombia involved errors in the implementation of the new model in three aspects.

First, Law 160 allotted a very high amount to the subsidy, around 70% of the total land cost, with a maximum of US $21,000. In itself this was seen by the WB as a problem, aggravated by the fact that this subsidy was only for the purchase of the land, with productive investments being left aside. The remaining 30% necessary for the purchase of the property came from their own resources or from loans from public or private banks. According to (Deininger, 2000, p. 219), the elevated subsidy stimulated a conspiracy between sellers and purchasers, and led to the overvaluation of property transfers, to such an extent that on average the price of land purchased via MALR was higher than the price paid via disappropriation. The systematic practice of overpricing was also detected by (Höllinger, 1999) and (Mondragón, 2003).

MALR was well received by landholders and their representative entities, interested in income from the land allowed by the high subsidy exclusively offered for the purchase of land. According to (Deininger, 2000, p. 219), the subsidy should have been reduced and also offered for productive investments.

In second place, there was a need to lower by a third the income ceiling for entering the program by a third, in order to focus on the poorest persons (Deininger, 2000, p. 219).

Third, MALR initially followed a centralized pattern of decision making and implementation under the responsibility of the Colombian Institute of Agrarian Reform (Instituto Colombiano de la Reforma Agraria - INCORA). The result was a very low level of operative decentralization, socialization of information, social participation, and the involvement of the private sector in the provision of services and the granting of credit (Deininger, 2000, p. 220).

According to the WB, the link with INCORA was one of the reasons for the 'resistance' which the implementation of MALR had faced within the state apparatus (Deininger, 2000, p. 218). This situation had placed the WB in a difficult situation, to the extent that the implementation of MALR depended on public funds administered by a central bureaucracy described as inefficient, corporative, and intractable to innovation.¹⁰

Furthermore, the preparation of productive projects before the purchase of land had not occurred, which according to (Deininger, 2000, p. 224) compromised the entire rationale of the program. In practice this situation reflected the absence of autonomy and the protagonism of those who entered the program as purchasers:

Without a clear understanding of the economic potential of productive units being established, the expected return and the alternative options (within and outside the agrarian reform program), the capacity of beneficiaries to engage in significant bargaining was quite reduced. It was simply natural that INCORA would assume leadership in negotiations with landholders... (Deininger, 2000, p. 224)

The poor results of MALR in Colombia forced the government to make adjustments. In 1997, based on a WB loan, a pilot program was begun in five municipalities.¹¹ According to (Deininger, 2000), the aim was to: a) formulate a "municipal agrarian reform plan," responsible for identifying situations in which the possible offer of land was three times superior to potential demand, so that a competitive land market could be created without risking the overvaluation of land prices, as had occurred; b) seek partnerships at a local scale with NGOs which could provide technical assistance and private financial institutions willing to finance the farmers; c) create new procedures for the selection of beneficiaries, which included greater publicity about the project, a local register of applicants, and the answering of questionnaires about their agricultural experience and socio-economic situation; d) formulate productive projects before land acquisition, in order to orientate potential beneficiaries about the economic potential of the new business, and only after this phase would applicants move on to the pre-selection of properties; e) monitor and assess project impact.

The introduction of these modifications revealed, according to the WB assessment, the lack of 'experience' and 'capacity' of the beneficiaries selected to perform the tariffs inherent to "negotiated land reform" (Deininger, 2000, pp. 224-225). Faced with this diagnostic, the WB began training for "pre-selected candidates." The collaboration of landholders was decisive for the pilot project; after all, "they were the party most anxious that it would advance" (Deininger, 2000, p. 233).

Even with all the operational changes, the WB's own indications undermined any type of euphoria. Despite the minuscule project carried out in five municipalities, the large majority of beneficiaries were found in a situation of chronic defaulting (Deininger, 2001, p. 89; WB, 2003a, p. 147).

