Evaluation of the Cédula da Terra project (1997-2002)

Abstracts

The article evaluates the pilot-project Cédula da Terra, inspired on the "market-assisted land reform" model of the World Bank, implemented since 1994 under various formats in countries marked by the deep agrarian problems and strong social tensions in the countryside, such as Colombia, South Africa, Guatemala and the Philippines. In Brazil, the project was implemented in five states of the federation and was acclaimed by the Bank as a successful and textbook experience. From it were developed other similar programs on a large scale. The article questions if the World Bank's diagnostics are indeed justifiable and if the Cédula da Terra was really a viable reference to be replicated in Brazil in larger scale.

World Bank; Cédula da Terra; Marked-assisted land reform; Rural poverty; Agrarian credit

O artigo avalia o projeto-piloto Cédula da Terra, inspirado no modelo de "reforma agrária assistida pelo mercado" do Banco Mundial, implementado a partir de 1994 sob formatos diferenciados, em países marcados por graves problemas agrários e fortes tensões sociais no campo, como Colômbia, África do Sul, Guatemala e Filipinas. No Brasil, o Cédula foi implantado em cinco Estados da federação e foi aclamado pelo Banco como uma experiência exitosa e exemplar. A partir dele surgiram outros programas similares em maior escala. O artigo interroga se o diagnóstico do Banco Mundial de fato se justifica e se o Cédula constituiu mesmo uma referência viável a ser replicada em maior escala no Brasil.

Banco Mundial; Cédula da Terra; Reforma agrária assistida pelo mercado; Pobreza rural; Crédito fundiário

NEW DEVELOPMENTALISM

Evaluation of the Cédula da Terra project (1997-2002)¹

João Márcio Mendes Pereira

João Márcio Mendes Pereira holds a Ph.D. in History from Universidade Federal Fluminense (UFF), is an associate professor at the Department of History and in the Graduate Program in History at UFRRJ, and coordinates the Research Group on Contemporary Agro (http://r1.ufrrj.br/geac/portal/). @ – joã[e-mail]

ABSTRACT

The article evaluates the pilot-project Cédula da Terra, inspired on the “market-assisted land reform” model of the World Bank, implemented since 1994 under various formats in countries marked by deep agrarian problems and strong social tensions in the rural areas, such as Colombia, South Africa, Guatemala and the Philippines. In Brazil, the project was implemented in five states of the federation and was acclaimed by the Bank as a successful and textbook experience. From it were developed other similar programs on a large scale. The article questions if the World Bank’s diagnostics are indeed justifiable and if Cédula da Terra was really a viable reference to be replicated in Brazil on a larger scale.

Keywords: World Bank, Cédula da Terra, Marked-assisted land reform, Rural poverty, Agrarian credit.

"If the pilot demonstrates the viability of the market-assisted land reform approach and cost estimates for this pilot were representative for the country [ ], [the] program could attend to 1 million families in a little over six years"

(World Bank, 1997b, p.7)

It must be said that this program [Cédula da Terra] has become, according to both the World Bank's parameter and the demand we have had from a very significant number of international delegations, an exemplary case of land reform program through land credit.

(Raul Jungmann, former minister of Agrarian Reform. Public hearing at the Senate's Social Affairs Committee on March 24,1999)

"In Brazil, [ ]a pilot program to allow market-based acquisition of land by beneficiaries has had impressive results, accomplishing the land reform faster than expected. The new approach is now being implemented nationwide".

(Deininger & Binswanger, 1999, p.268)

The "pilot project of land reform and poverty alleviation", better known as Cédula da Terra Project (Programa Cédula da Terra - PCT) was based on the World Bank's Market-Assisted Land Reform (MALR) model implemented since 1994 under different formats in countries marked by severe land problems and serious social tensions in rural areas, such as Colombia, South Africa, Guatemala and the Philippines. In Brazil, the PCT was implemented in five states of the federation (Pernambuco, Bahia, Ceará, Maranhão and northern Minas Gerais), and acclaimed by the World Bank (2003a) as a successful and exemplary experience of access to land via private market transaction through public funding. This pilot project gave rise to more comprehensive programs such as the Land Bank and Land Credit to Fight Rural Poverty, created in 1998 and 2002, and the National Land Credit Program, created in 2003 and in force since then.

Understanding the experience of the PCT, which ended in 2002, is indispensable for evaluating the agricultural credit programs existing in the country. After all, is the diagnosis of the World Bank justifiable? Was the PCT, in fact, a feasible reference to be replicated on a larger scale in Brazil? What was the purpose of this project and what results has it produced? This article seeks to answer these questions. Initially, it provides a summary description of the MALR, the intellectual matrix of the PCT. Then it outlines the background and political conditions that marked the implementation of the PCT in Brazil. Finally, it provides basic information about the project, to then evaluate its performance. The main points of the argument are resumed at the end.

The World Bank's model of market-assisted land reform

The MALR was conceived as an alternative to the difficulties and inefficiencies of the so-called "State-led land reform", based on the instrument of expropriation. The rupture with past experiences would lie in the fact that the new model was focused on stimulating and enhancing land markets rather than replacing them. According to its proponents, the goal of the MALR is to promote the purchase of land at low prices and enable new agricultural enterprises, in order to improve economic efficiency, foster social equity and fight exclusion (Deininger and Binswanger, 1999, p. 249). The historic opportunity for applying the new model in societies marked by a high concentration of land ownership would have arisen from the implementation of macroeconomic reforms and structural adjustment programs, in that they would have pulled down land prices and reduced the privileges (such as subsidies, low taxation and waiver of agricultural debt) offered to large landowners. Hence the possibility to implement a "land reform that is less detrimental to the functioning of markets" (ibid, p.267).

The MALR combines private equity transaction and distributive policy (Borras Jr. , 2003). Equity transaction because it is a voluntary purchase and sale transaction between private agents - i.e., a typical market operation - and because landowners are paid in cash and at market prices, while buyers assume wholly or in the most part the costs of the land and transaction costs. Distributive policy because it involves the transfer of non-repayable funds in a variable proportion for investments in infrastructure and production. It is therefore a purchase and sale relation between private agents, funded by the State, which provides higher or lower subsidies, as appropriate. Politically, this model has been included in the strategies to alleviate rural poverty that accompanied the Bank's policies for macroeconomic adjustment and liberalization of land markets.

The core principle of the MALR is a voluntary transaction between "willing sellers/willing buyers". Thus, the transaction occurs only when there is someone willing to sell and someone willing to buy land; if there is, ultimately, common will between the parties. This voluntary dimension would eliminate litigation and reduce land conflicts. The MALR is therefore a "negotiated" model based on market transactions and without conflicts of legal, social or redistributive nature (Burki & Perry, 1997, p.95).

The MALR presupposes the division of the social world between customers and suppliers. Market negotiation is seen as the most efficient mean of resource distribution. Therefore, the agents and their motivations are viewed from the "purely" economic angle, land is considered just another commodity, and land conflicts can be dealt with as technical problems that can be solved administratively (Bobrow-Strain, 2004). Market is separated from society, so that the power relations that make up social relations are relegated to the status of "externalities" in relation to the market as "distortions".

