Agrarian Reform and Workers Unions in Pernambuco Sugar Zone

7/22/2019 Agrarian Reform and Workers Unions in Pernambuco Sugar Zone

http://slidepdf.com/reader/full/agrarian-reform-and-workers-unions-in-pernambuco-sugar-zone 1/25

Agrarian Reform and the Rural Workers' Unions of the Pernambuco Sugar Zone, Brazil 1985-1988Author(s): Anthony W. PereiraSource: The Journal of Developing Areas, Vol. 26, No. 2 (Jan., 1992), pp. 169-192Published by: College of Business, Tennessee State University

Stable URL: http://www.jstor.org/stable/4192077 .

Accessed: 14/10/2013 02:40

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

College of Business, Tennessee State University is collaborating with JSTOR to digitize, preserve and extend

access to The Journal of Developing Areas.

http://www.jstor.org

This content downloaded from 202.92.130.224 on Mon, 14 Oct 2013 02:40:12 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=cbtsu

http://www.jstor.org/stable/4192077?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp





http://www.jstor.org/stable/4192077?origin=JSTOR-pdf

http://www.jstor.org/action/showPublisher?publisherCode=cbtsu



The Journal of Developing Areas 26 (January 1992) 169-192

Agrarian Reform and the Rural

Workers' Unions of the Pernambuco

Sugar Zone, Brazil 1985-1988

ANTHONY W. PEREIRA

Agrarian reform can be an important catalyst for economic develop-ment. The transfer of landownership or control to those who actually

work the land can reduce inequitable distributions of income and wealthin the countryside, increase domestic demand and savings, and increase

agricultural output.' It is therefore significant that the federal govern-ment in Brazil, the largest and most populous country in Latin America,

announced plans for a major agrarian reform in the spring of 1985.That reform has proved to be a failure, with very little of the land

planned for redistribution actually changing hands. Various explanationsfor this failure have been put forward. The political power of the landedelite (especially its Sao Paulo wing) and opposition from within the ranks

of the military high command have been cited.2 Other arguments men-tion the lack of resources and inefficiency of the land redistribution

bureaucracy and the absence of a strong urban base of support for thereform.3 Without discounting these powerful explanations, this article will

explore a relatively neglected cause of the reform's demise. This is the

skepticism and nonparticipation of the major representatives of the reform'sbeneficiaries: rural trade union leaders.

The data presented here are derived in part from a survey I conducted

in the sugar zone of the northeastern state of Pernambuco in 1988. In

this survey, conducted in a region in which the majority of rural laborersare wage workers with no access to land, I interviewed 50 trade unionleaders in 42 different unions.4 While the Pernambuco sugar zone is nottypical of Brazilian agrarian conditions (no region is), it offers a good case

Assistant Professor, Political Science Department, New School for Social Research, New

York, NY.

The author wishes to thank Marisol Oliveira de Sousa and Jose Lauryston for help in datacollection; the Inter-American Foundation, Organization of American States, and Har-

vard's Center for International Affairs for financial support; and Jorge Dominguez, Frances

Hagopian, Robert Fishman, Cliff Welch, William Nylen, John French, James Ho-Adler,and the anonymous reviewers of the JDA for comments on an earlier draft of this article.

? 1992 by Western Illinois University.







170 Anthony W. Pereira

study because it is a region with a relatively active and powerful ruraltrade union movement and a progressive governor, Miguel Arraes, who

was elected in 1986. If one expected rural union support for agrarianreform anywhere in Brazil, therefore, one would have expected it in thePernambuco sugar zone. This expectation is further enhanced when oneconsiders that land reform in this region has historically referred to accessto land on the part of wage laborers, rather than the transformation oflaborers into small farmers. Union leaders would therefore have had noreason to fear that agrarianreform would lead to the demise of the unionsthemselves by turning their members into independent smallholders.

The argument presented here is that rural trade union leaders did notpush strongly for the agrarian reform for three reasons. First, manyunionists simply did not believe that the federal government was com-mitted to large-scale land redistribution in 1985 and acted accordingly.Second, the regime change of 1985 did not profoundly alter local powerrelations in most rural areas; thus trade unionists saw little probabilityof defeating landlords in their area even if they had believed in the fed-eral plan. Finally, trade unionists were suspicious of other members ofthe proreform alliance, most notably church and left-party activists, whomthey saw as competitors in the struggle to mobilize the rural poor. Forthese reasons, rural union leaders tended to adhere to a low-risk, incre-mental strategy of representing members within the established laborbureaucracy rather than confronting landlords in an effort to force the

redistribution of land. Such behavior was common not only in thePernambuco sugar zone, but in the rest of rural Brazil.In order to understand the rural unionists' behavior on a deeper level,

one must examine both the relationship that exists between the unionsand the state, and the way that agrarian reform was enacted. This articledoes this in four parts. The first part describes the relationships of ruralpoverty, the unions, and the state with each other. The second notesthe failure of the central government to protect the rural poor (especiallyin the areas of labor law and landowner violence), while the third pointsout that a consequence of this nonaction is a widespread alienation fromthe state on the part of rural union leaders. Finally, the fourth section

examines the fate of the agrarian reform plan both nationally and in thePernambuco sugar zone.

Rural Poverty, Trade Unions, and the State

About 30 million people, or 73 percent of Brazil's total rural popula-tion, are considered to be poor.5 The majority of them are wage workersof one kind or another, the most numerous being the category of tem-porary worker (see table 1). Of this population, about 10 million wageworkers, tenants, sharecroppers, squatters, and small landowners aremembers of rural trade unions.6 Thus, when the newly established civil-ian president, Jose Sarney, after 21 years of military rule, announced in

1985 that millions of landless workers would be given land rights overthe next 15 years, union leaders were vitally interested. Most of theintended beneficiaries of what promised to be a massive land reform weremembers of their unions, and agrarian reform had long been a principaldemand of the rural labor movement. The slogan of the union umbrella







Agrarian Reform and the Rural Workers' Unions: Brazil 171

/ \N.S. MARAVILHASxIB

/ Maapatamam CRAG ALAC odd T

/ TERESA/ D' R, '

MATARYA /Iia.

Pdr~~~~~~~~~~~~~~~~~a

A S /, Hr;, < yll_araca~ianc

LI MOEI RO }ETRU / Mana Farinha

CoDSTILLERY / KG'S PER HARVEST G t

da Moa lda

*BASES UPON 19B71B8 HARVEST DATA AS OF 2115/BB / 1ULHOES tt'SUPPLIED BY THE iNSTITUTE OF SUGAR AND ALCOHOL.~- Pombos . ISTjI,r<9