Peasant organizations had little influence on the passing of Law 160 in 1994. According to ( Höllinger, 1999, pp. 161-168), the year previously they had presented an agrarian reform bill (no. 203), which ended up not being voted on in Congress. Only after widespread mobilization did the federal government incorporate them into the negotiations which resulted in Law 160. Due to their pressure, the subsidy rose from 50% to 70% of the price of land - something which was also of interest to the rural employers -, but other demands were not accepted, such as the definition of a fixed national budget and the establishment of a ceiling to prevent the increase of the price of land. Even though they rejected MALR and defended a redistributive agrarian reform under the responsibility of the state, various peasant organizations in practice acted as intermediaries in land transactions. For example, this was the case of the Asociación Nacional de Usuarios Campesinos (ANUC - National Association of Peasant Users) and others with a local base, competing between themselves.

MALR IN PRACTICE: THE CASE OF BRAZIL

Despite the criticism of MST and CONTAG, PCT commenced in December 1997 and two years later had funded in the five states in which it operated 6798 families (Buainain et al., 1999, p. 56). The preliminary assessment contracted by the federal government and paid by the WB showed that the project began in a period of drought, converting access to land into a means of immediate survival (Buainain et al., 1999, p. 27). Moreover, there was intense propaganda exalting the possibility of access to land 'without conflict,' aimed at an immense impoverished rural population, in a conjuncture of repression of land occupations. The research in question was clear:

In a region characterized by a high concentration of land ownership and social exclusion ... the possibility of immediate 'easy' access, without 'risk,' lacking in bureaucracy, to land stimulated adhesion to the Program ... Many interviewees referred to this aspect, counterpoising the facility offered by Cédula [PCT] to the risks of invasions and the difficulties of life in a landless workers encampment ... The interviewees reproduced the official discourse of solidarity agrarian reform, without conflict and in partnerships. (Buainain et al., 1999, pp. 27 and 271)

Moreover, there was also pressure to implement PCT coming from two directions, the WB on one hand, and state and federal governments on the other. More than half of the 223 projects accounted for in January 1999 had been implemented in the second half of 1998 (Buainain et al., 1999, p. 15), when the intensification of the electoral dispute, the increase in number of land occupations, and the explosion of sackings of food stocks in the Northeast all converged (Carvalho Fº, 2001, p. 208).

Meanwhile, in February 1998 the government majority in Congress authorized the creation of the Land Bank (Banco da Terra), by passing Complementary Law no. 93. This was a public fund capable of raising resources from various sources, including international ones, to finance the purchase of land by rural landless workers and poor peasants. In other words, without any assessment of ongoing experiences and despite the opposition of all national rural worker organizations, Congress passed a law allowing the creation of an instrument capable of implementing MALR throughout Brazil. The WB also worked to support the federal government, promising to release voluptuous loans (WB, 2001, p. 341).

The creation of the Land Bank was seen by organizations which were part of the National Forum for Agrarian Reform and Justice in the Countryside¹² as a sign that the replacement of land reform by MALR was underway. Reinforcing this diagnosis was the sensitive decline of INCRA disappropriation budget. During 1998, the Forum prioritized the struggle against PCT and the Land Bank, denouncing to the Federal Prosecution Service suspicions of overcharging and corruption in the purchase of land through PCT, also and submitting an investigation request to the WB Inspection Panel¹³ (Fórum, 1998a; 1998b).

The Panel accepted the Forum's submission and at the beginning of 1999 sent representatives to Brazil, who visited PCT areas in Bahia and in Pernambuco. The legitimacy of these visits was questioned by the Forum, under the allegation that conditions had been forged so that they would not appear flawed (Wolff; Sauer, 2001). In May 1999, the Panel deemed the Forum's arguments to be groundless and did not recommend to the WB board that the investigation be carried out. Furthermore, at the same time the WB also belittled the representativeness of the organizations which composed the Forum, considering its demands to be of a 'philosophical' nature (Inspection Panel, 1999a).