The second MALR principle is the "demand-driven approach". Only families demanding land and only those rural properties under demand should be subject to negotiation. Thus, it is assumed that market transactions would be based on the needs and interests of buyers. There is neither supply nor selection of land defined by the State, as is the case in the "traditional" model.

The MALR is also based on associativism. Thus, it requires the establishment of community associations for the selection and acquisition of properties which, in theory, would reduce individual assessment errors and facilitate the selection of higher quality properties, best suited for both the resources available to buyers and the characteristics of the productive project they wish to undertake (Deininger, 2001a).

Another principle is self-selection. Instead of the State selecting the beneficiaries, these would select themselves from full and previous access to information about eligibility and participation rules, and compliance with certain requirements such as proof of ability to pay. It is assumed that only those who have knowledge, workforce and some previous savings would accept the rules of the game. However, if the demand lacked sufficient "human capital", there would be the need for some type of training in order to ensure the payment capacity of borrowers (Deininger, 1998 and 2001a; Theophilus, 2003).

Another principle is decentralization. The basic cell of the model consists of autonomous resource management by community associations that negotiate the purchase of the land and would make productive investments. Understood as synonymous with "bureaucracy", decentralization would reduce the transaction costs of the properties acquired by eliminating the intermediation of central agencies. It would also enable strengthening relations among different local actors, encouraging dialogue and reconciling interests. Despite the risk of "capture" by local elites, decentralization would be justified in itself, in that the local space is identified not only with greater transparency, responsibility and accountability, but with fewer social conflicts (World Bank, 2003c).

The MALR also entails social participation at all stages of the process, since the potential beneficiaries should be able to select the property and negotiate the price directly with the owners, as well as to establish feasible production plans.

The MALR privatizes not only the land acquisition process, to the extent that the buying/selling relationship takes place between private agents, but also the services necessary for productive development (technical assistance, for example) (Binswanger, 1995a, p.93). According to its proponents, it cannot be effectively implemented by "centralized ministries or state bureaucracies" (Binswanger, 1995b, p.18). Hence the need to reduce as much as possible the participation of central government agencies and incorporate local spheres of government, NGOs and, in general, the "private sector" which, coordinated in decentralized arrangements would provide the best conditions for the flow and transparency of business operations (Deininger, 2001a, p.20).

The amount earmarked for the purchase of land is granted as a loan that must be repaid, and the amount for infrastructure construction and production development is granted as subsidy. Presumably, the logic of this mechanism would encourage bargaining for the price of the property, because the lower the price of the land is, the more resources will available for social and productive investments. Therefore, the new model would favor the purchase of marginal and unproductive lands at the lowest possible price (Deininger, 2001b, p.77). Also, this mechanism would reduce the chances of "collusion" between stakeholders, insofar as the amount of non-reimbursable funds would depend on the outcome of the bargaining for price of the land. Thus, the operating logic of the MALR would not stimulate the increase in the prices of rural properties. Furthermore, this scheme would be cheaper than expropriation, as it is decentralized and less bureaucratic, provides a subsidy equal to or less than that given by the "traditional" model, and above all because the beneficiaries are responsible for paying for the land. Therefore, there would be net savings of resources for society. To reach the poorest segments of the rural population, income ceilings could be set as criteria for access to financing.

According to World Bank economists, one of the problems of the "expropriation model" would have been the neglect of productive development and the increased competitiveness of settlers. In contrast, in the MALR productive projects should be developed before the purchase of the land, so that no real estate transaction would be conducted without the prior development of viable production plans (Deininger, 2001a, p.14). Moreover, part of the non-refundable resources available for productive investments could be used by borrowers to hire technical assistance services, which would improve the economic viability conditions of the new ventures.

The MALR is underpinned by the redefinition of the role of the State in relation to access to land by rural workers. Thus, the State would be solely responsible for: a) defining the "rules of the game", i.e., the broader regulatory frameworks, in order to stimulate the supply of land on the market; b) building a decentralized land management structure, in order to provide better conditions for the flow of market transactions (purchase and sale and lease); c) providing loans and credits to self-selected borrowers; d) providing training as needed; e) building infrastructure when the borrowers themselves are unable to do so (Burki & Perry, 1997, p.95).

According to the World Bank (n.d., p.2), there is the risk that the price of the land could increase due to the availability of credit for the purchase transaction, where these either did not exist or occurred on a smaller scale before, which would increase the cost of equity transactions and render them unfeasible. Hence the need to ensure that the supply of land were much higher than the demand. In other words, the MALR should only apply in locations where supply exceeds demand.

To increase land supply on the market, six actions should be considered: i) elimination of all subsidies, tax exemptions and protective tariffs that favor large landowners and are capitalized in land prices, raising it above agricultural profitability; ii ) end of legal restrictions that "distort" the operation of the land buying/selling/leasing markets, so that more "land assets" could be negotiated; iii) progressive taxation to discourage speculation in land ownership; iv) formalization and titling of property rights; v) establishment of market information systems to guide economic agents on the prices and characteristics of real estate; vi) reduction of transaction costs through administrative and legal simplification (Binswanger, 1995a and 1995c; Binswanger & van Zyl, 1995).

The political dimension behind the implementation of Cédula da Terra

For the World Bank, the macroeconomic measures of the Real Plan would produce a marked deterioration in living conditions in the rural areas, requiring, however, the creation of compensatory social programs focused on alleviating poverty (van Zyl et al., 1995 p.2). At the same time, the same policies have provided a historic opportunity to establish the land market as the central mechanism for reducing rural poverty:

The agricultural reforms which the government has implemented have contributed to a recent decline in land prices. [ ] This development may prove to be an important anti-poverty instrument for rural areas because it also should contribute to a more efficient land market. The government could take advantage of this opportunity and improve further the ability of the poor to buy land by providing grant resources to them" (World Bank, 1995a, p.xiii)

The PCT was preceded by the São José project (or "Solidarity Agrarian Reform"), a very small experiment started in Ceará by the Jereissati government in February 1997 under the Rural Poverty Combat Program (PCPR in the Portuguese acronym), which replaced a number of projects that had been funded by the World Bank in the Northeast since 1975. Redefined as actions to compensate for structural adjustment policies (World Bank, 1995b), the PCPR was included in the Solidarity Community program in 1995, with the aim to finance social and productive infrastructure projects. However, in Ceará a land component was introduced in the PCPR, which enabled financing the purchase of land, giving rise to the São José project. In 1996 the government of Ceará established a state fund in which it invested approximately R$4 million plus a WB loan of $6 million. The project was launched in April 1997, for the purpose of financing the purchase of 40,000 hectares by 800 families during one year. In practice, the project financed the purchase of 44 properties by 694 families, totaling 23,622 hectares. Where it was implemented, the project stimulated an increase in land prices (Brandão, 2000). This experience led to the emergence, a few months later, of the PCT, from a WB proposal to the Brazilian government. The loan agreement for the creation of Cédula da Terra was approved by the World Bank in April 1997 - with no review of the performance of its predecessor, which was still underway.

