Social Movements and Democracy in Brazil:
Patterns of Popular Contention and Their Impact on the Process of Re-Democratization, 1977-1989 _*

Salvador A.M. Sandoval _**

Center for Studies of Social Change
Working Paper No. 234

The process of re-democratization in Brazil displayed distinctive patterns of social movement contention which raise questions as to the role that these exercised on the process and their impact on the political outcomes. We find, in reviewing the distribution of contention data, depicted in Graph 1, important variations in the fluctuations of labor and social movement collective action which reflect the complexity of regime transformation and which is the subject of our study. During the period between 1977 and 1989, there were not only phases in which labor and urban social movements simulataneously mobilized in collective protest, phases in which only one or the other mobilized in public demonstrations, and years in which there was a decline in both types of movements. The diversity of these fluctuations makes it impossible to reduce an explanation to any single structural factor since during periods of regime transformation crises, authoritarian governments do not, in general, facilitate either labor or urban social movement contention in such a way that would result in the several patterns of fluctuations present in the Brazilian case.

This paper examines the role of popular contention in the transition from a military dictatorship to civilian democracy in Brazil in order to address the questions raised above. It focuses on two broad types of collective action as indicators of protest activities: collective mobilizations of popular urban movements and strike activities of the labor movement. Both types of collective mobilizations are considered protest activity since under military rule all collective action was regarded by the government and military elites as threats to national security and regime stability. The analysis presented here is structured around variation in the timing of working-class strikes and of urban social movement collective action as two facets of the protest cycle that characterized the transition period between 1977 and 1989. This period dates from 1977, in which the first collective protest against military rule re-appeared after 10 years of severe repression, to 1989, which marked the end of the first civilian presidency and the enactment of the Constitution of 1988 eliminating the vestiges of authoritarian rule in Brazil.

Urban popular contention has taken on diverse forms in Brazil depending on the degree of powerholder patronage of popular organizations, the degree of resistance by elites, and the locus of contention. Thus contentious actions appear in various forms: labor unrest manifested in local or sectoral workers' strikes and general strikes; mobilizations of popular social movements advocating a large array of political and social changes; and finally spontaneous riots and sackings of food markets, buses and commuter-trains. These forms of grass-roots protest resemble each other in as much as they use collective action to challenge powerholders and are based on the mobilization of substantial popular resources to press specific interests.

In distinguishing conceptually forms of popular contention, Charles Tilly has classified them according to on the one hand, the extent that they are autonomous or patronized by powerholders, and on the other, according to the locus (whether in the local or national political arena) of the targets of mobilization. ₁ In this conceptualization, each form of popular contention can be classified according to the intersection of these two variables. This classification scheme is illustrated, in an adaptation for the Brazilian case, in Graph 1. Popular contention changes as social movement actors become more autonomous of powerholders and movements more enduring and organizationally complex enough to advance from local political arenas to national ones. These may all be considered contentious collective actions varying in form, organization and dynamics depending on their relation to powerholders and the relevant political arena. Given the diversity of forms of popular protest in the Brazilian case, as indicated in Graph 1, the analysis of the temporal trajectory of contention and its impact on the democratic process is limited to labor strike activity and urban popular organizations' collective action.

In its most general form the social movement consists of "a sustained public challenge to powerholders in the name of a disadvantaged population living under the jurisdiction of those powerholders by means of repeated public displays of that population's numbers, determination, unity, and worthiness via means which go beyond the current prescription of the authorities." ₂ Today, social movements and the labor movement (as a specific form of a social movement) are standard vehicles for political action, but in periods of structural political transformations such as changes in regimes, the incidence of different forms of popular contention must be analyzed diachronically in relation with each other and with changes in elite behavior, if one is to assess how social movements influence the process of regime transformation.

Given our objective, the research question posed in this study is "to what extent and in which ways do social movements promote democracy in periods of political structural transformation?" We have reason to believe that there is an important correlation between the outcomes of democratization and the forms of social movement intervention, as demonstrated in the cases of Latin American countries, Spain and Portugal. This correlation, however, does not imply that social movements directly and single handedly cause specific forms of democratization since regime transformations, government reform, and even policy changes are the result of the intense interaction among multiple political forces including powerholders, polity elites, challengers and their elite allies. Thus we can expect that any response to this question will be less clean-cut and most likely neither unidirectional nor unidimensional since often setbacks for the social movement at one point in time can become an advantage in future contention, while at other times gains by the social movement actors at one conjuncture can result in disadvantages at another.

The paper focuses on three ways in which social movements influence democratization: These occur first, when social movements defending citizenship rights explicitly demand and gain the extension of one or more of the four elements of democracy (rights and obligations accorded to a large part of the population; their distribution with relative equity among citizens; binding consultation of citizens for public policies and government personnel; and protection from arbitrary action by state agents); ₃ second, when the movements cause an increase of participation in contentious politics; and third, when social movements introduce into the political arena proactive demands challenging the prerogatives of powerholders and their sustaining elites through direct confrontation of a sort seldom advocated by the more institutionalized opposition elites.

If one understands social movements as forms of participation through collective political action undertaken by segments of disadvantaged populations to attain political change by challenging powerholders, the results of the interaction between challengers-powerholders-allies makes searching for specific direct causal effects of social movements on the democratization an illusive task, for causal links are usually blurred and possibly lagged by the dynamics of the process of political contention. Nevertheless, it is possible to examine the impact of the social movement on the democratization process by analyzing: first, the variation in the collective action patterns during the protest cycle of the democratization period as reflecting different conjunctural factors and outcomes; second, the changes in the framing of claims and forms of movement participation as related to changes in social movement and opposition elites' strategies and goals, and authoritarian elites' responses; and third, the extent to which powerholders tolerate challenges in general and how power elites respond to social movement contention at different conjunctures.

Although the idea of establishing clear and distinct causality between aspects of regime transformation and the influence of the social movements seems unattainable, ₄ analyses of cases like Brazil serve to better characterize the complexity of the political process of contention and, in particular, the extent to which Brazilian democratization has been molded by the interventions of social movements. The relationship between popular contention and democracy is dialectical in as much as popular social movements and labor unrest impacted the democratization process through their proliferation and mobilization strategies, forcing powerholders and opposition elites to respond to popular claims through a variety of political strategies aimed at attaining some degree of control of popular protest. ₅

 

The Political Process Approach as We Understand It

Before presenting our analysis of the role of social movements in the transition to democracy in Brazil, first let us present schematically the general conceptual framework of analysis which guided our path through the historical and statistical data for the 13 years in question. The approach which predominates in the literature on social movements has taken political process as its main focus in the analysis of social movements. More recently, several political process scholars have claimed that social movements' collective actions are largely determined by a combination of structural variables which have been synthesized in the term the political opportunity structure. Doug McAdam has usefully synthesized the various definitions of political opportunity structure into five dimensions: (1) expanding access to political power; (2) instability of political alignments supporting the polity; (3) availability of influential elites as allies of social movements; (4) the institutional structures of the state; and (5) facilitation or repression strategies of elites regarding challengers' collective action. ₆ Recently several scholars have called attention to the tautological aspects of this concept as an explanatory variable and called for more precise definitions of the dimensions, since all situations in which social movements arise contain to some degree these dimensions, hence rendering the concept of little utility either theoretically or analytically.

The tautological tendencies in conceptualizing political opportunity structure as an theoretical category in explaining social movements is further demonstrated in the recent study by Kriesi, Koopmans, Dyvendak and Marco G. Giugni (1995), in which the authors define political opportunity structure as having the above five dimensions, but add to these the notion of cleavage structures (structures arising from the social cleavages in a society such as religious, class, and urban/rural cleavages; Kriesi et al., 1995: xv-xvii, 10-25). With the inclusion of social structural cleavages to the definition of political opportunity structure, the already present generality of the concept acquires an even more tautological aspect since the explanation of social movement mobilization can now be attributed to this set of structural factors which range from social cleavages to elite alignments and state institutional structures, indeed, almost every social or political variable commonly used in political analysis. It is noteworthy that one dimension has been systematically excluded from the ever expanding definition of political opportunity structure, the ideological variable, which has generally been understood as lacking the primacy as the structural dimensions.

In light of these broad generalizations in the definitions of political opportunity structure as an explanatory concept and the ambiguity of its meaning, as pointed out by McAdam (1996) and Gamson (1996), I propose an alternative conceptual perspective of the political process approach which redefines the concept of political opportunity and places it within the realm of political conjuncture, thus not a structural determinant as presented in the current literature.

In this alternative, political opportunity is understood as being the combination of structural and conjunctural circumstances which present social movement actors with alternative possibilities of collective action depending on the mobilization strength of their movement organizations, how these actors evaluate the political situation, how they frame their demands, and which strategies they pursue. In this respect, political opportunity is better understood as a historically specific potentiality for action rather than a structural a priori.

This approach also distinuishes between contingencies resulting from social and political structures, on the one hand, and conjunctural circumstances and characteristics of specific arenas of contention, on the other. Here both conjuncture and arenas of contention become relevant political facets in the analysis of social movements since opportunities to act collectively are generated in the contexts of specific combinations of political circumstances called conjunctures, which in turn create differentiated opportunities depending on the arena, the specific field in which social movements act. For example, during any given conjuncture, political opportunities are different in the labor arena as compared to other areas such as housing, public health, nuclear energy, civil rights, or anti-corruption. Not only do relevant institutional structures vary from one arena to another, but so do the interests and involvement of different segments of government and opposition elites, as do, similarly, the availability of resources and the organizational preparedness of the respective social movements. Thus the arenas of contention are the concretization of structural and conjunctural factors as they influence the possibilities of social movement mobilizations. By including these two mediating concepts, political conjuncture and arenas of contention in our analytical framework, I seek to avoid the tautology and vagueness found in the definitions and uses of the concept of political opportunity structure.

With its greater analytical emphasis on the notion of political conjuncture and the characteristics of the arenas of contention this approach also differs from the main currents which emphasize political opportunity structure, in as much as greater importance is attributed to the interactive aspects of the political actors as the primary determining factors in mobilization and contentious politics. While political opportunity is a potential to act collectively, the realization of collective action depends on social movement dimensions and collective actors' behavior in light of alternative opportunities.

Graph 2 illustrates this understanding of the political process approach and the main analytical categories used in bringing together the necessary data to explain the variation in protest mobilizations represented by the chronologies of labor unrest and social movement collective actions, depicted in Graph 3. This approach is structured around three broad concepts: the political structure, the political conjuncture, and the contention contexts or arenas. The framework conceives political process as interactive and dynamic, proceeding through sequential phases of contention contexts conditioned by historically specific political conjunctures, and resulting in outcomes that affect all the actors and redimension the conditions of subsequent political conjunctures.

