Stephan Schwartzman
Environmental Defense
January, 27, 2000

In 1997 the World Bank and the Government of Brazil embarked on an experiment in "market-based" land reform, proposed as a new approach to improving the lot of one of the poorest populations in the world -- Brazil’s rural poor. The pilot project, designed behind closed doors, was presented to the broadly representative grassroots and non-governmental organizations that represent the landless and rural workers, and that have worked for land reform in some cases for decades, as a fait accompli. Not surprisingly, the project came under intense criticism. The World Bank Inspection Panel has now rejected the second request for an investigation of the Bank pilot project from Brazil’s National Forum on land reform, the umbrella group of these organizations. Is the criticism, as Bank management maintains, merely a case of ideological bias against market mechanisms, or based in national-level organizations’ unawareness of the successes the project has already achieved? Hasn’t the Inspection Panel, after two field visits attested to the project’s basic soundness?

The following memo shows that, contrary to Bank claims, grassroots opposition to its land reform projects is based in well-grounded concern for the perverse effects on land reform that these projects are likely to cause. Bank staff has, through avoiding transparency and bypassing key actors in land reform in the project’s design, from the outset involved itself on one side of a political struggle. The most comprehensive independent analysis based in field studies of the pilot to date, the Preliminary Evaluation contracted by the Bank and the government, clearly supports many of the grassroots groups’ central concerns, while showing that various of management’s optimistic claims lack factual basis.

The following memo traces the recent history of land reform efforts in Brazil and the role of the Landless Rural Workers Movement, discusses the Bank project in this context and compares Bank management’s claims for the project with the major empirical study to date of its implementation, the Preliminary Evaluation.

Land Reform in Brazil and the Landless Rural Workers’ Movement (MST)

Brazil has the ninth largest GNP in the world and among the most unequal distributions of wealth — and land — in the world. Brazil’s current president, Fernando Henrique Cardoso (FHC) perhaps put it best: "Brazil is not a poor country, it is an unjust one." Social movements, government, and the Bank in principle agree that the redistribution of land is an urgent necessity for addressing rural poverty.

Current law regulating land reform in Brazil (the Estatuto da Terra of 1964) was authored at the outset of Brazil’s military dictatorship, based in the understanding that massive concentration of landholdings not only contributed to rural poverty but froze potentially productive assets and hindered development. The law established that the federal government could expropriate unproductive or over-large landholdings (latifúndia) for the purpose of land reform, and indemnify the owners with government bonds. The political power of Brazil’s large landowners, however, effectively precluded significant action on land reform during the military dictatorship. When the military stepped down and were succeeded by an indirectly elected President in 1985, the civilian government announced an ambitious goal of settling 1.4 million families on 43 million hectares over four years. Only a fraction of this objective was attained — some 82,000 families were settled on 4.3 million hectares by 1989.

By the mid-1990s, however this picture would change dramatically. While World Bank descriptions of the Cédula da Terra belabor the need to make land reform quicker and cheaper than the Constitutionally mandated expropriation/indemnification mechanism,

in reality both expropriation of unproductive areas, and settlement of landless families accelerated dramatically after 1994. The National Institute for Agrarian Reform and Colonization (INCRA) under the current government prides itself on having provided land to 372,866 families on over 8 million hectares between 1995 and 1999, over half again as many as the 218,000 families given land in the entire preceding 30 years since the promulgation of the Land Statute. President Fernando Henrique Cardoso’s four-year goal announced in 1995 was met and excceeded. The principal cause of this leap forward in land reform, is the mobilization of the Landless Rural Workers’ Movement (Movimento dos Trabalhadores Rurais Sem Terra — MST) and allied organizations (the National Confederation of Agricultural Workers — CONTAG, and others.)

Starting in southern Brazil in the mid-1980s’ MST built a grassroots movement of the rural poor and landless, drawing on the desperation of those marginalized in the consolidation of large scale capital-intensive export agriculture and displaced by large scale infrastructure projects. The MST was able in various specific instances to contest the seemingly insurmountable political power of large landholders through the confrontational tactic of land occupation. Local groups would identify lands qualifying under the law for expropriation and dozens or hundreds of families would move in together, set up makeshift encampments and begin to farm. Growing numbers of MST adherents assembled in tent cities along roads and highways, planning new occupations. Confrontations with landowners, their hired guards, and the Military Police proliferated, as did groups identified with the MST, with varying degrees of relationship to the national coordinating body. In 1995 in Rondonia and 1996 in Pará, landless groups were massacred by military police sent to prevent or expel occupations. With a frankly leftist political orientation, the movement soon became identified as the most dynamic opposition to the FHC government, all the more clearly as MST groups repeatedly occupied INCRA offices to force negotiation or action on expropriations. MST marches and demonstrations regularly mobilized tens of thousands of people, even by the government’s count.