The MALR proposal therefore was introduced in Brazil as a land component of compensatory programs already underway in rural areas aimed at alleviating the socially regressive impact of economic liberalization. Moreover, the federal government and the WB found it necessary to reduce the pressure caused by land occupations organized mostly by the Landless Movement (MST in Portuguese), by introducing a mechanism that could encourage the accession of landless workers. A region with the highest concentration of rural poverty in the country (Rocha, 2003, p.89) and where the WB had been operating for over 20 years was then selected, so that the implementation of the PCT could achieve immediate results (World Bank, 1997a, p.3).

The national and international impact of land conflicts and rural violence, coupled with increased land occupations, led the Brazilian government to implement the PCT (World Bank, 1997b, p.7). According to the assessment of its operators, the introduction of the MALR would disrupt the connection between occupations and expropriations and would allow State agencies to fund and facilitate buying/selling transactions between workers and landowners. The goal, therefore, was to prioritize the MALR to the detriment of expropriation:

The government model of land reform through land distribution is a vicious cycle: land is redistributed where is a social conflict, and social conflicts put pressure on the government land redistribution program [...] As new alternatives star to take effect, the government may be able to reduce the emphasis on expropriations and consequently break the link between its land reform policy and rural conflicts (World Bank, 2003a, p.127)

Despite the modest goals of the pilot project - to finance the purchase of land by 15,000 families in four years, its results should legitimize the extension of the MALR on a large scale in Brazil, as shown in the headings of this article. In turn, the federal government maintained the discourse that it was a complementary action to the agrarian reform.

Immediately, the PCT was criticized by the MST and CONTAG, and identified as part of the State's withdrawal from social issues and an instrument incapable of democratizing the agrarian structure. Despite the criticism, the PCT began in December 1997 and two years later had benefited 6,798 families in five states (Buainain et al., 1999 p.56). For the federal government these figures demonstrated the "demand" for this type of initiative.

However, the fact that the project was launched in the drought season had a direct impact on demand, converting access to land into a means of immediate survival (Buainain et al., 1999 p.27). Furthermore, there was intense propaganda extolling the possibility of access to land "without conflict", addressed to a huge and impoverished rural population living in a region with no job opportunities, in a context of repression against land occupation by social movements. The preliminary assessment commissioned by the federal government and financed by the World Bank is clear in this regard:

In a region characterized by high concentration of land ownership and social exclusion [...] the possibility of "easy", "risk-free", de-bureaucratized and immediate access to land favors and encourages adherence to the Program [...]. Many interviewees referred to this aspect, contrasting the facility offered by Cédula da Terra with the risks of invasions and difficulties of life in a camp of landless workers [...]. It is noticed that the interviewees reproduce the official discourse of solidarity agrarian reform without conflicts and through partnerships. (Buainain et al., 1999 p.27 and 271)

Finally, there was also a double pressure to accelerate implementation of the PCT from state governments - with an eye in the 1998 elections - and the federal government and the WB - which were interested in legitimizing the new model (ibid, p.272).

More than half of the 223 projects recorded in January 1999 were implemented in the second half of 1998 (ibid, p.15), a period of intensification of electoral competition, increased land occupations and outbreaks of looting in the Northeast, which conveyed a negative image of the federal and state governments (Carvalho Filho, 2001 p.208-9).

The alliance between the federal government and the World Bank in favor of extending Cédula da Terra to the entire country marked the implementation of the project since its inception. Social movements and trade unions organized in the National Forum for Agrarian Reform and Justice in the Rural Areas to protest against the project, submitted two requests for investigation to the WB Inspection Panel.² It makes no sense to review this experience without taking into account the political context in which it is inserted (Medeiros, 2002; Pereira, 2007, 2010b).

Rules and operation structure of Cédula da Terra

The objectives of the PCT were to reduce rural poverty, raise agricultural income and test an alternative model to the "expropriation-oriented" land reform. Its target audience were landless rural workers (wage workers, partners and tenants) and farmers (landowners or not) with insufficient land to provided for the family's self consumption. Access to the project was restricted to community associations either legally established or existing or created for that purpose.

With a four-year implementation scheduled, the PCT was estimated at $150 million, to be provided by the following sources: $90 million by the WB; $45 million by the federal government; $6.6 million by the state governments; and $8.4 million by community associations in the form of labor. These funds should be spent as follows: 30% ($45 million) in the purchase of land; 56.2% ($84.3 million) in complementary investments; 2.6% ($3.9 million) in technical assistance and training; 6.7% ($10.1 million) in the project monitoring, supervision and management; 4.5% ($6.7 million) in evaluation and advertising by the federal government (World Bank, 1997b, p.2). The WB loan would finance complementary investments, while the federal government would fund the purchase of land.

Following the same logic of the São José Project, the PCT was divided into two subcomponents: one aimed at land purchase (SAT) and the other at complementary investments (SIC). The SAT offered funding repayable in ten years, with a three-year grace period, and financial indexation based on the TJLP (Portuguese acronym for Long-term Interest Rate).³ In turn, the SIC provided non-reimbursable funds for community investments in three basic areas: infrastructure (electrification, road improvement, water supply, etc.); social improvements (schools or health clinics, community centers, etc.); and production (irrigation, tractors, small-scale processing, etc.).

The credit ceiling per family was $11,200, including expenses related to land purchase, registration, measurement, taxes and community investment. Each family received $1,300 in grants for establishment purposes. There was a maximum subsidy of $6,900 per family, including the establishment allowance, the subsidy of up to 50% embedded in the land loan, and other subsidies on the amount of the SIC.

The logic of this financing scheme was that the smaller the amount spent on land purchase (refundable) were, the higher the amount of subsidy and resources for complementary (non-refundable) investments would be. The loan was taken out jointly and severally, so that the associations were legally responsible for paying the loan installments. Once the loan had been repaid, each family would receive a property deed.

In theory, the operation of purchasing rural properties should follow as main criteria: i) the search for properties with economic potential and a low level of additional investment; ii) the compatibility of the negotiated price with the market parameters; iii) compliance with the legal requirements for registration and transfer of rural properties; iv) good access conditions, water supply and reasonable infrastructure; v) sufficiently large area, as a general rule as large as or larger than the minimum land subdivision in the region (Buainain et al, 1999 p.20).

The PCT should work as follows: i) rural community associations would select the property they wanted to buy and negotiate the price directly with the owner, submitting to the agency responsible for managing the PCT, the property purchase proposal and the list of community subprojects; ii) the agency would review the proposal regarding the eligibility of beneficiaries and the property, preparing a technical appraisal report of the property; iii) once the proposal was approved, the agency would guide the development of a detailed land purchase project and issue a technical opinion; community investment projects, in turn, would be developed by third parties or government institutions participating in the PCT; if the land purchase proposal was rejected, the agency would advise the association to continue negotiating; iv) if the proposal was approved, Banco do Nordeste was authorized to finance the association, paying the owner and the providers of title transfer services and billing and collecting payments from the loans of each beneficiary; Banco do Nordeste should also transfer funds for community investment directly to the associations (ibid, p.24-5).

Following the principle of decentralization, a particular institutional arrangement was established in each of the five states, widely using the WB structure set up to implement the PCPR. In the design of the PCT, local councils - usually created to manage the PCPR - also appeared implementation and social participation forums.

