Autoworkers and the working class in Brazil - John Humphrey

Introduction

In May 1978 a series of stoppages occurred in the auto industry in Brazil, which is concentrated in the southern industrial belt of Sao Paulo. After a decade in which the semblance of industrial peace had been guaranteed by a combination of legal controls over the trade unions, repression of rank-and-file organizations in the plants, and strict control over political opposition to the military government, the working class once again defied legal restrictions on its activities. The stoppages initiated a period of labor opposition to the state and its policies towards the working class which was to last for the remainder of 1978 and stretched into 1979. In March of 1979, the auto workers went on strike again, but this time in an all-out strike of the three metalworking unions of the southern belt that lasted for two weeks and provoked a major confrontation with the state.

After a decade of apparent calm that lasted from the end of the 1968 to the strike wave in May 1978, the sudden development of the stoppages in May and the determination shown by workers at that time might appear surprising. An examination of the analyses of the working class in Latin America made in the seventies would increase rather than diminish this surprise, because the strike movement received its major impetus from workers in the dynamic industries - those which expanded as a result of investment of foreign capital in Brazil from the 1950s onwards. It had been argued by some theorists that workers in the dynamic industries, of which the motor industry is one of the most important sectors, would be more likely to be incorporated into the political alliance of the dominant classes than other groups of workers (Cardoso, 1972: 93), or that they would possess qualities that would mark them off from the rest of the labor force (see, for example, Quijano, 1974: 407- 408 and 419). All the more surprising, then, that the stoppages in the auto industry in 1978 led to stoppages in many other industries, including traditional ones. At the same time, the opposition of the auto workers to the state and the general support that their demands for free bargaining, the right to strike, and trade union reform have encountered among other sections of the working class would seem to belie theories that modern-sector workers seek changes that would be of limited benefit to other sections of the working class (see, for example, Almeida, 1978).

Therefore, the events of 1978 and 1979 point to the need for a re-examination of the position of workers in the dynamic industries and of their relation to the rest of the working class. Elsewhere I have sought to show that workers in these sectors are neither a privileged elite nor incorporated into the political bloc of the dominant classes in Brazil (Humphrey, 1977 and 1979). I have argued that the evidence from the auto industry suggests that workers there share many of the problems faced by the working class in general - stagnant or declining real wages, bad working conditions, excessive pressure of work, insecurity in employment, etc. - and that their higher-than-average wages do not constitute the basis for a general situation of privilege. However, certain fundamental questions still need to be posed and answered. If it is correct that auto workers are not part of a labor aristocracy in Brazil, it still remains to be asked why it is that auto workers rather than any other group of workers have played a dominant role in the period of industrial and political conflict which started in 1978. What is it about their position that has led auto workers to take the first steps of opposition in 1978, heighten the conflict in 1979, and provide the dominant part of the union which is now the unchallenged leader in the battle for change? At the same time, if certain differences can be seen in the attitudes, situations, or demands of auto workers that would explain their leading role in the current confrontation between the unions and the state, do such differences imply that serious divisions between the auto workers (and dynamic-sector workers in general) and other sections of the working class are arising or will arise at some future date? This paper examines the reasons for the auto workers taking a leading role in 1978 and 1979, the events of the period, and the kinds of changes being demanded by the leading unions in order to answer these two questions.

Modern industry and its workers

The workers in the motor industry are part of a wider group of workers in the metalworking, mechanical, electrical, and transportation equipment industries who are organized into metalworking unions.1 The workers in these industries have experienced the contradictions of industrial growth more than any other group. The industries have grown rapidly, productivity and profits have risen sharply, and yet the apparent prosperity of the employers has not been translated into higher wages. Workers in these dynamic industries have experienced the same decline or stagnation of real wage rates as other sections of the working class.2

The rapidly expanding mechanical, electrical, and transportation equipment industries have come to occupy an increasingly important place in Brazil's industry. Whereas in 1949 these three industries only employed 4.3 per cent of the total labor force in the manufacturing sector, by 1970 their share had expanded to 17.5 percent (IBGE, n.d., and 1974a). In the major industrial state, Sao Paulo, the 1970 figure was 24.9 percent (IBGE, 1974b), and by 1974 the three industries employed 29.2 percent of the total manufacturing labor force in the state (IBGE, 1977), or over half-a-million workers. Starting from small bases, these industries expanded rapidly in the latter part of the fifties. After a difficult period in the early sixties, they benefitted greatly from the policies pursued by the new military government after 1964.

The motor vehicle industry, for example, produced 174,000 cars, vans, pick-ups, jeeps, and trucks in 1963, the year before the military coup. By 1974 this figure had reached 858,000, presenting an annual average rate of expansion of 15.6 percent. The motor vehicle industry, like many other of the new industries implanted in Brazil, grew up outside of the traditional areas of industrial growth. The foreign companies that dominated the fast-expanding sectors built factories in the peripheral areas of the city: in Osasco to the west and in Santo Andre, Sao Bernardo do Campo, and Sao Caetano to the south and southeast. In these new industrial belts, large firms dominated the metalworking industries. On the eve of the May 1978 strike movement, the 15 largest factories in Sao Bernardo employed 68 percent of the metalworkers in the area, while in Osasco 12 firms employed 17,000 of the 27,000 workers in the same category. In 1977, half of the metalworkers in Santo Andre were employed in just 10 large firms employing more than one thousand workers.3

The workers employed in these fast-growing industries were brought into contact with the high growth, high productivity, and high profit areas of the Brazilian economy, but they did not see the benefits of the "economic miracle." Far from being an exceptional, privileged group, the labor force in these industries suffered the same privations as other groups, and they saw the wealth of their employers rise as their own incomes at best kept pace with the rate of inflation. At the same time, the rapidly expanding demand for consumer durable goods meant that there was considerable pressure to increase productivity and extend the working day. By the end of the "miracle" period, an examination of trade-union demands and concerns could show that workers in the dynamic industries had serious grievances (for a full discussion of the grievances of workers in the auto industry, see Humphrey, 1977: 89-119). Of course, many other groups of workers suffered the same problems, but there are reasons for expecting workers in the auto industry, more than other groups, to take action as a result of this situation.