Shortly afterwards, the federal government released documents to the Forum about the implementation of PCT. This material pointed to numerous irregularities and indications of corruption, such as: a) the purchase of land that could be disappropriated; b) increase in the price of land and the purchase of areas above the price paid by INCRA via disappropriation; c) the bad quality of the land purchased; d) existence of technical reports (intended to guide the negotiating of the price of the property) signed after the date of the purchase of the property or unsigned; e) sale of various properties by the same landholder (Wolff; Sauer, 2001).

Based on this documentation the Forum requested a new investigation from the Panel in August 1999. Four months later, another negative response was received. However, this time the merit of the argument was not considered, rather it was stated that the Forum had not fulfilled all the required procedures to make the request eligible, such as exhausting all channels of dialogue with the Brazilian government (Inspection Panel, 1999b).

Despite these defeats, the Forum's actions were crucial in blocking the loan the WB was supposed to give the federal government to finance the Land Bank throughout the country (WB, 1999, p. 3). An impasse was thereby created, since the implementation of MALR at a national scale could not find enough political support. Apart from the support of all rural employers entities, there was only local support from rural worker trade unions in the Northeast, from Força Sindical in the state of São Paulo, and trade union federations in the southern states, whose central demand had never been the struggle for land, but policies favorable to 'family farming'. However, the principal national organizations linked in the Forum were against.

This unity was broken when CONTAG decided to negotiation the creation of a new program - the Land Credit to Combat Rural Poverty (Crédito Fundiário de Combate à Pobreza Rural - CFCP), similar to previous ones, though with some modifications -, allowing the loan, until then not approved by the WB board, to be authorized. Therefore the WB did not finance the Land Bank, but rather the CFCP.

From then on, CONTAG continued to link PCT and the Land Bank to MALR, but differentiated them from the new CFCP, considering the latter as only a land credit program complementary to agrarian reform. At the same time, CFCP prohibited the purchase of areas that could be disappropriated and, more importantly, created participation and co-administration mechanisms for rural worker trade unions. Following the creation of CFCP, the fight against MALR disappeared from the Forum's agenda.

COMPARISONS AND ASSESSMENT

In theoretical and methodological terms, the relationship of the WB with national states and subnational authorities should not be seen as a mere external imposition. Although there exist differentiated forms and mechanisms of pressure used by WB according to circumstances, the agency's actions have historically been implemented via a dense and expanding network of relations involving national and international, public, private, non-governmental, philanthropic, and business agents, who, with distinct means and levels of gravitation, support, propose, adapt, bargain, and transmit the ideas and prescriptions of the institution. From this perspective, the effectiveness of the WB's actions requires the construction of world vision and mutual interests, and depends on points of support and diffusion, both within and outside national spaces. Clients states are not equal, neither are they all victims of the institution.

Furthermore, the implementation of MALR varied according to national particularities which involved the correlation of political forces, the contradictions of economic liberalization, the external insertion of economies, the variable priorities of governments, budget restraints, and the relation between the federal government and subnational spheres, amongst others. Therefore, the Colombian and Brazilian experiences can be compared in accordance with their level of correspondence to the directives of MALR. At the highest level, the governments of the two countries: a) adopted policies with a voluntary and mercantile nature, a decentralized method of implementation, and the privatization of services; b) they sought to legitimate the adoption of MALR through a critique of the "disappropriationist model"; c) income ceilings were adopted as criteria for access to focus it on the poorest segments of the rural population; d) associationalism was stimulated as a criteria of access to programs; e) land policy was subordinated to the neoliberal agenda, adapting it to the imperatives of the fiscal adjustment, decentralization, and privatization, and lowering its status, inserting it in the list of compensatory policies; f) MALR was internalized through the promulgation of laws.