At local level, the legally established community associations should operate as the basic cells of any project, since they would be responsible for selecting the rural properties for sale, negotiate the price directly with the owner, identify the necessary investments to be funded and manage the funds disbursed.

Performance and results of Cédula da Terra

The PCT began in December 1997 and ended in December 2002. It is known that its implementation was partially interrupted in late 1999 and part of 2000 for lack of federal government funds, due to the currency crisis and the two requests submitted to the Inspection Panel (World Bank, 2003a, p.13; Buainain et al, 2003, p.17).

The PCT used $121.3 million as follows: from the World Bank ($68.5 million); from the federal government ($45 million); from the state governments ($1.9 million); and from associations ($5.9 million). In turn, the actual cost of the project was distributed into five items: land purchase ($45 million); complementary investments ($66.4 million); technical assistance and training ($2.6 million); monitoring, supervision and management ($2.1 million); evaluation and advertising ($5.2 million) (World Bank, 2003a).

The devaluation of the Real in 1999 helped reduce the final cost of the project, which is why the $21.4 million loan was cancelled by the federal government (ibid, p.26). Until 2000, the funds for land purchase under the PCT came from INCRA - the same source of the agrarian reform program. Thereafter the funds came from Banco da Terra (Land Bank), totaling R$38,465,000.00.

The spatial coverage of the PCT was large, except in MG, where it was restricted to the north, comprising 50 municipalities. In the states if Ceará and Maranhão, virtually all municipalities were covered. The PCT coverage was large also in Pernambuco and Bahia. The selection of municipalities was driven by factors such as the existence of land conflicts, situation of extreme poverty, existence of rural workers' unions favorable to the project, operational capacity of the technical unit and support of local governments and local political leaderships (Buainain et al., 1999, p.22).

The coverage level achieved demonstrates the intention of spreading the incidence of the project as much as possible, which alone overthrows the argument that Cédula da Terra was a complementary program to expropriations - the number of which, incidentally, fell sharply in all states where the project was implemented (Victor & Sauer, 2002), particularly considering that, back then, areas subject to expropriation could be purchased through the project. Moreover, the criteria that guided the state agencies in the implementation of the PCT suggest that it was intended to be used as a tool of competition with social movements, in that it sought to prioritize conflicting areas and extremely poor segments.

The implementation of the PCT was not subject to any social control. Municipal councils linked to the PCPR had, in practice, a minor role in the project management (Steil, 2000, p.21). As a rule, community investment subprojects were submitted directly to the state technical units, thus annulling the role of municipal councils, the only institutional space provided for any type of social participation in the management of Cédula da Terra.

The PCT is underpinned by the principle of voluntary transaction between willing sellers/willing buyers. No research has defined the socioeconomic profile of sellers, but the data indicate that the rural properties sold were, in most cases, underutilized or abandoned because of drought and the crisis of livestock and traditional cultures (Buainain et al., 1999 p.31). Misused and abandoned properties were distributed as follows: 76% in Bahia; 81.3% in Ceará; 81.3% in Maranhão; 91.7% in Minas Gerais; and 83.3% in Pernambuco (Buainain et al., 1999, p.118, 2003, p.105). Therefore, there are strong indications that many of those establishments were bankrupt.

As a rule, the properties purchased were of medium size with an average area of 815.3 ha. There were few projects over 2000 ha, and projects with very small areas were also rare (the smallest is 68 ha). About 31% of the projects had up to 500 ha and 71.7% up to 1000 ha. There were only twenty projects with an area above 1200 ha (Buainain et al., 1999 p.131). The preliminary evaluation did not check whether the larger properties acquired were subject to expropriation or not. However, Victor & Sauer (2002, p.84) detected the purchase by the PCT of areas subject to expropriation in all states.

Generally, buyers had a very low level of education (37.7% were illiterate and 47.1% had gotten as far as the 1st and 4th grades) and lived in extreme poverty, so that the total annual income before joining the PCT was estimated at R$2,057 for a family of 5.2 members on average (somewhere around R$32.90 per person per month, less than one third the minimum wage at the time) (Buainain et al., 1999, p.273). The vast majority was estimated as consisting of waged rural workers and non-owner farmers (partners and tenants), usually with additional occupations. Nearly all borrowers had been living in the area for many years, and a significant number had been born in or around the region. Approximately 90% of the population worked in rural areas (ibid, p.63-9 and 104). It is no surprise, then, that the appalling living conditions of the population in the rural areas affected the "demand" for the PCT: "In this context of extreme poverty, participation in Cédula da Terra has played also an emergency option role, giving it, inevitably, a welfare role" (ibid, p.30). Victor & Sauer (2002, p.34-5) also found a similar socioeconomic profile, reinforcing the perception that the lack of alternatives was one of the most powerful reasons for joining Cédula da Terra.

Given this diagnosis, it is not surprising that many "customers" were willing to accept higher prices and therefore higher levels of debt in order to have immediate access to land. Example: half the associations purchased the first property offered, without seeking any further, and only for 13.5% the price of the property was a relevant fact in the real estate transaction phase (Buainain et al, 1999, p.120).

Political and cultural factors also influenced participation in the PCT. As already seen, the project implementation gained momentum in the second half of 1998, precisely the period of intensification of the electoral run. Moreover, media campaigns disqualified land occupations while criminalizing this form of social struggle. In turn, the "dream" of access to land was also a motivating factor for joining the project (Buainain et al, 1999, p.278; Victor & Sauer, 2002 p.35).

As seen above, the willing sellers/willing buyers principle seems to neglect the existing context, as the commercial buying/selling relationship operated in a hypothetical and nonexistent situation of social void, in which the free will of rational economic agents prevailed. It can be stated, therefore, that the main category supporting the MALR and the PCT will not survive a reality check.

Another MALR principle also incorporated into the PCT was the "demand-driven" or " community-driven" approach (Theophilus, 2003). Under this principle, land acquisition should be in accordance with the effective demand, in such a way that only the rural properties demanded should be subject to equity transactions. Under Cédula da Terra demand had to be organized into associations, as this would reduce errors of individual evaluation and facilitate the selection of higher quality properties that were best suitable for local conditions, for the resources available for the group of buyers and for the characteristics of the productive project one wished to implement. Furthermore, the commercial transaction involved negotiation between stakeholders, so that the whole land acquisition process would be accomplished with minimal interference by the state bureaucracy and maximum participation of stakeholders.

For the "demand-driven" focus to work, it would be necessary to draw up a "governance structure" (mechanisms of coordination, incentives and sanctions) that enabled the proper selection of land assets and negotiation between stakeholders (Buainain et al. 2002, p.75-6). It was assumed, therefore, that the customers would act according to a maximizing economic rationality. This explains why the governance structure of the program was set up on microeconomic incentives, as well as why the evaluation financed by the WB widely resorted to microeconomics to analyze the performance of the PCT.