The initial expansion of the dynamic sectors, when relatively high wages were paid in order to attract large numbers of suitable workers, created the strong and lasting expectation that multinational companies should pay the top rates for any particular job of work. As the industries matured, however, their wages policies underwent some changes. In the case of skilled workers, shortages of skilled labor meant that the motor industry in particular could not maintain its position as the highest payer, other firms would match auto industry rates in the attempt to secure adequate supplies of skilled labor. In 1963 an auto industry study found that all but one of a group of skilled workers thought that the auto company employing them paid above-average wages for their kind of work. In 1975, only 2 out of 20 skilled workers in the same factory were of this opinion (Rodrigues, 1970: 46; Humphrey, 1977:103). A less dramatic decline in evaluation can be seen for semi-skilled workers, where the "above-average" evaluation dropped from 80 to under 50 percent. These subjective evaluations of relative wage rates can be reinforced by other data. A number of sources confirm that wages in the auto industry rose more slowly than wage rates in other sectors. A comparison of the Industrial Census of 1970 (IBGE, 1974b) with the Industrial Survey of 1974 (IBGE, 1977) shows that average wages in the Sao Paulo transport-materials sector rose by 131 percent in the four-year period. For the manufacturing sector as a whole, the increase was 172 percent. An unpublished analysis by the Departamento Intersindical de Estadistica e Estudos Socio-Economicos (DIEESE) of wage rates in the metalworking and textile industries revealed that in the decade 1966-1976 (and in each of the half decades (1966-71 and 1971-1976), the median wage in five large textile firms rose by more (or fell by less) than the median for five large firms in the auto-assembly and components sectors (DIEESE, 1977: 80 and 89).4 Finally, data for the metalworkers of Sao Bernardo for the period 1972-1978 show a significant narrowing of differentials between workers in the auto-assembly sector and mechanical and metallurgical sectors of the union. In other words, the workers in one of the most dynamic industries lost ground to other workers. Their employers used the state's wages policy to deny them a share in the benefits of industrial expansion, narrowing differentials between the auto industry and other sectors at a time when the industry was in a boom phase. It is to be expected, then, that there would be demands from within the ranks of auto workers for higher wages, wage raises according to the rate of increase in productivity, and an end to the wages policy imposed by the state.

The existence of discontent does not, however, guarantee that workers will be able to express such discontent in an effective and collective manner. Workers may feel powerless, or they may act in unorganized or individual ways. For example, workers may leave work, lower production, engage in isolated acts of sabotage, or merely feel that their power is so restricted in the face of the employers' domination that such resistance is unwise. In the case of the auto industry, there are four reasons for suggesting that resistance to management domination could take a collective form. Firstly, the relatively good pay and conditions found in the industry militate against leaving it as a solution to dissatisfaction. Unskilled and semi-skilled workers cannot expect to earn higher wages outside of the motor industry, and even skilled workers generally prefer to stay within the industry and fight for better pay and conditions. It was noticeable in 1973-1974 that skilled workers in the larger firms used labor shortages as a weapon to improve their situation within the same job, while skilled workers in smaller firms often moved in order to better their situation (see Humphrey, 1977: 143-165, for a discussion of these differences). Secondly, the skill distribution in the metalworking sectors is important. In metal-fabricating industries the number of skilled workers is often higher than in more traditional sectors such as food, textiles, and clothing. In the motor industry, as much as 30 percent or more of workers could be classified as being skilled, working in the tool rooms, maintenance departments, and, to a lesser extent, on the production lines. These skilled workers are important for workers' organizations within the plants because they are less subject to dismissal. Unskilled and semi-skilled workers are much more vulnerable, and this vulnerability, combined with high turnover rates, makes organizing among them quite difficult. In the case of skilled workers, their greater security, longer length of employment, better education, and better standard of living allow them to organize more easily. At the same time, the combination of significant numbers of skilled workers and a large mass of unskilled and semi-skilled employees can provide a volatile situation. Skilled workers will be aware of the needs and problems of the other workers in the plants, and they will also become aware that unless all the workers unite, the skilled workers can be isolated by the employers and defeated. For these reasons, the skilled workers have an interest in mobilizing other employees. Once mobilized, the unskilled and semi-skilled workers have many grievances that will fuel confrontations with the employers.

A third factor which influences labor relations in the auto industry is the rapid growth of the industry itself. This has two distinct effects. On the one hand, the growth of industry and the development of large factories employing thousands of workers breaks down paternalistic industrial relations and makes the division between capital and labor more apparent. In smaller plants workers are under the direct supervision of the top management and the owners, and their union and organizing activities are more subject to pressure. The breakdown of labor-force submissiveness and the massification of the workforce in larger plants, combined with the formalized and bureaucratic employment structures necessary in large firms, tend to produce collective and formalized responses from the workforce.5 The bureaucratic, rule-governed structures in the modern firm reflect the imposition of general priorities on lower management which inhibit their capacity to follow paternalistic policies. In the case of the auto industry in particular, the rigid imposition of financial norms forced plant management and labor-relations staff to adopt policies of labor rotation (hire-and-fire) in the seventies, which led workers to be very disillusioned about the benevolence of their employers.6 On the other hand, the growth of the dynamic sectors gave the trade union an expanding base on which to develop its organization and an easy target for allegations of high profits and low wages. The health of the dynamic sectors, and particularly of the booming motor vehicle sector, gave the union the opportunity to develop arguments about the capacity of the big firms to pay and about the potential for success of a form of trade unionism oriented towards plant demands and direct pressure of the employers. The development of the Metalworkers' Union in the Sao Bernardo industrial area is an important part of the growth of workers' struggles in the seventies, and it deserves a section to itself.

Trade unionism in the auto-industry

The development of the Metalworkers' Union in Sao Bernardo cannot be taken as a simple direct reflection of the nature of the workers that it represents, although there is a clear link between the union's base and the form of its politics and union strategy. The union's development has taken place over a decade, and in this time the membership has responded by mobilizing support, and the leadership has been renewed through the co-option of militants from the rank and file in the plants. Although it will be argued that the strategy employed by the union can be taken up by unions in other industries with different characteristics, the initial development of such a strategy in adverse circumstances was favored by the conditions found in the dynamic industries.