At an intermediary level, it was found that: a) the leasing of the lands acquired through MALR was legally prohibited in Colombia, but not in Brazil, although these relations have not been stimulated; b) in Brazil, it can be financially accessed by associations (PCT and CFCP) or individually (Land Bank), but in Colombia only by associations; c) only Brazil adopted a variable combination of loans and subsidies; in Colombia the subsidy was fixed and covered 70% of the price of the land, and no subsidy was authorized for productive investments; d) in Brazil, MALR operated as something separate and in direct competition with the agrarian reform program, while in Colombia - due to pressure from INCORA, and not the WB -, the disappropriation instrument appeared in Law 160/94 as the final option, if market negotiations failed.

Finally, some components of MALR never got off the paper. In neither country was progressive taxation or systematic private land ownership adopted. Nor was a decentralized market information and land registration apparatus created. In other words, the support mechanisms considered necessary for increasing the offer of land and the fall in its price, as well as the legal security for property rights, remained only at the level of recommendations. Nor was MALR inserted in a wide-ranging and effective rural development strategy in either country.

The two countries carried out loan operations with the WB, although in very distinct proportions: for Brazil, US $90 million in 1997 for PCT and EUR 218.2 million in 2001 for CFCP, with the possibility of additional sums until 2012 with a total of US $1 billion, with an equal national counterpart; for Colombia, US $50 million in 1996. Of all the countries which implemented MALR, Brazil was by far the one which got most loans from the WB for this, while there was also political opposition of great international visibility.

In Colombia, the drafting of Law 160 fundamentally met the interests of rural employers. First, because the law offered a voluntary exit from the agricultural sector to landholders with financial difficulties or affected by armed conflict, through a high subsidy for the purchase of the land. Second, because the law did not allow for effective mechanisms to expand the offer of land. Progressive taxation was not established and though disappropriation was available as the final recourse, if the voluntary negotiation failed, it was slow and bureaucratic, for which reason it was not used. Landholders could decide whether or not they wanted to negotiate and under which terms (Höllinger, 1999, p. 160). For these reasons, their monopolistic positon in the land market remained unscathed. Instead of contributing to lower the price of land, the subsidy helped to keep it high.

Despite the bias in favor of landholders and the efforts of the WB and the federal government to make MALR a success in Colombia, the results achieved contrasted with the supposed advantages of the model and were very much below the expectations of its proponents and local operators. Why? First, the process did not gain sufficient scale, since between 1995 and 2001 only 19,397 families were financed, who acquired 286,939 ha. (Balcázar et al., 2003, p. 312), in a universe whose demands for land were estimated in the middle of the 1990s at around 721,000 families, of whom 166,000 were landless peasants and 555,000 peasants with insufficient land (Urbina, 1996, p. 190). Second, the number of families who solicited access to credit was much higher than the number accepted, while demand was greater than the offer of land (Mondragón, 2003). In third place, the program was focused on a restricted segment of the land market, composed of mid-sized landholders ruined by the commercial opening, and never the large ranchers, for which reasons the commercial transactions did not involve the best located land with the highest quality (Mondragón, 2003; Höllinger, 1999, p. 187). Fourth, a significant part of those who benefited defaulted, even with the high rate of subsidies given for purchasing land (WB, 2003c, p. 150). Furthermore, the prices paid for land were generally arbitrated by landholders and technicians from INCORA, showing the political fragility of peasants in the negotiation (Höllinger, 1999, p. 191-193). Moreover, the implementation of MALR led to the overvaluation of the price of land, at least in the first four years, the most documented period (Deininger, 2000; Mondragón, 2003). Finally, in the overwhelming majority of cases, agricultural production was limited to family subsistence, not resulting in an effective mechanism for the reduction of rural poverty (Borras Jr., 2003, pp. 382-383).

The allocation of funds to MALR was significant only between 1996 and 1998. After 2000, funds were more or less divided in the same proportion between direct purchase by the federal government and MALR, though at very reduced levels (Grusczynski; Jaramillo, 2002). The project ended in 2003 and the (WB, 2004) faltered in the defense of MALR for Colombia, but continued to insist on the market as a preferential means of access to land for poor peasants and landless workers. At the same time, from 2000 onwards, agrarian reform disappeared not only in terms of public policy for the countryside, but also from the actual official lexicon.