AQUA, B.S.NCA R M O \ , e) AireS,,BeAadtOE

4~~~~~~~~~~ 5 e o 5 20 30 d 40es

< g qool~~~~~~~~~~~~~~~~~~~~~~~~~rnao

=~ " zI' ALAOARS KILIMTe

~~~~~~~~~~~~~~~~~~~~~~a*/_Rbrinhao

Fig. 1. The Sugar AonE of PernaMUco.

SOURC:NMa byLnna.Per;usdwthprisin

(CONTAG), had beenMN democracyithouStEagrrianreforhm" siNcethe

The rural poor struggle to survive in an agrarian system d n db

l' RACtifenedito\ (lrg etae) Brai' attrnfln ocnrato isroneo

MILLF7o DITLLR

do GoH

OVER ~ ig 1.ILL VSugrar doe. Pernambuco.M

(CONTAG), had been "No dmocracy withoutagraria JABorm"sicethbAeSinnPON h9870s.RETAA F /58 U

latifundios (largeestates).Gravaazi' BOMerflndcneSUainisoeo







172 AnthonyW. Pereira

TABLE 1

WORKERSWITHOUTLAND OR WITH LIITLE LAND, BRAZIL

1978 AND 1984NUMBER

(In Thousands)

CATEGORY 1978 1984

MinifundistasOwners 1,469 1,872Occupiers 505 644

Sharecroppers 273 433Renters 122 180Permanent wage workers 1,104 2,147Temporary wage workers 2,560 4,260Other nonsalaried workers 713 1,104

Total 6,746 10,640SOURCE:Ministerio da Reforma e do Desenvolvimento

Agrario (MIRAD), Proposta para a Elaboracao do 10 PlanoNacional de Reforma Agraria da Nova Republica (Brasi-lia: MIRAD, May 1985), p. 13.

the most unequal in the world. Holdings exceeding 1,000 hectares (ha)or more account for only 1 percent of all rural properties but cover 45percent of the total cultivated area. Those farms of less than 100 ha (min-ifundios) equal 90 percent of all holdings but have only 20 percent ofthe total area.7 Many small cultivators, an estimated 39 percent of all

farmers, farm the land under some form of precarious tenure.The structure of Brazilian agriculture is dualistic.8 On one hand, large

estates produce export crops requiring heavy capital investment and/orhigh levels of inputs (soybeans, coffee, sugar, tobacco, cocoa, citrus). Onthe other hand, minifundios (small farms) with under 100 ha produce thebulk of the domestically consumed food crops such as beans, rice, man-ioc, potatoes, and corn. Within this structure there is little room formedium-sized, capital-intensive family farms of the type that, until recently,played such a vital role in U.S. food production. The empresa rural, theBrazilian equivalent of the family farm, accounts for only 3.3 percent ofall properties and 5.0 percent of the total cultivated area in Brazil.9 Rural

Brazil, therefore, is polarized between an elite of large landowners anda great mass of small landowners, tenants, and the landless.

Despite the existence of acute poverty in rural areas, land in Brazil isnot scarce. An estimated 160 million ha of fertile land, or four times thetotal area occupied by the minifundios, is unused.'0 Much of this is heldpurely for speculative purposes.

Important transformations in the Brazilian countryside began to takeplace in the 1950s and early 1960s. As the commercialization of agri-culture proceeded and the value of land increased, landlords ousted ten-ants who had previously enjoyed land-use rights on the latifundios. Mostof these tenants migrated to towns and became wage laborers in agri-culture or the urban informal sector. This breakdown in patron-client tiesled to new forms of organization among workers, and rural trade unionswere founded and recognized by the government. There were about 500recognized rural trade unions at the time of the 1964 military coup."The Rural Worker Statute had also been passed in 1963; this statute








simply reemphasized the 1943 Consolidated Labor Laws and stipulated

that rural workers should enjoy all the legal rights of urban workers.

The military government that came to power in 1964 in its first monthsof office passed a Land Statute with the ostensible goal of eliminatingthe latifundio-minifundio complex and creating a system of efficient fam-ily farms. In fact, the policies of the military regime (1964-85) chieflybenefited the latifundiarios, who received a disproportionate share of thepublic goods provided to agriculture by the state, and from whom taxes

were often not collected.'2 Land concentration and rural poverty bothincreased during this time.13 The government became highly involved in

all stages of production on the large estates, from research on seeds andplanting techniques to the management and subsidy of exports. Through

a network of state-owned banks, large estateswere

grantedcheap credit;

rural credit increased fivefold in real terms in the years from 1968 to1978.14 With credit, the large estate owners bought machinery, pesti-

cides, fertilizers, and other inputs. At the same time, the state built uprural infrastructure including roads, electricity grids, dams, and irrigationnetworks. This policy of capitalizing large estates meant that the latteroften displaced tenants, squatters, and smallholders.15 Toward the orga-nizations representing these latter groups, rural unions, the military

adopted a policy of selective rewards and punishments.Immediately after the 1964 coup, activists, especially Communists, were

expelled from the unions. In Pernambuco, scores of unions had their

leaders replaced by workers handpicked by the Ministry of Labor. Anunknown number of rural workers were killed by landlords or the army;others were jailed and/or tortured. The Peasant Leagues, civil associa-

tions that had been active throughout the Northeast, were abolished.

Competition for the right to represent rural workers was prohibited and

replaced by a unitary system in which the Ministry of Labor grantedexclusive rights of representation to unions in given areas. The unions

were rigorously controlled and permitted to concern themselves only with

social welfare and wage issues. Collective bargaining was not allowed;

wage settlements were issued unilaterally by labor courts. Strikes were,in effect, prohibited.

In the late 1960s, the military government expanded FUNRURAL, asystem in which unions provided medical and dental services to mem-

bers with funds provided by the government. This encouraged the foun-

dation of unions, and rural unionization accelerated rapidly in the 1970s.

In 1971, PRORURAL, a rural social security system, was set up, pro-viding further impetus to the formation of unions. By 1987, CONTAGwas the largest confederation in the Brazilian labor relations system. Rural

unions under the military regime, however, functioned more as welfare

agencies than as conduits for collective worker demands in the areas of

land and working conditions.

The Central Government's Neglect of the Rural Poor

Under the military regime, the state mobilized an impressive array of

resources in modernizing the export-oriented large estates. The state's

welfare programs for the rural poor, however, were modest. Rather than

supply medical services itself, the central government instructed unions








to do all of the provisioning themselves. In the poverty-stricken North-east, the burdens imposed by this injunction were severe. Government

funding for the program was often late and inadequate. Similarly, ruralsocial security reached a small number of people because relatively few

rural workers survived until 65. The pension, 50 percent of the mini-

mum wage for workers, 30 percent of the minimum wage for survivingspouses, was probably below subsistence level.

Nevertheless, the military regime tried to appease, if only symboli-cally, the land hunger that the commercialization of agriculture was cre-

ating. In 1966, the first national agrarian reform plan, based on the 1964

Land Statute, was put in place. It amounted to a few, scattered colo-

nization projects.'6 In 1968, a second plan was promulgated; it focused

primarily on colonization in the Amazon as a substitute for land redis-tribution in settled areas such as the Northeast. By 1985, the National

Institute for Colonization and Agrarian Reform (INCRA) had given titles

to 83,000 families in the Amazon.'7 This was a small number comparedto the number of migrants settling in the Amazon without titles, and tothe total number of landless in Brazil.

These gestures toward land reform underscore the fact that the Braziliancentral government has delivered little to rural unionists. The army andthe Ministry of Labor combined to repress workers' organizations at the

time of the 1964 coup. At almost all other times, the central governmenthas been conspicuous in the countryside by its absence and inaction.

Rural welfare measures were enacted much later than in urban areas andon a diminished scale. Land reforms were promised by INCRA but notenacted. And a legal order was not enforced.

For example, a system of labor law has still not been consistentlyimposed in the Brazilian countryside. Although Brazil's first labor lawconcerned rural labor, it was not generally implemented.'8 Later, theConsolidated Labor Laws (1943) were not effectively applied to ruralworkers. This is partly because most rural laborers at that time were notfull proletarians, but had some kind of dependent relationship with alandowner, exchanging labor or part of a crop, for example, for land-use

rights. In addition, rural labor was not well-organized enough to demandthe legal protection of its rights; vertical patron-client ties were strongerthan horizontal class ties. Even in 1963, with the passage of the RuralWorker Statute, enforcement of labor law was patchy and exceptional.

Today, the most significant violation of labor legislation concerns theillegal hiring of seasonal employees, called boias-frias (literally, "cold-lunches") in the south, and clandestinos in the Northeast and north. Thisis usually done through a labor contractor (empreiteiro), who is in a posi-tion to exploit workers. Daily wages paid to illegal workers are not nec-essarily lower than those paid to legal workers, though they often are.By hiring "off the books," however, the employer saves taxes and makes

none of the extra payments required by law for social security, holidaypay, and worker union dues, for example. By the same token, whileillegal workers have great flexibility in employment (although their direpoverty often negates the freedom of choice inherent in this arrange-ment), they enjoy none of the benefits and security of workers who have








a carteira assinada (worker's card signed by both the worker and thefederal Ministry of Labor).