The Metalworkers' Union of Sao Bernardo represents all metalworkers in that area of Sao Paulo, and because of the industrial location of the motor industry, the majority of auto workers are represented by this union. The executive of the union has been dominated by auto workers, and membership rates have been relatively high in the big auto plants. Of the 125,000 metalworkers in Sao Bernardo in March 1978, 65,000 worked in five large auto plants and another 1,500 workers were employed in another five plants. Actual union membership (see note 1) varied from around 45 to just over 30 percent for hourly paid workers.7

In the seventies the union has attempted to overcome the problems faced by almost all Brazilian unions: low membership, an inability to organize in the plants, and a lack of credibility in the eyes of workers because of the control of the state. The union strengthened its welfare provisions, started to produce a monthly newspaper, and then tried to further its image as a competent organ by establishing some presence in the plants. It did this by taking advantage of the structure of the unions in Brazil, which allows up to 16 members of the executive to have the stability of employment associated with election to union office while at the same time continuing full-time work. From the early seventies these diretores de base,(union directors who continue to work in the plants) were encouraged to act as union organizers in the plants, giving out union membership forms, advising workers of meetings, representing workers in disputes with management, etc. (For an account of the penetration of these directors in one auto plant, see Humphrey, 1977: 176-178). In the course of time, this work was extended to factory-gate leafleting and meetings, the organization of union congresses, and mass meetings of workers in advance of wage claims. Slowly the credibility of the union was built up among the rank and file. In some factories the union directors devoted themselves almost full time to union work and, depending on the receptiveness of management, could act as mediators between workers and their employers. Later, the union started to organize meetings of workers in different plants and sections to discuss the problems they encountered at work, and the union would attempt to discuss these problems with management.8 Although managements in the large firms were generally reluctant to discuss plant issues with the union, seeing this as an invitation for the union to gain credibility and interfere with the normal running of their plants, opportunities slowly emerged. In 1975, Ford negotiated with the union over a new medical plan, and in 1977 Saab-Scania set up a union committee in the plant as a result of pressure from the union affiliate in Sweden. By law, firms were obliged to consult the union over issues such as the alteration of shift schedules and extra work in lieu of extra holidays. By pushing to the limits of the state system, the metalworkers' union was able to become considerably more active and mobilizing than many other unions in Brazil. At the annual negotiations, the union tried to negotiate separately from other unions in the state,9 and mounted campaigns for an extensive series of demands. In 1977, for example, the union had a list of 33 demands to be presented to the employers, which covered aspects of wages, fringe benefits, stability of employment, union organization, and other matters.

However, these attempts to use the official trade-union structure had limitations. The threat of being suspended by the Ministry of Labor was always present, and little progress could be made on the following fundamental demands of the union: end of Ministry control, free collective bargaining, the right to strike, and the right to have union delegates in the plants. The development of the union up to 1977 merely provided the base for a much more aggressive attitude later. This was partly made possible by a relaxation of state control and the verbal commitment to liberalization and democracy that emerged towards the end of President Geisel's period of office. The first step towards a new period of union activity was undoubtedly the campaign for the reposigio salarial (salary replacement), which started in 1977. In that year there was official confirmation of the general belief that the rate of inflation in 1973 had been underestimated by the government and that, as a result, wage increases had been lower than they should have been. Although this was not the first time the government's inflation figures had been questioned by the union movement, the size of the difference and the fact the government's own figures confirmed the "error" created a stir. The Metalworkers of Sao Bernardo used this to start a general campaign for an immediate wage increase to compensate for the manipulation of the inflation index, and a lot of propaganda work was carried out. This campaign was supplemented by an outright rejection of the formal negotiating machinery at the 1978 negotiations. In previous years the union had mobilized in advance of the negotiations, but to no avail. The negotiations were always a formality, and after a "failure to agree" had been registered, compulsory arbitration followed. In the words of one auto industry executive, the Regional Labor Court "always makes its decision within the narrowest parameters," which favored no change to the status quo. In the run-up to the 1978 negotiations, the union decided to do nothing and boycott the negotiations completely in order to show its members that the formal procedures were a sham and that nothing could be accomplished by continuing to participate. In April 1978 the workers in the union were rewarded the same increase as the other unions in the negotiations. This increase was not considered very good, and it did not take into account the demand for the reposição.

The message to be put across to workers was that improvements could only be obtained by going beyond the formal channels. In the course of the reposição campaign, the word "strike" had been used, and the union president, Luis Inácio da Silva ("Lula"), had made it clear that only the workers' own efforts would improve their situation. In March 1978, a new union executive was due to take office, and the old leadership had taken advantage of the elections to renew the executive and bring into it some of the militants in the factories that had played a part in the reposição campaign. In his speech upon taking office for the second time, the reelected president stressed the need for an active unionism that functioned in the factories. By expressing the feelings and contradictions of the workers in the dynamic industries, the union had built up its strength. In March 1978 the union was calling for more militant action in a political situation that appeared to allow more leeway for the working class. In this context, only the exact time and place of the strike movement in May 1978 could cause much surprise.

The importance of the recent strike movements

The decision by the day-shift toolroom workers in Saab-Scania to stop work on May 12, 1978, sparked off a series of events that are still having a profound effect on industrial relations and politics in Brazil. From the Saab-Scania plant, the stoppages spread to the other major plants in Sao Bernardo and then to neighboring industrial areas. Within two weeks over twenty factories and 45,000 workers had stopped work in pursuit of wage increases. In the southern industrial belt, factories employing over 125,000 workers were affected by total or partial stoppages (Veja, May 31, 1978). In the following two weeks the movement affected many more factories in the area and also spread to Osasco and Sao Paulo itself. In the motor industry only General Motors, whose plants lie outside of the Sao Bernardo area, was unaffected, and all the companies agreed to make a common offer of an 11 percent increase to 13.5 percent in anticipation of the April 1979 wage agreement. The Sao Bernardo union gained prestige as it was called in by workers to present their demands to management and communicate responses. The union did not, however, ask workers to go back to work, nor did it make recommendations to workers; only after a contract was signed did the union encourage workers not to break it. In other areas, where unions were weaker or lacked the confidence of the workers, the strikers set up factory committees or negotiated directly with their employers.10

The movement starting in May demonstrated that management would negotiate directly with employees if faced by a work stoppage. The movement also demonstrated that there was a considerable degree of discipline and determination to be found among the workers of the southern industrial belt. In the case of Saab-Scania and Mercedes, the workers went back to work pending an offer later in the week and then stopped work again when that offer turned out to be inadequate, while at Ford the stoppage lasted a whole week in spite of management pressure. Across a wide range of firms, the tactic of going to work but refusing to carry out normal duties proved highly successful, but in addition to immediate gains on the wages front, two further developments merit attention.