In relation to Brazil, PCT's targets were met and 15,267 families bought land between July 1997 and December 2002, when the project ended.¹⁴ Hyped as 'exemplary' by the (WB 2003c), Cédula (PCT), however, did not achieve very encouraging results. Why? In first case, the majority of properties acquired were abandoned or underused, due to drought and a crisis in livestock farmers and traditional cultures (Buainain et al., 1999, p. 31). Second, the 'demand' for the project was strongly conditioned by drought, impoverishment, and the lack of perspectives for work, the directives which state agents and local politicians imposed on the entire process, the criminalization of land occupations, and the 'dream' of access to land fed by poor rural workers (Buainain et al., 1999, p. 27; Victor and Sauer, 2002, pp. 34-35). In third place, the funds allocated to productive investment were consumed in the majority of cases as purchase costs, being exhausted before covering the minimum investment package (Buainain et al., 2003, pp. 100 and 150). Finally, income projections made by the final assessment did not indicate the consolidation of an efficient commercial agriculture, and even in the cases considered most positive, monoculture was maintained (Buainain et al., 2003, pp. 157-170).

In both countries rural employer organizations adopted positions in favor of MALR, for political and economic reasons. While the organizations from the world of labor criticized MALR from its beginnings, there were important national variations. In Colombia, local trade union groups participated in the administration of the pilot project in a marginal form. In Brazil, some local trade union organizations were favorable, but the principal national organizations remained en masse against MALR between 1997 and 1999, with this unity being broken in 2000. The comparison shows that in the two countries, in contexts of neoliberalization of the countryside and the criminalization of social struggles, there were local or national peasant organizations which endorsed MALR.

There is a basic criticism valid for both cases. In capitalism, land is a factor of production and a net asset in the portfolio of financial investments. For this reason, both its productive and its speculative functions are reflected in the price of land. However, in addition to economic factors, there exist extra-economic elements related to agrarian property, such as political power and social prestige, which are 'capitalized' in the price of land and which can have a weight which is greater than economic factors. It can be said that in the land market three distinct logics intersected: productive (for licit and illicit goods), speculative, and extra-economic, determined by socio-cultural factors and territorial power (Höllinger, 1999, p. 149). In theoretical terms, MALR was based on a mercantile vision of rural land, as if it were only a factor of production, a commodity, tradeable like any other merchandise. Ignored was the fact that land has a multidimensional nature, for which reason property rights and control over it express, above all, relations of power between social classes and groups (Borras Jr., 2003). According to this mercantile vision, the theory of MALR stated that the gap between the market price and price based on productive profitability was a merely conjunctural phenomenon, resulting from 'distorted' economic and sectorial policies and 'erroneous' institutions, not from something structural. However, this theory did not explain why there was no substantial increase in the offer of land in the market as a response to the dismantling of the protection mechanisms for national agriculture and the injection of credit for the purchase of land. According to the economicist assumptions of MALR, there were various 'incentives' for this offer to increase. In fact, the model did not take into account the weight which the monopoly of land ownership had in political power and in the configuration of relations of power between groups and classes in radically unequal societies.

Despite the immense differences between the Brazilian and Colombian societies, the principal empirical research done (Höllinger, 1999; Mondragón, 2003; Buainain et al., 1999; 2003; Deininger, 2000; 2001; WB, 2003c; 2004; Victor; Sauer, 2002) has shown that the implementation of MALR had similar characteristics and results. In relation to land acquisition, the prices paid for rural property were either not as low as had been projected (Brazil), or were higher than the price paid via disappropriation (Colombia). Moreover, on average the land acquired was of low quality. It can thus be affirmed that MALR rewarded landholders who used it to sell abandoned, bankrupt, or badly located property. Furthermore, in neither of the two countries was there a substantial offer of land on the market from large indebted landholders, so that the predominant profile of sellers was limited to mid-sized producers ruined by neoliberal policies and not ranchers. On the other hand, the implementation of the programs caused an increase in the price of land in many places, although this effect was neither homogeneous nor constant. In turn, borrowers had unequal power in negotiations with landholders. Finally, there was no transparency in the implementation of programs and there were numerous accusations of corruption and illicit favoring of landholders and public agents.

