Until recently, most historians and social scientists writing about twentieth-century Brazil have attributed programs for social reform and concomitant social control to the state, particularly the regime that came to power with the “revolution of 1930” under the leadership of Getúlio Vargas. The corporatist vision of Vargas and his closest advisors has been credited with producing the social reforms (in the form of legislation) and mechanisms for labor control (in the form of government-sponsored unions) that laid the groundwork for rapid industrialization with minimal social disruption. Though many students of the Vargas period characterize these reforms as ultimately benefiting the employers, not the workers, according to most historical accounts the industrialists themselves played either a peripheral or retrogressive role in the process. Lacking a more sophisticated or forward-looking project for reform and control, the industrialists allegedly clung to the notion that “the social question is a question for the police,” and bitterly opposed the labor reforms instituted by the Vargas regime.1

There is both historical evidence and a variety of theoretical positions that encourage this interpretation of the industrialists’ role (or the lack thereof), especially with regard to those manufacturers based in Brazil’s burgeoning industrial center, the state of São Paulo. Most of the prominent figures in the Centro de Indústrias do Estado de São Paulo (CIESP) supported, and even participated in the leadership of, the 1932 Paulista revolt against the Vargas government. Presumably, a major motive for doing so was their hostility to Vargas’s labor reforms.2 On a more theoretical level, important figures on both the right and the left in Brazil have argued that a strong state was necessary in order for Brazil to make a full transition to a modern, industrial society since the “national bourgeoisie” was either nonexistent or too weak to play its historic role.3 Even scholars who have attempted to incorporate the industrialists into the process of state formation in the 1930s portray their representatives as politically weak and unable to forge a “hegemonic project,” subordinating themselves, instead, to an estado de compromisso.4 It is this “compromise state,” allegedly autonomous from domination by any specific social group, that forges a viable project for industrialization and social control.5

The “strong state-weak bourgeoisie” dichotomy, as well as the estado de compromisso model, have been criticized in a number of recent works that perceive a new political and social orientation among industrialists, particularly in São Paulo, even before the advent of the Vargas regime.6 Indeed, there is mounting evidence to show that leading figures in São Paulo’s emerging industrial bourgeoisie (and their technocratic allies) were trying to develop a more coherent approach to the problem of industrialization and class conflict as early as the 1920s. The writings of Roberto Simonsen (especially O trabalho moderno, 1919) are most often cited as a sign of the new orientation within the industrialists’ ranks, but lesser-known figures, such as Euvaldo Lodi, Armando de Arruda Pereira, Roberto Mange, Mariano Ferraz, and Antônio de Souza Noschese—all of them, like Simonsen, trained as engineers—were writing and working in a similar vein. The very founding of CIESP in 1928, with the full collaboration of such powerful industrialists as Francisco Matarazzo, can be viewed as a sign of the Paulista industrial elite’s growing commitment to an economic and social project of its own making.7

To be sure, the “project” that they were gradually developing was not identical to the policies later implemented by Vargas. Most of the major industrial figures continued to be hostile to government attempts to “intrude” itself into relations between capital and labor on the factory level—unless in the repressive role of breaking strikes and “keeping order.” Prominent members of CIESP, for example, vowed to resist the courts’ intermittent attempts to enforce the child labor law in the late 1920s.8 This campaign of resistance found its most articulate promoter in Octávio Pupo Nogueira, a minor industrialist who served as secretary of both CIESP and the Paulista textile manufacturers’ association during these years. Well into the 1930s, Pupo Nogueira argued against the need for extensive labor legislation. His strikingly reactionary positions, such as his argument that social legislation was the “true incubator of class struggle” in Brazil, seem to confirm the idea that the industrialists were opposed to any and all innovations in the area of labor relations.9

Pupo Nogueira’s arguments may have been representative of a certain faction of the industrial bourgeoisie, but he was a marginal figure in the group being formed by Simonsen, and there is considerable evidence of alternative points of view within the CIESP leadership.10 Documents written by lawyers retained by the center, and distributed to the CIESP membership in January 1929, argued that without protective labor legislation workers would always be at such a disadvantage in negotiations with employers that peaceful resolution of differences would be impossible; thus such legislation was necessary, even if it restricted freedom of contract.11 Moreover, leading members of the industrialists’ ranks expressed considerable interest in new strategies for social control on the factory floor emanating from the United States. Taylorism, Fordism, industrial psychology (psicotécnica in Brazil) offered these industrialists “scientific” methods to increase production, regiment labor, and reduce workers’ control over the work process, and implied a more complex approach to labor discipline than the paternalistic or overtly repressive methods then in practice.12