In my survey of 50 trade union leaders in the Pernambuco sugar zone,44 said that illegal workers are "a problem for the union." The Pernambucorural workers' federation, FETAPE (Federacao dos Trabalhadores naAgricultura do Estado de Pernambuco-Federation of Agricultural Workersof Pernambuco State), claims that more than 50 percent of the workershired to cut cane during the harvest are illegals, many coming from theinterior of the state. Although illegality is a problem, only 10 of theunion leaders surveyed reported, in an open-ended question followingup on the preceding one, that they had contacted the local Ministry ofLabor office or another government office about it. Two of these saidthat the labor ministry had been effective in reducing illegal hiring, whiletwo said that ministry officials had repeatedly failed to appear as prom-ised in their area. What these responses show is the low expectation onthe part of union leaders that the central government will enforce laborlaw, an expectation reinforced by the Ministry of Labor's poor perfor-mance in monitoring reported violations. The majority of leaders sur-veyed, 33 of 50, answered that they were trying to solve the illegalityproblem solely with their own organizing efforts by talking with workers,making agreements with local owners, coordinating with nearby unions,and so on.

In the Pernambuco sugar zone, unions struggle most for the enforce-

ment not of labor law in general, but of a statewide collective contract(dissidio coletivo). The collective contract is signed each year by repre-sentatives of employers and workers in the sugar sector and is supposedto be enforced by the local office of the Ministry of Labor. This contractdetails the piece rates at which cane cutters should be paid. It also con-tains numerous provisions relating to working conditions. Of the 50 lead-ers interviewed, 47 cited violations of the collective contract as one ofthe most common complaints that their members bring to the union

(especially in the area of piece rates). Furthermore, in an open-endedquestion about the union's principal problems, 48 percent of respondentscited the owners' lack of respect for the collective contract. Similarly,

only 6 percent of the officials said that all the provisions of the collectivecontract were being enforced at the time of the interview.

Another problem for the union officials concerns the labor courts, where

disputes between employers and employees over such matters as sev-erance pay are adjudicated. Many leaders complained that the courtswere too slow, taking up to five or six years to resolve a case. Othersquestioned the impartiality of the labor court judges and pointed out thatthe latter were in the same social class and often had frequent socialcontacts with landlords. This is what was implied by one leader whenhe said of the judges, "The river always runs to the sea."

While 82 percent of the respondents felt that Brazilian labor law shouldbe changed (the majority of these saying that it operated against theworkers), it seems that the simple enforcement of existing labor law bythe Ministry of Labor would create a revolution in social relations in the

sugar zone.Landlord violence against rural labor is another area in which the








Brazilian central government has failed to enforce a legal order. Violenceagainst rural workers and peasants has increased in recent years (see table

2). Reports have documented this violence, which is typically carried outby landlords' hired gunmen against workers with whom the landlord hasa labor or land dispute.'9 One of these reports, published by AmnestyInternational, stated that despite "notorious, widespread and persistent"killings, "the Brazilian Government has failed at every level to take effec-tive measures to stop them." The human rights group was concernedabout "allegations that full and impartial investigations by the competentauthorities into such killings are not being carried out" and that, con-sequently, "those responsible for killings, threats, and other brutal actsof intimidation against the rural poor enjoy de facto immunity from pros-ecution, which encourages further killings and violence."20

Responsibility for this state of affairs rests with the three main lawenforcement agencies in Brazil: the civil, military, and federal police.The response of the civil police to killings of rural workers, many ofthem allegedly carried out on the orders of local landowners, has beeninadequate. Members of both the military and civil police have beenimplicated in acts of violence against rural workers. The federal police,relatively free of local political pressures, represent a force capable ofproperly investigating these acts. The federal government, however, hasnot ordered its federal police to do this. Government officials explainthat this is against the Constitution and that "the restoration of democ-

racy . . . [can] best be achieved by trying to strengthen localinstitutions. "21

The judiciary has also been responsible for the escalation of violence.It is dependent, both financially and administratively, on state govern-ment, and judges are susceptible to political pressure. A common expla-nation for the failure of judicial authorities to investigate assassinationsof rural workers, understandable owing to the current fiscal crisis in Brazil,is lack of resources and staff.22

Despite the fact that the collective contract prohibits plantation fore-men and supervisors from carrying arms without special authorization,

TABLE 2KILLINGS OF RURAL WORKERS AND PEASANTS

IN BRAZIL AND THE STATE OF PERNAMBUCO, 1982-1987

1982 1983 1984 1985 1986 1987 Total

Pernambuco 3 4 8 5 7 9 36Brazil 58 96 123 222 20(+a 200+a 899+

SOURCE:Movimento dos Trabalhadores Rurais Sem Terra, Assassinatos No Campo: Crimee Impunidade 1964-1986 (Sao Paulo: Global Editora, 1987); Comissao Nacional dos Bisposdo Brasil (CNBB), Regional Nordeste II, Terra: Nossa Necessidade e Nosso Direito (Recife:CNBB, 1984); Diario de Pernambuco, various dates; Amnesty International, Brazil:

Authorized Violence in Rural Areas (London: Amnesty International, 1988); and U.S.Department of State, Country Reports on Human Rights Practices for 1988, report sub-mitted to the House Foreign Affairs and Senate Foreign Relations Committees (Wash-ington, DC: Government Printing Office, 1989).

aPrecise data from 1986 and 1987 are unavailable owing to discrepancies in reported kill-ings for those years.








this provision is broken regularly. In the fields, cane cutters are accus-tomed to working under the eyes of armed foremen who, in the words

of one of them, consider walking unarmed to be "a sentence of death."23In such a setting, workers who demand rights or participate in unionactivities risk being threatened, harassed, beaten, and ultimately killed.

To an open-ended question about rural violence, 15 of 50 officials men-tioned that a worker or workers in their area had been killed in recentyears in conflicts with landlords. At one union the president said thathis predecessor had been gunned down in 1985, while at another theirlawyer had been murdered in 1982, and their current lawyer was beingthreatened. Almost everyone in the sugar zone is aware of incidents suchas these. Seventy-four percent of respondents said that there was vio-lence against rural workers in their area. Forty-eight percent said thatthey knew a member of their union who had lost his job for union activ-ity, while the same percentage knew a member who had been jailed forthe same reason. The potential costs of union activity, therefore, are con-stantly presented to workers, dampening the zeal of all but the stout-hearted. As one leader said, "I don't want to end up on the side of theroad with flies in my mouth." For others, the violence provokes a greaterdetermination to resist. "If they kill one of us," said a young union sec-retary, "we only get stronger."

The central government is seen by rural unionists as against them,because its Ministry of Justice cannot protect them and their members

against violence. This, in turn, affects trade union strategy; having littlefaith in ability of the central government to address their concerns, unionleaders spend relatively little energy pressuring it directly. The followingcase illustrates this point.

Several days before his labor court hearing against his former employer,on 4 April 1988, rural worker Jose Batista da Silva was abducted froma Recife market by civil policemen. The latter, who had no arrest war-rant, said that they were taking the worker to his former employer forquestioning. Batista was not seen again. Hearing about the case, leadersof Batista's rural union in Cabo decided to take action. Significantly, theaction was not a legal one. Instead, the union decided to use public

space to bring attention to its plight.24 It planned a march to protestwhat it considered to be another act of violence on the part of land-owners, this time in conjunction with the police. The march was sched-uled to end in front of the police station in Cabo. In a meeting withthe trade unionists, however, the state Secretary of Public Security indi-cated that if the march passed in front of the police station, he wouldnot be responsible for anything that happened.5 The workers thus decidedto change their route. On May 15, supported by various organizations,the march took place, avoiding the heavily guarded police station.

The Cabo case reveals a lack of faith in the legal system and the cen-tral government on the part of the union leaders. Although the abductionof their member was blatant and illegal, the union's main response wasto try to convince the public of the rightness of its position rather thanto seek official redress. While the state governor appointed a special del-egate to investigate the kidnapping, this was done at the request of the

employer and his association, the Sindicato dos Cultivadores de Cana de








Pernambuco (Pernambuco Union of Cane Planters). In addition, after unionofficials denounced the landlord as responsible for the kidnapping, they,

along with the worker's wife and brother, were sued for calumny anddefamation by the landlord.26

Rural violence increased markedly around the time of the transition tocivilian rule in 1985, and the new regime was unable to contain it. Thisis because the transition, and the proposed agrarian reform plan thataccompanied it, created a climate in which it seemed that the demandsof the landless might be met. Landlords, facing direct action on the partof workers and peasants, feared the loss of their property and privileges.Employing private gunmen and retaining their traditional influence overthe local police, they answered the workers with violent confrontation.