In this approach political opportunities for social movements to act are not only a function of the specificities of the dimensions of political structure as these assume historical content in concrete political conjunctures; opportunities also are determined by the actions of authorities, elites and the social movements themselves. While political structure sets the parameters of conjunctures, actions and outcomes are determined by the contentious interaction of actors employing their resources in political arenas to promote their interests.

Consequently, the relations between the social movements and their allies, on the one hand, and their adversaries on the other, form a key facet of the analysis. These relations are critical aspects in relating political opportunities to collective action and outcomes, in as much as elite-social movement relations determine in part: 1) the array of opportunities available to the social movements in a given conjuncture; 2) which of these opportunities can best be used to the advantage of the social movement; 3) how opportunities are perceived and framed in terms of costs and benefits; 4) how mobilization strategy is structured to profit from the opportunities; and 5 what outcomes are expected by the social movements in taking advantage of specific political opportunities.

As in the case of the redemocratization of Brazil, elite competition in struggles for regime change is an important facet of the political process, and the relations of the social movements to both situational and opposition elites are an important aspect in understanding fluctuations in popular mobilization. Consequently, we have defined these relations, for analytical purposes, in a manner that allows for comparative analysis of political conjunctures. The posture of adversary elites is conceptualized as a three point scale consisting of the following types of behavior: cooptational, negotiational, or confrontational. The posture of allied elites can be classified along another three point scale: assimilative, cooperative or competitive. ₇ The intersection of these two typologial scales is illustrated in Graph 4 which depicts the variability of elite postures that social movements face in political conjunctures. This graphic depiction permits us to map the shifts in elite-social movement relations over the period of democratic transition in order to understand, at least partially, variations of elite behavior in collective action.

The political process approach adopted here posits that protest and its outcomes are part of a broader process of political bargaining, ₈ and not simply the result of elements of the political structure or of characteristics of elite actors. Outcomes, as understood here, are the consequence of the process of contention and bargaining, and as such are a function of each actor's capacity to mobilize resources relative to other actors, within the historical parameters set by political structures as concretized in specific political conjunctures. Thus analyses of movement success or failure in achieving their demands should first focus on the broader outcomes of interactions that play out in contention arenas where the social movement encounters adversaries and allies.

Outcomes are closely related to social movement goals. All social movements pursue their goals, some explicitly stated and others more implicit. Yet ascertaining the content of these goals is problematic since social movements have many participants, with different levels of commitment (militants, activists and followers), and each participant has specific views of what are the movement's goals. In addition, movements often involve several social movement organizations, which may also have their own versions of goals. Finally, since contention is dynamic, social movements' goals change over the course of a movement's political engagement, a result of confrontations and bargains with adversaries and allies. Thus a working definition of social movement goals should not only include claims and demands articulated by social movement organizations, but also the changes in the content of these claims and demands as an outcome of the political process.

While many of the studies of social movement outcomes have focused on the question of whether or not social movements attain their stated goals, this approach ignores the array of outcomes which are fundamental in any understanding of the role of social movements in the political process and ascertaining their historical significance. By defining outcomes only in terms of policy reforms congruent with demands, outcomes are overlooked that point to structural transformations and changes in the capacity of social movements to remain key political actors after initial policy gains (or concessions). A broader conception of outcomes is necessary in order to enable the analyst to examine the consequences of contention over a protracted period in which there are gains and setbacks, and stated demands are only achieved after years of contention.

We have devised a classification scheme of outcomes which allow us to analyze outcomes in this broader sense since our task has been to analyze contention over a 13 year period of the re-democratization process in Brazil. Our typology, presented in Graph 5, consists of three categories of outcomes: 1) Social Movement Outcomes: those outcomes that impact the internal dynamics of the social movements; 2) Policy/Legislation Outcomes: those with impacts related to policy or legislativve changes; ₉ and 3) Structural Outcomes: those related to changes in the political structure. In all contentious actions, we assume that there will always be impacts on the social movement since contention implies not only social movement extra-institutional activities to achieve their claims but also responses on the part of adversaries and allies to social movement mobilizations. Less certain are impacts on matters related to policy, legislation or implementation, on the one hand, or impacts on the political structure, on the other. Impacts on implementation, policy and legislation would seem to require less force on the part of social movements than impact on political structure. It is for this reason that our scaled typology of outcomes takes into account the notion that outcomes vary according to the strength of social movement intervention. The outcomes typology permits us analytically to summarize data based on the consequences of contention at different moments of protest activity.

 

Data on Collective Protest in Brazil : A Methodological Note

Many factors contribute adversely to the accurate reporting of contentious action events. In Brazil, as in many other countries which have experienced recent prolonged periods of dictatorial rule (often claiming to defend democracy), data on the collective action of social movements are problematic because of the contradiction between the autocratic nature of the state and its pretense of democratic participation and legality. Within this context, newspaper reporting has been one of the few sources for obtaining data on collective protest indicative of general trends in contention. This source admittedly has major shortcomings precisely because of the regime's authoritarianism. Reports of social movement collective actions are influenced by: (1) government censorship and self censorship by the press; (2) the ideologically determined editorial policies of newspapers; (3) journalists' biased view of elite actions as politically more important than popular activities; (4) limited geographic coverage of newspaper reporting (which influences variation in factors 1 through 3, on the regional newpaper bureau level; and (5) the difficuties in covering collective action in times of government repression, when the newsworthiness of events is difficult for reporters to judge in advance, yet they occur in very short time spans in relation to newspaper deadlines. Authoritarian rule also affects government collection of these types of data, again making them both problematic and certainly not readily available for public consultation. Hence government-sponsored data collection necessarily more accurate than newspaper accounts, except in terms of completeness in the number of events recorded, because of the sensitivity of authoritarian regimes about broad circulation of reports of popular discontent. Most likely the better data, at least in terms of number of events, is gathered by the intelligence and national security agencies of these regimes, but these are precisely the agencies that have most to hide from public scrutiny since frequently their data collection activities are closely coupled to infractions of the law, especially under dictatorships. This makes public access to these data most difficult even after the end of dictatorial regimes.

Further, data collected from less ideologically and politically compromised government sources, like the regular police, labor ministry organs and the like, are not flawless in comparison to newspapers accounts, except for their broader coverage of the number of events. This is generally the case because even under dictatorial regimes, the antagonisms between intelligence agencies and civilian bureaucracies over their roles in repression contribute to the problematic aspect of data collected by civilian bureaucracies. Because collective contention is illegal under such political conditions (even in non-dictatorial regimes much contention is extra-institutional) participants likewise have little interest in having many their actions noticed by the government's security agencies. In this case newspapers, militants in collective protests count on public visibility through press reporting as a tactic for impacting their targets. Thus social movement activists are more likely to inform reporters about planned collective actions and offer details about occurrences at these events, especially in times of repression; however, self reporting of facts by activists also raises questions as to accuracy.

Much of the doubts about newspaper reports of popular social movement actions apply as well to strike reports, even though unions are generally legal in modern authoritiarian regimes. Ministries of labor are often better equiped to record and report labor unrest, but they may not make these data available for public use. Further, since labor unrest is closely associated to questions of government economic policy, prospects of growth and development, and investment risks, government reporting of labor unrest is often plagued with political considerations. Unrest reflects badly on govermental elites, especially with regard to foreign and national investors, and multilateral funding agencies. Though data collection may be quite complete, data reporting is often fragmented because of these considerations. In this context, newspaper reporting has an edge over goverment in as much as an important segment of newspaper readership, like investors, bankers and foreign representatives, want information as to labor unrest. Consequently, we suggest that newpaper reporting is more likely to be biased in favor of over reporting incidences of labor unrest as compared to those of popular social movements. But this bias does not necessarily have to be detrimental to reporting social movement action in periods of conflict over regime transformation. In this case, increased worker opposition to autocratic rule results in a corresponding increase in reports of events, which in turn causes greater newspaper attention to other forms of anti-regime activities, even from the popular sectors. Thus, in times of major regime changes, newspaper reporting may become one of the most valuable sources of documented information about contention, especially if contention takes the form of collective action by labor and popular social movements.

The question of the suitability for social inquiry of newspaper accounts about collective action during periods of broad political changes, may ultimately be reduced to nothing more than the question of availability of data. If these are the only data available, and many times they are, then the researcher is faced with the old Brazilian adage "quem não têm cão, caça com gato" (he who doesn't have a dog, hunts with a cat). Hence we proceed below on the trail of political transformation with the cat on our theoretical leash.

The quantitative data on collective action used in this study are derived from several sources. Data on the collective protest of urban popular movements were derived from newspaper accounts collected by a national Brazilian non-governmental organization, IBASE, which has gathered this information since 1983. The quantitative data cover all types of extra-institutional forms of action in the following arenas of contention: housing, education, health, food prices, public transportation and human rights. They do not include mass mobilization campaigns organized by the political parties in which social movements had a major role in their realization, but were not authors of the actions. These data, in spite of the drawbacks suggested by the above discussion, seems accurately to reflect the general trajectory of popular contention between 1977 and 1989, judging from other accounts of specific years and comparing the pattern of the yearly distribution of events indicated by our data and the pattern of the distribution of mobilization documents (used as an indicator of social movement organizations' activity) from a recent study. ₁₀ Even though the two types of data point to different, but closely related phenomena, the correspondence of the trends suggests that our data serve the purpose intended.

The quantitative strike data are those analyzed in my study Social Change and Labor Unrest in Brazil Since 1945. ₁₁ They were taken from several sources. For the years between 1978 and 1882 they were compiled from reports in the Folha de São Paulo, one of the country's major newspapers. For the years between 1983 and 1990 most strike data came from two institutional sources. The first, the Nucleo de Estudos de Politicas Publicas of the University of Campinas, compiled data from reports from the DIEESE, Departamento Intersindical de Estatistica e Estudos Socio-Economicos, a union based research institute which began to gather strike data systematically in 1983. The second source is the Ministry of Labor's Sistema de Greves/SIGREVE which began in 1985 to compile strike statistics from reports of the regional offices of the labor ministry.