For the “progressive” industrialists, the key goals with regard to production itself were mechanization and rationalization of the work process, which, according to Simonsen, would greatly increase productivity and allow for a simultaneous decrease in the prices of manufactured goods and an increase in workers’ wages. This, in turn, would ensure greater “social peace,” a particularly pressing concern following the widespread general strikes in São Paulo in the years from 1917 to 1919. Simonsen assured his readers that “rationalization has profound social effects and clearly acts against the fundamental ideas of Marxism.” Moreover, “rationalization” could be applied to more than just the work process: the modern factory, with an in-house cafeteria organized according to scientific notions of nutrition, with a medical staff in residence and greater attention to safety issues, would produce a hardier, more efficient, and less dissatisfied labor force. All this, of course, was to be accomplished by a newly conscious industrialist class on a voluntary basis, and not as a result of government legislation.13

Hostility toward certain government-sponsored social reforms persisted into the early 1930s. The industrial slump caused by the Great Depression made it particularly difficult for industrialists to swallow legislation that would cost them money—for example, the Lei das Férias, which required factory owners to indemnify workers for vacations not granted in 1931. Indeed, the government’s insistence on this indemnification, at a time of widespread strike activity, directly contributed to the CIESP leadership’s decision to join the abortive Paulista rebellion against Vargas in 1932.14 Not that the industrialists opposed all such laws or were excluded from the lawmaking process. CIESP and other industrialist organizations were sent copies of legislative proposals by the Ministry of Labor, which gave them the opportunity to suggest alterations or amendments that would make the bills more amenable to business interests. And certain legislative proposals—such as the revised (and weakened) child labor law and restrictions on women workers—met with complete approval from CIESP.15 Still, throughout the early 1930s, and perhaps the entire period preceding the Estado Nôvo (1937-45), São Paulo’s industrial spokesmen harbored an abiding suspicion of Vargas’s social legislation. And not surprisingly, since Vargas’s policies during the first few years of his administration gave no clear indication of his commitment to industrialization. Further, the industrialists feared that Vargas’s apparent need to mobilize popular support would lead him to promote social reforms without regard for the employers’ interests.16

The Paulista industrialists’ organizations also continued to exhibit an almost pathological aversion to intervention into the factory routine, however mild, whether by government officials or union leaders. Pupo Nogueira, writing in 1935, claimed that the Departamento Estadual de Trabalho in São Paulo had imposed “unnameable violations” on the factory owners.17 And an attempt by union leaders to serve as “inspectors” in the factories to ensure implementation of a particular labor law provoked an enraged protest from the owners that persisted even after the government sided with them against the union officials. Indeed, the Federação das Indústrias do Estado de São Paulo (FIESP), as CIESP was now known, circulated a notice on this subject in which it vowed to fight “the greatest evil that could harm us, that is, the weakening of the employer’s authority.”18

Despite these earlier suspicions and conflicts, by the late 1930s the leading industrial spokesmen had shifted to a position of almost unqualified support for the Vargas administration. There were still areas of dispute between the industrialists and the government, and some employers may have accepted the new order as a fait accompli, resolving to make the best of it.19 Nevertheless, it is important to stress that the manufacturers’ new position was not a complete reversal or break with the past, since major figures like Simonsen, Horácio Lafer, and Euvaldo Lodi (first president of the Confederação Nacional da Indústria [CNI]) had supported elements of Vargas’s social program from the outset. The shift in industrialist opinion reflected, rather, an acceptance of the role of the state in promoting a particular project for social control. In other words, those manufacturers who collaborated with Vargas during the Estado Nôvo had abandoned the essentially Fordist idea that rationalization of work processes and working-class life could occur solely on a voluntary basis in individual factories, or that the problem of social control could be left to the factory owners without the mediating role of the state.