The death toll, and the lack of protection from the Ministry of Justice,created a reaction on the part of the rural workers' movement, a reactionto which we shall now turn.

Union Leaders' Alienation from the Central Government

The fact that neither the federal Ministry of Labor nor the justice min-istry adequately enforces the law in rural areas has consequences for theattitudes and behavior of rural trade union leaders in the Pernambucosugar zone. Put simply, it engenders a sense of alienation toward thecentral government and a belief that positive social changes will comeabout only through the efforts of rural workers themselves. While ben-

efits from the government would naturally be welcome, they are notreally expected, and the request for commitment to any particular govern-ment program is viewed with suspicion. Second, because the state is mar-ginal to the day-to-day struggle of the union leaders, regime change-amodification in the personnel and institutional arrangements of the cen-tral government, such as occurred in 1985-are also seen as relativelyunimportant.

This view stems from the nature of work relations in the sugar zone.The cane workers were originally bound to plantation owners in a kindof quasi-feudal system of paternalistic relations, bonds of peonage thatwent far beyond the sphere of work and extended to family relations(compadrice, or godfatherism) including ceremonies of birth, death, har-vest, and the like. While most workers became full proletarians in the1960s, a fully capitalist attitude toward work-contractual, monetary, short-term, impersonal-has not permeated the region, not least because asmall but significant portion of the work force still lives on the plantation.

The landlords' abandonment of traditional obligations created a senseof injustice among those abandoned. The landlords, most of whom nowlive in the city, see the proletarianization of the labor force as inevitable,necessary, and progressive. For them, it is part of the capitalist mod-ernization of the plantation, a process from which they have benefited.The workers, on the other hand, see their expulsion from the plantationsas neither necessary nor progressive.27 They were not consulted in thedecision to change the relations of production; they experienced thedecision as something done to, not for, them. Their nostalgia for a goldenage when they were on the plantation, growing their own food within asmall, pastoral, and essentially nonmonetized community, contrasts sharply








with their frequent criticism of life in the towns, with its stress, con-fusions, and demands for cash.8 Although some among the younger gen-

eration now prefer the town, a surprising number share their elders'belief that times were better on the land.

Because the workers feel a sense of loss due to the destruction of patron-client ties, their image of the landlord is ambiguous. The landlord is

criticized as a wealthy man who, despite his wealth, is willing to trampleon any worker's right in pursuit of his own gain-a "fox among the chick-ens" whom only a guard dog (the union) can restrain.30 But this criticismis based on an image of an ideal landlord who personally protects and

rewards his workers. Such an image is implied in comments about thelandlord such as "his heart is in his wallet," "he doesn't treat the worker

with compassion," and "he chews up the worker and throws away hisskin as he would an orange."-31 What the landlord could and should be,the concept of the ideal employer, is held up to rebuke the real landlordfor his perceived harshness and insensitivity.

This dualistic image of the landlord is analogous to the workers' image

of the central government.32 The union leaders surveyed made frequentcomments that the government should help them, reflecting an assump-tion that the federal ministries are neutral agencies capable of acting ontheir behalf. This contradicted a radical skepticism, also frequently

expressed, that the government only protects tubaroes (literally sharks:

capitalists, landlords), exploits the poor, is headed by corrupt and dis-

honest politicians, and would not help them.' This latter view led manyleaders to declare that the situation would only change if workers didsomething for themselves. It also manifested itself in an "us vs. them"

mentality, and a belief that government would only become benevolent

when some of "us" (rural workers) were elected to office. In the 1988

municipal elections in Pernambuco the rural movement acted upon this

belief when a record number of union officials-over 100-ran for local

political offices.34Trade union officials' criticism of the state and the social order over

which it presides can be sweeping. One secretary, for example, said that

"in this country, justice is money."35 In this view, the legitimate rightsof the poor are denied and the rich and powerful are able to act on everywhim, no matter how unlawful or immoral, with impunity. Such a social

order is upside-down, radically unjust, because "he who has a right does

not have a right, and he who does not have a right has a right," as one

worker said.36 Such a vision approaches the evangelical's millenarian pic-ture of the world before the Second Coming of Christ, a world of uncon-

strained evil in which the weak are set upon by the powerful and the

depraved, sustained only by a faith that the social order will eventuallybe turned right-side-up and the "meek shall inherit the earth." Althoughpotentially a basis for political action, this attitude can also become a

justification for a resigned, otherworldly passivity in the face of oppres-sion. It is not surprising, therefore, that evangelical Protestantism has

found many converts among the rural poor.37Rural unionists' anger toward the central government, like their anger

toward the landlord, stems from its perceived inability to consider and








protect workers' interests, and takes a personal form. In expressing dis-gust for President Sarney, they implied that because he was a large land-

owner, he would be unlikely to sympathize with the rural poor. Theirswas not a liberal conception of the state, in which the latter's role ismerely to enforce certain rules of the game, although, as has been argued,they decry the fact that rules are unenforced. Similarly, their view ofcitizenship goes beyond that implied by a representative model ofdemocracy, in which their right as citizens is essentially that of choosingbetween competing elites.' Instead, they see the necessity of federalofficials listening to their demands, then intervening on their behalf, andgranting them material benefits. The unionists experience alienationbecause this so rarely happens.

The survey results reflect these attitudes. Of 41 out of 50respondentswho claimed that the Sarney government was not democratic, only one

gave as a reason that "it was not elected by the people." But 22 repliedthat it does not side with the people; 13 that it favors landlords and othercapitalists; and 10 that it did not practice democracy in making decisions.?9

Of 34 respondents who were asked whether it "is important to con-sider the consequences for the government when making demands," 47percent said no, and 12 percent said, "Only the state government.' Inthe words of one respondent, "For us the struggle is to think of ourmembers, and not the government." Similarly, when asked what recentpolitical changes had given them the most hope, 15 percent of the 39

answered 4"nothing,"and 26 percent described something to do withincreased union organization. Of the latter group, typical responses were"the force of the union itself is our hope"; "who betters the situation ofthe worker is the worker and the union"; and "we don't work with theunion waiting for party political benefits."

A common feeling among rural union leaders was that the union is theunion, the government is the government, and never the twain shouldmeet. In this view, union contacts with politicians or government officialsare always harmful to the union. Thus, 28 percent of the 50 respondentssaid that they had no contact with local politicians (mayors, city coun-cillors), and 36 percent said that they had no contact with state politi-

cians (only 10 percent claimed they had contact with federal politicians).The fears expressed here were that politicians, and especially parties,could divide the union and might try to manipulate it for their own ends.

Furthermore, asked whether "the political change marked by the inau-guration of President Sarney [was] important and positive for the unionshere," 77 percent of the respondents said no. Most telling of all, 96percent did not believe that the Sarney government had the will to carryout a land reform.40 And when asked if there was a country that theyadmired for its political system, 21 of the 33 who said yes mentionedCuba, typically giving land reform and a high standard of living for ruralworkers as reasons for their choice.4'

The responses of those surveyed also reveal an ambivalence about theimportance of regime change. Asked if the situation of their union hadimproved after direct elections for governor were reintroduced in 1982(after 18 years of military-appointed governors), an equal number (46percent) said no as said yes. Similarly, 41 percent saw no difference








between military and civilian rule at the federal level, while 54 percentdid. Asked what political change had given them most hope in recent

years, only 4 of 39 respondents mentioned the election of Tancredo Nevesin 1985 (the first election of a civilian president since 1960). Of far greaterimportance were the union's own activities (15 of 39).

Documentary evidence suggests that state policies, and in particular

the 1985 land reform proposal, were also viewed skeptically by rural union

supporters. A proland reform pamphlet published in 1985 described the

civilian regime as "the same dog with a different collar." Speaking forrural workers, it asks, What changed for us in the New Republic?" andcautions that the change of regime was a "maneuver of the powerful toweaken our movement and postpone the solution to our problems." Thepromise of land reform, it argued, merely confused workers by encour-

aging them to wait for government handouts, rather than to fight for theright to land themselves. The New Republic offered only the vain hopethat "you can have a society of a fox with a chicken."42