After the acquisition of the land, the economic development of families also did not have better results. In practice, the preparation of feasible productive projects did not occur before the purchase of the land, as MALR stipulated on paper. Moreover, when it existed technical assistance was precarious and irregular, hindering the productive performance of families. Generally speaking, productive projects were characterized by subsistence agriculture, and not by highly profitable commercial agriculture, and in the cases considered positive by the WB in Brazil, monoculture was reproduced. The subsidy granted was revealed to be insufficient to boost agricultural production and private credit markets remained inaccessible to those who entered MALR.

Finally, MALR programs were not inserted in wide-ranging strategies of the reduction of rural poverty and social inequalities. Rather they followed a unique and individualist approach, centered on the stimulation of demand via credit. Due to the force of latifundiários in the economy, in politics, and in the state, not even in the last resource did governments use the threat of disappropriation (with compensation below the market price) to increase the offer of land and lower its price.

CONCLUSION

During the 1990s, the agrarian agenda of the WB was linked to neoliberal meta-politics, for which privatizing and deregulating has universal validity. In relation to the principal item on this agenda, MALR, it can be concluded that projects linked to this model in Colombia and in Brazil did not contribute to the democratization of landholding structures in rural lands. In fact, its implementation served more to offer a politically conservative form of access to land in light of the increase of social contradictions in the countryside.

The experiences analyzed question the WB's discourse about the feasibility of MALR as a reference for the formulation of land access programs in societies with elevated indices of landholding concentration. Like all public policies, MALR could have been technically improved - e.g., increasing participation and transparency mechanisms, lowering interest rates, etc. Nevertheless, the model had insoluble limits and contradictions, such as dependency on the offer of land by landholders and the incapacity to democratize landholding structure or to achieve a social scale, due to the payment in cash and the market price. Moreover, based on the idea of homo economicus and an economicist vision of the land, the model assumes that landholding speculation will be something merely temporary, and not structural, for which reason it does not offer instruments to understand - and attack - the forms through which the relations of power configure land ownership in highly unequal societies.

Not by chance, in the two countries rural social movements continued to demand a profound agrarian reform which would democratize the landholding structure and guarantee conditions of social reproduction for peasants, who depended on the redistributive action of the state and public policies linked in the wide-ranging development strategies.