It is understandable that the Vargas regime of the Estado Nôvo was considerably more appealing to industrial leaders than the Vargas regime of the early 1930s. Whereas Vargas’s attitude toward industrialization had seemed nebulous, at best, during the early period, he now displayed definite intentions of promoting industrialization, advocating roles for the state and private initiative that echoed the views of São Paulo’s industrial elite.20 The blatantly authoritarian nature of the Estado Nôvo, meanwhile, meant that Vargas would not be obliged to implement social reforms to mobilize working-class votes in his favor. The rank-and-file of the manufacturing class must also have been mollified by the increasing recognition that Vargas’s labor legislation afforded many loopholes to the employer. For example, the eight-hour-day law allowed workers and employers to negotiate a four-hour extension of the work day on a “voluntary” basis. The later legislation of the Vargas era, such as the creation of mediation boards that formed the basis for the system of labor courts, also offered a highly attractive alternative to open class conflict, especially since the composition of the boards gave the manufacturers an obvious edge in the negotiations.21 Similarly, Vargas’s emphasis on unionization was far less alarming to employers in the late 1930s and early ’40s—a time when federal police agents attended union meetings and a federal ban prohibited Communists from holding union office. Finally, the two main labor ministers of the Estado Nôvo years—Waldemar Falcão and Alexandre Marcondes Filho—were both considered to be very friendly to the interests of the industrialists, and especially those in São Paulo.22

At the same time, the industrial leadership may also have become more skeptical of its ability to persuade its less socially conscious brethren to implement even very meager reforms and participate in social programs on a voluntary basis. The circulars distributed to all FIESP members during the 1930s contain various indications of the leadership’s exasperation or disappointment with the responses of the “rank-and-file.” For instance, during the 1932 São Paulo revolt, in which FIESP actively participated, the industrialists created a series of medical and dental clinics to provide free or cheap care for industrial workers. Their obvious intention was to stimulate greater support among the workers, who were generally cool toward the campaign against Vargas, by demonstrating the elite’s generosity and benevolence, and the original plan was to disband the clinics as soon as the revolt ended. However, the program was so well received by the workers that the FIESP leadership attempted to keep the posts in operation well after the end of the conflict by soliciting voluntary contributions from its membership. As it turned out, very few employers were willing to pledge funds for the endeavor, and the clinics had to be closed. Other social projects, such as the attempt to form an insurance cooperative to cover work-related accidents, produced similarly disappointing results. Indeed, the FIESP leadership repeatedly voiced its irritation with its membership for not returning questionnaires it distributed, even when the purpose was clearly in the employers’ best interests.23

The campaign to promote more rational work processes and working conditions in the factory also met with frustration for the most part. In 1933, the Instituto de Organização Racional do Trabalho (IDORT), a private organization in São Paulo that had many connections with FIESP, initiated a study of industrial lighting that the FIESP leadership strongly endorsed. Thus, a circular went out to the membership urging factory owners to welcome IDORT representatives into their firms and to assist its research in any way possible. Given the rather innocuous issue at hand —lighting on the factory floor—and the unthreatening nature of the organization doing the research, one would expect the FIESP request for support to be merely pro forma. Yet the language of the circular indicates otherwise. It all but pleaded with members to allow the IDORT representatives access to their factories, and assured them that the recommendations made would not have to be implemented. Moreover, the final results of the research, when made public, would not discuss individual factories. In short, every effort was made to portray the IDORT project as beneficial assistance to the factory owner, and not as a form of intervention or negative publicity.24

Both the changing perception of the Vargas regime and the realization that the average industrialist would not voluntarily participate in more ambitious programs for social reform and control contributed to a new orientation within the industrial leadership. This consisted of a greater acceptance of the legitimate role of the government in helping to reorganize industrial life and oversee relations between capital and labor. But even those industrialists who collaborated most closely and enthusiastically with the central government—Simonsen, Lodi, Armando de Arruda Pereira, and Morvan Dias de Figueiredo—strove to limit government influence and control in certain sectors of special interest to the industrial bourgeoisie. In contrast to the trend in other industrializing (and industrialized) nations, the Brazilian manufacturing class struggled to keep two key functions usually delegated to the state under the direct control of the industrialists themselves: worker training and industrial social services.