 

Brazil 1977-1989: Democratic Transition and Consolidation

Looking at the period between 1977 and 1989 from the perspective of popular sector collective action, this period of re-democratization can be subdivided into phases according to the distinctive timing of collective mobilizations of the labor movement and the urban popular movements. As shown in Graph 6, the differences in the fluctuations in the mobilizations of labor and popular social movements produce a protest cycle characterized by four phases of popular protest, each having a specific pattern of mobilization/demobilization for each form of popular protest. The first protest phase is between 1977-1980; the second, between 1981-1984; the third, between 1985-1987; and the fourth phase, between 1988-1989. Each phase displays different dynamics of the political process as authoritarian elites, opposition elites and the popular social actors interacted in different political conjunctures.

Phase 1: 1977 to 1980 - Sectorial Contention for Democratization

The revival of collective protest action against the continuation of the military regime began in 1977. The early part of the year was characterized by signs of "political decompression" under the more moderate military presidency of Ernesto Giesel, although the forces of repression continued to play a major role in political affairs. A mass ecumenical religious demonstration opened this phase, led by the Catholic Church and the Sao Paulo Jewish Rabbinate in memory of Valdimir Herzog, a journalist killed by military intelligence while under interrogation, and the subsequent murder of a leading labor leader under similar circumstances. Other acts of opposition emerged (characterized by varying sectorial situations depending on how repression affected their specific category) as different social groups mobilized to express their opposition to authoritarian rule.

Under the close scrutiny of the armed forces, organized and collective opposition first appeared in the relatively guarded environments of meetings of professional associations or religious ceremonies under the auspices of the church. At its annual congress, for example, the national press association protested the Herzog killing and continued severe press censorship, and issued a call for the end of press censorship and political safeguards for journalists. In a similar manner, on the occasion of the 150th anniversary celebrations of the founding of the first law faculty in the city of Sao Paulo, 200 of the most prominent Brazilian jurists issued a manifesto calling for the end of arbitrary authoritarian rule, a return to constitutional democracy and immediate reinstatement of habeas corpus. However, the Brazilian Society for the Progress of Science (SBPC), the country's oldest scientific association, faced government obstruction of its 37th annual meeting as the government sought to avoid anti-government criticism by the university community. When the Ministry of Education succeeded in forbidding the Federal University of Ceara to host the 1977 congress, the Catholic University of Sao Paulo, with the support of the city's Archbishop, sponsored the congress in clear defiance of the regime. Two months later, in reprisal, Sao Paulo military police conducted a raid on the Catholic University campus, which caused considerable damage to buildings, student offices, the university library and class rooms, as well as serious physical injury to many students attending classes at the time. The raid was justified as an attempt to arrest student leaders present at a clandestine meeting aimed at re-organizing the National Student Union, abolished by the government after the 1964 military coup.

1977 also marks the resurgence of public demonstrations as university students in the major state capitals and the Federal District took to the streets to demand educational reforms, more funding for education, university autonomy, an end to government intervention in the universities and the direct election of all university officials. Since the students dared to leave the confines of the university auditoriums to participate in street demonstrations, the government was swift to repress these actions through police force and subsequent summary expulsions of several university students leaders under the Law 477 (which allowed the Minister of Education to banish--not merely expel--students from the university for political reasons). Further, some faculty members were dismissed or forced to retire on account of their support of student protest.

There was clearly no facilitation on the part of the state for this reemergence of protest activity in 1977. While President Geisel was believed to belong to the moderate wing of the armed forces, his commitment to the regime was unwavering. Using dictatorial powers provided by junta decrees, Geisel closed the national congress and created presidentially-appointed senators from each state as a preventive measure against the loss of the congressional majority, since earlier elections had indicated growing support among voters for the only opposition party, the Movimento Democratico Brasileiro (MDB). Geisel also used his dictatorial powers to abrogate the political rights of the Federal Deputy Alencar Castro for a speech which the military commanders deemed offensive to the armed forces.

In the same year, the International Monetary Fund made public a report that the Brazilian government's calculations of the inflation rates for the previous years had been significantly falsified; indeed, the government acknowledged the report as fact.

Early in 1978, an unprecedented decision by a Sao Paulo federal judge declared the military government responsible for the "suicide" of Validmir Herzog and ordered the state to pay damages to the journalist's family. This was the first time that the judiciary imposed accountability on the military regime for its security activities.

A second round of mass protest followed from an unexpected quarter, the metal, automotive and chemical workers of the industrial belt of Sao Paulo. Organized in huge unions with hundreds of thousands of members, and unhindered by draconian legislation which obstructed the possibility of legal strikes since all labor disputes were adjudicated by labor courts presided over by government appointed judges, the union leaders threatened to strike. They demanded higher wages to compensate for the previous year's inflation and for lost wages resulting from the falsification of the inflation rates by the finance minister. The labor judge who was assigned to adjudicate the dispute and set the wage adjustment rate, declared himself unqualified to rule on the matter since the existing labor and wage legislation did not contemplate the question of government falsification of inflation indices by a federal minister, a matter for other government authorities.

This unexpected crack in the autocratic regulatory apparatus was immediately perceived by labor militants as an opportunity to confront employers and the state within the authoritarian rules of the regime. Concerned that the legality of a strike would be challenged by the regime, and aware of the strong possibility of government repression of strikers, union leaders acted as "mediators" as shop floor activists closed down the largest industrial enterprises of the country. These strikers demanded immediate restitution of their lost wages resulting from the inflation rate falsification. The government response was to sent police into the factories to arrest strikers and break the will of the workers. Within 30 days the first strike in ten years was over, as strikers returned to work without having achieved their demands. Throughout the rest of 1978, other regional unions in the region struck, following the example of the ABC metal workers of the metropolitan area of Sao Paulo and making similar demands. ₁₂ The main outcome of these strikes was to thrust progressive labor leaders from Sao Paulo, Rio de Janeiro and Minas Gerais into the national political arena, and force opposition elites--who were already attempting to gain democratizing concessions from the military government--to consider the labor movement as potential allies.

In 1979, labor succeeded in bringing to the foreground not only the organized working-class but also the popular sectors as key actors in the re-democratization process. After the experience of the 1978 strikes, union leaders began to prepare for the coming confrontation with the state at the time of wage readjustments in early 1979. Certain that further strike actions would bring government intervention, the possible arrests of union leaders and repression of strikers, unions, in collaboration with Catholic community leaders, began to prepare their respective organizations to resist the regime.

A strike began in April, 1979, and the labor court promptly held that the strike was legal and the wage demands justified, given the fraudulent manipulation of the inflation index. At this time strikers were also demanding the right to negotiate any work agreement directly with management. Workers in the surrounding cities struck as well, and by mid-April the largest industrial region of Brazil was paralyzed by the striking metal workers. The government quickly forced the labor judge to reverse his decision, declare the strike illegal, and order state intervention. Union headquarters were occupied by Labor Ministry officials, the police prevented workers from meeting in the local stadiums, and curfews were imposed. At first, labor leaders were not arrested, in hopes that they would call off the strike.

The unions were able to resist these initial repressive measures, as church based neighborhood organizations mobilized their members to collect food and strike funds, obstruct police patrols at union headquarters by women and children, and open churches and parish halls to workers for union and strike meetings. In response to this popular sector and church support, the government ordered the army into all the affected cities. Soldiers were posted in front of churches to prevent workers from meeting, and working-class neighborhoods and factory districts were patrolled by millitary forces on the ground and in the air. Union leaders were indicted under the National Security Law for subversion, arrested and held incommunicado for interrogation. By May, other workers had replaced the imprisoned leaders as strike leaders, and together with other sectors opposing the regime, organized mass demonstrations of workers, students, university professors, clergy, jurists and journalists to demand an end to the repression and a negotiated contract. After 60 days, industrialists began to pressure the government authorities to allow for a harmonious return to work since they were fearful that worker discontent could endanger factory premises after the strike. In response, union leaders were released from prison on the condition that workers return peacefully to work.

Workers returned their jobs, their union still vulnerable and their leaders threatened with conviction under the subversion laws; none of their demands had been met except the expected wage increase provided by law to compensate for the inflation of the previous year. Yet the strikes and the events that accompanied them achieved several outcomes: they forced upon the traditional opposition elites (politicians, church hierarchy, intellectuals and entrepreneurs) new political actors (labor and social movements) who henceforth would have to be consulted in the negotiations for redemocratization; they altered the political agenda under negotiation to include demands of shortening the timetable for the return to democratic rule and an end to the authoritarian provisions in the labor legislation; and they brought about reluctant acceptance by traditional opposition elites of the use of direct collective action by labor and social movement organizations as a major tactic for pressing for change.

The year-to-year variation in the national patterns of strikes clearly reflects these political factors that characterized the transition from dictatorship to civilian rule. ₁₃ Here we will be observing strikes along three dimensions which vary independently of one another: total strikes per year (strike rates), average strike size, and duration of strikes. The coefficient of overall strike strength is the product of strike rates times duration times size.

Whether we consider the number of work stoppages or the average size and duration of strikes, there was a gradual decline in the volume of strike activity between 1978 and 1984. In fact, two patterns of strike activity can be identified in the years of the last two military governments. In the first pattern, characterizing the years 1978-1981, there were few strikes, due to governmental controls and the reduced mobilizational capacity of most worker organizations under the military government. Strikes in these years typically involved large numbers of workers per strike from among the largest and better organized occupational categories typically employed in the larger enterprises located in the major industrial centers of the country. These work stoppages were marked by a longer average duration which reflects not only the capacity of these workers to mobilize but also to resist the intransigence of employers and the pressures of the autocratic authorities. This pattern has been characteristic of strike activity in periods of greater governmental intolerance for working-class mobilization, as depicted by the narrow spikes of strike activity illustrated in Graph 7 for the years 1978 through 1981; it is also similar to patterns found in previous periods of increased government repression of labor protest.

Over these three years, the annual pattern of strikes did not alter their basic form but gradually decreased in volume. This decreased volume represents a decline in the strength of the strike actions under the last military government as repression persisted, and workers' organizations shifted strategies toward support of concerted action within the broader democratic coalition. Other indicators of this shift were 1) the sharp decline in the number of workers striking each year (the average number of strikers per strike in 1981 was 6107, but by 1984 this rate had fallen to 2,946); and 2) the significant yearly decrease of the average strike duration from 8.8 days in 1981 to 3.9 days in 1984.

At first glance the decrease in the average size of the strike could be attributed, in part, to the technical fact of an increased number of strikes in 1983 and 1984, coupled with increased governmental repression of strike activity. A more detailed look at the interactions with the opposition as they responded to military intervention, however, reveals shifts in activities aimed at strengthening mobilization structures and broadening their following within the working-class. The decline in labor protest was not only a consequence of the conditions in the political environment, but also the result of the labor movement's own choices in strategy as they joined the broader opposition coalition.