REFERÊNCIAS

• BALCÁZAR, Álvaro et al. Colombia: lecciones de la reforma agraria. In: TEJO, Pedro (Comp.) Mercados de tierras agrícolas en América Latina y el Caribe: una realidad incompleta. v.1. Santiago de Chile: Naciones Unidas; Cepal; GTZ, 2003.
• BINSWANGER, Hans; DEININGER, Klaus. South African Land Policy: The Legacy of History and Current Options. World Development, v.21, n.9, p.1451-1475, 1993.
• BINSWANGER, Hans et al. Power, Distortions, Revolt and Reform in Agricultural Land Relations. Handbook of Development Economics, v.3, n.4, p.4-42, 1995.
• BRANDÃO, Wilson. A reforma agrária solidária no Ceará. In: LEITE, Pedro Sisnando et al. (Org.) Reforma agrária e desenvolvimento sustentável. Brasília: MDA/CNDRS/NEAD; Paralelo 15, 2000.
• BORRAS JR., Saturnino M. Questioning market-led Agrarian Reform: Experiences from Brazil, Colombia and South Africa. Journal of Agrarian Change, v.3, p.367-394, 2003.
• BUAINAIN, Antônio M. et al. Estudo de avaliação de impactos do programa Cédula da Terra. Brasília: MDA/NEAD; Unicamp, relatório final, nov. 2003.
• _______. et al. Relatório preliminar de avaliação do Projeto Cédula da Terra. Brasília: NEAD/MEPF; Unicamp, jun. 1999.
• BURKI, Shahid Javed; PERRY, Guillermo. The Long March: A Reform Agenda for Latin America and the Caribbean in the Next Decade. Washington, DC: The World Bank, 1997.
• CARVALHO F, José Juliano de. Política agrária do governo FHC: desenvolvimento rural e a Nova Reforma Agrária. In: LEITE, S. et al. Políticas públicas e agricultura no Brasil. Porto Alegre: Ed. UFRGS, 2001. p.193-224.
• CHERNICK, Marc. La industria y el desarrollo de la droga en la región andina y el conflicto armado en Colombia. In: _______. Acuerdo posible: Solución negociada al conflicto armado colombiano. Bogotá: Aurora, 2008.
• CONGOST, Rosa. Tierras, leyes, historia: estudios sobre "la gran obra de la propiedad". Barcelona: Crítica, 2007.
• DEININGER, Klaus. Política y administración de tierras: lecciones recogidas y nuevos desafíos para la agenda de desarrollo del Banco Mundial. Washington, DC: World Bank, 2001.
• _______. Fazendo a reforma agrária negociada funcionar: experiência inicial da Colômbia, Brasil e África do Sul. In: LEITE, Pedro Sisnando et al. (Org.) Reforma agrária e desenvolvimento sustentável. Brasília: MDA/CNDRS/NEAD; Paralelo 15, 2000.
• DEININGER, Klaus; BINSWANGER, Hans. The Evolution of the World Bank's Land Policy: Principles, Experience and Future Challenges. The World Bank Research Observer, v.14, n.2, p.247-276, 1999.
• FAJARDO, Darío. Las guerras de la agricultura colombiana (1980-2010). Bogotá: ILSA, 2014.
• FAO. El mercado de tierras y la formación de propietarios en Colombia. Roma, 1994.
• FÓRUM Nacional pela Reforma Agrária e Justiça no Campo. Segunda solicitação ao Painel de Inspeção. Brasília, 27 ago. 1999a.
• _______. Solicitação ao Painel de Inspeção. Brasília, 10 out. 1998a.
• _______. Complementação ao Pedido de Inspeção. Brasília, 21 dez. 1998b.
• GARCÍA, Antonio. Sociología de la reforma agraria en América Latina. Buenos Aires: Cruz del Sur, 1973.
• GONZÁLEZ, Fernán. Poder y violencia en Colombia. Bogotá: Odecofi/Cinep, 2014.
• GRUSCZYNSKI, Diana; JARAMILLO, Carlos F. Integrating land issues into the broader development agenda: Colombia. Land Reform, v.3, p.73-102, 2002.
• HOBSBAWM, Eric. Rebeldes primitivos. Barcelona: Ariel, 1968.
• HÖLLINGER, Frank. Del mercado de tierras al mercado de reforma agraria. In: MACHADO, A.; SUARES, R. (Org.) El mercado de tierras en Colombia. Bogotá: CEGA/IICA/TM, 1999. p.137-195.
• INSPECTION PANEL. Report and Recommendation on Request for Inspection - Brazil: Land Reform and Poverty Alleviation Pilot Project. Washington, DC: May 1999a.
• _______. Report and Recommendation on Second Request for Inspection - Brazil: Land Reform and Poverty Alleviation Pilot Project. Washington, DC, Dec. 1999b.
• KALMANOVITZ, Salomón; LÓPEZ, Enrique. La agricultura colombiana en el siglo XX. Bogotá: FCE, 2006.
• MEPF. Ministério Extraordinário de Política Fundiária. Banco da Terra. Brasília, 1999.
• MONDRAGÓN, Hector. Colombia: o mercado de tierras o reforma agrária. In: SCHWARTZMAN, S.; SAUER, S.; BARROS, F. (Org.) Os impactos negativos da política de reforma agrária de mercado do Banco Mundial. Brasília: Rede Brasil, 2003.
• PEÑA, Mario A. Contrapoder y justicia guerrillera: fragmentación política y orden insurgente en Colombia (1952-2003). Bogotá: IEPRI; Debate, 2014.
• RESTREPO, Vilma F. Orden contrainsurgente y dominación. Bogotá: Siglo del Hombre, 2009.
• URBINA, Hernando. Política de reforma agraria en época de apertura económica - caso de Colombia. In: REYDON, B.; RAMOS, R. (Coord.) Mercado y políticas de tierras. Campinas: Ed. Unicamp, 1996. p.181-211.
• VAN ZYL, Johan et al. (Ed.) Agricultural Land Reform in South Africa: Policies, Markets and Mechanisms. Cape Town: Oxford University Press, 1996.
• VICTOR, Andréa; SAUER, Sérgio (Coord.) Estudo sobre a política do Banco Mundial para o setor agrário brasileiro com base no caso do Projeto Cédula da Terra. Brasília: CPT/MST; Rede Brasil; FIAN-Brasil, 2002.
• WOLFF, Luciano; SAUER, Sérgio. O Painel de Inspeção e o caso do Cédula da Terra. In: BARROS, Flavia (Org.) Banco Mundial, participação, transparência e responsabilização: a experiência brasileira com o Painel de Inspeção. Brasília: Rede Brasil, 2001. p.159-198.
• WORLD BANK (WB). Colombia: Land Policy in Transition. Report nº 27942-CO, 2004.
• _______. Land Policies for Growth and Poverty Reduction. Washington, DC, 2003a.
• _______. Rural Poverty Alleviation in Brazil: Toward an Integrated Strategy. Washington, DC, 2003b.
• _______. Implementation Completion Report (CPL - 41470). Report no. 25973, 2003c.
• _______. Country Assistance Strategy - Brazil (2000-2003). In: BARROS, Flávia; MINEIRO, Adhemar dos S. As estratégias dos Bancos Multilaterais para o Brasil. Brasília: Rede Brasil, 2001.
• _______. Rural Development: from Vision to Action - a Sector Strategy. Washington, DC, 1997a.
• _______. Project Appraisal Document to Brazil for Land Reform and Poverty Alleviation Pilot Project. Report no. 16342-BR, 1997b.
• _______. Land Reform and Poverty Alleviation Pilot Project. Report no. PIC 4974, 1997c.
• _______. Staff Appraisal Report: Rural Poverty Alleviation Project - Ceará. Report no. 14395-BR, 1995.