Firstly, the stoppages provoked a general period of labor unrest, mainly in the Greater Sao Paulo area. As well as the extension of the form of the Sao Bernardo stoppages - workers going to work and staying in the factories for the normal period without operating the machines - a series of all-out strikes and industrial disputes took place later in the year. Nonindustrial sectors, such as schools, banks, hospitals, and public services were affected, as were a variety of industries. It is certain that these later movements would not have taken place if the metalworkers of the southern belt had not tested the determination of the state and the employers in the May-June confrontations. In November 1978 the giant Metalworkers' Union of Sao Paulo went on strike, and the union president was hard put to maintain control of the workers and obtain a settlement. In this case, too, the mobilization of the rank and file and their attempts to reject a leadership too inclined to accommodate to the state were a direct result of the general climate of labor unrest generated by the May-June movement, Secondly, the May events were only the first step for the metalworkers of Sao Bernardo. Having taken the companies by surprise once, it was obvious that they would not encounter such a lack of preparation again, and in the build-up to the negotiations in March 1979, both the unions and management were aware of the issues at stake. For management it was clear that some organization would be necessary and that some attempt would be needed to halt the "run of success" of the union so as to diminish its popularity among workers, although the control of the union over the unofficial strikers in May had been impressive and "responsible" enough for some sectors of management to see the possibility of encouraging a business trade-unionism along U.S.-style lines. Managements in some firms tried to fire the activists in their plants, and all the auto industry management entered into general consultation about the next moves under the umbrella of the auto industry employers' union, Sindicato Nacional dos Fabricantes de Veiculos Automotores (SINFAVEA). Meanwhile the union prepared the ground for the next round, knowing that it needed to establish the principle of direct negotiations and attempt to build up its plant organization by securing the acceptance of union delegates in the plants with security of employment.

In spite of apparently reasonable attitudes on both sides, the 1979 negotiations could hardly have led to anything other than conflict. There was too much at stake in the long term. On the management side, the dominant tendency wanted to contain the advances of the union, and, if possible, claw back some of what had been lost by the hasty concessions made in May 1978. For the union, it was important to move forward, obtain new gains, and confirm the leading role of the southern belt metalworkers within Sao Paulo unionism as part of the challenge to the older, pro-government leaderships. In March the negotiations finally broke down when the unions of the interior accepted a pay deal that offered 63 percent to the lower-paid workers earning up to three times the minimum wage and 57 percent to those earning more than three, but less than ten, times the minimum wage. Workers earning more than ten times the minimum wage would get the official government wage increase (supposedly binding as a wage settlement for all workers, but ignored in the negotiations) of 44 percent. Therefore, although the state's wages policy was ignored in the negotiations, the settlement accepted by the unions of the interior was not good for the workers of the southern industrial belt. Most workers in the auto industry earn more than three times the minimum wage, and so they would only have received the 57 percent increase, and in addition to this, the increases gained by strike action in 1978 would be discounted from the settlement. In other words, what was conceded in May 1978 was being taken back in March 1979. An auto worker earning above three times the minimum wage would only get a increase of 41.5 percent above his June 1978 wage. Furthermore, once the additional gradual 13.5 percent increases also conceded in May 1978 are taken into account, the actual increase in earnings in April 1979 would have been 26 percent.'11

Faced with an offer that would see the gains of 1978 eroded, make no headway of the issue of management recognition of union delegates in the plants, and not allow the southern belt unions to establish their position at the head of the union movement as a whole, three unions - in Sao Bernardo, in Sao Caetano, and in Santo Andre - went on strike in March 1979. The strike lasted two weeks in a climate of great confrontation. Mass meetings and mass pickets at some of the major factories demonstrated the mobilization of the workers. The employers refused to budge, although as the strike went on, there were considerable splits developing within their ranks. Most importantly, however, the state played an increasing role. Backed by many, although not all, of the employers, the state first tried to control the effectiveness of the strike through policing of the pickets. Then the Minister of Labor played an increasing role in the negotiations. The government was against any significant improvement in the terms to be offered to the striking workers, and after exhaustive negotiations in the second week of the strike, the only offer made to the workers was a return to work and further discussions. When this offer was rejected by the workers at mass assemblies on March 22, the state forcefully took over the three unions. The offices were occupied, the arenas for mass meetings were taken over, and the union leaders were deposed from office. However, this did not end the strike, even though it made the continuing organization of pickets, publication of strike bulletins, etc., almost impossible. In fact, interventions in the unions merely increased tension in the area, and in spite of mass policing, there were public manifestations of support for the workers and for the deposed leaders. Faced with a situation in which the strike could get further out of hand, the state was forced to negotiate a compromise. If the workers would go back to work under the terms rejected on March 22 (further negotiations over a period of 45 days), then the union would be handed back to the elected officers within in the same period. Until that time, the negotiations would be carried out by a committee nominated by the deposed union leaders. This compromise was accepted at a mass meeting of workers the week after the intervention.