Urban popular movement protest was much less frequent, as labor unrest dominated the latter part of the 1977-1981 phase. Since strike activity galvanized resistence to the regime at this time and the urban social movements participated actively in strike support activities, there was little collective action by the social movements, numbering on average less than 10 events per year, focused primarily on specific demands of direct benefit to their constituencies. But the lack of any significant popular mobilization in this period did not mean that the urban social movements were unimportant actors on the political scene. Their logistical support and collective action in support of the 1979 and 1980 strikes proved their capacity to mobilize large contingents of the population in confrontational activities, while demonstrating their willingness to join coalition protest against the regime. Thus the period between 1977 to 1980 was a precursor of later more intense engagement of the labor and urban popular movements in the struggle for democratization.

Phase 2: 1981 to 1984 - Coalition and Sectoral Contention for Democratization

After the experiences of the strikes of 1977 to 1981, labor and social movement leaders decided that to achieve a return to democratic rule while at the same time consolidating the position of the labor and urban popular movements required the creation of national political organizations capable of adequately representing their interests. Labor leaders undertook discussions among themselves leading to the creation of a national labor central. The formation of a coalition of labor union leaders with a broad ideological outlook made strike actions by more militant unions an issue of much debate in the national labor arena since these actions alienated moderate union leaders and more liberal entrepreneurs who might be won over to the struggle for democracy. In addition, continued strikes threatened the consolidation of a national labor movement with government intervention, since national labor organizations were banned under the military regime. As a result, more progressive labor leaders began to set aside sectoral forms of opposition politics and move toward coalition collective action.

During the years since 1960, there had been a significant expansion of unions, many of which increased in militancy in the early 1980s and contributed to the emergence of a genuinely national labor movement. ₁₄ This evolution in working-class organization is further demonstrated by the rise in 1981-83 of the first national labor organizations with a broad bases in the unions. The early steps in forming such an organization took place in 1981 in a national conference of the working classes (Conferência Nacional das Classes Trabalhadoras--CONCLAT), which in 1983 split over ideological cleavages to form two major national confederations--the Central Unica dos Trabalhadores (CUT) and the Central Geral dos Trabalhadores (CGT). These were founded in the wake of a much contested general strike calling for democratization, labor law reform, union autonomy, and consultation over government economic policy with the International Monetary Fund. ₁₅ Although the general strike, the first since 1963, brought out approximately 2,000,000 workers, thus demonstrating progressive labor organizations' strength, its organizational outcome was to deepen the split between progressives and moderates within the labor movement. Despite the ideological cleavage within the labor movement, however, a consensus persisted that the struggle for democratization should make common cause uniting the various sectors of opposition.

Around this time, labor and popular movement leaders, along with groups of the progressive clergy, intellectuals, and opposition politicians began laying the foundations for the creation of a working-class based political party, taking advantage of a government attempt to dismember the only opposition party, MDB, by dissolving both legalized parties, thus permitting the formation of other parties. While opposition politicians protested the party reform bill, realizing that it was aimed at disrupting a unified political opposition, labor and popular movement leaders saw the government measure as an opportunity to form an more ideologically committed workers' party. The formation of a working-class-based opposition political party in 1982 required systematic coordination between labor and popular organizations, while avoiding a major fissure in the broader opposition alliance. Henceforward, collective action would have to be calculated within the broader strategy of achieving redemocratization--and contingent upon the commitment from traditional elites that they would abide by demands for far reaching change in the political structure, not merely a return to electoral politics. Consequently, the second phase of contention was characterized by increased mobilization, not only of labor organizations advocating sectoral interests, but of other social movements, as opposition elites were obliged to tolerate direct action from the more proactive labor and popular organizations.

In spite of the failure to sustain a unified national labor organization, by 1983 labor unrest ressumed its upward climb, as strikes almost tripled in number in comparison to the previous five years. The 1983 annual strike rate rose to 1.86 from .74 the previous year, and the number of strikes yearly increased for the rest of the decade. ₁₆ Strike demands were directed not only toward work place issues, especially wages and inflation, but more importantly around demands for union autonomy, redemocratization and consultation.

1983 marked the beginning of an important change in the trajectory of strike activity which continued through the decade. Even though the overall strength of strikes continued to decline (reflecting the smaller size and shorter duration of strikes), there was three to fourfold increase in the frequency of strikes over that of the 1980-1982 period.

These changes in the strike dimensions represent a second type of strike pattern, characterized by small, frequent strikes in contrast to the larger and longer ones common in more autocratic periods. The change reflected as well a geographic and sectoral expansion of strike activity, as newly involved segments of the working class joined those already active in previous years. These tendencies were repeated in 1984, when there was a continuing decline in overall strike strength but an increased number of smaller and shorter strikes.

Taking the combination of the three dimensions of strike actions as an indicator of the strike strength for each year, we find that between the previous phase and this one, strike strength declined from a coefficient of 49.34 in 1980 to 29.18 in 1984. ₁₇ This represents a decrease in the intensity of strikes despite the fact that there was indeed an important increase in the number of work stoppages; it most likely reflected the introduction of new workers and their unions to the experience of labor mobilization. Between 1981 and 1984, this modification in the mix of the strike dimensions and the overall decline of strike strength became characteristic of the Figueiredo years, when the general strike set the tone for labor militancy oriented toward a united front strategy against the military regime.

The trajectory of popular protest in this second phase followed an opposite tendency from that of strikes, with a very significant increase in the mobilization and collective action of urban social movements as they contributed their resources to the democratic coalition. 1983 marks a resurgence of social movement militancy with an important rise in both average numbers participating and number of collective action events. This differed from the period 1977-1981 when such events were very few in number, and labor protest dominated political collective action. In the previous 1977-1981 phase, social movement protests could be counted in the tens, but by 1983 these had risen to 305, with an average participation of 654 persons. Unlike the previous period, when there were more generalized sectoral demands for democratization, the mobilization of this second phase (1982-1984) made more specific sectoral demands (food prices, housing, transportation, education and health) but--more importantly--framed these demands around democratization as a means to achieve them and claimed these interests as part of a struggle in defense of human rights. Thus we find social movement organizations, like their labor counterparts, adopting a coalition strategy of opposition to the military regime.

In 1984, there was a significant decline in the collective action events of social movement organizations, but interestingly, a doubling in the average number of participants in each event. This development can be attributed to the fact that in forming their opposition coalition in that year the various sectors had resolved to press the military to transfer the power of the state to civilian control. As social movement organizations became more involved in coalition-based mobilization, there was a corresponding decline in the number of popular sectoral actions.

The process of negotiation with the military was marked that year by the call for the continuing mass mobilization by progressive segments of the opposition alliance in favor of the ratification of the Oliveira constitutional amendment, which would restore direct elections for president in that fall's election. This nationwide campaign, called the Diretas Jà, consisted of mass demonstrations held in the major urban centers coordinated by the coalition of opposition parties, labor unions, and social movement organizations advocating the end of military rule. The opposition coalition had launched the Diretas Jà campaign in June, 1983, with a street demonstration in Goiania, and these continued until April 1984, when the congress rejected the amendment. ₁₈ The importance of the social movements is evident in the unexpected size of the turnout at each demonstration: for example, Goiania, two hundred thousand; Rio de Janeiro, five hundred thousand; Belo Horizonte, three hundred thousand; Porto Alegre, two hundred thousand; and Sao Paulo, one million people attending. ₁₉

The magnitude of the Diretas Jà demonstrations contributed to the reduction of other collective action events. Organizations shifted collective action from individual sectoral actions to participation in the broader demonstrations in support of the Oliveira Amendment. Even though the opposition coalition failed to bring about passage of the constitutional amendment, it succeeded in consolidating massive mobilizations against the resistance of the regime.

After this failure to achieve re-democratization through constitutional reform, the united opposition faced a presidential election at year's end under the rules of the military regime. The democratic coalition succeeded in hammering out a consensus around the opposition candidacy of Tancredo Neves, a traditional opposition congressman with a progressive past, and a running mate who had bolted from the faction supporting the military to run against the official candidate. With the support of popular social movements, labor organizations, the church, and professional associations, and the reluctant support of the more moderate opposition politicians the Tancredo campaign mocked the indirect presidential election process. It repeated the mass mobilization tactics of the Diretas Jà campaign, staging rallies in the state capitals which in turnout approximated the earlier demonstrations.

On the eve of the convening of the electoral college, rumors spread that hardline military factions would stage another coup if the opposition won, leading Trancredo Neves and his collaborators to make plans of escape from Brasilia to their home states to resist. On January 15, 1985, the day of the presidential election, the Federal District was militarized to prevent any demonstrations pressuring the congress, and the armed forces were placed on alert throughout the country. The government's candidate received 180 votes, while the opposition obtained a total of 446--280 from delegates of the opposition parties, and 166 votes from defecting government delegates. In light of the overwhelming victory of the opposition, the already evident mobilization capacity of labor and the social movements, and the readiness of Tancredo Neves and his allies to resist any repressive action, the military establishment conceded defeat. Thus came to an end 20 years of military rule. ₂₀

Phase 3: 1985 to 1987 - Sectoral Contention and Demobilization in Transitional Politics

The victory of the opposition alliance in electing Tancredo Neves to the presidency was not long lived. On the eve of his inauguration, Tancredo--who was suffering from a painful abdominal disorder--was admitted to Brasilia's main public hospital for surgery. As Tancredo Neves' illness dragged on, however, concern for the president-elect's survival and the prospects of democratization mounted. With the president-elect in intensive care at the University of Sao Paulo medical center hospital, rumours about a right-wing conspiracy to eliminate him circulated. The government was paralyzed over the uncertainty of presidential succession throughout the three months which Tancredo's illness lasted. Neither Tancredo nor his vice president, Josˇ Sarney, had been sworn in. Worse still, the vice president-elect had been one of the military regime's loyal supporters over most of his career. Thus the opposition parties, along with labor and social movement leaders, questioned the legitimacy of Sarney taking Tancredo's place in the presidential palace. Sarney assumed the day-to-day operations of the government, declaring that he would not make use of any of the authoritarian legal provisions remaining from the previous regime, and appointed a temporary cabinet composed of prominent opposition politicians. Popular reaction to illness of Tancredo Neves and the crisis in the newly elected civilian government was best illustrated at the president-elect's death, vividly described by Thomas Skidmore.