NOTES

• 1
  For a theory criticism of this focus, dominated by economists, and its implications for social history studies of land ownership, see Congost, 2007).
• 2
  A program created in 1995 with the aim of coordinating governmental actions aimed at combatting hunger and poverty. It was intended to involve collaboration between public agencies, private entities, and NGOs.
• 3
  Until that moment, it was a WB norm not to directly finance land purchases. However, this position was made more flexible in 2002, when this type of operation was authorized in Malawi.
• 4
  San Benito Abad (Sucre), Montelíbano (Córdoba), Puerto Wilches (Santander), Rivera (Huila), and Fuente de Oro (Meta).
• 5
  Created in 1995 as a space for discussion and the organization of collective actions, in 1997 it housed more than thirty entities, of which the principal ones were CONTAG and MST.
• 6
  Created in 1994 as an independent agency to listen to social agents who felt directly or indirectly harmed by projects funded by the institution. Complainants had to demonstrate that the negative effects resulted from the WB's non-compliance with its own norms.
• 7
  For reasons of space and due to the fact that the Land Bank only began to operate timidly in 2000, and CFCP in the second half of 2001, this analysis will only focus on PCT.

Publication Dates

• Publication in this collection
  17 Nov 2015
• Date of issue
  Jul-Dec 2015

History

• Received
  13 May 2015
• Accepted
  10 Sept 2015