In summary, rural union leaders have an alienated and bifurcated viewof the central government, in which the reality of repression and neglectprevails over hopes deriving from an ideal state. This view results in adefensive and rather passive political strategy. The leaders can mobilizemembers strongly to defend themselves against possible losses, as theydo at the peak of each year's cane harvest, when a salary campaign pre-

cedes negotiations with landlords over the collective contract. But they

are far less likely to mobilize in response to possible gains. They areskeptical of government programs, especially when, as in the case of land

reform, pushing for those programs in the past has resulted in politicaldefeat (i.e., after the 1964 coup), and the programs require sustainedand intensive government commitment in order to work.43 The unions

can avoid defeats, but they have difficulty winning in the political arena.4The unions' choice of political strategy is complicated when, as in the

case of Pernambuco after 1986, the state government is regarded more

positively than the federal government. As will be argued in the next

section, the transition to civilian rule in Pernambuco saw the unions enter

a political coalition that included landowners and sugar millowners,

weakening their stated commitment to land reform and putting them atodds with church and party activists who placed land redistribution above

loyalty to the ruling coalition. The unions' political bargain also created

dissension within its own ranks, and in 1988 leaders' attitudes toward

Governor Arraes, a politician who identified himself as a protector of the

rural poor, ranged from unquestioning devotion to bitter disappointment.

Regime Change and the Agrarian Reform Plan

Agrarian Reform on the National Level. At the national level, it is

tellingthat the 1985 agrarian reform plan was issued as a decree-law, in

the style of the old military regime, rather than as a bill requiring a

majority of votes in Congress. The plan had been announced in May,with much fanfare, at a CONTAG congress of thousands of rural union-

ists in Brasilia. It projected the expropriation of almost 450 million ha

and the granting of land title to 7.1 million people by the year 2000.








The plan was supported by the progressive wing of the church, partic-ularly its comissoes pastorais da terra (CPTs; pastoral land commissions),left-wing parties such as the PT (the Workers' Party, an important advo-

cate of land reform), the Movimento dos Sem Terra (Landless Workers'Movement, MST), and CONTAG.

The agrarian reform plan officially approved on 10 October 1985 dif-fered substantially from the May proposal. Whereas the plan presentedto the CONTAG delegates made the reform the task of the central gov-ernment, the October plan gave regional government agencies the rightto select the land for expropriation, opening up the reform to all sorts

of local pressures. Also, only properties that neglected their "social func-

tion" would be liable for expropriation. Issuing an official note in response

to the October plan, CONTAG called it a repudiation of the promisesmade to the workers' representatives in May. One of its complaints was

that sugar plantations were not made a priority for land redistribution.

On 18 October 1985, the superintendent of INCRA, Jose Gomes da Silva,and his entire staff resigned, alleging that the final text of the plan had

disfigured the original proposal.The revision of the plan between May and October reflected the power

of agrarian interests and the opposition of elements within the military,but it also revealed the weakness and division of the proreform move-ment. Ostensibly united in a national campaign, middle-class reformists,

church and left-party activists, and rural trade union officials formed anuneasy alliance. CONTAG, in particular, had a difficult relationship withother supporters of land reform. It regarded church organizations of thelandless such as the CPTs as potential rivals to the unions.45 In addition,CONTAG was affiliated with the CGT, the General Workers' Central,the politically more moderate of the country's two major central labororganizations. The other central, the CUT (Unified Workers' Central),heavily influenced by the PT, had a more radical strategy on agrarianreform.46

That strategy was to encourage direct, and at times illegal actions, such

as the occupation of unused land and government offices, rather than towait for government programs to distribute benefits from above. A CUTdocument published in 1986 declared that "the agrarian reform meansto struggle against the New Republic, wresting victories in confrontationswith the powerful and reaffirming . . . the power of the workers, theonly type of power capable of. . . realizing a real agrarian reform underthe control of workers."47 In contrast, the CGT favored a more legalisticapproach, urging proponents of land reform to wait for the government'sland management bureaucracies to do their work. It felt that illegal actionshad provoked a violent rightist backlash and endangered the transitionto democracy.

Church activists and PT members in the countryside tended to agreewith the CUT's strategy, while old-line union leaders most often sidedwith the CGT position and supported the ruling Partido MovimentoDemocratico Brasileiro (PMDB; Brazilian Democratic Movement Party).Sometimes, these differences would flare up during union elections when








leaders with opposing positions would compete, or during land occupa-tions when CONTAG officials kept a distance and church and PT activ-

ists took up the cause of the occupiers.The proreform lobby, while counting on the passive support of millions

of rural poor, lacked strong urban and party political support. Most par-ties of the left and center, including the PT, were based primarily onthe support of the urban working and middle class, to whom land reformoffered only indirect benefits. In rural areas, nationwide elections forstate offices in 1986 meant that most left and center parties were sub-dued on the land reform issue, because they feared alienating landown-ers whose campaign contributions they wanted. One INCRA official,exasperated by the slowness of the reform, commented that "the mobi-

lization of those interested in agrarian reform has to be . . . balanced,because any false step could serve as a pretext for a retrocession. It islike walking in a mine field."48

The reform as envisaged was expensive. Owners of expropriated landwere to be compensated-in cash for their home and surrounding plot,and in agrarian debt bonds (titulos da divida agraria, TDAs) for all otherlands. The federal government would also incur costs for mapping andsurveying services, transport for staff, issuing land titles, and so on. Thehigh costs of the plan, combined with the fiscal crisis of the government,bothered its supporters from the outset.49

In December 1985, 25 state agrarian reform plans were prepared bythe regional directorates of INCRA and given to President Sarney. Theseplans were drawn up without the input of potential reform beneficiaries,rural workers, or their trade union representatives. To oversee the reform'simplementation, agrarian commissions, of which 3 of 9 members weretrade union officials, were set up in the states in August and September1986. But this was long after the state plans had been approved by thepresident, in May. The institutions meant to ensure trade union partic-ipation in the planning of the reform were thus set up too late, after theplans were already complete. Furthermore, the government minister whohad first announced the plan to the CONTAG congress, Nelson Ribeiro

of the newly created Ministry of Agrarian Reform and Development(MIRAD), was fired in May 1986. Altogether, MIRAD had four differentministers in the 1985-88 period.

The land expropriation procedure employed by INCRA was complexand highly centralized, capable of being blocked by landlord pressure atnumerous stages. Although the original plan had stressed that the "par-ticipation . . . of rural workers" was a "vital requisite" for the plan'ssuccess,50 such participation did not occur in practice. INCRA selectedbeneficiaries for land, using its own confidential surveys. Land expro-priation decrees always came from Brasilia, where they required the sig-nature of the president, and listed the number of

parcelsof land into

which the expropriated plot must be divided without any local input.Once granted a parcel, beneficiaries usually had to install themselves onthe land at their own expense, although in some cases INCRA providedmoney for three months, until the first harvest.51

The state agrarian commissions, designed to increase participation in








the reform, had nothing to do. A little over a year after their establish-

ment, in October 1987,the federal

governmentabolished INCRA and

transferred its responsibilities to MIRAD. CONTAG interpreted this asan abandonment of the reform by the government and withdrew itsmembers from the commissions.

A vital blow to the reform was dealt on 10 May 1988, when the Con-stituent Assembly, then writing a new constitution, placed a clause inthe document stating that "productive" land could not be expropriated.By withholding "productive" land from the reform, critics argued, anyproperty with a few head of cattle or some machinery could be exemptedfrom expropriation; land speculation would be protected. This was oneof the most bitterly fought issues in the making of the Constitution. While

CONTAG officials argued against the clause, a well-funded organizationof landowners, the Uniao Democratica Ruralista (UDR, Democratic RuralUnion), lobbied for it. After seven separate votes in the Assembly, amajority was attained. The exemption of "productive" land was upheldon 29 August 1988, in a second round of voting, with 83 of 168 PMDBcongressional deputies voting for it.52 Before this vote, the governmenthad the law, but not the will, to enact agrarian reform; now, it lackedthe law as well.

The rural unions considered the "productive" clause of the Constitu-tion a betrayal. The federation in Pernambuco published posters withthe title "Traitors of the People" after the August vote, showing pictures

of congressional deputies who had voted for it. A union song with thesame title was sung in meetings in the sugar zone. "The deputies whovoted against agrarian reform betrayed the Brazilians," lamented the lyr-ics, "leaving the masses saddened, crushing us who are rural workers."'5