But attention was still riveted on the daily medical bulletins relayed by Tancredo's press secretary. Could a late-inning miracle somehow save the man meant to lead Brazil into a new democratic age? That miracle was not to be. Tancredo died on April 27, after seven operations and thirty-nine days in the hospital. The public, although increasingly skeptical of the many 'successful' surgeries, was stunned by the news. In life Tancredo had loomed as a political savior. In death he assumed saintly proportions. All the stored-up hopes, centered on the man who had not lived to fulfil them, came spilling out.

Tancredo's last journey began on a fire engine in Sao Paulo, where two million paulistas jammed the elevated superhighways to mourn and accompany his departure. His body traveled first to Brasilia, where 300,000 grief-stricken onlookers watched his coffin ascend the ramp of the presidential palace, a climb he never got to make in life. Tancredo's next stop was Belo Horizonte, his state capital, where a million grieving mineiros surged forward to pay their last respects. His final stop was Sao Joao del Rey, where he was buried in his native soil.

As national television networks transmitted the funeral procession, work stoppages were called throughout the country by labor leaders to allow workers to take part. In Sao Paulo, as in other major industrial centers, unions set up television sets in the workplace and production came to halt on the 27 April, 1985.

The public outpouring of grief which the death of Trancredo Neves provoked was not unlike that which occurred ten years earlier with the death in 1975 of another popular civilian president Jucelino Kubitschek. At that time, Brazil suffered under the yoke of the most repressive period in its history during the government of General Medici. Kubitschek's death in an automobile accident (since he was a leading opposition figure, it was strongly suspected that it had been deliberately staged) stirred public outrage against the regime, manifested in mass participation in the funeral cortege from Rio de Janeiro to Belo Horizonte, his home state. The family rejected all military honors at his funeral and refused to permit the military police to escort the cavalcade, while mourners sang anti-government songs and carried signs associating Kubitschek with democracy. As in the Kubitschek funeral, the popular outpouring of grief over the death of Tancredo Neves publicly displayed widespread support for a return to democratic rule. ₂₁ Thus the beginning of the new civilian government under Tancredo Neves' vice president-elect made Sarney himself, and the rest of the old regime's elite, aware that re-democratization was inevitable, and that the popular social movements remained key actors at this turning point.

In comparing labor and social movement mobilizations in this period, then, we find two distinct patterns. On the one hand, the popular social movements--less well organized and more dispersed--contracted under the intensification of partisan competition of government and non-government elites. As social movement militants sought to achieve sectoral gains through alliances with party leaders and governmental institutions, collective protest declined. On the other hand, the pattern of labor movement activism was characterized by a strong and steady increase in strike activities. Labor movement organizations, stronger than the popular social movements, enjoyed relative autonomy from party loyalty and the capacity for concerted action against government economic policies in favor of specific work place demands.

The first years of the civilian presidency saw major explicit and implicit changes in political structures and in the nature of the political opportunities available for social movements. The political agreements made by the various segments of the opposition coalition in their concerted thrust against the authoritarian regime, and the political-institutional crisis provoked by the death of the president-elect shaped the political conjuncture in which a new government would be formed. The most important political agreement was that substantive structural changes would be formalized exclusively through revision of the national constitution, which was promptly scheduled for 1988. In the meantime, reform and the removal of authoritarian pratices were accomplished on an ad hoc basis, through either legislative action or simply by administrative inaction. Thus, for example, the legally prohibited national labor organizations and the National Student Union became institutionalized within the polity without formal alteration of their legal status.

At all levels of government, civilian party interests dominated government actions: politicians from a broad spectrum of ideological tendencies--aware of the potential of the social movement constituencies--initiated practices, policies and administrative reforms intended to assimilate, negotiate with or, outright coopt movement leaders and activists through the so-called "participatory governments." These variously constituted popular consultative councils, social action programs, and advisory staffs were committed to assimilating social movement leaders. Consequently, the tendency was for popular social movements' collective action to decline significantly in the years immediately following the return of civilian rule.

These changes in the trajectory of social movement mobilization are re popular social movements, enjoyed relative autonomy fromflected by the decrease in the number of 1985 collective action events to 34, as compared to the 305 events registered in 1983. Although the number of events increased in 1987, the average number of participants per event remained far below the levels of 1983-1984. Along with the demobilizing effect of political party competition, the resurgence of union autonomy and progressive labor political action had its effects on social movement demobilization as dedicated militants from grassroots organizations turned to union militancy. The national political organizations (such as the Partido dos Trabalhadores and the national labor central) which the social movements had helped to create in the early 1980s likewise recruited many of their militants from among social movement activists. ₂₂ At the same time, the Catholic Church which had been a mainstay of popular social movements during the harsh years of military rule now faced internal conservative opposition to the progressive clergy's support of popular collective action forcing these actors to reduce their role in organizing demonstrations. The outcome of changes in the resources of social movements and political opportunities was that social movement collective action declined substantially between 1985 and 1987.

In contrast, labor activisim took an upward turn with the end of the military government, making impressive gains in mobilization in Phase 3 (1985-1987) and Phase 4 (1988-1989). In 1989, the final year of the Sarney government of President Sarney, there were 3,164 strikes as compared to the 144 strikes in 1980.

During the 1984 transitional phase, the revival of strike activity revealed important changes in the strength and form of working-class protest, often motivated by the effects of the chronic economic instability and the desire for democratization of union life which dominated the concerns of labor leaders at this time. Remember, the number of strikes had remained relatively small during the last years of military government, when the number of strike actions rarely rose above the 1978-1979 levels. This is best understood as the result of the mixed effects of government repression and shifts in opposition coalition strategies in which the labor movement was an active participant.

Yet by Phase 3 (beginning in 1985), the increased strike activity can be attributed to the process of democratic transition under the Nova Republica, ₂₃ when there was a greater tolerance of worker protest. In spite of the low level of unionization--estimated at around 20 percent between 1986 and 1988-- strike activity grew steadily, aimed against rising inflation/government economic policies and aided by the unions' increased mobilizational capacity. ₂₄

In fact, the resurgence of labor militancy occurred only in 1985, the first year of the civilian government, when all three dimensions of strike activity increased markedly, as shown in Graph 7. (Strikes not only became more frequent as the strike rate rose from to 2.54 to 3.28 over the period 1984-85, but strike duration also increased from 3.9 days to 6.3 days, and the average strike size rose to almost four times that of the previous year, from 2,946 to 11,016. The significance of these changes in strike activity can best be visualized by comparing the annual strike strenth coefficient--which combines the three dimensions in one measure (see Graph 7). 1985 was the year in which strike intensity surpassed all years since 1978; the strike strength coefficient then equaled 227.63, as compared to 49.34 in 1980 and 29.18 in 1984. Thus 1985 can be defined as a strike wave year. ₂₅

After the 1985 strike wave, work stoppages again increased in frequency, but decreased in both average size and duration. Hence, in comparison to the previous year, the strength of strike activity in 1986 declined. One factor that likely contributed to the decline in strike activity was the reduction of the inflation rate from 235.11 percent in 1985 to 65.03 percent in 1986 was the introduction of a new economic program, Plano Cruzado. Although strike strength on average was lower than in 1985, strike activity accelerated through the year, as inflation rose. Since the strike strength coefficients are annual averages for the year, changes within the period are not readily apparent. In this respect, it is worth noting that the decline in overall strike strength in 1986 was merely a prelude to greater strike activity in 1987.

The second major strike wave in the decade came in 1987. The annual strike rate more than doubled from the previous year and reached new historic heights. Similarly, strike duration rose again to an average of 8.2 days, reminiscent of the 1978-1980 period, as employers became more resistant to negotiation. The average number of participants per strike however was slightly lower than that in 1985, suggesting that more workers in smaller establishments were entering the ranks of strikers protesting the then record annual inflation rate of 415.83 percent. This marked increase in strike activity is summed up in the 160 per cent rise in the 1987 strike strength coefficient over the 1985 strike wave year.

The recurrence of a strike wave two years after the 1985 wave reflects workers' increasing capacity for mobilization. More workers were becoming involved in labor disputes motivated by the uncontrolled inflation spiral; they demonstrated their increased resistance to employers by holding out for longer periods. The 1987 strike wave marked a new trend in strike activity, for there was a significant change in the mix of strike dimensions from the previous years. From this year until the end of the decade, yearly strike activity was characterized by a larger number of strike events with a lower average number of participants but a longer duration, as opposed to the prior pattern in which there were a larger average number of participants per strike and proportionately fewer work stoppages.

While the tall spike was characteristic of strike activity in years of greater government intolerance lower frequency, broader based strike activity has been more typical of periods in which there was less government interference and repression. In the latter periods, greater tolerance permitted the spread of strikes among a wider range of worker types and establishment sizes. The strike wave of 1987 not only represented an upsurge of labor protest against deteriorating economic conditions, but also marked a new phase in the labor movement in which work stoppages reached new levels in scope and strength.

In contrast, 1988 saw less intense strike activity and a reduction in overall strike strength, as generally happens in the years following a strike waves. Nevertheless, the 1988 strike strength coefficient was larger than that in average years, larger even than that of the 1985 strike wave. In this respect, we find that each strike wave ushered in successively higher levels of strike activity throughout the decade, reflecting the increasing capacity of unions to engage workers in collective action.

Inflation declined in 1986, worker strike activity modulated to a non-strike wave level; social movements, however, rose again to the 1984 level. 1987, the second strike wave year, marked labor relations reaching a higher level of strike strength than any previous strike waves. Inflation took a significant upward turn about then, and the rate of social movement protest increased above the 1984, 1985 and 1986 levels, but not the even higher level of 1983.

The democratic transition phase brought a return to sectoral politics by both labor and social movement organizations, with new priority placed on specific sectoral interests as opposed to the coalitional politics of the previous phase. The political conjuncture and the attendant new opportunities were molded by the major structural change begun in the course of democratic transition. This change included, first, the recomposition of the governmental elite, as military and civilian politicians from the dictatorship were eased out of government and replaced by members of the opposition; second, the reformulation of authoritarian laws; third, the placement of both civil and military security and intelligence organs under civilian surveillance, and access to government significantly increased through an intensification of multi-party competition; and fourth, the new openness of high government officials to negotiation with and assimilation, or even cooptation of social movement leaders and activists.