At the economic level, the eventual result of the strike was not a great success for the union. The final terms for a settlement, negotiated in May 1979, gave to all workers earning up to ten times the minimum wage a 63 percent increase. So for many workers, the only result of the strike was an extra 6 percent, as opposed to the extra 16 percent or more originally demanded. Furthermore, the employers made no concessions in the area of union delegates, which was one of the major demands at the start of the negotiations. In addition, the union's rank and file were somewhat demoralized by the terms of the settlement and also by the terms under which they were forced to return to work in March. Certainly the effort demanded by such a strike led to a significant downturn in union activity in the plants once the settlement had been made. The union, in fact, had come close to a serious defeat and setback that would have gravely damaged its prestige among its members. But, if the union had come out of the strike with considerably less than total victory, the employers and the state suffered a serious defeat. As the negotiations reached stalemate and the unions refused to accept ultimatums from the Ministry of Labor, the attitude of the employers and the state hardened. By the end of the strike, the government wanted to defeat the strikers so as to establish the authority of the state and put an end to the radicalization of the union movement. They failed to do this. In spite of employers' resistance and the use of the full powers of the Ministry of Labor, the government was forced to negotiate a compromise with the deposed union leaders. The intervention itself was a failure, except insofar as it made the continuation of the strike impossible. Three large unions were able to undertake and maintain an illegal strike for over two weeks, and after the full penalties for such behavior had been imposed on the union leaders, the Ministry of Labor was forced to negotiate the withdrawal of those sanctions. The union maintained its organization and prestige intact. In addition, the negotiations took place outside the formal framework of the labor courts and Regional Labor Office, and the issue at the heart of the dispute was the degree to which the settlement would exceed the state's pay norms.

For these reasons, the March strike and its aftermath represent an important political step for the working class, but the difficulties encountered at the time of the strike point to the need for further reforms and further struggles. The partial victory in March was only won by an immense effort in the face of a series of institutional obstacles that need to be dismantled if the working class is to grow in strength. The first of these obstacles is the formal negotiating procedure itself. The official negotiating machinery was neutralized in 1978-1979 by the pressure for higher wages, but the system still exists. Indeed, the only major "concession" by the government to the demands of workers in 1979 was the announcement that September of a modification to the wages policy that would give lower-paid workers increases in wages beyond that necessary to compensate for the effects of inflation (as officially calculated). While this alteration in the policy recognizes the need to increase the wages of low-paid workers, it is an attempt to maintain the state-controlled system of wage determination by reducing some of the pressure for wage increases beyond the limits established by law. The government was not prepared to allow direct bargaining between unions and employers or to concede the necessary corollary of such bargaining, the right to strike. The unions feel that unless they are able to bargain directly, they will have to accept government control, which always favors the employers. The type of bargaining found in 1979 will remain an exception unless the whole system of wage determination is reformed. The second problem that faces the unions is the question of trade-union organization in the workplace. In 1978 and 1979, many firms responded to the strike movements by firing those workers considered to be the organizers, and as a result of this policy, the unions were weaker by mid-1979 than they had been a year earlier. Unless able to defend their activists, the unions will not be able to establish strong bases in the major factories and will continue to have difficulties in establishing their presence in the smaller ones.

The third obstacle that needs to be removed before the unions can develop their strength much further is the power of the Ministry of Labor to take over unions and to suspend union officers. The intervention in the three unions of the southern industrial belt in March proved to be only the first of a series of interventions in strike movements in 1979. In the case of the metalworkers, the intervention effectively made the strike impossible. They were denied a place to meet en masse; they could not organize effective pickets or keep in touch with the rank and file because they were also denied the use of their communications facilities; they could not use union funds; and their leaders were prevented from effectively running the strike between the day of the intervention and the day of the vote to return to work under threat of imprisonment. Although the unions gradually adapted to these conditions in the course of 1979, there can be no doubt that the extensive powers of the state place major obstacles in the path of workers' mobilizations. By the time of the bank-workers strike in Porto Alegre in September, the union was able to carry on in spite of intervention and the imprisonment of its president, but even here the strike was defeated.12 As further strikes took place, it also became clear that the legal limitations on inter-union support were a problem for unions. The existing labor legislation forbids direct contact between unions, inter-union organizations, and financial help or solidarity strikes by one union in support of another. While the government has not yet enforced all of these restrictions, the threat remains.

It becomes apparent that the pursuit of higher wages and better conditions of work implies a series of changes that are quite extensive. If the unions no longer wish to depend on the state for the determination of their wages and working conditions, then they need free bargaining power. This implies changes in the wages policy and the anti-strike law at the very least. It also implies changes in the limits placed on union activity, which means alterations in the role of the Ministry of Labor and the laws relating to trade union activities, as well as changes in the law on labor stability. In other words, the demand for free collective bargaining is a demand for radical changes in labor legislation and a complete change in the role of the unions from that envisaged in current legislation. In 1978 and 1979, for a variety of reasons it has been possible to suspend the operation of some of this legislation and replace the formal system of industrial relations by a series of ad hoc reactions to labor unrest. However, there has been no change in the system as such. As union leaders have pointed out, the political liberalization has not reached the unions as yet. Therefore, union leaders have seen political changes as a major prerequisite for further progress in industrial relations. They do not hope for reforms from the military government. Instead, many union leaders are looking to the struggle for democratization as the major area of change. With democracy, it is argued, the state will be more responsive to workers' needs, and the space to organize workers and reform industrial-relations practices will be opened up. Therefore, the struggle for wage increases is transformed into a political struggle with a much wider significance.

Auto-workers and the working class

In the preceding section, I have discussed the implications of the events of 1978 and 1979 in which workers in the auto industry played a dominant role. The reasons for their playing of such a role were dealt with earlier. What remains is to discuss the effects of the events of 1978 and 1979 on other sections of the working class and to ask whether or not the kinds of changes put into motion and sought by auto workers will be in the interests of the rest of the working class. If auto workers are more likely than other groups of workers to lead the union movement in the pursuit of changes in the existing system, might not these changes be less desirable for some sections of the working class than for those sections which are seeking them actively?

The effects of the auto workers struggles are not all of a kind, nor do they remain constant over time, and so it is necessary to discuss a series of effects and their implications for the unity of the working class. The immediate effects of the current struggles can be seen as beneficial for wide sections of the class. The stoppages in May 1978 and the strike of 1979 provided the impetus for other struggles over wages, and it has been noted that the example of auto workers was quickly followed in Sao Paulo. In 1979 the strike movement was strongly evident in other industrial centers such as Belo Horizonte, Rio de Janeiro and Porto Alegre, and although the major confrontations were dominated by workers in a limited number of unions - metalworkers, bank workers, teachers, construction workers - it is worth noting that many of the sectors involved have not had histories of militant unionism. In a number of cases the striking workers have been in opposition to their own union leaders who have not wished to express the grievances of their members. The opposition of the major unions, then, stimulated the activity and organization of many other groups. At the same time, it is important to note that the president of the Sao Bernardo metalworkers, "Lula," has acquired the stature of a national trade-union leader because the demands raised by the union under his leadership encounter a resonance among broad groups of workers. He is seen in many areas as a representative of the workers against a government that represents the bosses.