The severe inadequacy of social services, whether state or private, for industrial workers was (and still is) painfully apparent, and compounded the impact of persistently low wages for most factory workers. The FIESP leadership viewed the shortage of such services as contributing both to discontent among the workers and to the deficiencies of the Brazilian workforce with regard to health and education. Studies produced during the 1930s by students of the Escola Livre de Sociologia e Política— which Simonsen helped found—and other institutions in São Paulo cited the need for more rational and “hygienic” working-class housing and recreation. (Indeed, Simonsen himself had begun his business career constructing “casas operárias” in the port city of Santos.) Generally speaking, the objective of such studies was not to reveal injustice or argue for better wages, but to suggest supposedly efficient solutions to these problems and remove them from the arena of class struggle.25

The inadequacy of the existing system of worker training and apprenticeship was also apparent to the industrialists by the 1920s. Typically, young workers acquired skills not through a formal apprenticeship, but by observing older workers on the factory floor. Employers and educators complained that such informal, on-the-job training militated against innovations in the work process. The leading exponent of standardized vocational instruction, Swiss-born engineer Roberto Mange, claimed that the traditional manual training process tended to reproduce “bad habits” among the workers, and could not be expected to instill the sense of discipline demanded by the new, more “rational” industrial methods.26 Articles in the proletarian press, meanwhile, also lamented the limited opportunities available to working-class children interested in acquiring manual skills. Industrialists, they charged, often hired minors as poorly paid “apprentices,” who, instead of receiving systematic instruction, were simply assigned to the most unskilled and menial tasks.27

Whether from the industrialists’ or the workers’ point of view, the few existing professional (i.e., vocational) schools offered a poor solution to the problem. Few working-class families could afford to send their adolescent sons or daughters to school full time, or even provide them with the educational preparation necessary for entry into such “professional schools” as the Liceu de Artes e Ofícios de São Paulo or the Escola Profissional Masculina. Employers, meanwhile, accused the schools of giving their students too formal a preparation and of being too remote from the actual routine on the factory floor. At the very time that Simonsen and Mange were extolling rationalization and standardization in all aspects of industrial life, the director of São Paulo’s Escola Profissional Masculina steadfastly refused to reorganize the school’s workshops (which sold their wares to the public) along the lines of a factory. Such an approach, he argued, meant that

[i]n order to earn more money, we should adopt a plan of absolutely separate parts: one student will make the legs of a table; another, the drawers; another, the handles; another will do the sanding and staining. In this way, by the end of the year, the school will have made an enormous profit, but this will have been accomplished by exploiting child labor and by doing just what the industrialists do.28

Finally, the vocational schools, as they existed in the 1920s and ’30s, could not possibly supply São Paulo’s factories with the necessary number of skilled workers. During the 1930s, manufacturers protested against the labor law requiring two-thirds of industrial workers to be Brazilian citizens, arguing that the shortage of highly skilled workers made such restrictions intolerable.29

Despite the complaints of leading industrialists, and demands from labor unions for more access to vocational instruction, few alternative forms of training emerged. The one development in this area that did earn high praise from industrialists and like-minded educators was the center for training railroad mechanics founded by Mange. Initiated in the 1920s as a special mechanics course in the Liceu de Artes e Ofícios, by the mid- 1930s the Centro Ferroviário de Seleção e Ensino Profissional (CFESP) had become a key source of skilled workers for São Paulo’s vast railroad network. Mange and his colleagues in IDORT, meanwhile, treated the center as a laboratory for rational methods of worker selection and training. Violating the customary prerogative of railroad workers to enroll their sons as apprentices, Mange based entrance into the courses on “aptitude” tests administered by the center. The courses themselves, taught by special instructors, utilized the “methodical series” approach to training perfected in Germany whereby apprentices begin with the simplest process or piece, repeat the task until perfected, then move in a highly regimented fashion to more and more complicated tasks. Mange assiduously publicized CFESP’s efforts, staging a competition in 1931 between his apprentices and those trained by traditional means to demonstrate the superiority of the center’s rational methods.30 Throughout the 1930s, CFESP trained only a few hundred apprentices each year, making a minimal impact on the pool of skilled labor in São Paulo. Its importance lies, rather, in the alternative it offered to the traditional methods of worker training— one that emphasized standardization and the reduction of worker control over the training process.