Nevertheless, despite the increased opportunities for mobilization, many popular movement organizations opted for collaboration with governmental and party elites, which resulted in a major demobilization. In contrast, the labor movement chose more confrontational strategies in spite of its greater access to government elites. This difference in the patterns of collective action reflects the specific aspects of the political structure and conjuncture encountered by each type of social movement, and how respective social movement actors evaluated their strengths and weaknesses in relation to the available political opportunities. ₂₆

The contraction of contentious collective action did not mean that popular movement activists had abandoned developing new organizational resources, however. For example, the ANAMPOS - Artciulação Nacional de Movimentos Populares e Sindiciais, an organization intended to unite the labor and popular movements in a single national association, was formed. The project failed, as labor's national confederations gained force, but the popular movements were unable to match the progress of their labor counterparts. tNevertheless, one lasting result of this attempt was the Instituto Cajamar, an organization dedicated to social movement leadership training with close ties to the Workers' Party and the CUT. ₂₇

Popular movement organizations were more vulnerable to external political party influences and partisan factionalism resulting from increased interelite competition, much less capable of sustained mobilization. They commanded fewer and less stable resources than did labor, pursued more generic, less-sharply-focused goals, and faced an often diffuse and ill-defined adversary. Consequently, popular movement leaders saw collaboration with sympathetic politicians as a way to advance their constituents' specific interests. Labor, unlike the popular organizations, was confronted with a more articulated and better organized adversary in the employers' associations and their strong political allies in government. To offset this, labor counted on stronger local, regional, and national organizational structures, a well defined political agenda of structural reforms, a more stable constituency, and greater mobilization resources capable of pressing their demands at different levels of government. In short, labor had the capacity to act on its interests independently of government or opposition elite patronage. Thus the key segments of the labor movement chose to follow a more confrontational strategy in pursuit of their claims. This resulted not only in increased levels of labor conflict through sectoral strike actions, but also in two major general strikes, each of which attracted a national following of approximately 10 million workers (see Graph 7).

Phase 4: 1988-1989 - Sectoral Contention and State Reforms

In 1988 and 1989 (phase 4), social movement protest was primarily characterized by sectoral demand-making conducted as groups struggled to make their impact on the constitutional reform process. During the congressional elections during the second half of 1987, the political enviroment was excited by prospects of constitutional reform. In 1988, the principal mandate of the congress was the revision and reform of the military dictatorship's federal constitution. Although many of the authoritarian aspects contained in the old constitution had ceased to be observed and less arbitrary practices instituted, the formalization of these democratic measures was postponed until 1988 when the constitutional structure of the state underwent a major overhaul. The revision sought not only to eliminate autocratic features left from previous regimes, but also to alter provisions seen by many sectors as obstructions to the development of a modern industrial society. Hence, the political conjuncture which initiated this phase was centered around the activities of the congress as it acted as the national constituent assembly.

Conservative and moderate politicians and their supporters had mobilized their energies with the aid of the presidency to avoid the election of a separate constituent assembly, fearing that this would provide progressive political groups, such as the social movements, the socialist and social democratic parties, and the Church an opportunity to dominate the constitutional reform process. After the congressional elections of 1987, when it became clear that these groups had not achieved a majority, the president and the new congress went about preparing the task of constitutional reform.

As a result, the social movement actors were confronted with the task of devising strategies to pressure members of congress in support of their claims. More importantly, they sought to bring together those with different positions about what the role of these claims ought to be in the consensual formulation of constitutional proposals which would unite diveerse movement organizations. Precisely this process had a strong effect on the way contention developed in this period. Certainly, the political conjuncture of democratic reform set the backdrop before which contention was to be acted out. However, the forms, strategies and options of which political opportunities would be taken, and how and when they would be used, were an outcome of the capacities of each type of social movement.

In general, popular social movements returned to the political struggle after a three year slump in the level of collective activity ever more determined to achieve major reforms. From 1987 to 1989, average participation in collective action events tripled from 636 to 1903. This increase in participation was noteworthy, but even more so was the fact that this occurred in a smaller number of collective events, indicating that social movements were staging fewer, but larger and strategically placed demonstrations with a focus on mobilizing larger numbers of supporters. This type of strategy is understandable in the context of a constitutional reform process in which social movement participants lobbied members of congress and political parties to support their proposals in the constituent assembly. Because of the strong pressure from the labor and social movements, the constituent assembly devised and instituted mechanisms through which popular sectors could submit their proposals to the various committees. ₂₈

Further, the significant increase in mobilization reflects as well a growing disenchantment in social movement militants with elected politicians' attention to movement claims in spite of past collaboration. After a few years of close ties with government, social movement organizations were able to reject both the goverment elites' efforts to coopt them and opposition elites' assimilative efforts. Social movement discourse became markedly anti-partisan, advocating an autonomous role in the political process and a greater predisposition to engage in confrontational politics. ₂₉ Movements displayed their strength through public demonstrations, and their militants once again became active in coalitional action around their specific constituencies' interests, with the goal of presenting unified proposals for constitutional reform and pressuring for their enactment. A variety of forums, bringing together professional specialists, politicians and social movement activists emerged to debate, formulate, and propose reforms to the national parliament. Their proposals included housing, health, childcare, education, transportation, human rights, and agrarian reforms, revision of labor relations, and institutional changes.

However, the political conjuncture--pressed by the urgency of political reform--shaped the position of political elites toward the social and labor movements. Conservative government elites took a firmer stand against social movements' demand-making as politicians came to perceive the importance of the constituent assembly's proceedings. Furthermore, given that 1989 was a presidential election year, independent social movement activity was seen as threatening, for the Partido dos Trabalhadores had gained support steadily and would be a major opponent in the coming presidential race. Social movement discontent with government came not only from its incapacity to attend to their demands, but also from the ongoing patronage which the Sarney administration was extending to the propertied classes, while neglecting programs to improve the social conditions of the lower classes.

Non-government elites of the center and center-left also wavered between cooperative and competitive postures towards the social movements. Some of these elites had at times worked with social movements pressing for common demands, but the enhanced conditions for political change present in this period often led these elites to view social movements as competitors over the leadership of a reform-oriented agenda. Consequently, social movement activists found that these elites had became less interested in cooperation, especially in matters of political reform.

Indicative of these changes in the attitudes of political elites are the rates in this period of police repression of collective action--reminiscent of the final years of military rule. While in 1983 and 1984 (Phase 2), on average 37% of protest events prompted police intervention, in 1985 and 1986 (Phase 3) the percentage decreased to 16%; in Phase 4, repression once again rose, to 42%, as shown in Graph 8. Certainly, as social movements became more intent on having their demands met and as political elites became more adversarial towards contention, police repression once again became probable. There was a substantial reduction in the number of incidents of collective action, from 124 in 1987, to 73 in 1988 and 52 in 1989. Neverthless, urban popular movements were able to mobilize three times as many participants as in previous years, averaging 1900 people per demonstration in 1988 and 1989 as compared to 600 in 1986-1987 (Graph 6).

With the rise of a third, more moderate labor confederation, the Força Sindical, these national organizations were central to the of forging worker solidarity and protest. Of the three confederations, the CUT was the most aggressive in defending working-class interests and consequently more strike-prone. ₃₀

In 1988, in spite of inflation reaching record heights (1037.62 percent for that year), strike rates did not reach the levels of the previous year. In fact, both the strike rate and average number of strikers decreased slightly in 1988 from their 1987 level, while duration increased only slightly. Consequently, the strike strength coefficient declined to 296.43, from 369.08 in 1987, but certainly higher than any of the other years considered here. This decline can best be understood in the context of the political conjuncture. The previous year's strike wave and general strike and the failure of government to bring spiraling inflation under control led it to implement yet another economic stabilization plan, the Plano de Verao, which postponed discontent as workers waited to see the results. Another political factor which affected the strike rate was that the constitutional reform process redirected the political activities of labor leaders, like their urban popular movement counterparts. They turned away from strikes to the task of forging consensus within the ranks of labor about the needed constitutional reforms that would guarantee workers' jobs and establish a base for future labor relations. The complicated task of arriving at constitutional proposals by an ideologically-divided labor movement was time-consuming, as labor organized itself locally, regionally and nationally to reach a common position. At the same time, they devised joint strategies to pressure for its enactment. Thus the decline in strike activity at this time was more the outcome of the interaction of the organizational and ideological characteristics of the labor movement and the political conjuncture dominated by the constitutional reform process, than the lack of opportunity for contention.

This becomes more clear when one looks at the following year, the last of the Sarney administration (1989), which was marked by yet another strike wave and the largest general strike in the country's history, representing a significant expansion of worker participation in strikes. This third strike wave in the decade was the working-class response to the hyperinflation which continued to grip the country, threatening severe economic consequences for workers. Labor disputes almost doubled from the previous year (1988), a 1.5 increase over the prior strike wave year of 1987. Average strike duration increased to 11.3 days, the highest national average to date, but the average number of strikers per stoppage decreased slightly. The strike strength coefficient at 507.30 was 1.7 times greater than the previous year; 1.4 times greater than the 1987 strike wave year; 2.2 times greater than the 1985 strike wave and 10.3 times greater than the strike activity of 1978-1980, when labor opposition to the military regime began. This new upsurge in unrest was largely provoked by the economic instability brought about by hyperinflation, but it was also influenced by the 1988 demands for broad changes in the national labor relations system during the constitutent assembly. By 1989, the situation was ripe for yet another general strike against the government's economic policies, a show of force in anticipatation of the presidential elections. One of the two leading presidential candidates was the Workers' Party nominee, Luiz Ignacio da Silva (Lula), the leader of the 1978-80 metal workers' strikes. By any standard, the general strike was successful. It reached a new high in mobilizing approximately twenty-two million workers across the various regions of the country.

The intense activities of the labor and urban popular movement organizations attempting to influence the constituent assembly paid off in the inclusion in the 1988 Constitution of many of their claims. Labor succeeded, in one fashion or another, in obtaining important changes in labor relations and union autonomy, first advocated in the 1979 strikes. The popular social movements achieved the adoption of many constitutional provisions on social issues which govern the formulation of ordinary legislation and social policy. Special provisions guaranteed political and civil rights, undoing decades of authoritarian rules and practices.

Concluding this section, I mention another legacy of the social movements, the rise of the Workers' Party to national prominence. Founded in the early 1980s under authoritarian rule, by 1989 the Partido dos Trabalhadores had become sufficiently consolidated to run its own presidential candidate. In the running from the first round of voting to the run-off elections against a center-right candidate, Lula lost the latter by about 4% of the national vote. In spite of his loss, the fact that a leftist working-class candidate was able to rise to such political prominence within such a short time, attests to the capacity of the labor and urban popular social movements as agents of change in Brazilian society.