For this same reason, the demands for trade-union reform and the demand for democracy are seen as demands which express the opposition of workers in general to a regime that has denied the working class and the peasantry the benefits of rapid growth. The demand for the right to strike, or the right to elect Presidents and Governors, is seen as a demand for a transfer of power to the people. The Metalworkers of Sao Bernardo and their president obtain widespread support because they have expressed the pent-up grievances of masses of people. However, the fact that in the short term the resistance of the workers of Sao Bernardo may express and be a symbol of the rejection of military rule and low wages by many groups of workers should not hide the fact that there are differences between various groups of workers in Brazil and that the Metalworkers of Sao Bernardo are not merely calling for a return to the situation that existed before the military coup in 1964. Much to the contrary, the theory and practice of trade-unionism in Sao Bernardo is built on a rejection of the populist structures inherited from the Estado Novo period. It is, then, legitimate to ask about the changes sought by the workers in the dynamic industries.

It can be assumed that for the mass of workers, the call for a return to democracy is in itself uncontroversial. What matters is the relations between the trade unions and the state in the democratic period. The aims of the Metalworkers of Sao Bernardo and the other unions grouped around it are quite clear. The general demand is for an end to the control of the Ministry of Labor and the arbitration of the labor courts so that workers and their employers may determine wages and conditions of work by direct collective contract. Along with such a reform would go a greater stress on rank-and-file organization in the plants and the need to unionize a much higher percentage of workers. An example of the reforms that are being demanded can be found in this extract from a series of demands made by the presidents of the confederations of workers in agriculture and in credit establishments when meeting a leading government politician in 1977:

revision of the trade-union structure transplanted from Italian fascism and, therefore, inadequate for a democratic society; an end to the restriction of trade-union freedom by factors such as obligatory dues13 and, principally, by the link between the union organizations and the Ministry of Labor; participation of workers in the fixing of wage increases; permission for free and direct bargaining between workers and employers, which current legislation makes difficult; setting official wage increases as minimums, rather than as the only fixed possibility; reformulation of the FGTS14 , "the real AI-515 of the workers"; liberalization of the right to strike as a "last resort" of the workers (Movimento, October 24, 1977).

This statement of demands is interesting for a number of reasons. The fact that it comes from union leaders outside of the industrial sectors shows that support for reform is not restricted to modern-sector workers, and the time of the statement shows that the desire for change existed before 1978, although it was less effective. The effect of such reforms would be to force unions to be more self-reliant. The exact nature of demands made by union leaders varies - some union leaders would not want to see the trade union abolished completely or all at once, while others would want no minimum wage settlements to be established by law or the labor courts, for example. The general thrust, however, is to abolish both the control and the protection offered by the current legislation.

It has been argued that only certain sections of the working class would gain from such a change. Such arguments are based on two premises. The first is that some workers are less able to organize than others and that therefore they will be prejudiced by a trade-union system that relies on the organization and bargaining power of different groups of workers in separate negotiations. The second is that a form of centralized bargaining, probably linked to the state and probably of a populist form, will better guarantee the standard of living and conditions of work of the weaker-organized sections of the working class. Both of these premises are incorrect.

There are two reasons to suppose that weaker unions will not be prejudiced by the introduction of free collective bargaining. Firstly, the strength obtained by the stronger unions in the course of direct negotiations with the employers will help the working class as a whole to get a better deal from the state in such areas as the right to strike, legislation on stability of labor, the legal enforceability of contracts, etc. Just as in the current period, general advances for the class are made as a result of the ability of certain sections of the class to take the first steps against the present system, so in a democratic period there are a large number of issues upon which decisions will have to be made by the state. Although the state will play a less interventionist role if the current demands for union reforms are met, it will still have considerable influence on industrial relations, and the decisions taken will affect all groups of workers. If it is true that the strength of the most advanced sections of the working class creates general conditions that benefit all of the class, then supposedly weaker sections will gain from free bargaining as well. Secondly, this first point is emphasized and reinforced by the fact that supposedly weaker unions have organized and taken successful strike action in 1978 and 1979. The demonstration effect of the larger unions has led to a resurgence of many of the smaller ones. For example, in July 1979 a small Sao Paulo union in the leather and plastic-goods industries, secured an agreement with the larger employers in the category for wage increase of between 58 and 61 percent for most workers, without any discount of increases secured by strike action in the previous year. This agreement was imposed by the threat of strike action, even though the employers had only offered 44-46 percent originally, and the claim had been decided in the labor court. In spite of the resistance of the labor courts, the Regional Labor Office, and the Employers Federation, the increase obtained for the larger firms in the category was generalized to the smaller firms by actual strike action.16 Similarly, the abilities of unions whose members work in small firms in traditional sectors can be seen in the case of the clothing workers in Porto Alegre. This category is dominated by female employees subjected to bad working conditions in small firms, and yet in 1978 the union was able to obtain factory delegates (one per factory) because of the employers' need to negotiate a new contract for altering shift work. Once the delegates had been elected by the workers, they were able to start to force improvements in working conditions in many factories:

The working conditions for our category are terrible.... the worst that exist ... But in most factories you can see the delegate taking care of working conditions. There are factories where the delegates have already forced changes in conditions. Bathrooms have been put in; water fountains have been put in. All this has been done already. If there isn't a canteen, the delegate claims the legal right to have one.17