By the end of 1988, it was reported that the agrarian reform plan hadfailed to meet its objectives. Of the 27 million ha targeted for the1985-88 period, only 7 percent had been expropriated. Of the 850,000families slated to receive land, only 40,000, or 4.7 percent, actually had.M4Nelson Ribeiro, the ex-minister of MIRAD, had written optimistically in1987 that land redistribution, by "atomizing . . . [the] concentration of

power" of the large landowners, would "give the most decisive contri-bution to the strengthening of democracy for all."' Instead, the powerof the large landowners, and the weakness of its supporters, had atom-ized the reform.

Nonreform in the Pernambuco Sugar Zone. In Pernambuco, a north-eastern state, the transition to civilian rule was quickly followed by thereturn of Miguel Arraes to the governorship. Arraes had been governorin 1963/64. During his brief administration, rural workers had won theright to a written collective contract, and a successful general strike inthe sugar zone had resulted in a dramatic increase in their wages. Arraes,simply by recognizing the unions' right to bargain on behalf of their

members, had altered the balance of power in the sugar zone. Becauseof improvements in worker welfare under his administration, he won frommany workers a loyalty and gratitude that borders on worship, reflectedin the popular saying, "Deus no Ceu, Arraes na terra" (God in Heaven,Arraes on earth). On the other hand, he earned the enmity of manylandlords for allowing "chaos" and "agitation" to reign in the fields, and








he was deposed by the military in the 1964 coup. After a 15-year exile,he returned to Pernambuco politics in 1979.

The 1986 governor's race pitted Jose Mucio, a millowner of the PartidoFrente Liberal (PFL, Liberal Front Party), against Arraes of the PMDB.

Campaigning on behalf of Mucio and running for federal deputy was

Francisco Juliao, a lawyer/politician who had advocated agrarian reform

in the early 1960s and, like Arraes, been forced into exile after the 1964

coup. Juliao wrote an open letter to the sugar zone's millowners urging

them to set aside 10 percent of their lands and, out of a sense of Chris-

tian charity, give it to their workers. He later claimed that, had Mucio

been elected, he would have been appointed state secretary for agrarian

reform, with the power to confiscate land from owners.'

Rural workers were unsympathetic to the campaigns of both Mucioand Juliao. Both candidates lost badly in the election. Of those trade

union leaders surveyed, 94 percent responded that they had voted for

Arraes against Mucio, and only 6 percent said they voted for Juliao. The

latter was widely seen to have become a "turncoat," allying with a mill-

owner for his own political advantage.The winning Arraes coalition in 1986, therefore, received the strong

support of FETAPE, but it also received the support of a segment of

the state's landowning elite. It is estimated that 30 percent of the cam-

paign money spent by sugar millowners in 1986 went to Arraes's Frente

Popular (Popular Front).57 One of the Frente's PMDB candidates was

Antonio Farias, a sugar millowner who won a seat in the federal Senate.And 17 of the 23 PMDB state deputies were former members of ARENA

(Alianca Renovadora Nacional), the dominant party during the military

regime.58

The rural union leaders thus knew that this would be a different Arraes

government from the one elected in 1962.59 Many were cautious, recall-

ing those earlier days when activists had militated in favor of union rights

and land reform and paid heavily for it after the 1964 coup (when about

300 rural workers were purged from the movement, some losing their

lives). Although the prospects of military intervention looked remote, union

officials adopted a wait-and-see attitude toward the Arraes government.Not all of that government's early appointments and policies met with

the approval of the unionists. While the new secretary of labor was a

former FETAPE lawyer, the important secretary of public security, in

charge of the civil police, was a general in the army reserve named Evilasio

Gondim. Gondim had been head of the Servico Nacional de Informacoes

(SN, National Intelligence Service) when that agency was accumulatingevidence that Arraes, then exiled in Paris, was a subversive. Further-

more, Arraes's appointment for chief of the military police, Fernando

Gonzaga Pessoa, was head of the infantry regiment in Socorro, where

Arraes was imprisoned after being deposed in April 1964. Finally, Arraes's

secretary for administration had been a driver transporting political pris-oners after the coup.f6 Clearly, in composing his second administration,Arraes made substantial concessions to political forces that had strongly

opposed his first government in 1963.

Arraes's position on agrarian reform was ambiguous, partly because he








had ambitions for higher office and did not want to antagonize right-wingers. While rhetorically supporting the reform, he maintained that it

was a federal matter and should not involve the state government. On4 May 1986, when President Sarney met with the presidents of CONTAGand FETAPE at Recife's airport to discuss the state's newly drafted agrarianreform plan, Governor Arraes did not attend the meeting.

In the sugar zone, reactions to the reform plan were mixed. TheFETAPE president, Jose Rodrigues, said that "the plan leaves much tobe desired," and accused the federal government of having caved in tolandlord pressures and given away substantial concessions in relation tothe initial proposal.61 The president of the state Sindicato dos Cultiva-dores de Cana, the cane growers' association, said that the growers sup-

ported reform, but only if it were done on "unproductive"land. GovernorArraes, for his part, continued to avoid the issue and instead initiated alocal program of voluntary land donations on the part of the sugar mills.These donations, heavily publicized by the governor's press office, didnot involve more than symbolic amounts of land or people.62 They allowedArraes, however, to be seen as supportive of agrarian reform withoutgetting entangled in the messy conflicts that the federal reform entailed.The millowners, on the other hand, seemed to participate in Arraes'sprogram in the hope that it would obviate the need for them to partic-ipate in the national plan.

In the second half of 1986 the state agrariancommission for Pernambuco

was set up, ostensibly to assist in the enactment of the national plan. Asin many other parts of Brazil, this body never really exercised its func-tion. When INCRA was abolished in October 1987, FETAPE, along withthe other state federations, withdrew its members from the commission.

If at this point the union leaders were not expecting agrarian reformin Pernambuco, some landless families, acting out of desperation, tookmatters into their own hands. On 23 August 1987, 199 families movedon to an area of unused land belonging to the powerful Lundgren family,about 50 miles from Recife on the border of Abreu e Lima and Igarassucounties. They demanded title to the property. The case went to court,the squatters were ordered to be removed, and the state governmentsettled each of the families on 10 ha of land belonging to the Ministryof Agriculture, in Igarassu. The original plot of land, however, called bythe squatters Pitanga II (a previous land occupation in another area hadbeen called Pitanga I), became the focus of a legal and political battlethat lasted for over a year.

The squatters were determined and patient, occupying the grounds ofthe MIRAD building in Recife three times during the ensuing year. Theyhad no faith that they would receive land in the official reform.63 Theysucceeded in getting some help from local trade unions and later fromthe state government, which sent food and provided medical care to thesquatters on a weekly basis. Despite this, the squatters' leaders werecritical of FETAPE. They claimed that the federation was not advocatingtheir cause strongly for fear of embarrassing the state government andalienating supporters in the ruling PMDB. The church's CPT, based inRecife, became the squatters' most steadfast political ally.

In February of 1988, FETAPE paid for a group of squatters and union








representatives to go to Brasilia for a meeting with the MIRAD minister,Jader Barbalho. Barbalho promised to recommend expropriation of Pitanga

II to President Sarney. The squatters' leaders returned to wait with theircompanions at their bleak settlement, where 11 children died of illnessbetween August and April.

On April 11, Governor Arraes declared himself against the expropri-ation of Pitanga II, citing the opinion of the federal Forestry Institute(IBDF) that settling the squatters on Pitanga II would cause unaccept-able ecological damage. The next day he gave in to pressure from hissecretary of labor and FETAPE's President Rodrigues and sent a tele-gram to MIRAD asking for the "immediate expropriation" of the area.On April 20, President Sarney signed a decree expropriating 1,120 ha

of Pitanga II. Rodrigues hailed the decision, crediting it to the "firmposition of the state government" when, in reality, Governor Arraes onlytook action in support of the squatters' claims a full seven months afterthe land was first occupied.64

On July 4, carrying pots and pans, mattresses, and food, the PitangaII squatters occupied the MIRAD building as they had earlier in theyear, this time to demand title to the expropriated land. They denouncedthe "lack of support from FETAPE" and declared that only the churchactivists of the CPT were behind themi65 The squatters were tired ofpromessas da boca (oral promises) and wanted a written guarantee thatthey could settle on the land. One squatter explained: "I don't under-

stand laws, I understand right. The laws were made by men; right comesfrom God."' Finally, on 25 November 1988, the federal government issueda title of 840 ha for 102 families at Pitanga II (other families had aban-doned the struggle or been resettled elsewhere). One year and threemonths after the initial land invasion, the squatters appeared to haveachieved their goal.

The Pitanga II case illustrates several points about land reform inPernambuco. First, the small amount of land that got redistributed tendedto be the result of direct action on the part of the rural poor themselves,rather than the work of the unions or of the government's land reformbureaucracy. Second, because of his unwillingness to offend landowners,the state governor was reluctant to back the demands of the landless.He was willing to provide food, in part to avoid an embarrassingencampment in front of the Governor's Palace, but he did not push forfederal expropriation until it seemed that the squatters simply would notgo away. Finally, the trade unions, as supporters of the governor, sharedhis dilemma. Criticized by the squatters for not doing enough, they couldnot be seen to be doing nothing. They did provide essential help in theeventual resolution of the problem, but land for the landless was,throughout the period, a secondary issue for them. The wages and work-ing conditions of the mass of their members remained the top priority.