In the 11 years considered here as the period of redemocratization, then, social movements had a major impact on the democracy which emerged in Brazil. Nevertheless, it is difficult to pinpoint a direct causal relation between mobilization and political transformation because, as this study shows, the causal relation between popular contention and democratization is intertwined with the interaction of other actors and diverse political conjunctures. As the combined effects of social protest, opposition elite pressures, and authoritarian elite maneuvers transformed political circumstances from one phase to another, labor movement and popular social movement actors followed variable and complex trajectories over the period, as illustrated in Graph 10. Here social movement trajectories are traced within a space demarcated on the one hand, by the degree to which movement actors were autonomous or patronized by elites (orientation to elites), and on the other hand by the scope of mobilization, whether local or national; the graph summarizes the trajectory of each movement as well as allows us to highlight the differences in each trajectory over the period under study.

In the first phase, both the popular movement and the labor movement, were autonomous collective actors with their actions directed to the local political arena demonstrating proximate ties and collaborative activities. In the second phase, the movements became closely allied with each other in their efforts for redemocratization and creation of new political organizations which would better represent their interests in the national politial arena. In the phase in which the first civilian government began, the movements moved in different directions; labor continued in a more autonomous and national vector while the popular social movements returned to a local scope of action and closer collaboration with political elites. It is this phase which has concerned most of the studies on social movements and democratization. These works have decried the failure of popular organizations to maintain their independence from established political elites. In the fourth phase, popular collective actors returned to a more autonomous, nationally focused position and renewed collaboration with the labor movement which maintained its autonomous and national level orientation. The two trajectories demonstrate movement dynamics that are quite distinct from one another. In the case of Brazil, all the democratic demands made by the labor and urban popular movements eventually achieved institutional form in the 1988 Constitution, but their concretization was a task which required 11 years of contention, chipping away at resistance, at first from military rulers and later from former civilian allies. The Brazilian case does not point to unidirectional or unidimensional explanations, but demonstrates the complexity of any analysis of social movements and political change. The data may be seen as the component tiles fitted into an interpretative mosaic that deciphers the trail of political change.

 

Conclusion: A Method for Comparative Study of Social Movements and Democratization

Comparing the Brazilian experience of redemocratization to other cases of democratization among the Latin countries of the Americas and Western Europe, several variables appear to be relevant in cross-national comparisons. First is the relative autonomy of social movements from either regime elites or opposition elites. In Brazil, such autonomy meant that social movement actors not only possessed the conditions for independent intervention, but more importantly, such autonomous capacity made them effective actors both in confronting authoritarian elites and in bargaining with traditional opposition elites as they pressed to broaden the terms of the transition to democracy. In like manner, the autonomy of the social movements allowed them to remain key actors in the post-dictatorship period, thus contributing to the nature of the political reforms necessary after the long dictatorial rule.

A second factor relevant to understanding the role of social movements in democratization is the capacity for coalition building of established opposition elites. In this case we conclude that the capacity of coalition building encompasses two central features: cohesion with opposition elite coalitions and the capacity to incorporate new actors (especially independent popular movements) into the democratic coalition confronting authoritarian elites. This elite cohesion/incorporation capacity not only contributed to a clearer agenda for democratic change but also permitted the emergence of a consensus about the scope of subsequent political reforms in which popular collective actors' claims would be addressed. Taking the risk of oversimplification, we depict the relationship of these two dimensions in Graph 10.

This graph situates along two axes several countries which underwent the transition from dictatorship to democracy in the last two decades, locating them in relation to two variables: social movement autonomy, and opposition elites' level of internal cohesion and capacity to incorporate new actors. As Graph 10 indicates, there is a correlation between on the one hand, the degree of democratic regime stability, on the one hand, and on the other, each country's experience with social movement autonomy and cohesion and coalition building of opposition elites. The graph affords a simplified expression of both the 1) top-bottom theories of democratic transition of contemporary theorists like O'Donnell, Schmitter, Linz, and Valenzuela in the cohesion/coalition building variable; and 2) bottom-top theories of Bendix, Stein Rokkan, and Moore on the variable autonomy of social movements.

Countries like Argentina and Chile have endured constant threats of authoritarian intervention since democratization; in these countries, social movements have been very much subordinated to opposition parties which were less successful in building an opposition coalition; indeed, such coalitions were often on the verge of collapse, instead of confronting authoritarian elites. Peru has been an unstable democracy because civilian elites have traditionally lacked cohesion and failed to form anti-authoritarian elite coalitions as social movements became more independent actors in the political arena. The combination of fragmented political elites and autonomous social movements has meant the recurrent military tutelage of civilian politics. A somewhat similar combination of factors has occurred in Venezuela, as its democratic institutions underwent profound crisis in recent years.

In the other side of the graph, one finds the countries which have had more success in establishing stable democratic institutions. Here are, on the one hand, the cases of Spain and Uruguay in which opposition elites were able to create stable coalitions to confront the Franquist establishment and social movements were closely tied to these opposition elites, intervening only in moments when coalition politics demanded popular support. On the other hand, there are the cases of Brazil and Portugual. There opposition elites were highly cohesive and effective in coalition building and incorporating emerging new political actors during the process of transition. Social movements were significantly autonomous of these opposition elites but were included in the opposition coalition. This permitted the formation of stable democratic institutions in which social movement collective actors became key players in politics without destabilizing effects.

Graph 10 then serves two purposes: first, it simplifies a more complex argument about the interrelation of social movement mobilization and elite negotiation of democratic transitions; and second, it portrays the two main theoretical perspectives on the role of social movements in political change. The bimodal distribution of the cases the graph suggests that social movement involvement in the democratization process under conditions of cohesive and coalition-building oppositions promotes the formation of stable democracies, as the older school of democratization theory proposed, and contrary to more recent interpretations. The graph also suggests interesting hypotheses as a point of departure for further investigation of the role and impact of social movements on democratic transition in comparative perspective.

 

For their helpful comments, I would like to thank Louise Tilly, Charles Tilly, Joel Stillerman, Behrooz Moazami, Themis Chronopoulos, Marco Giugni, Mauricio Font, Alexei Waters and the participants of the Contentious Politics seminar at Columbia University and the Think, Then Drink seminar at the Center for Studies of Social Change, New School for Social Research.

---

*: . Back.

**: Salvador A.M. Sandoval
Pontifícia Universidade Católica de São Paulo
Universidade Estadual de Campinas.
Back.

Note 1: Charles Tilly, "The Nineteenth-Century Origins of Our Twentieth-Century Collective-Action Repertoire," paper presented at the Conference on Economy and Society in the Twentieth Century in Honor of Tom Burns, University of Edinburgh, Edinburgh, Scotland, September 1981, reissued as the Center For Research on Social Organization, The University of Michigan, CR50 Working Paper no.244; "Britain Creates the Social Movement," CR50 Working Paper no.232, Center for Research on Social Organization, The University of Michigan, 1981, pp. 5-9; also published in James Cronin and Jonathan Schneer, eds., Social Conflict and Political in Modern Britain  (London: Croom Helm, 1983). Back.

Note 2: Charles Tilly, "Social Movements and National Politics," paper presented to the Conference on Organizing Women (Stockholm, Sweden, February 1978) and reissued as CR50 Working Paper 168, Center for Research on Social Organization, The University of Michigan, p.23. Back.

Note 3: Charles Tilly, "Social Movements as Historically Specific Clusters of Political Performances" (New York: New School for Social Research, Center for Studies of Social Change, The Working Paper Series, No.162, June 1993), p.10. Back.

Note 4: In discussing precisely the question of the relation between social movements and democratization Charles Tilly writes: "let me suggest that at best the proliferation of social movements only promotes democracy under limited conditions: it only occurs when movements organize around a wide variety of claims including explicit demands for democracy and the state gains capacity to realize such claims at least as fast as the claims increase. But the conditions around achieving these goals are problematic as well. Indeed, democratic theorists have always posed precisely these two questions: under what conditions ordinary people actually demand democracy; under what conditions does state capacity increase to meet these demands? Perhaps we can take some comfort from the observation that the analysis of relationships between social movements and democratization leads us straight to major unresolved problems of democratic theory." Charles Tilly, "Social Movements as Historically Specific Clusters of Political Performances" (New York: New School for Social Research, Center for Studies of Social Change, The Working Paper Series No.162, June 1993) p.11. Back.

Note 5: The statistical data for this study area are derived from a number of sources: for strike activity I have relied primarily on strike data from my study of 45 years of strike actions in Brazil, the Brazilian Ministry of Labor and the Departamento Inter-Sindical de Estatistica e Estudos Socio-Econoinicos-DIEESE, published in Labor Unrest and Social Change in Brazil Since 194 5 (Boulder, CO; Westview Press, 1994; forthcoming in Brazil, Editora Atica); for the social movement data I have used data collected from newspapers sources compiled by the Instituto Brasileiro de Analise Sociais e economicos-IBASE and the Centro de Documentação e Pesquisa Vergeiro in a research proiect under the supervision of Prof. Salvador A. M. Sandoval and with preliminary data analysis reported in Samira Kauchakje, "Movimentos Sociais Populares Urbanos no Brazil de 1983 a 199O", Master of Education Thesis, Universidade Estadual de Campinas, 1992. Back.

Note 6: Doug McAdam, "Conceptual Origins, Current Problems, Future Directions", in Doug McAdam, John D. McCarthy and Mayer N. Zald, eds. Comparative Perspectives on Social Movements  (Cambridge: Cambridge University Press, 1996) 27. Back.

Note 7: This typology of elite behavior was derived from another classification scheme presented in Donatella della Porta and Dieter Rucht, "Left-Libertarian Movements in Context: A Comparison of Italy and West Germany, l965-1990", in J. Craig Jenkins and Bert Kiandemians, eds. The Politics of Social Protest; Comparative Perspectives on States and Social Movements  (Minneapolis: University of Minnesota Press, 1995) pp.235-236. Back.

Note 8: The bargaining perspective is discussed in an article by Paul Burstein, Rachel L. Einwoliner and Jocelyn A. Hollander, "The Success of Politic: Movements: The Bargaining Perspective", in Jenkins and Klandermans, eds., The Politics of Social Protest , pp.279-281. Back.

Note 9: The classification of policy outcomes used here is based on the types of policy responsiveness' discussed in Paul Burstein, Rachel L. Finwohner and Jocelyn A. Hollander, "The Success of Political Movements: The Bargaining Perspective", in Jenkins and Klandermans, eds., The Politics of Social Protest , pp. 282-284. Back.