It is impossible to imagine that the union concerned would have organized delegates and taken advantage of the employers' problems to obtain a negotiated agreement unless the issue had been raised by stronger unions in the preceding period. The example also demonstrates how even relatively weak sections of the working class can benefit from the implementation of demands made by the leading sectors. Such mutual benefit obtained by unions in their struggles would be further increased if trade unions were reformed and direct contacts between unions were allowed. The second premise also lacks foundation. The widespread support among trade unionists for reforms and the vigor with which many sections of the working class have sought to improve their wages and working conditions indicate that the ability to gain from a trade-union structure tied less to the state may exist among wider layers of the working class than the first premise would suppose. If workers in small industries and the president of the agricultural workers confederation appear to favor reforms, it is not clear which groups would not. However, let it be assumed that there are certain groups of workers whose organization, even within the context of the development of other unions, would not allow them to deal directly with the employers of their category without prejudice to them. Would such groups of workers benefit from a centralized, populist system? The evidence from the fifties and early sixties indicates that they would not. Centralized wage bargaining, or state determination of wages, does not appear to prevent wage differentials from emerging or widening, and after a long period of centralized wage determination stretching back to the 1930s, the plight of the low-paid remains a severe problem. At the same time, there is little evidence that the populist unions were concerned about or able to resolve problems of working conditions. In 1963, Jorge Miglioli was able to write an agitational pamphlet on strikes in Brazil in the series Cadernos do Povo which argued that strikes over conditions of work were no longer necessary because "nowadays working conditions are much better" (Miglioli, 1963: 102). Within the area of working conditions, Miglioli included hygiene, safety, intensity and form of work, hours of work and rest, and personal relations, and in spite of Miglioli's assertion, there is no evidence that either the unions or the labor courts were able to guarantee adequate working conditions for the mass of the working class.

The currently existing structure of Brazilian trade-unionism militates against the solution of both low pay and bad working conditions because it prevents effective rank-and-file organization in the factories and it places obstacles in the path of the generalization of gains from one union to another. At the same time, there is no evidence that even when the state is attentive to the demands of the working class that workers in badly organized categories are helped by the existence of the state-dominated structure. On the contrary, all the evidence suggests that within the state-dominated system, the trade union tax and the efforts of the Ministry of Labor have supported weak union leaders who do not attempt to mobilize their categories and who accommodate the interests of the employers. If workers are too weak to help themselves, there is little chance that the state will help them. In other words, the current system of trade-union organization merely holds back the strong without helping the weak, even during periods when the government is in the hands of populist groups apparently sympathetic to the working class. This is the best-case hypothesis for the second premise. Given that the government in Brazil is unlikely to be sympathetic to the working class in the foreseeable future, the argument against trade-union reform becomes even weaker. If the developing freedom of the union movement does not involve a sudden and dramatic end to state recognition, or the abolition of state financing of unions without some transition period, then the reforms listed earlier in this section would seem to be unlikely to prejudice even the weaker sections of the working class.18

Conclusions

In this paper an attempt has been made to outline some of the reasons why workers in the auto industry, rather than workers in any other section of industry, were the first to openly oppose the state's policies on wages and strikes in the 1970s. It was argued that this had to do with the concentration of workers in large plants, the rapid expansion of the industry in the preceding decade and a half, the erosion of wage differentials between the auto industry and other industries, the influence of skilled workers, and the manner in which the union representing the majority of auto workers - the Metalworkers of Sao Bernardo - had developed. In other words, the causes of the behavior of auto workers were found to be their special situation that allowed them a greater possibility of opposing management and the state. The leading role played by auto workers cannot be put down to the specific problems of the auto industry and the specific solutions sought by auto workers to them. If the latter had been the case, it would have been reasonable to expect to see a split between the activities of auto workers and those of the rest of the working class. In fact, an examination of the strike movements in 1978 and 1979 confirmed the earlier arguments in the paper by showing that the demands raised by auto workers were taken up by other groups of workers and that the struggle for free bargaining and the end to being under the tutelage of the state involved wider struggles for democracy and freedom for workers to organize themselves that would be generally supported by the working class. In order to argue that this confluence of interest between auto workers and the rest of the working class is based on more than a temporary sharing of opposition to dictatorship, the final section of the paper discussed the issue of the effects of union reform on the supposedly weaker-organized groups, and it was shown that the demand for liberalization of state control over the unions and free collective bargaining might well be of benefit to all workers.

It can never be stated that conflicts and divisions will not arise within the working class. Indeed, working-class history in many countries suggests the opposite. However, it has been shown in this paper that some of the arguments for expecting a cleavage between workers in the dynamic sectors - of which auto workers are a large and salient part - and workers in the more traditional industries are unfounded, and that at the present time there exists a large area of common interest. If divisions do appear within the Brazilian working class - and indeed there are serious splits at the present time within the ranks of the union leaderships - there is little reason to expect that the line of cleavage will be between the traditional and dynamic sectors of industry.

References

• Almeida, Maria H. T. de 1978 "Desarrollo capitalista y acción sindical," Revista Mexicana de Sociología, LV (April- June), 467492
• Bacha, Edmar L. 1975 "Hierarquia e remuneração gerencial," pp. 124-155 in Tolipan and Tinelli (eds.), A controversia sobre a distribuicao da renda e desenvolvimento, Rio de Janeiro: Zahar
• Cara a Cara 1978 "Os operarios tomam a palavra," I (July-December), 8-53
• Cardoso, Fernando H. 1972 "Dependency and development in Latin America," New Left Review, 74 (July-August), 83-95
• DIEESE (Departamento Intersindical de Estatistica e Estudos S6cio-Economicos) 1975 10 anos da politica salarial, Sao Paulo: DIEESE 1977 "Distribuiçao salarial em Sao Paulo segundo gúia de contribuicao sindical," unpublished document
• Gouldner, Alvin 1954 Wildcat Strike, New Jersey: Antioch
• Humphrey, John 1977 The Development of Industry and the Bases for Trade Unionism: A Case Study of Workers in the Car Industry in Sao Paulo, Brazil, unpublished Ph.D. dissertation, Sussex University 1979, "Labor in the Brazilian Motor Vehicle Industry," unpublished paper
• IBGE (Instituto Brasileiro de Geografia e Estadistica) n.d. Censo Industrial, Brasil, 1960, Rio de Janeiro: IBGE 1974a Censo Industrial, Brasil, 1970, Rio de Janeiro: IBGE 1974b Censo Industrial, 1970, Rio de Janeiro: IBGE 1977 Pesquisa Industrial, Centro-Sul, 1974, Rio de Janeiro: IBGE
• Lobos, Julio A. 1976 Technology and Organization Structure: A Comparative Case-study of Auto and Processing Firms in Brazil, unpublished Ph.D dissertation, Cornell University
• Miglioli, Jorge 1963 Como sao feitas as greves no Brasil?, Rio de Janeiro: Civilização Brasileira
• Quijano, Anìbal 1974 "The Marginal Pole of the Economy and the Marginalised Labor Force," Economy and Society. III (October-December), 393-428
• Rodrigues, Leoncio M. 1970 Industrialização e atitudes operarias, Sao Paulo: DIFEL Latin American Perspectives: Issue 23, Fall 1979, Vol. VI, No. 4