Despite the promises that accompanied regime transition, and the suc-cesses of isolated groups of squatters such as the people of Pitanga II,little change in the pattern of landownership in the Pernambuco sugarzone occurred. Excluding the Pitanga II case, the government had

expropriated a mere 1.78 percent of the land targeted in its plan and

granted land to 1.4 percent of the projected beneficiaries by the end of








1987. In absolute terms, this amounted to 21 families receiving 400 ha,

a small number in an area containing hundreds of thousands of ruralpoor.67

Conclusion

The skepticism and limited participation of rural trade union leadersin the 1985-88 Brazilian agrarian reform constitute a neglected factor inexplaining the failure of that reform. The rural trade union leaders didnot push strongly for land redistribution because they doubted the gov-ernment's commitment to its own plan, were never asked to participatemeaningfully in its enactment, and did not want to neglect daily strug-gles in order to fight for the program. Not only was their expectation of

success low, but the expected cost of failure-a landlord backlash suchas occurred in the wake of the 1964 coup-was high.

A recent study of 16 modern land reforms concludes that reforms onlysucceed when the beneficiaries, workers and peasants, gain land throughpressure exerted by their own organizations, and "when these organi-zations have substantial input into the conditions and structure of thereform."' Given this condition, it is fair to say that as a result of theway its implementation was designed, the 1985 Brazilian reform plan wasdoomed to failure from the outset.

The roots of the rural trade union leaders' attitude toward the reformcan be found in both the nature of Brazil's central government as it oper-ates in the countryside, and the nature of the 1985 transition to civilianrule. The Ministries of Labor and Justice have failed to protect ruralunionists, not only historically, but by their contemporary inability toenforce a legal order. In the areas of labor law and violent crime, mostcritically, the Brazilian government has not consistently enforced rulesin the way that it has in other areas. Because of this failure, and theexploitation and violence against the rural poor that it permits, rural tradeunion leaders are extremely critical and suspicious of the federal gov-ernment, directing relatively little of their energy toward getting benefitsfrom it. As a consequence, they are frequently unimpressed by regime

change. Government proposals such as land reform have limited credi-bility in their eyes, whether made by military or civilian ruler's.Finally, the 1985 Brazilian regime was marked by considerable con-

tinuity of both policy and personnel at the elite level. The supporters ofagrarian reform faced a strong and well-organized opposition and sufferedfrom internal division. In the case of the Pernambuco sugar zone, thewinner of the 1986 gubernatorial election, while drawing massive supportfrom the rural trade union movement, was also backed by a significantminority of large landholders. This meant that even when the state gov-ernment proved sluggish on the land question, rural unionists werereluctant to criticize or pressure it. This in turn led to conflict between

the unions and other representatives of the rural poor, such as churchand left-party activists, who favored more vigorous direct action in favorof the landless. But the union federation did not break with the stategovernment. Instead, like most rural union federations in Brazil, it con-centrated on its chief field of interest, the wages and working conditionsof its members.








NOTES

1. There is a vast literature on the role of agrarian reform in development. Forassertions of the benefits of agrarian reform in Latin America, see Everett E. Hagen, TheEconomics of Develpment (Homewood, IL: Richard D. Irwin, 1975), p. 122; Ignacy Sachs,Studies in Political Economy of Development (Oxford: Pergamon Press, 1980), p. 76; MichaelP. Todaro, Economic Development in the Third World (London: Longman, 1977),pp. 214-15; and F. E. Trainer, Abandon Affluence (London: Zed Books, 1985), p. 140.

2. Frances Hagopian and Scott Mainwaring, "Democracy in Brazil: Problems andProspects," World Policy Journal (Summer 1987): 502-4.

3. Dirceu Pessoa, ed., Politica Fundiaria no Nordeste, vol. 1, Pano de Fundo (Recife:Fundacao Joaquim Nabuco, Instituto de Pesquisas Sociais, Departamento de Economia,1986), pp. 121-22.

4. I conducted this research between April and November 1988, using a written

questionnaire containing over 50 questions. Respondents were all elected members ofunion directorates: presidents, vice-presidents, treasurers, and secretaries from unions inboth the northern and southern sugar zones. Interviews covered 42 of the region's 45sugar unions so as to incorporate unions of different size, degree of militancy, and politicalorientation in the sample. A copy of the questionnaire and aggregated responses is avail-able from the author on request.

5. Food and Agriculture Organization (FAO) of the United Nations, Rural Poverty(Rome: FAO, 1988), Annex 2, p. 17. Figures are from 1980. The FAO defines povertyas the "incapacity to become inserted in the socio-economic environment in a way thatcontinually allows for the satisfaction of basic necessities of life." For measurement pur-poses, absolute poverty is "that income below which a set of basic necessities cannot beafforded." From ibid., p. 7.

6. The National Confederation of Agricultural Workers (CONTAG) has approxi-

mately 9,700,000 members, with 22 state federations and 2,822 unions (Jornal do Brasil,26 July 1987, p. 4). A different source lists CONTAG membership as 8,000,000: see BiornMaybury-Lewis, The Debate over Agrarian Reform in Brazil, Institute of Latin Americanand Iberian Studies Paper no. 14 (New York: Columbia University, 1990), p. 11.

7. Ministerio da Reforma e do Desenvolvimento Agrario (MIRAD), Proposta paraa Elaboracao do 10 Plano Nacional de Reforma Agraria da Nova Republica (Brasilia: MIRAD,May 1985), p. 4.

8. Caio Prado, Jr., Historia Economica do Brasil (Sao Paulo: Editora Brasiliense,1953), pp. 231-87.

9. Angela Kageyama, "Algumas Caracteristicas das Categorias de Imoveis Rurais noBrasil em 1978," Reforma Agraria 16 (December 1986-March 1987): 50. The data arefrom a 1978 study by the Instituto Nacional de Colonizacao e Reforma Agraria (INCRA)that was not released until 1985.

10. Ibid., p. 52. The existence of this quantity of unused land has been disputed,rather unconvincingly, by agronomist Francisco Graziano Neto; see "Reforma no Brejo,"Veja, 7 November 1990, pp. 5-7.

11. Robert E. Price, Rural Unionization in Brazil, Land Tenure Center Research

Paper Number 14 (Madison: University of Wisconsin, 1964), p. 68.

12. The value of the debt of latifundiarios who had not paid the land tax (impostoterritorial rural) between 1980 and 1985 was estimated in the latter year at 3.6 trillioncruzeiros. From Campanha Nacional Pela Reforma Agraria (CNPRA), Reforma Agraria,Para Que? (Rio de Janeiro: CNPRA, August 1985), p. 11.

13. Between 1970 and 1980, the number of rural poor rose from 27.6 to 28.8 million

(FAO, Rural Poverty, p. 7). Brazil's Gini coefficient for land concentration went from 0.85in 1960 to 0.86 in 1980; ibid., pp. 48-49. Properties of more than 1,000 ha increasedtheir share of the cultivated land from 47 percent in 1967 to 58 percent in 1984. Small

properties of less than 100 ha had their share of land decrease through these years, from19 percent to 14 percent (MIRAD, Proposta, p. 4).

14. David Goodman et al., "Agro-Industry, State Policy, and Rural Social Structures:

Recent Analyses of Proletarianisation in Brazilian Agriculture," in Proletarianisation in the

Third World: Studies in the Creation of a Labour Force under Dependent Capitalism,ed. Barry Munslow and Henry Finch (London: Croom Helm, 1984), p. 198. Accordingto the 1975 Agricultural Census, rural properties of 100 ha or more got 68.1 percent of








all credit; some of this credit was not used for agriculture at all, but was instead investedin urban property. See Instituto Brasileiro de Analise Socio-Economico (IBASE), Alguns

Dados sobre a Estrutura Agraria Brasileira (Rio de Janeiro: IBASE, 1981), pp. 3-4.15. Many of these displaced peasants went to the cities; rural-urban migration is esti-mated at 30 million for the 1960-80 period.

16. An FAO mission that came to Brazil to monitor the reform in 1968 reported thatthe number of families that had received land until that date did not exceed 329. FromCarlos Enrique Guanziroli, Politica Agraria do Regime Pos-64 (Rio de Janeiro: IBASE,1984), p. 1.

17. FAO, Rural Poverty, p. 74.18. Price, Rural Unionization in Brazil, p. 5.19. Americas Watch, Rural Violence in Brazil (New York: Human Rights Watch, 1991).

See also Amnesty International, Brazil: Authorized Violence in Rural Areas (London:Amnesty International, 1988); and U.S. Department of State, Country Reports on HumanRights Practices for 1988, report submitted to the House Foreign Affairs and Senate For-

eign Relations Committees (Washington, DC: Government Printing Office, 1989).20. Quoted in U.S. Department of State, Country Reports, p. 471.21. Amnesty International, Brazil: Authorized Violence in Rural Areas, p. 67.22. Ibid., p. 10.23. Jornal do Brasil, 9 September 1984.24. For a discussion of urban trade unionists' use of public space in Barcelona, Spain,

see Robert Fishman, "Protests in Public Places: Underpinnings of Working Class Oppo-sition in Franco Spain" (Paper delivered at the Conference of Europeanists, Washington,DC, 30 October -1 November 1987), pp. 6-11.

25. The president of the union complained about this veiled threat, because the statesecretary had been appointed by a governor (Miguel Arraes) who had received strongsupport from rural workers. "We find the position of the secretary very strange," he said,"because all the rural workers voted for Arraes, and it isn't possible that his police now