The federal government was faced with a serious dilemma. On the one hand, MST direct actions generated highly visible conflicts and pressure to act on the government, and the President’s, repeated commitments to social equity. On the other, the government’s major parliamentary coalition partner, key to the passage of its economic reform program in the Congress, the conservative PFL (Liberal Front Party), includes most of the representatives of the landed elite. That INCRA settled more families between 1995 and 1998 than in the preceding 30 years was politically costly, controversial within the ranks of the government’s parliamentary coalition, and could not have occurred without the continual, large-scale, public pressure applied by the MST strategy of land occupations. Unsurprisingly, while government has often negotiated with MST leadership, locally and nationally, it has as often denounced the movement as irresponsible, politically motivated, violent, given to illegal methods, and has brought a series of legal charges against leaders at various levels. The government and the MST are political adversaries.

Regardless of the view one takes of the MST or the strategy of land occupation as a means of achieving government action on land reform, it is indisputable that the movement has been the driving force behind the advances in land reform via expropriation that have taken place in the 1990s. Bank staff themselves have noted that it is owing to the mobilization of the MST that there is a land reform process occurring in Brazil for the bank to support.

The Cédula da Terra in the context of land reform in Brazil

When the Bank embarked upon its land reform venture, it intervened in an ongoing political conflict. Rather than taking a neutral posture that supported Land Reform as such, the Bank’s project supported the government and landowners against the Landless Rural Workers’ Movement and the civil society organizations aligned with it. First, the Bank planned its pilot project without consulting or informing any of the national representative organizations of the beneficiaries of land reform. While bank staff may claim that MST and other national level groups are ideologically opposed to "market-based" land reform, the fact is that no effort was made to engage these groups in dialogue at the project’s inception. Second, the Bank and government refused to negotiate when rural workers’ organizations made proposals (eg, the National Confederation of Agricultural Workers’ (CONTAG) 1997 proposals to use Cédula funds for credit in areas of high concentrations of landholdings too small to support families.) Third, having bypassed the representative organizations of rural workers and the landless, the Bank designed a program, ostensibly driven by freely organized local associations of rural poor, which in fact funded and empowered state government agencies to organize or enlist associations into the program. The state government agencies are typically much more susceptible to influence and manipulation by the landed interests most intransigent in their opposition to land reform than federal agencies. The Preliminary Evaluation of the Cédula da Terra Report , in fact found that the Cédula associations typically define themselves in opposition to the MST. The evaluators state, " . . . the Cédula da Terra Program introduces a political and ideological dispute with other social movements and their mediators (principally the MST, sectors of the Catholic Church, and civil society organized in NGOs), which today have the political initiative in this area and support access to land through expropriation."

In principle, nothing prevented the Bank from supporting the land reform process as such instead of taking sides in the political battle. The Bank could have pursued alternatives raised by CONTAG in 1997; it could have supported the Constitutionally mandated land reform process in addition to a "market-based" pilot; or it could have linked the pilot to specific targets for expropriation and settlement. One difficulty here appears to be that, while management maintains that the project was a pilot for an alternative, complementary land reform mechanism, some sectors of government defend the approach as a substitute for existing land reform. Thus, the Bank has sided with the political agenda of particular government agencies and their supporters that ultimately seek the replacement of existing means of land reform with "market-based" mechanisms, and against the social movements and sectors of the government that seek land reform via the Constitutionally sanctioned means of expropriation of unproductive land.

"Market-Based" Land Reform — What the data say

It is of considerable importance to sort through the claims management has made for the likely prospects of this approach. If the Cédula approach can be shown to better the situation of the rural poor more cheaply and quickly than traditional land reform, based on participating families ability to buy land in the free market and pay off their loans, as the Bank claims, there might be little reason for concern. The only independent data on progress to date and potential for success or failure of the pilot is the Preliminary Evaluation Report, commissioned by Ministry for Agrarian Reform as one of the conditions of the project, and elaborated by a team from the University of Campinas (UNICAMP) and the University of São Paulo (USP). This study is referenced in the Management Response to the first inspection panel claim. The research team surveyed 116 of the 223 sub-projects in the five states where the pilot is underway, that is, the sample is over 50% of the projects. The data and conclusions of this report are critical to understanding what the project has done to date, and what its future results are likely to be.