• 1'The trade-union structure in Brazil was established in the 1930s and 1940s. The state defines both the categories of workers to be grouped into unions and the geographical area which a particular union of a given category will cover. Once designated as the union for a given category in a given area, the union has sole rights. In other words, there is no plural unionism. The Ministry of Labor not only recognizes unions, it also has the power to suspend union directors, veto candidates for union office, establish the general pattern of union constitutions, check on union expenditure, and collect a levy on all workers who belong to the category in the relevant area and pass part of this on to the union concerned. The latter is called the contribuicao sindical (union contribution), although it was known for a long time as the imposto sindical (union tax). By an accident of geographical location, most auto plants are in one area of Sao Paulo, called Sao Bernardo do Campo, and auto workers dominate the local metalworking union. Although all workers pay the imposto sindical, they are only members of the union if they join and pay a separate monthly due.
• 2The information on real wage rates is complicated by the existence of different price deflators and the lack of many reliable long-run series of figures for industrial wages. However, the work of the Trade Union Research and Statistical Department in Sao Paulo (DIEESE, 1975) and of Bacha (1975) seem to indicate that real wages for most industrial workers at best kept pace with the rate of inflation after 1964 and probably fell in the case of unskilled and semi-skilled workers. Recently computed and as yet unpublished figures from DIEESE indicate that real wages for many workers fell in the decade 1966-1976 (DIEESE, 1977).
• 3The source of information about the metalworking unions are calculations made by DIEESE. I am grateful for their considerable help. Latin American Perspectives; Issue 23, Fall 1979, Vol. VI, No. 4
• 4The median has been taken because an average wage figure would be distorted by the earnings of top management. The median figure gives a rough indication of the earnings of the better-paid semi-skilled workers.
• 5A study in the United States which examines the effects of a shift in management outlook and behavior on the attitudes of workers is Gouldner (1954). The bureaucratic management attitude is not, of course, confined to large firms.
• 6One senior manager from the auto industry described the effects of financial targets on labor relations and plant management personnel. There are target figures for the total wage bill for given periods and for wages as a proportion of total costs. Meeting these targets imposes two necessities on plant and labor relations personnel. Firstly, labor costs have to be reduced when output is reduced, which means that labor employed follows the fluctuations in output, which follow the volatile market for new car sales. Secondly, the overall target for the total wage bill limits the number of workers who can gain seniority and the extra payments that are associated with it. To keep the total wage bill under control, or within the specified limits, long-service workers have to be dismissed. These policies are referred to as "rotation of labor," and they are intensely disliked by the unions and the workers in the auto industry (see Humphrey, 1977: 112-114). In general, labor relations has been a low priority for auto firms in Brazil, and the prestige and importance of Industrial Relations Departments were found to be ranked in last place (along with, significantly enough, Research and Development) in a hierarchy of importance, behind finance, sales, manufacturing and services, according to a survey of managers in three Brazilian auto companies (Lobos, 1976: 198-201).
• 7The rates of unionization are taken from union figures for December 1978 and management figures for three of the five large auto companies for July 1979.
• 8This information has been obtained through discussion with union directors and consultation of the union's archive.
• 9In the state of Sao Paulo, the metalworkers of the city of Sao Paulo and the neighboring industrial areas of Osasco and Guarulhos have agreements that run from November to November. Almost all the other metalworking unions in the state have agreements that run from April to April, and in.the past they have negotiated jointly. The metalworkers of Sao Bernardo did try to start separate negotiations with the employers in 1975 so that they could obtain a contract that attended to the need of the auto industry, but this attempt was unsuccessful because of Ministry of Labor pressure. The unions that negotiate in April are collectively referred to as the "Metalworkers of the Interior."
• 10A good series of accounts of the strike movement in many factories described by workers who took part can be found in "A Greve na Voz dos Trabalhadores," published by Alfa-Omega in 1979 in their series of booklets entitled Historia Imediata. Another useful source for information and analysis is a debate with union leaders in the journal Cara a Cara (1978). See the interview with Luis Inacio da Silva, this issue.
• 11For an analysis of the negotiations and the effects of the wage settlement proposed, see Movimento (May 14, 1979). Accounts of the strike and subsequent events can be found in many newspapers and journals. The weekly magazines, Isto E, Veja, Visao, Movimento, and Em Tempo provide possibly the most informative coverage.
• 12The bank workers of Porto Alegre were in a rather different position. There are relatively few of them, for example, but they did learn from the experience of the metalworkers.
• 13This is a reference to the contribuicao sindical, which is mentioned in footnote 1 above.
• 14The FGTS is the Fundo de Garantia do Tempo de Servico, which is the law introduced in 1966 regulating stability of employment and compensation for dismissal without due cause. It is widely regarded as a measure which encourages the rotation of labor (see footnote 6 above).
• 15AI-5 refers to Ato Institutional 5, the basis of repressive military rule introduced in December 1968. It gives extremely wide powers to the President and suspends many civil rights. The Institutional Act has been abolished since October 1977, but many of its provisions remain elsewhere on the statute book.
• 16This information is taken from the newspaper of the union concerned.
• 17This quote is taken from an interview with the president of the union some two months after delegates had been introduced, in October 1978. I am grateful to Werner Wirtele of the Latin American Institute at the Free University of Berlin for making this and other interviews available to me and other researchers.
• 18At the present time, one of the strategies of the bourgeoisie will be to either claim that weaker workers need the state-controlled system of industrial relations, or seek such a rapid dismantling of the current structure that unions will not have time to adapt to the new situation.