However, neither the rural workers acted according to these assumptions, nor the property selection process involved these individuals. Under the plot of actually existing social relations in the rural Northeast, the selection of properties acquired through the PCT was far from involving the participation of the "customers":

Like in the definition of prices, the workers had no part in the negotiations, because most of them were conducted in absentia of beneficiary families [...]. Families had little, if any, power in decisions about the areas acquired. All persons interviewed in three states (MG, BA and MA) - including the presidents of the associations - said they had no direct participation in the selection or purchase of the land. (Victor & Sauer, 2002 p.109)

A preliminary evaluation indicated that in 60% of the cases the initiative was taken by the associations, and 34.6 % by the landowners (Buainain et al., 1999 p.115). However, one cannot subtract from these percentages the reliable evidence of the active role of the associations, since their participation was marginal in the negotiation of the properties purchased: "The vast majority of associations, even those that claimed to have negotiated directly with the owner, played a secondary role in the process, limiting themselves to receiving the proposal, submitting it to the agency in charge, coming back with a counter-proposal and so on" (ibid , p.120-1).

The evidence of the weak role of the associations in the real estate negotiation process reveal that the role attributed to them in fact did not exist. In practice, the process was led by government agencies:

In all cases considered the negotiation is unequal. Stakeholders are divided into unequal rights in relation to the land market, and the alleged full information is a fiction. Even in situations in which the negotiation occurs [...] between owners and members without direct government involvement, what occurs in most cases [is] a negotiation between government agencies and landowners. For these, the potential buyer is the State and they do not care if the members will be able to pay or not. The owners know that the weight of the government is crucial in the negotiation and that it is the technical staff of the government agencies [...] that prepare the technical report, set the level of the land value, veto the sale of the property due to existing irregularities and oftentimes "convince" the parties. Ultimately, the deal is closed by the government and not by the association. (Buainain et al., 1999 p.121)

One year earlier an official evaluation had already diagnosed the same situation:

IDACE, as a government agency, has significant participation in the price and form of payment of the land in question, mainly due to the lack of initiative or lack of knowledge of market prices by beneficiaries. [...] IDACE emerges as a direct negotiator with sellers, contradicting its role of arbiter or mediator in the process. (Groppo et al., 1998, p.3-4)

Another study also confirmed the socio-political weakness of the associations, which contrasts with the ideal participation schemes conceived by the WB staff:

When the World Bank designs and implements its participation policy, it presupposes the existence of social subjects with minimum conditions for training and social organization. As in Brazil these subjects are generally precarious, the ideal world of participation designed in World Bank project documents often collide with the real world of social relations marked by patronizing and pork barrel practices. (Steil, 2000, p.55)

Furthermore, Steil detected the weakness of the associations, showing that the experience of the PCT had not engendered the leadership and free initiative of rational economic agents:

Cédula da Terra has failed to produce a significant experience that could be replicated on a larger scale project in terms of workers' training in management and autonomy. On the contrary [...], the most common was for project coordinating agencies in the states to lead the negotiation process. (Steil, 2000, p.26)

Victor & Sauer also pointed out the weakness of the associations in playing the role assigned to them by the MALR, the power inequality in negotiation between buyers and sellers, and the leadership, patronizing and protection of government agencies:

There was no effective participation of families and associations in the processes of selecting the purchased areas [...]. Proposals for community participation and empowerment are totally compromised because in fact Cédula da Terra is a program fully led by the State itself [...]. The field research also confirmed the total imbalance in negotiation processes, noting the absence or submission of buyers to processes conducted by owners, government agencies or other "forces" alien to the interests of the families [...]. The difficulties of dealing equitably with large landowners open possibilities to reproduce the historical processes of manipulation and exploitation of workers. These end up being forced to accept "the terms of the deal." (Victor & Sauer, 2002 p.80 and 87)

Not only the associations had little or no initiative and autonomy, but their own establishment resulted from external pressures:

There is a significant participation of external agents and other institutions in the process of establishing and running the associations. The official proposal characterizes the participation of small farmers in the PCT as a process whose initiative would come from the stakeholders themselves. However, [...] the creation of associations is not as "natural" as expected. In all states, [...] there is clear intervention of actors and institutions outside the group, such as mayors, local politicians, "well-meaning" people, landowners, technical staff related to governmental institutions, etc. (Buainain et al., 1999, p. 279)

Half the associations were created exclusively to participate in the PCT, primarily from the initiative of government agencies directly or indirectly linked to the project, as well as mayors, local politicians and landowners, revealing a purely administrative procedure (ibid., 1999, p.223). Hence its low organicity and lack of representation (Navarro, 1998; Steil, 2000). In extreme cases the association was established after the property had been purchased and the first board was defined by outside agents (Victor & Sauer, 2002 p.69). However, even the existence of previous organizational forms was not considered able to overcome the weight of the balance of power existing in the northeastern rural area, strongly marked by the "populist tradition that defines an area of representation based on the cooptation, subjection and social and political control of the poor" (Buainain et al., 1999 p.281).

As for the nature of foreign interference, the preliminary assessment distinguished three general situations: in the first, the external agent merely informed rules and provided guidance on how to proceed; in the second, it had a more direct and effective presence in the organization of the association (for example, participation in meetings, assistance in getting documents and financial contribution); in the third, the external agent controlled directly and explicitly the decisions made. Most cases involving the action of government agencies were included in situations of guidance and organization, while the cases of organization and direct control were observed more in relation to the action of local politicians, mayors and former owners (ibid, p.234).

A factor that added to the weakness of the associations was their isolation from broader organizational networks, such as those created by the action of social movements and trade unions (Steil, 2000). In fact, by implementing Cédula da Terra against the will of all national entities representing rural workers in the country, the federal government and the World Bank isolated even further associations that already suffered from an original artificiality.

The extremely low level of exchange of basic information about the project also weakened the associations, limiting the borrowers' autonomous participation ability:

60% of the population declared they were not aware of the interest rates they would pay on the loan to purchase the land. 11% answered incorrectly and only 0.2% gave the right answer [...]. 35.9% to 49.6% of the beneficiaries [...] do not know that the property purchased is used as collateral for land credit. Only 3.1% to 9.8% [...] know that [...]. Around 19% of the population of beneficiaries, corresponding to approximately 1270 people, declared that no guarantee had been required. Around 30% said they had not taken out a loan to join Cédula da Terra [...]. Virtually no one knew the precise terms of the loan. (Buainain et al., 1999, p.105-6)

Another pillar of the MALR reproduced in the PCT was the self-selection of borrowers. This mechanism would have the advantage of avoiding selection errors by the State and reducing the costs of identifying and monitoring borrowers (Theophilus 2003, p.218). However, once again reality did not fit the model. The self-selection process simply did not exist:

The selection of beneficiaries should reflect a self-selection process with low interference from the government agencies responsible for implementing the Program in each state. In practice, this process has not happened in such a "pure" way. Despite the wide dissemination of the program by the mass media [...], a relevant part of the beneficiaries was actually "selected" - or rather "chosen" - to participate. (Buainain et al., 1999 p.50)

Well, if the establishment and operation of the associations was not an autonomous process, how could their members be "self-selected"? Indeed, the selection of borrowers was the object of political bargaining among different public and private actors:

At the institutional level there is selectivity in the origin of the accession of rural landless workers and small family farmers. This occurs through the establishment of inclusion and exclusion rules developed both officially and unofficially. In general, these agencies are not merely supporting actors in the creation of community associations [...]. Even in "purely technical" decisions such as, for example, the selection of candidates, the procedures of legalization of associations, technical appraisal of the property, and the opening of current accounts for receiving funds, the "weight of personal relations" is felt [...]. (Buainain et al., 1999 p.279-80 and 223-24, emphasis in the original)

Figure 1

Government agencies were instructed to select the poorest families (Buainain et al, 2003, p.18), which helps to understand the speed of the "demand" for the PCT. In turn, this instruction entailed the risk of incorporating people so devoid of investment capacity, who later on could become delinquent with Cédula da Terra (Buainain et al., 1999 p.271).