The Vargas government took the first serious step toward creating a more systematic form of worker training in 1934 with the formation of an interministerial commission charged with making proposals on this matter. All nine members of the commission were professional educators, but at least three (Mange, João Lourenço Filho, and Horácio Silveira) had close ties to the Paulista industrialists as well as being active in IDORT. The minutes of the commission’s meetings, and the documents produced by its members, demonstrate a concern for industrial training that went well beyond its narrow economic or functional value. There seemed to be universal agreement that the low educational level of the Brazilian worker, who had, on average, less than two years of schooling, meant not just a lack of skill, but a lack of discipline. The statist technocrats who dominated the commission also emphasized the workers’ poor “moral and civic formation,” a phrase that seems to have been their shorthand for an array of perceived deficiencies within the proletariat, ranging from personal vices to insufficient identification with national goals. Therefore, the country urgently needed apprenticeship and training programs that transcended the goal of strictly manual instruction: “The education of the worker, whether ministered in government schools, or those maintained by industry, must assure training that is moral, intellectual, physical, social and civic, as well as technical.”31 Predictably, the commission looked to more developed nations to locate a program that could be adapted to Brazilian conditions; the program that first emerged as a model for Brazil was the apprenticeship system then in operation in Nazi Germany.

Commission member Rodolpho Fuchs, an unabashed admirer of the Third Reich, visited Germany in July 1938 to participate in an international congress on vocational education, and returned with a formal report that all but waxed euphoric over the German apprenticeship system. Under this system, all male workers who entered the factory workforce at the conclusion of their primary education (usually age 14) also entered a compulsory apprenticeship program that consisted of on-the-job training and six to eight hours of formal classroom instruction per week. The program included case-by-case vocational guidance to direct apprentices, as much as possible, into professions where there was a shortage of skilled laborers. Though carefully controlled by the state, most of the funding for the program came from a tax on industrial employers. Fuchs argued that this program was an excellent model for Brazil since it avoided the high costs and nonindustrial settings of the vocational high schools, while the factory school would serve as the “school of discipline par excellence.”32 The latter remark, and several other comments included in the report, make it evident that Fuchs found the German apprenticeship system attractive not only because of its efficiency in producing skilled workers, but also because it presumably produced an industrial workforce that closely identified with the goals of the state.

Four years later, when Minister of Education Gustavo Capanema spoke at the inauguration of the Serviço Nacional de Aprendizagem Industrial (SENAI)—or National Industrial Training Service—he referred to Fuchs’s report as the “seed document” for the organization.33 Indeed, the central notion of combining formal instruction with factory-floor experience, at the heart of Fuchs’s report, did emerge as a key characteristic of SENAI. Other aspects of Fuchs’s recommendations, however, met with more critical responses. Much as Fuchs had predicted, industrialists’ organizations (and their close allies, such as IDORT) objected to the idea of obligatory apprenticeship for all workers. In the late 1930s, IDORT published a report, probably authored by Mange, that argued that only 15 percent of industrial workers needed apprenticeship-type training; the rest were either crude manual laborers or machine tenders who could learn the required “skills” in a matter of weeks.34 Apparently, the IDORT group was primarily concerned with the functional aspects of apprenticeship training, whereas Fuchs focused first and foremost on training as a vehicle for discipline, social control, and worker integration into the state-directed project for national development.

In May 1939, the Vargas government issued the first decree-law for vocational training, based on the commission’s recommendations. Clearly a trial balloon, and not a law that the government was determined to enforce, Decreto-Lei No. 1.238 required all factories with five hundred or more workers to provide free training programs that would involve part-time work and part-time instruction. Again, opponents of the decree-law (both industrialists and educators) argued that the majority of workers in large factories needed no formal training. Furthermore, they argued that it was precisely in the larger plants, most of them textile factories, that mechanization and assembly-line production had reduced the need for highly skilled workers.