Note 10: In her analysis of mobilization documents of social movements in Brazil, Ana Maria Doimo notes a similar trend in the realization of collective actions by popular organizations for the same period, see her A Vez e a Voz do Popular: Movimentos Socials Populares e ParicinacaLo Politica no Brasil Pós-70 , (Rio de Janeiro: Editora Relume Dumara, 1995), pp. 118-122. Back.

Note 11: Sandoval, Social Change and Labor Unrest in Brazil , pp.2-3. Back.

Note 12: The strikes took place, not in the city of Sao Paulo, but rather in the industrial belt composed of the cities of Santo Andre, Sao Bernardo and Sao Caetano, commonly referred to as the ABC region. Back.

Note 13: Two partial but informative analyses of strike activity for the period between 1982 and 1985 are CEDI-Centro Ecurnenico de Documentação e Informação, Trabalhadores Urbanos no Brazil/82-84 , (São Paulo: CEDI, Acontecen Especial 16, 1986); Marcia de Paula Leite, "Tres Anos de Greves em São Paulo 1983-1985", Revista São Paulo em Perspectiva , 1 (2) (July-September 1987): pp.50-64; For a brief description of the situational circumstances around which this strike activity occurred for the period 1980 to 1986, see Brazil 1986: Relatório sobre a Situação Social do Pais  (Campinas: Nucleo de Estudos de Politicas Públicas, Universidade Estadual de Campinas, 1987), pp.54-86; for 1987 see Brazil 1987: Relatório sobre a Situaçao Social do Pais (Campinas: Nucleo de Estudos de Politicas Públicas, Universidade Estadual de Campinas, 1989) pp.11l-145. Back.

Note 14: Of the non-agricultural workers' unions, 42.7 percent were founded after 1960, raising the number of labor unions from 1,670 before 1960 to 2,916 in 1988. Of these post-1960 unions. 69 percent were created between 1971 and 1988. Similarly, 40.5 percent of industrial workers' unions were founded in the period after 1960, as were 46 percent of bank workers' unions and 65 percent of teacher, health and other service workers' unions. Back.

Note 15: For a more detailed discussion of the initial rise of the CUT and CGT, see CEDI-Centro Ecumenico de Documenta¸ão e Informa¸ão, Trabalhadores Urbanos no Brazil/82-84 , (São Paulo: CEDI, Aconteceu Especial 16, 1986) pp.41-47; Margaret F. Keck, "The 'New Unionism' in the Brazilian Transition", in Alfred Stepan, ed., Democratizing Brazil  (New York: Columbia University Press, 1990). Back.

Note 16: The annual strike rate is the proportion of the number of strikes per one hundred thousand workers in the non-agricultural work force. For a more detailed analysis of strike activity see Sandoval, Labor Unrest and Social Change in Brazil . Back.

Note 17: The combination of the three dimensions of strike activity can be represented by the cube of the dimensions, i.e., the strike shape; see Chapter 4 of Sandoval, Labor Unrest and Social Change in Brazil . A numerical indicator of the average strength of strike activity for each year allows for comparative analyses. This indicator of strike strength, the coefficient of strike strength is the area of the strike shape divided by 1000: volume or strike strength's coefficient is equal to the product of average duration (D), multiplied by average size (W), multiplied by strike rate (S), divided by 1000. Strike Strength Coefficient = (D x W x S)/1000. For further discussion of strike dimensions indicators see Edward Shorter and Charles Tilly. Strikes in France 1830-1968  (London: Cambridge University Press, 1974) and P.K Edwards, Strikes in the United States 1881-1974  (Oxford: Basil Blackwell, 1981) Appendix C: "The Statistical Description of Strike Activity," pp.316-320. Back.

Note 18: For a discussion of the autonomy-patronization relation to powerholders as a characteristic of social movements and other collective action, see 'Filly, "The Nineteenth-Century Origins of Our Twentieth-Century Collective-Action Repertoire." See especially diagram on pages 7-8. Back.

Note 19: This political campaign was from tile beginning part of the strategy of the opposition parties' coalition to pressure for approval of the constitutional amendment. While the Diretas Já demonstrations resembled those of social movements, the fact remains that the Diretas Já ; collective action was organized and coordinated by the opposition parties, which in many cases already controlled state and municipal government. In this respect, we have opted not to include these demonstrations as social movement collective action even though the massive popular turnout can only be explained by the involvement of a large array of social movement organizations. For a rich detailed factual account of the Diretas J´ mobilizations see Ricardo Kotscho, Explode Um Novo Brasil: Diário da Campanha das Diretas  (São Paulo: Editora Brasiliense, 1984). Back.

Note 20: Thomas E. Skidmore, The Politics of Military Rule in Brazil, 1964-85  (New York: Oxford University Press, 1988), pp 250-253. Back.

Note 21: We are inclined to understand another massive outpouring of grief, in 1994 about the death of Airton Senna, the Brazilian world champion in Formula 1 as demonstrating political frustration over the detours which the democratization process had taken at that time when national political institutions were still plagued with widespread corruption, inefficiency and incompetence. Back.

Note 22: Leoncio Martins Rodrigues, "A Composi¸ão Social das Lideranças do FT,"in ibid., Partido e Sindicatos: Escritos de Sociologia Politica  (São Paulo: Fditora Atica, 1990). Back.

Note 23: The Nova Republica is the name given to the José Sarney administration 1985-1989, the first civilian government following the military dictatorship. Back.

Note 24: Instituto Brasileiro de Geografia e Estatística-IBGE, Pesquisa Nacional por Amostra de Domicilios , Suplemento no.1, Associativismo (Rio de Janeiro: Funda¸ão Instituto Brasileiro de Geografia e Estatística, Departamento de Emprego e Rendimento, 1986) Table 3, p.46; IBGE, Participa¸ão Politico-Social 1988: Brazil e Grandes Regiões  (Rio de Janeiro: Funda¸ão Instituto Brasileiro de Geografia e Estatística, Departamento de Estatistícas e Indicadores Sociais, 1990) Table 4.1, p.8. Back.

Note 25: According to Shorter and Tilly, Strikes in France , a "strike wave occurs when both the number of strikes and the number of strikers in a given year exceed the means of the previous five years by more than 50 percent," pp.106-107. I have defined strike waves by the Shorter and Tilly definition, but included in that definition change in the coefficient of strike strength which exceeds the mean of tile previous five years by 50 per cent. Back.

Note 26: This fact has been clearly documented by a wide variety of social movement analysts as a phenomenon occurring in all types of social movements regardless of geographic region. With reference to some studies on the recruitment and cooptation of social movement militants into political party machines or municipal administrations, see Pedro Jacobi, "Movimentos Sociais Urbanos no Brasil: Reflexão sobre a Literatura nos Anos 70 e 80", in ANPOCS Bib - Boletim Informativo e Bibliográfico de Ciencias Sociais , no.23 (Januaiy-June, 1987), 26; Julio Assis Simões, 0 Dilema da Particip¸ãio Popular: A Etnográfia de Um Caso  (São Paulo: Editora Marco Zero/ANPOCS, 1992), see Chapters 3 and 5; Renato Raul Boscbi, Arte da Associa¸ão: Politica de Base e Democracia no Brasil  (Rio de Janeiro: IUPERJ-Editora Vertice, 1987) pp. 46-58; Maria da Gloria Marcondes Gohn, A Forca da Periferia: A Luta das Mulheres por Creches ern São Paulo  (Petropolis: Editora Vozes, 1985) 139-152; Ruth Corrêa Leite Cardoso, "Os Movimentos Populares no Contexto da Corslida¸ão da Democracia" in Fablo Wanderley Reis and Guillermo O'Donnell, eds., A Democracia no Brasil: Dilemas e Perspectivas  (Rio de Janeiro: Editora Vertice, 1988) 374-377; Maria da Gloria Marcondes Cohn, Movimentos Sociais e Luta pela Moradia  (São Paulo: Ediçoes Loyola) 161-164; Pedro Roberto Jacobi. Movimentos Socials e Políticas Públicas  (São Paulo: Cortez Bditora, 1989) see especially pages 157-159; Maria da Merces Comes Somarriba and Mariza Rezende Afonso, "Movimentos Urbanos e Fstado: Novas Tendências" in Pompermayer, ed Movimentos Sociais , especially p. 106-109; Mariza Rezende Afonso and Sergio de Azevedo, "Cidade, Poder Público e Movimento de Favelados" in Pompermayer, ed. Movimentos Sociais , especially p. 135-138; Sonia E. Alvarez, Engendering Democracy in Brazil: Women's Movements in Transition Politics  (Princeton New Jersey: Princeton University Press, 1990) Chapter 6. For references on the recruitment of social movements activists into the labor movement see Leoncio Martins Rodrigues and Adalberto Moreira Cardosos, Forca Sindical: Uma Analise Sócio-Politica  (Rio de Janeiro: Editora Paz e Terra, 1993) 70; Michel Marie LeVen, "Movirnento Operano e Sindical, 1972-85", in Pompermayer, ed. Movimentos Sociais , especially p.38-41; Hamilton José Barreto de Faria, "A Experiência Operária nos Anos de Resisténcia: A Oposi¸ão Sindical Metalurgica de Sáo Paulo e a Dinamica do Movimento Oper´rio (1964/1978)" (Master's Thesis in Social Sciences, Pontificia Universidade Catolica de São Paulo, 1986), Chapters 3 and 4; Ivanise Leite de Miranda, "0 Individuo Representativo: Considerações Sobre as Relaçóes entre Valores Individuais e Consciencia Social" (unpublished doctoral dissertation, Universidade de Sao Paulo, 1992) vol.2. Back.

Note 27: Doimo, A Vez e a Voz , pp.176-177. Back.

Note 28: This procedure resulted in a massive information base comprising thousands pages of documents from the initial proposals through the committees' reports and records of its' debates to the final versions of proposed amendments submitted by each committee to the plenary of the congress where, in term, these proposed amendments underwent further reformulations as a result of the debates and negotiations between the federal deputies. Back.

Note 29: Eder Sader, Quando Novos Personagens Entram em Cena  (Rio de Janeiro: Editora Paz e Terra, 1,988). Back.

Note 30: For a fuller analysis of the two major national labor organizations, the CUT and the Força Sindical, see Leôncio Martins Rodrigues, CUT: Os Militantes e a Ideo1ogia , (São Paulo: Editora Paz e Terra, 1990) and Leôncio Martins Rodrigues, Forca Sindical: Uma Análise Sócio-Folitica , (São Paulo: Editora Paz e Terra, 1993). Back.