One can therefore say that the self-selection of borrowers never left the realm of abstraction, since "the most common practice was to announce the existence of the program and encourage - or induce - the group to join in" (Buainain et al., 1999, p.52). The diagnosis produced then was clear: "The pre-selection of particular groups disrupts the self-selection process, which is one of the most important pillars for the generation of efficient governance structure by the program" (ibid.).

Moreover, contrary to what is established by the MALR, the development of productive projects did not occur before the purchase of rural properties, but rather after, including on grounds of the urgency of government agencies to implement the PCT (ibid, p.271 and 291). The difficulty of the associations in receiving regular technical assistance that enabled them to formulate appropriate proposals - whose remuneration, initially, had not been provided for in the PCT - was also harmful (ibid, p.195). Therefore, the acquisition of properties ended up not being preceded by any significant technical evaluation of the potential of the project to be started.

There were systematic delays in the release of land as well as of resources for additional investments, which resulted in considerable damages for PCT borrowers (Buainain et al., 1999, p.28; Victor & Sauer, 2002, p.46). However, no similar delays were observed in the payment to those who sold the properties. One cannot say, therefore, that the decentralized arrangement had been immune to serious operational problems, as the discourse that mechanically associates decentralization with administrative efficiency would have us believe.

As for productive development, the preliminary evaluation did not investigate the quality of the land purchased under the PCT, and therefore found nothing about the economic sustainability of the projects. However, before the PCT started its second year of operation, it was found that half of the SIC funds had been spent on family financing (Buainain et al., 1999, p.196-8), thus revealing the emergency nature of the project.

Another significant finding was that the prices of half of the land purchased was ​75% below the average price of total contracts, which would leave to most families an amount of some R$5,000 for community investments (ibid., p.138). However, these funds were spent on various items besides funding - infrastructure like roads, water, electricity and housing, not necessarily with satisfactory results - "not sufficient to complement/adapt existing improvements to the new property use or to establish a solid production base from which beneficiaries could generate enough income to improve their lives and pay the debt" (ibid, p.290). So, it was seen that most projects with fewer families would require complementary loans to ensure their viability, which in turn would create a new source of debt and differentiation between projects.

Another important finding was the poor condition of the technical assistance services offered to borrowers. Besides uninformed about the general conditions and operation of the PCT, the technical staff restricted their activities to specific times, without the necessary continuity (Buainain et al., 1999 p.109).

Victor & Sauer (2002, p.40-4) found an even more negative scenario. Social infrastructures - including electricity and water, roads and access roads, schools and health clinics - were evaluated as very poor in most of the projects visited, although many of them had been in place for more than four years. Therefore, there is no surprise in the high rate of families who give up and leave the areas purchased under the PCT - on average more than 50% - although the projects were still in the grace period. The need for precarious wage employment had not been overcome either, due to production and marketing difficulties. Another finding was the weakness or even absence of technical assistance (ibid. p.49-50).

Victor & Sauer (2002, p.51-61) also noted the occurrence of authoritarian practices by the technicians responsible for implementing the PCT. It was observed, especially, the imposition of areas for community production based on monoculture, with no regard for family plots, with the goal of producing marketable surplus to generate enough income for paying the loan installments. Moreover, this type of service was privatized and should be paid with SIC funds, financed by the World Bank. In practice, this situation linked spending on technical assistance to the management of the SIC budget, which was subject to cuts. Secondly, the provision of that service had been planned for one year, renewable for another year, therefore a period of time insufficient to train the families, especially taking into account the low frequency and poor quality of the service in most cases.

Agricultural production was considered as subsistence production in most projects. Only two of the 16 projects surveyed indicated the existence of marketable surplus. For Victor & Sauer (2002, p. 54), the difficulties precluded commercial viability.

According to the same survey, the land purchased under the PCT was of low quality and concentrated on less dynamic and more impoverished regions of the states, which represented another difficulty for the productive development of the projects. According to the authors, this would be part of the very logic of the PCT, since "the limit of resources for the acquisition of areas in general leads to the implementation of projects in less dynamic regions, with the purchase of less valued land, which are therefore weak and have serious production constraints" (ibid, p.55-6).

Conducted still in the grace period of many projects, the Victor & Sauer (2002, p.63) survey did not get to record information about the payment of the real estate debt. However, it found that the overwhelming majority anticipated a situation of default. Indeed, the logic of programs such as the PCT imposes an insoluble problem: the purchase of cheap properties leads to a decrease in the quality of the land and improvements and, therefore, the amount required for the construction of infrastructure and productive investments needs to be increased; if more expensive properties are purchased, the loan exceeds the borrowing capacity of the families, thus generating default.

Victor & Sauer (2002, p.82-5) also compiled several signs and allegations of deviation from purpose, favoritism and corruption in countless PCT projects, such as fraudulent technical reports; overpricing of rural properties; purchase of various areas from the same owner; acquisition of properties located in Atlantic Forest areas; imposition of procedures by local politicians; collusion between local governments and landowners; and purchase of areas potentially subject to expropriation.

In all states there was a high dropout rate in the biennium 2000-2001 and family replacement in 2002, as shown on Table 1:

More than half of the projects were created from January 2002, indicating an acceleration in the implementation of the PCT similar to that seen in the second half of 1998. In this second phase, the expansion of the projects to new areas initiated in 1999 was already in progress, increasing the presence of the PCT in the states (Buainain et al., 2002 p.87). Apparently, because they were initially concentrated in poor and inadequate areas - as suggested by the high dropout rates, the projects were later expanded to new areas. Moreover, the implementation of the PCT was adjusted after 1999, reducing the number of families per project, with the aim of expanding it to the largest number of micro-regions possible. This reduced the amount of resources available for additional investments, contradicting the logic of the MALR and making families dependent on additional public resources.

Regarding earned annual income, the sample taken between August 2002 and July 2003 arrived at an average gross income of R$5,777.05, representing a monthly income of R$483.64 per family, about 2.1 minimum wages of R$240.00 (amount in force from May 2003 to May 2004). When distributed by strata, it is noticed that: a) 25% of the families were still earning monthly gross income equivalent to a minimum wage; b) another 25% of the families had gross monthly income of 1 to 1.4 minimum wages; c) the rest of the families had monthly income above 1.4 minimum wages, ranging from the R$ 4,253.00 to R$19,894.00 per year. In all probability, for a small group (about 10%) income was far higher than for the others, so that the overall average income per family was pulled upwards. The current average monetary household income, in turn, was estimated at R$3,947.00 per year, or 1.37 monthly minimum wages of R$240.00, slightly above the poverty line set at R$70.00 per person per month (Buainain et al. 2003, p.63-8).