To investigate this claim, the entire commission visited twelve factories in the state of São Paulo. In eleven of them they found mechanization and production “in series” to be the rule; in only one factory—with less than five hundred workers—did they find the proportion of skilled workers in the labor force to be substantial.35 Reflecting these findings, the “Conclusions and Suggestions” drawn up by the commission after the visit speculated that only 10 percent of the factory workforce could be characterized as skilled, and therefore requiring apprenticeship. What preparation was necessary for the remaining 90 percent of the workforce to become “good workers”? “It is evident that one cannot expect more than primary schooling, and from that the creation, at the proper age, of habits of hygiene, of work, of discipline, of cooperation, of thrift.” As something of an afterthought, the commission further suggested that reading, writing, and elementary arithmetic would be “useful instruments” for such workers.36

By 1940, a new proposal had been made to replace Decreto-Lei No. 1.238. Following the guidelines of the earlier IDORT report, it recommended that all factories (regardless of size) enroll a certain percentage of their workers in apprenticeship or adult training programs, with the understanding that the percentage for each industry would be gauged according to the present and future needs of that sector. The state would be primarily responsible for the operation of the program, although the burden of financing would be shared by the government, the employers, and the workers, whose representatives would sit together on the national commission.37

Both Euvaldo Lodi, president of the Confederação Nacional da Indústria (CNI), and Simonsen of FIESP responded positively to this new government initiative. Simonsen, however, did suggest some further alterations, the thrust of which was to greatly increase the role of the industrialists, reduce the role of the state, and eliminate the workers’ participation altogether. Simonsen even suggested that the industrialists take sole responsibility for funding the new program, although it would mean an “onerous burden” for the employers. Above all, he stressed the need for greater engagement or interaction between factories and training centers.38

The highly influential minister of education, Capanema, immediately protested the new arrangement, which became official on January 22, 1942, with a decree-law creating SENAI In a letter to Vargas he argued that reducing the role of the state in worker training violated Article 125 of the 1938 constitution, which affirmed the principle that “education is the duty of the parents and the state,” and Article 129, which asserted that vocational instruction is “primarily the duty of the state.”39 Fuchs, in a letter to Capanema written at the time of SENAI’s inauguration, was far more vociferous in his criticism of the dominant role granted to the private sector. “The training of industrial workers,” wrote Fuchs, “has not become … the great protective armor of the Brazilian adolescent who works, but only an institution for the technical improvement of the manual labor required by industry.”40 Despite such protests, Simonsen’s position triumphed.

To be sure, the structure of SENAI did not completely eliminate the state from its operations. A government decree created SENAI, and government entities would receive and disburse the compulsory payroll tax on industrial employers that would fund SENAI. The actual management of the service, however, was to be subordinated to the CNI on the national level, and the state industrialist federations, such as FIESP, on the local level. This meant that the content and design of SENAI instruction would, indeed, be determined by the needs of industry.

The way in which the question of worker training evolved, and the ultimate “victory” of the industrialists, tells us a great deal about the way in which the relationship between the manufacturers and the state evolved during the Estado Nôvo. The close collaboration between the industrialists, led by Simonsen and Lodi, on the one hand, and the Vargas government on the other, as well as the government’s increased sensitivity to the industrialists’ demands, shows their increasing compatibility, in terms of both worldview and interests. However, one can also detect a clear intention on the part of the industrial leaders to maintain some distance and autonomy from the central government. Students of industrialization have normally assumed that industrialists will want to “socialize” the costs of vocational education and other services by delegating responsibility for their funding and functioning to the state.41 Leading Brazilian industrialists, though, clearly believed that the extra costs associated with direct control of industrial training would be compensated by their ability to adapt that training to the needs of industry, without government interference. Industrialist control would also mean that even symbolic worker or trade union participation in the program could be eliminated.