The Preliminary Evaluation Report contradicts or casts doubt on various central claims Bank management has made for the project. The following are some of management’s claims, contrasted with the factual evidence of the Evaluation Report.

Is the Cédula da Terra a success?

"New Approach to Land Reform Proves Successful" is the characterization of the Cédula Project that appears on the Bank’s web page. The description goes on to say that "the three year project achieved its objectives one year ahead of schedule. A follow-up $1 billion adaptable program loan is being prepared for this fiscal year to capitalize on the current project’s success and maintain the momentum that this kind of land reform has generated." This public relations document reproduces arguments that appear in the Management Response to the first inspection panel request: that implementation is running ahead of schedule and that based on "results of the pilot", government wanted to rapidly expand the program; that the economic and financial viability of the program has been verified, "based on the experiences of ongoing subprojects", that the project "places beneficiaries in the drivers’ seat and its success depends entirely on their active participation at, in all stages of the Project cycle."

But the evaluators state clearly and repeatedly that it is too early to determine what the actual results of the program are. "It should be underlined that the Cédula da Terra program is still too recent to permit an evaluation of its socio-economic impacts, either on the beneficiaries or on the communities." Or again, " This study . . . should not be confused either with an evaluation of [the Program’s] socio-economic impacts or with a study of its viability or sustainability. Or, once again, " The questionnaire applied had as its objective to trace a socio-economic and technological profile of the beneficiaries, as well as showing the productive use of their property in its first year of occupation. For the reasons noted above, the second objective was considerably prejudiced, since there is practically no production on the properties in the project." For these among other reasons, the evaluators state, "It is an open question whether these beneficiaries . . . will be able to take charge of the management of their properties and generate sufficient income to meet the commitments they have undertaken and improve their living standards."

Even were information available on the productivity of the land acquired in the project and the management capacity of the beneficiaries, it would still be extremely precipitous to call the program a "success", since there is a three-year grace period for project loans, and the first beneficiaries will only have to begin repaying their loans in 2001.

Does having "met its objectives early" mean that the project has succeeded?

That the program has "met its objectives one year early" means that project funds were spent. This is unsurprising and no indication of its lasting effect on land reform or poverty alleviation.

It would be surprising if the project funds were not spent, given that 1997 and 1998, the years the program began were severe drought years in most of the project area, which reduced the already poor to destitution and desperation, restricted employment opportunities, and made any chance for access to land and money an exceptionally attractive option. In the evaluators’ view the project, " . . . took on a role similar to emergency drought-relief programs . . . in a context marked by a succession of bad and very bad agricultural years access to land is seen, more than anything, as a means of survival . . . As is well-known, one of the principal effects of drought in the semi-arid Northeast is to drastically reduce job opportunities . . . Part of . . . the traditional clientele for emergency relief programs sought support in the Cédula, regardless of considerations of personal aptitude for participating in a program of this kind of economic calculation of the viability of the projects. " It is consequently perfectly understandable that as Bank staff constantly emphasize, there is great local demand for project funds, and a surfeit of families that want to participate in the project.

In what sense is the Cédula a "participatory" project?

The context of the drought and the attractiveness of the Cédula program as a survival strategy are important to understanding what are perhaps management’s most profoundly misleading claims for the project. "The [Cédula] program represents the ultimate participatory process, says its manager, Luis Coirolo, because it revolves completely around its beneficiaries: They choose the land and negotiate its purchase. They decide how to use the land, what investments are required to make it productive and what technical assistance they will need." The idea that associations of the rural poor and landless negotiate directly with landowners, select the lands that best suit their needs, and purchase them at their market value, is indeed at the heart of the program.

The Evaluation Report however documents that large part of the beneficiaries lack even the most basic information on the terms of the program in which they are involved: "Another extremely revealing point is the almost complete lack of knowledge of the terms of functioning of the Cédula da Terra program itself. When asked if they had taken out a loan in the last year, the spontaneous response of the great majority was "no". Reminded of the credit to buy land, and asked about the source, value, interest rate and type of guarantee given, practically no one knew how to answer." Moreover, even when reminded of their participation in the program, nearly a third (30%) of the beneficiaries are unaware that they have taken out loans; fewer than 1% could correctly state the interest rate on their loans; and fewer than 10% recognize that they have guaranteed their loans with their land.

It is difficult to see under these circumstances how anyone could consider this a "participatory" process.

Are the beneficiaries "in the driver’s seat"?