As for income composition, the report indicated that the weight of non-monetary income remained high, accounting for 32% of total gross income and 46% of monetary income. In turn, it pointed out that the weight of agricultural activity in the composition of income had increased significantly, decreasing the relative importance of temporary wage employment, despite the persisting dependence on income from retirement and pensions (ibid, p.177). If the results indicated an increase in income compared to the situation prior to entering the program, the picture outlined was far from the prognostics of excellence regarding the PCT:

The settlements visited, which had been established between 1997 and 1999, today are between 4 and 6 years old. Families [...] today in general succeed in getting from agricultural an income higher than the one they had before the project, but that is not always sufficient for their subsistence. Many settlers supplement their agricultural income by selling their labor and with various government transfers [...]. Some settlements visited were in a very difficult situation, with few families settled and very low levels of agricultural income. (ibid, p.172)

The precariousness of the technical assistance service and of access to credit was identified as the main limiting factor to production (ibid, p.151). Almost 66% of respondents reported having never or only occasionally received technical assistance the in biennium 2002-2003 (ibid, p.135). The survey did not inform how the respondents evaluated the quality of the service provided.

It was found again that SIC resources had been spent on infrastructure (construction of houses, power grid and water supply), depleting the non-reimbursable amount without covering the minimum package of productive investments (ibid, p.100-1 and 150). Besides insufficient, these funds were misused, due largely to the lack of social control:

Besides clearly insufficient to leverage productive activities, the money from SIC was misused. This was due greatly to the lack of social control in the associations over the performance of their boards and the lack of commitment of management units [...]. In the settlements that are more in more serious conditions (some don't even have houses), this factor is always present and often hinders the development of the settlement. (ibid, p.174)

Since SIC funds did not translate into productive investments, access to rural credit ended up being of vital importance to PCT projects. However, access to this credit line proved difficult and, in some cases, the delay in releasing the funds dragged for a long time a situation of precariousness that probably would hinder the payment of debts and increase dropout rates (ibid., p.174-5).

Regarding the profile of agricultural production, there was a scenario of improvement over the previous situation, which is not surprising considering that "in all states most of the properties acquired were abandoned or underutilized by the previous owners" (ibid, p.105). In general, the production pattern implemented remained focused on crop production (basically temporary crops), with very low level of technological development and carried out individually, contrary to initial expectations that the agricultural production agenda would change, the associative dimension would gain momentum and the technological level would take a qualitative leap.

Although the report indicated an increase in household incomes, it also pointed to the existence of considerable restrictions to the productive development of PCT projects, affecting most families:

If on the one hand the production data are consistent and indicate a variety of products and different production strategies, on the other they show the serious difficulties faced by the vast majority of beneficiaries, who once again suffer from a difficult combination of resource constraints - of access to public services and capital, and especially the lack of technical support for the implementation of projects. (ibid, p.138)

The report did not inform about the payment of the debt from the purchase of the land, although many projects had been in place for over six years when the filed survey was conducted. It merely presented projections on the evolution of the annual agricultural household income, all of them far behind the scenario of growing prosperity touted by the program managers, with the aggravating factor that the production practices reproduced monoculture (ibid, p.157-70). In the following years the federal government was forced to renegotiate the debts of Cédula da Terra borrowers and of the subsequent land credit programs, clearly indicating the economic weakness of these groups (Pereira & Sauer, 2011).

Conclusions

The results of the PCT did not corroborate the discourse on the alleged advantages of the MALR, since: a) even with cash payment and the fall in land prices in some regions, the prices paid for rural properties were not as low as initially announced, and the poor quality of the land acquired in a significant number of cases underscores this fact, so that we can say that the PCT effectively rewarded owners who used the program to sell abandoned, underutilized or bankrupt properties; b) purchases under the PCT strayed too far from the circuit of larger and better rural properties, showing the total inability of the project to change the rates of land ownership concentration; c) the implementation of the PCT was neither guided by any territorial planning designed to identify areas of concentration of properties not subject to expropriation for sale, nor was it linked to any broader rural development strategy, thus being reduced to merely a specific program, whose implementation was strongly influenced by the political situation; d) most of the projects led to subsistence production or monoculture, and not commercial and diversified agriculture; e) the subsidy proved insufficient to boost agricultural productivity and competitiveness, so that PCT borrowers remained strongly dependent upon a consistent rural credit policy; f) there was no access to private credit markets for investment in production; g) neither progressive taxation nor systematic titling was implemented alongside the PCT; there was no advance either in processes to modernize land registration systems or in measures to lower transaction costs; h) the most important components and phases of the PCT experienced a huge social participation deficit; i) borrowers had unequal power in negotiating with landowners in all cases, demonstrating that the MALR does not take into account the actually existing relations of economic exploitation, political domination and social clout contained in the land ownership monopoly of countries with a highly concentrated agrarian structure; j) it was the state agents who actually led the rural property purchase process, often involved in pork barrel practices and networks; l) there was no "self-selection" of borrowers because, in different ways, there was influence or interference from ‘external' forces (such as government agents, landowners, local politicians, NGOs, etc.); m) the decentralized implementation of the PCT was marked by lack of transparency and accountability, due to the absence of mechanisms of social control over its dynamics; n) the borrowers lacked denser organizational networks, capable of promoting their interests in view of the "artificiality" of most associations and to the lack of connection with social movements; o) judging by the projections presented, the hypothesis that most borrowers would be able to pay off the debt in the following years was difficult to materialize.

The experience analyzed, which is considered highly successful by the WB, questions the discourse about the viability of the MALR model as a reference for the design of land access and rural development programs in societies with high levels of land concentration. In all probability, specific technical improvements can possibly improve programs of this type - increasing, for example, participation and transparency mechanisms - but are unable to overcome the structural problems and limits of the model, such as dependency on land supply by owners and the inability to democratize the land structure and achieve social scale, in view of the cash payment at market prices. Designed and implemented as an alternative instrument to agrarian reform, the Cédula da Terra project did not prove able to provide an effective means to promote economic development and social justice in the rural area.