Although the government retained token representation on the governing boards of SENAI, the service became the bailiwick of the industrialists and such technocrats as Mange, João Luderitz, and Joaquim Faria Góes Filho. The SENAI schools were named, almost without exception, after leading industrial magnates, and the SENAI literature repeatedly emphasized that workers now had the opportunity to get training or retraining, while on the job, due to the generosity and progressive spirit of the industrial elite.42 The structure of SENAI also allowed it to be more responsive to employers’ needs than the existing vocational institutions. For example, because of the pressure of the war years, when the shortage of skilled labor in certain strategic industries became particularly acute, SENAI temporarily shifted its priorities from two- and three-year programs for young apprentices to rapid adult retraining courses. After the war, SENAI-São Paulo regularly created special programs at the behest of particular firms, such as the glassmaking course offered in the early 1950s for employees of Nadir Figueiredo, S.A., a course for furnace stokers prompted by a fuel shortage, or the school for textile workers founded in the interior city of Jundiaí at the request of local businessmen.43 SENAI could also certify new or existing training programs in large factories as meeting the agency’s standards, and thereby exempt their owners from payment of the 1 percent payroll-tax contribution.

The early reports and literature of SENAI-São Paulo devote considerable space to distinguishing SENAI training from that offered by the traditional, government-funded vocational schools. Students in the latter institutions supposedly received an “eclectic” preparation more suited to the needs of supervisory personnel or small, independent producers, whereas SENAI students received highly specialized training that prepared them for a specific role in the production process.44 SENAI had also been organized to erase any “false” division between school and factory. The apprentices were to be designated by employers from among juvenile workers; then, once in the program, the SENAI student would alternate between classroom and factory floor. This had the additional advantage of allowing young workers to keep their jobs while being trained, but employers had the right to pay apprentices only half of the minimum wage.45

Aside from the actual instruction in manual skills, SENAI (especially in São Paulo) offered several ancillary services and activities: regular medical and dental checkups, courses in Portuguese and general culture, “moral and civic formation,” extracurricular sports and patriotic celebrations, free meals and snacks, etc. Mange, who was director of SENAI-São Paulo from its inception in 1942 until his death in 1955, argued that it would be a mistake to think of it simply as an agency for industrial training. Rather, “the moral, civic, and social education of the apprentices should be a constant concern of those who collaborate with SENAI.”46 In his writings, Mange frequently proclaimed the agency’s commitment to an “integral education” that would produce a worker who was healthy, disciplined, and patriotic, as well as competently skilled.

In the social and familial atmosphere inhabited by the SENAI student, he finds little encouragement to improve his general culture and to elevate his civic and moral concepts … [thus the student’s] preparation will be incomplete unless the school’s coursework is placed within a context of order and discipline, of morality and happiness in work, that would seek to create in these youths, who are the workers of tomorrow, a sense of responsibility and an elevated interest in the progress of technology and the national labor force.47

Despite the heavy emphasis Mange and his acolytes placed on social services, moral and civic education, and wholesome extracurricular activities, “socialization” for SENAI apprentices stemmed mainly from the training courses themselves. SENAI made vocational instruction into an extension of the factory, rather than the school, and designated its instructors, rather than other workers, as the appropriate source of manual training. It defined skills—lathe operator, tool and die maker, adjuster, welder, leather cutter, molder, woodworker—as narrowly as possible, to reflect the advanced division of labor characteristic of “rational” forms of production. It reinforced existing gender divisions in the workplace, not only by channeling female apprentices almost exclusively into textile courses, but by barring them from courses intended to prepare foremen for the textile industry.48 And it promoted an industrial hierarchy rigidly composed (in ascending order) of unskilled labor (braço anatômico), semi-skilled labor (braço atento), skilled labor (braço pensante), and supervisory labor (braço pensante e dirigente).49 In the words of Euvaldo Lodi, spoken at the inauguration of SENAI’s Escola “Roberto Simonsen”: “The perfect order, the exact punctuality, the irreprehensible cleanliness, the constant obedience, the sense of hierarchy in the industrial schools of SENAI constitute living lessons that all our young students can imbibe.”50

Ironically, the final conception of SENAI, for which Mange bore substantial responsibility, limited its usefulness as a socializing institution even under the best of conditions. Mange, Simonsen, and others had struggled to advance the idea that only 15 percent of the factory workforce needed apprenticeship-type training, or retraining. One of Mange’s colleagues from IDORT went so far as to argue (in an article reprinted in the Boletim SENAI) that “excessive” vocational training would make workers unfit for most jobs in industry since only unskilled, “ignorant” workers would tolerate the monotony of assembly-line production.51 Yet these assumptions meant that SENAI, even if operating at its intended capacity, would be recruiting and socializing only a small minority of urban workers. True, this skilled minority could be seen as playing an economically and politically crucial role in the process of industrialization. But could the industrial elite afford to neglect the socialization of the remaining 85 percent of their labor force (and working-class housewives and children, for that matter)?52