Generalized lack of basic information is only one of the factual findings of the evaluation that contradicts management claims. It is also not the case that the beneficiaries "choose the land and negotiate its purchase." "The great majority of the associations [of the beneficiaries], even those that declared that they had negotiated directly with the landowner, played a secondary role in the process, limited to receiving an offer, taking it to the government agency responsible for the program, taking back the counteroffer, and so on. . . . what occurs, in the majority of cases, is a negotiation between government agencies and landowners. . . . As far as the latter are concerned, the potential buyer is the Government and whether the associations are able or unable to pay is of no importance. The landowners know that the weight of the government agencies is fundamental in the negotiations and that the staff of the agencies responsible for the program elaborate inspection reports (o laudo técnico), establish the value of the land, veto sales because of irregularities, and often "convince" the parties. It is the government, not the associations, that closes the deals."

The project further supposes that the program will attract the entrepreneurial, whose informed negotiations will net them land at price that reflects its productivity. But the study strongly suggests that the beneficiaries regard securing land as the primary goal, do not consider price important (as long as it is below the maximum permitted by the program), and frequently take the first property identified. The principal selection criterion considered by 80% of the interviewees was "quality", as opposed to 13.5%, who considered price the primary selection criterion. The majority of the associations (52%) acquired the first and only property identified without seeking another. "The beneficiaries do not have paying a good price as a central concern, but rather assuring access to land "at any price." For them the negotiation is not about price or conditions of payment, but about access to land. In various cases, the association pressured the government to accept the landowners’ offer. Everything indicates that these . . . attitudes . . . reflect a mixture of anxiety to get land, and fear of losing the opportunity. It is necessary to recognize that this attitude compromises the capacity of many associations to carry on good negotiations to acquire property."

Are the beneficiaries’ Associations independent?

The independent character of the associations, a key premise of the project, is often dubious. The evaluation report finds that, "the majority of the associations originated from the populist tradition, in which representation is based on co-optation, subordination, and social and political control over poor populations. Even many of the associations that are being created exclusively to participate in the Program appear not to escape this context. It is however necessary to call attention to the fact that programs such as Cédula da Terra presuppose autonomous associations, with the capacity to take and implement strategic decisions on the use of assets under their control . . ."

State government agencies, local politicians, landlords and other members of the elites, in fact have much more of a role in the creation of the Cédula da Terra associations than Bank accounts would lead one to believe: "The official proposal characterizes joining the Cédula da Terra program as an initiative taken by the beneficiaries themselves. However, the interviews show that the creation of the associations is not as "natural" as one might hope. In all the states . . . there is clear intervention of actors and institutions that are external to the group, such as municipal governments, local politicians, "well intentioned" individuals, landowners, staff of government agencies, etc. . . .In general these agencies are not merely supporting actors in the creation of community associations."

The evaluation goes on to note that " in spite of the application and interest of the beneficiaries, the associations are not establishing themselves as independent actors, with real capacity for dialogue with Government, relative to local political elites and the landowners. Their members perceive the complexity of the tasks for which they are responsible but are not succeeding in carrying them out."

Are land purchases negotiated in the open market?

The very notion of free negotiation in the marketplace among landless buyers and landed sellers on which the project’s claim to represent an alternative approach to land reform rests is shown by the Evaluation Report to be seriously flawed in several important respects.

Land prices are not set in an open market. "The low liquidity of the land market, the small number of transactions and the still highly concentrated structure of landholdings favor … a process of price formation that is not transparent and is strongly influenced by extra-market factors." The evaluators conclude their review of process of negotiating land acquisition by stating, "(i) local land markets are incipient and do not define reliable parameters for the transactions concluded; market prices that could serve as reference points for the value of the land locally and for transactions carried out through the Cédula, practically do not exist. The conclusion is obvious: in large measure the price paid depends on the bargaining power of the beneficiaries — which is small — and the negotiating capacity of the parties; ( ii) the majority of the associations do not appear prepared to deal with the negotiation process, such that intermediation (until now exercised by the state government agencies) is fundamental."

In short, the Cédula da Terra is not in practice predicated on direct negotiation between buyers and sellers in the open market, which does not in any case exist in large parts of the project area, and the beneficiaries’ associations are not the primary actors in the negotiations over land acquisition. Many of them are unaware that they have taken our loans at all, and almost none know the actual terms of the financing for which they are now responsible. This in reality is not a "market-based" land reform project at all, but a land reform project that devolves responsibility for land reform from the federal government to state governments — precisely those more susceptible to pressure and manipulation of local and regional elites. It is then hardly surprising that, as the report goes on to detail, the associations define themselves in opposition to the MST. Whatever else it may prove to be, Cédula da Terra is an effective source of support for local and regional interests ideologically opposed to the MST and organizations aligned with it, to counter MST organizing and potentially undermine its membership base at a local level.