Notes

References

• BANCO MUNDIAL (BIRD). Brazil: a poverty assessment. Report nş 14323-BR, 2 volumes, 1995a.
• _______. Rural poverty alleviation project – Ceara. Report nº 14395-BR, 1995b.
• _______. Land reform and poverty alleviation pilot project. Report nş PIC 4974, 1997a.
• _______. Project appraisal document to Brazil for Land reform and poverty alleviation pilot project. Report nº 16342-BR, abril, 1997b.
• _______. Country assistance strategy – Brazil (1997-1999). Washington DC, 1997c.
• _______. Implementation completion report. Report nş 25973 Brazil Country Management Unit., 2003a.
• _______. Rural poverty alleviation in Brazil: toward an integrated strategy. Washington DC, 2003b.
• _______. Land policies for growth and poverty reduction Washington DC, 2003c.
• BANCO MUNDIAL (BIRD). The theory behind market-assisted land reform S. d. Disponível em: <www.worldbank.org/landpolicy>
• BINSWANGER, H. Rural development and poverty reduction. In: VAN ZYL, J. et al. (Ed.) Policies, markets and mechanisms for agricultural land reform in South Africa. Draft Manuscript, 1995a.
• _______. Five patterns of rural development: painful lessons. In: VAN ZYL, J, et al. (Ed.) Policies, markets and mechanisms for agricultural land reform in South Africa. Draft Manuscript, 1995b.
• _______. The political implications of alternative models of land reform and compensation. In: VAN ZYL, J. et al. (Ed.) Policies, markets and mechanisms for agricultural land reform in South Africa. Draft Manuscript, 1995c. (publicado em 1996 por Oxford University Press).
• BINSWANGER, H.; DEININGER, K. South African land policy: the legacy of history and current options. In: VAN ZYL, J. et al. (Ed.) Policies, markets and mechanisms for agricultural land reform in South Africa. Draft Manuscript, 1995.
• BINSWANGER, H.; VAN ZYL, J. Market assisted rural land reform: how will is work? In: VAN ZYL, J. et al. (Ed.) Policies, markets and mechanisms for agricultural land reform in South Africa. Draft Manuscript, 1995.
• BOBROW-STRAIN, A. (Dis)Accords: the politics of market-assisted land reforms in Chiapas, Mexico. World Development, v.32, n.6, p.887-903, 2004.
• BORRAS JUNIOR, S. M. Questioning market-led agrarian reform: experiences from Brazil, Colombia and South Africa. Journal of Agrarian Change, v.3, p.367-94, July 2003.
• BRANDÃO, W. A reforma agrária solidária no Ceará. In: LEITE, P. S. (Org.) Reforma agrária e desenvolvimento sustentável Brasília: MDA/CNDRS/Nead, Paralelo 15, 2000.
• BUAINAIN, A. M. et al. Relatório preliminar de avaliação do Projeto Cédula da Terra Brasília: Unicamp/Nead/MEPF, jun. 1999.
• _______. Estudo de avaliação de impactos do programa Cédula da Terra Relatório síntese. Brasília: Unicamp/Nead/MDA, nov. 2002.
• _______. Estudo de avaliação de impactos do programa Cédula da Terra Relatório final. Brasília: Unicamp/Nead/MDA, nov. 2003.
• BURKI, S. J.; PERRY, G. The long march Washington DC: World Bank, 1997.
• CARVALHO FILHO, J. J. de. Política agrária do governo FHC: desenvolvimento rural e a Nova Reforma Agrária. In: LEITE, S. (Org.) Políticas públicas e agricultura no Brasil Porto Alegre: EDUFRGS, 2001.
• DEININGER, K. Making negotiated land reform work: initial experience from Colombia, Brazil, and South Africa Washington D.C., 1998. Disponível em: <www.wordbank.org>
• _______. Land markets and land reform Texto apresentado na International Conference on Acess to Land: Innovative Agrarian Reforms for Sustainability and Poverty Reduction, Bonn, março 2001a.
• DEININGER, K. Política y administración de tierras: lecciones recogidas y nuevos desafíos para la agenda de desarrollo del Banco Mundial. Washington DC: World Bank, 2001b.
• DEININGER, K.; BINSWANGER, H. The evolution of the World Bank's land policy: principles, experience and future challenges. The World Bank Research Observer, v.14, n.2, ago. 1999.
• GROPPO, P. et al. Avaliação sintética do Projeto Cédula da Terra Fortaleza: Convênio Incra/FAO, maio 1998.
• MEDEIROS, L. S. de. Movimentos sociais, disputas políticas e reforma agrária de mercado no Brasil Rio de Janeiro: CPDA/UFRRJ e Unrisd, 2002.
• NAVARRO, Z. O projeto-piloto Cédula da Terra: comentários sobre as condições sociais e político-institucionais de seu desenvolvimento recente. Brasília, 1998.
• PEREIRA, J. M. M. The World Bank's "market-assisted land reform" as political issue: evidence from Brazil (1997-2006). European Review of Latin American and Caribbean Studies, v.82, p.21-49, April 2007.
• _______. O Banco Mundial como ator político, intelectual e financeiro (1944-2008) Rio de Janeiro: Civilização Brasileira, 2010a.
• _______. A política de reforma agrária de mercado do Banco Mundial: fundamentos, objetivos, contradições e perspectivas. São Paulo, Hucitec, 2010b.
• _______. Conflitos e parcerias em torno de projetos socioambientais. Tempo Social, v.23, n.2, p.177-206, 2011.
• PEREIRA, J. M. M.; SAUER, S. A "reforma agrária assistida pelo mercado" do Banco Mundial no Brasil: dimensões políticas, implantação e resultados. Sociedade e Estado, v.26, n.3, p.587-612, 2011.
• ROCHA, S. A pobreza no Brasil Afinal, de que se trata? Rio de Janeiro: Editora da FGV, 2003.
• STEIL, C. A. Cédula da Terra. In: SOARES, M. C. C. (Coord.) Estudo sobre fundos sociais e ambientais financiados ou administrados pelo Banco Mundial no Brasil Rio de Janeiro: Banco Mundial, mar. 2000.
• TEÓFILO, E. Brasil: nuevos paradigmas de la reforma agraria. In: TEJO, P. (Org.) Mercados de tierras agrícolas en América Latina y el Caribe: una realidad incompleta. Santiago do Chile: Nações Unidas/Cepal/GTZ, 2003. v.1.
• VAN ZYL, J. et al. Decentralized rural development and enhanced community participation: a case study from Northeast Brazil. World Bank, Policy Research Working Paper, n.1498, agosto 1995.
• VICTOR, A. D.; SAUER, S. (Coord.) Estudo sobre a política do Banco Mundial para o setor agrário brasileiro com base no caso do Projeto Cédula da Terra Brasília: Comissão Pastoral da Terra/MST/Rede Brasil sobre Instituições Financeiras Multilaterais/FIAN-Brasil, out. 2002.

1

This article resumes some reflections widely developed in Pereira (2010b).

2

In response to the pressure from environmental NGOs, the World Bank created the Inspection Panel in 1994 to provide an "independent" forum for social agents that feel directly or indirectly affected by the implementation of Bank-financed projects. The complaint should demonstrate the negative effects arising from non-compliance with WB rules and procedures in the preparation, implementation and evaluation of said projects. Rather than investigating independently a project by direct request of the affected parties, the Panel merely recommends to the WB Executive Board, based on a preliminary assessment, if the allegation should be investigated, but the decision is made by the Board. On the topic, see Pereira (2011 and 2010a).

3

In 1999 the government changed the conditions for funding the program, extending the amortization period to twenty years and establishing a fixed interest rate of 4% per year plus indexation by the General Price Index (IGP) of the Getulio Vargas Foundation. In February 2002, the interest rate was raised to 6% per year.

Publication Dates

• Publication in this collection
  25 July 2012
• Date of issue
  Aug 2012

History

• Received
  20 Apr 2010
• Accepted
  09 Sept 2010