line up against us." From Diario de Pernambuco, 14 May 1988.26. The Cabo union president told me on 22 September 1988 that there was a pend-ing legal action against the policemen thought to be responsible for the abduction of Batista,but it is not clear whether this was brought by the union or another party.

27. Lygia Sigaud, Os Clandestinos e Os Direitos (Sao Paulo: Livraria Duas Cidades,1979).

28. Eighty-two percent of those surveyed felt that the standard of living of the ruralworkers was higher during Miguel Arraes's first term as governor (1963/64), when themajority of workers had access to subsistence plots (sitios), than it is today.

29. The opinion that the standard of living of the workers has declined since theexpulsion from the plantations is just as prevalent among the young (who had no adultexperience of the former time) as it is among the old. In the survey, 85 percent of those40 or younger and 80 percent of the over-40 category, hold it.

30. Fifty-seven percent of those who said they were in favor of the right to strike(which was 92 percent of all respondents) justified their stance by saying that landownersonly make concessions in the face of strikes or threats of strikes.

31. Author's field notes, April-November 1988.32. For an analysis of the dualistic view of the national government among rural workers

in Sao Paulo state in the early 1970s, see Verena Martinez-Alier and Armando Boito, Jr.,"The Hoe and the Vote: Rural Labourers and the National Election in Brazil in 1974,"Journal of Peasant Studies 4 (April 1977): 160. For an analysis of a different, and highlydependent view of the national government among urban workers in the early 1970s, seeYoussef Cohen, "The Benevolent Leviathan: Political Consciousness among Urban Work-ers under State Corporatism," American Political Science Review 76 (1982): 46-59.

33. For example, one union official said to me: "The government isn't [democratic]because it gives a lot to the rich. It only looks after the rich, even the middle class itdoesn't look after. A democratic government would look after everyone." From field notes,Camutanga, Pernambuco sugar zone, 17 November 1988.

34. Few workers seem to consider the danger that union officials may abandon theirroots once in local government, developing new interests and bases of support, i.e., thatone of "us" may turn into one of "them" if given the chance. The fact that this dangeris a real one is reflected in the rueful observation of one union official, commenting onthe performance of a former union president who became mayor, that although "we had








access" to him, "he wasn't the mayor we had hoped for." From field notes, Paudalho,Pernambuco, 18 July 1988.

35. Interview in Vitoria de Santo Antao, Pernambuco, 2 May 1987.36. Rural worker in Maraial, Pernambuco, 30 November 1988.37. In Pernambuco, the total population has doubled from 1950 to 1980, from about

3 to 6 million. During the same period, the Protestant population (traditional and Pen-tecostal) has quadrupled, from 80,000 to 325,000. From Instituto Brasileiro da Geografiae Economia (IBGE), Servico Nacional de Recenseamento, Censo de 1960, Volume II,Tomo VI, Parte Ia, Pernambuco (Rio de Janeiro: IBGE, 1961); and IGBE, RecenseamentoGeral do Brasil 1980, Volume I, Tomo VI, No. 12, Censo Demografico, Pernambuco (Riode Janeiro: IBGE, 1981).

38. Carole Pateman, Participation and Democratic Theory (Cambridge: CambridgeUniversity Press, 1970).

39. This was an open-ended question to which most people gave more than one answer,the reason that the total is greater than 41.

40. Admittedly, this question was asked in 1988, three years after the agrarian reformplan had been first announced. But documentary evidence confirms the view that manytrade union leaders were skeptical of the government's commitment to reform in 1985.

41. One respondent said of Cuba, "They say that there, the children of the workerdon't die, the sugar mills don't have an owner, the conditions of work are better and thewages are better."

42. Centro Educacao Popular do Instituto Sedes Sapiente (CEPIS), Reflexao sobre aViolencia no Campo (Sao Paulo: CEPIS/Movimento Sem Terra, 1985), pp. 14-15.

43. Rural union officials, in talking about land reform, usually said that land itselfwas not enough, but that the small farmer needed government credit, technical assistance,

help with marketing, and so on. A 1988 study by Brazil's development bank, the BNDES,revealed that this kind of help was noticeably lacking in the Sarney land reform. In a

study of 1,517 small farmers who received land in 12 different states, the bank concluded

that in many cases, recipients of land ended up selling their lots or only growing enoughfor subsistence, working as wage laborers for someone else. The government did not pro-vide the smallholders enough inputs to become viable commercial farmers, so they fellback into their previous condition of poverty. From Relatorio Reservado 24, no. 1173,24-30 July 1989.

44. This is the point made by Alain de Janvry and Elisabeth Sadoulet in "The Polit-ical Feasibility of Rural Poverty Alleviation" (University of California, Berkeley, 1989),p. 22. The authors suggest that the rural poor can only win politically if positive gainsfrom a particular policy also accrue to the nonpoor. Groups from the latter category can

then take the initiative in mobilizing the poor in support of the policy. In the case of

Brazilian land reform, the nonpoor did stand to benefit from land redistribution in the

form of cheaper food prices and lower rates of urban crime, but these benefits were quiteindirect. This accounts, perhaps, for the relatively low level of political pressure for the

policy on the part of the nonpoor.45. At the Fourth CONTAG Congress in Brasilia in May 1985, delegates expressed

the fear that religious groups had the objective of "dismantling the union movement" in

the countryside. From Yves Chaloult, "Uma das Contradicoes da Nova Republica: 0 Pro-

jeto Nordeste" (Brasilia, 1985), p. 12.

46. For more on the differences within the proagrarian reform alliance, see Maybury-Lewis, Debate over Agrarian Reform in Brazil, pp. 11-14.

47. Maria Lia Pandolfi, "Forcas Sociais e Articulacoes no Encaminhamento da Reforma

Agraria" (Fundacao Joaquim Nabuco, Departamento de Sociologia, Recife, 1987), p. 182.

48. Pessoa, Politica Fundiaria no Nordeste, pp. 121-22. For the author of those words,the minefield exploded. In September 1987, he and three other INCRA officials were

killed in a plane crash in southern Para in mysterious circumstances. Although the official

government investigation of the crash ruled out foul play, southern Para is an area of

intense land conflict, and some commentators believe that the INCRA plane was sabo-

taged by landowners who felt threatened by the agency's apparent commitment to land

redistribution in the area.

49. CNPRA, Reforma Agraria, Para Que? p. 6.

50. MIRAD, Proposta, p. 26.

51. Pessoa, Politica Fundiaria no Nordeste, pp. 115-16.








52. "O Que Esta na Constituicao," Folha de Sao Paulo, 11 September 1988,p. A-10.

53. "Traidores do Povo," union song lyrics distributed in Nazare da Mata, Pernambuco,2 October 1988.

54. Jan Rocha, "Landowners Claim Total Amazon Rights," Manchester Guardian

Weekly, 29 January 1989. INCRA data uses a longer time frame and differs slightly from

those presented here. Between May 1985 and May 1988, it reports, 2,966,902.4 ha (11

percent of the target) had been expropriated, and this land was capable of sustaining78,229 families, or 9.2 percent of the beneficiaries planned for settlement by the end of

1988. There was no data on how many of this latter figure had actually been settled on

land. From Instituto Nacional de Colonizacao e Reforma Agraria (INCRA), Superinten-

dencia Estadual de Pernambuco, Divisao de Projetos de Colonizacao, Secao de Progra-macao e Controle, Areas Obtidas para Assentamento (Recife: INCRA, 1988), p. 1.

55. Nelson de Figueiredo Ribeiro, "A Reforma Agraria na Constituinte: Uma Opcao

pelos Pobres," Reforma Agraria 16 (December 1986-March 1987): 30.

56. From author's interview with Francisco Juliao, Cuernavaca, Mexico, 24 May 1988.57. Roberto Aguiar, "Os Financidores das Campanhas," manuscript for the Folha de

Pernambuco, September 1988.

58. From "Esquerda Vou Ver," Folha de Sao Paulo, no, date.

59. Because illiterates did not have the vote in 1962, mosf rural workers did not vote

in that election. By 1986, illiterates had the vote, and rates of illiteracy had declined inthe sugar zone, ensuring the participation of almost all rural workers in the election.

60. From Veja, 18 March 1987. Gondim was the same man who threatened the ruralworkers' march in Cabo, described earlier.

61. Pandolfi, "Forcas Sociais e Articulacoes Politicas," p. 180.

62. Primeria Pagina, a newspaper published by the state government, reported that

1,200 ha had been donated in the program as of 4 April 1988. Given that INCRA defined

the minimum plot size to sustain an average family in the sugar zone as 15 ha, this would

allow 80 beneficiaries to get land. Yet there are said to be 100,000 to 150,000 unemployedcane cutters in the sugar zone during the slack season.

63. The following exchange between a landless demonstrator occupying the MIRAD

grounds and a policeman was reported in the local press. Demonstrator: "We want landto cultivate." Policeman: "But President Sarney wants to enact an agrarian reform." Dem-onstrator: "Only in another state, sir; here, no." From Diario de Pernambuco, 14 April1988.

64. Jornal de Comercio, 22 April 1988.

65. Diario de Pernambuco, 5 July 1988.

66. Jornal de Comercio, 7 July 1988. The difference between leis (laws) and direito(right) is an important one among the rural poor. Leis are seen as unfair, easily manip-ulated, and stretched. Direito is what is right in accordance with popular notions of justiceand credited with divine origin. There is a parallel between lei and direito and E. P.

Thompson's "market" and "moral" economy; see "The Moral Economy of the EnglishCrowd in the Eighteenth Century," Past and Present 50 (February 1971).

67. Comissao Nacional dos Bispos do Brasil (CNBB), Regional Nordeste II, Setor dePastoral Rural, Os Descompassos da Reforma Agraria em 1987-Pernambuco (Recife: Centrode Estudo e Acao Social, Setor Rural, 1987), pp. 2-3.

68. John P. Powelson, "Land Tenure and Land Reform: Past and Present," Land UsePolicy 4 (April 1987): 120.

Documents

Agrarian Reform and Workers Unions in Pernambuco Sugar Zone