Conclusion

It is puzzling, in light of the foregoing data, why the Bank would want to commit itself to a $1 billion program based in this model at this point if its goal is to construct a viable alternative approach to land reform in Brazil. Given the very serious issues raised by the official evaluation, prudence would dictate at very least that the Bank and Government wait until 1) it were possible to evaluate agricultural production on the land acquired in the pilot project over a minimally representative time period, rather than relying on projected simulations; 2) that the Bank judge the actual effects of some part of the program on beneficiaries’ incomes and living standards and discuss them with governments and other actors in land reform in Brazil before deciding what course to follow. Actual effects include inter alia the beneficiaries’ ability to repay the debts they have incurred, over a minimally representative time period, as well as the projects’ effects on poverty. The Bank has produced impressive farm models, which project excellent results in virtually all scenarios. But these are models, based, as are all models, on the assumptions and presuppositions built into them. Since loan repayment only begins for the first participants in 2001, there are no actual data about the results of the program as a "market-based" land reform and poverty alleviation program. Yet management argues that this is a model pilot program. The Bank has worked in Brazil for nearly 50 years, and only supported land reform beginning in 1997. Why this sudden haste to process a $1 billion dollar, ten-year program, before any empirical results of the supposed pilot are available for review?

The Bank’s approach here, intentionally or not, clearly fits well with the political agenda of the sectors of Brazil’s federal and sate governments that oppose the MST, CONTAG and the other representative organizations most active on land reform, as well as opponents of land reform more generally. Large scale funding would be provided through state government agencies to create or enlist rural workers associations in land purchase schemes that are subject to political manipulation by local elites at the moment of creation of the associations, selection of beneficiaries, and in the negotiation and purchase of land. In the process, part of the MST membership base can potentially be co-opted. While a $150 million pilot program may be relatively limited in its impact in a country as large as Brazil, a $2 billion, ten-year program is another matter. For the sectors of Brazilian government that seek to halt pressure for land reform, this program is an excellent tool.

The point here is not that land reform via expropriation is better than market-based alternatives, or vice-versa. As Bank staff have stated, that there is a process of land reform in Brazil for the Bank to support at all is owing to the mobilization of the grassroots groups — MST in particular. In supporting one side in the dispute, the Bank risks undermining the very process its project alleges to be supporting. Regardless of whether one views expropriation or purchase as better, worse, or complementary means of achieving redistribution of land, it is obvious who the actors in the political dispute that is land reform in Brazil are. It is also evident that in bypassing the major representative organizations behind land reform in the design of the Cédula, as the Bank clearly did, the Bank opted to participate in, as the independent evaluation put it, "a political and ideological dispute with other social movements and their mediators (principally the MST, sectors of the Catholic Church, and civil society organized in NGOs), which today have the political initiative in this area and support access to land through expropriation. The Bank cannot in this case plausibly accuse the MST or any of the other organizations that requested an inspection of the program of being any more "ideologically" or "philosophically" motivated than it has shown itself to be. As noted above, the Bank had various other options before it when it embarked on its venture in land reform in Brasil that would have allowed it a more neutral role, in support of land reform as a process, rather than picking sides in the fight. That it has incurred criticism for doing so should surprise no one.

Recommendations

• Suspend preparation and negotiations of the Adaptable Program Loan (APL) intended to replicate the Cédula da Terra nationally;
• Release relevant data on the Cédula program and the APL, including the Preliminary Evaluation Report, project supervision reports, appraisal reports, loan agreements and loan contract;
• Negotiate terms for the discussion of a participatory evaluation process for the Cédula project, involving the major representative organizations of the beneficiaries of land reform, the government, and recognized independent researchers, to begin in 2001.
• Commit to the principles that further support for "market-based" land reform will only follow adequate, open and participatory evaluation of the pilot program, the timing and terms of which are to be decided jointly by Government, NGOs and the Bank, which empirically shows positive results including better cost efficiency than traditional land reform, and that;
• Land Reform projects of any description will be conditioned on specific, quantitative and time-bound goals for expropriation of unproductive landholdings according to Brazil’s Constitution, and resettlement of the landless and rural poor including the technical assistance and credit required for them to farm viably.

Stephan Schwartzman
Environmental Defense
1875 Connecticut Ave. NW
Suite 1016
Washington DC 20009
202-387-3500
202-234-6049 - fax
steves@edf.org