Apparently, the men who had emerged during the Estado Nôvo as the leading industrialist spokesmen decided that they could not afford to do so. During the final year of the Estado Nôvo, Simonsen began to formulate plans for another nationwide entity, the Serviço Social da Indústria (SESI). Officially inaugurated in 1946, this organization, like SENAI, was initiated by a federal decree, but the industrialists assumed sole responsibility for its funding and direction. (As in the case of SENAI, funding came from a payroll tax, in this instance of 2 percent.) Far more than SENAI, SESI was the brainchild of Simonsen and his coterie, representing the successful realization of the long-running campaign by the Paulista industrial leadership to set up privately funded and controlled medical and social services for workers. Also, unlike SENAI, SESI’s intended clientele was the entire working-class community, and its orientation was explicitly political. According to Simonsen, “SESI … will enable the Brazilian working masses to cross the Red Sea of oppressive and inhumane totalitarianism without wetting their feet in it, and, after the undoubtedly arduous journey, [the workers] will breathe the clean Brazilian air, purified by our civic spirit and by our vocation for democracy.”53

The services SESI offered during its early years can be divided into two types. First there were the directly “assistance-oriented” activities, that often represented quick responses to problems causing worker “unrest.” Even before SESI became an official entity, it began to set up discount food posts throughout the city of São Paulo and its metropolitan area to help resolve the problem of postwar food shortages that in some areas had caused riots by the workers. The pôstos de abastecimento were also strategically located to compete with similar entities founded by Communist party militants. In one of the earliest meetings of SESI-São Paulo’s regional board, Acting President Armando de Arruda Pereira announced that “with the installation of discount food posts in certain zones of the city, we have managed to close six Communist posts.”54 Other assistance-oriented projects undertaken by SESI included a chain of medical and dental clinics, recreational centers, and several central kitchens to provide hot meals to factories that did not have their own cooking facilities.

Nearer and dearer to the hearts of SESI’s founders were its more explicitly educational functions. Within a year of its founding, SESI-São Paulo had trained hundreds of “social educators” who were charged with giving lectures and courses in factories, union headquarters, and SESI centers on such matters as labor legislation, unionization, rational organization of work processes, safe work habits, and social relations in the workplace. SESI also trained literacy teachers to offer courses in factories, and home economics teachers who offered working-class housewives and daughters instruction in “proper” techniques of childcare, nutrition, cooking, home decoration, etc. in the Centros de Aprendizado Doméstico. SESI’s directors always gave primary emphasis to such activities, and even sought to emphasize the educational aspects of assistential services. For example, the balanced meals served by SESI’s central kitchens supposedly taught the workers the rudiments of a nutritional diet. And social workers at the medical and dental posts were expected to help SESI’s clients incorporate proper health habits into their daily routine.55

Not all industrial employers proved eager to make SESI’s educational services available to their workers, and many of the early social educators found themselves physically barred from workplaces by the very people who were contributing to SESI.56 But SESI did provide the more “enlightened” Paulista industrialists some means to influence conditions inside the factories of the less “enlightened. ” Aside from the actual programs and services sponsored by SESI, any firm that could show, on inspection, that it already provided similar types of facilities (discount food stores, low-cost cafeterias, literacy classes, minimal safety conditions, and resident medical and dental personnel) would receive a reduction in the monthly payment it owed SESI, up to a maximum of 50 percent.57 And, in a broader sense, the payroll surtax collected by FIESP and other state industrial federations to fund SESI activities certainly increased the power and influence of these organizations.58

As happened in the case of industrial training, once again the industrialists took it on themselves to perform functions that in other societies are delegated to the state and, especially, the public education system. Accepting the limited schooling experience of the average worker as a fait accompli, SESI took it on itself to compensate for the resulting “deficiencies” in the worker’s socialization. The labor unions were, of course, another potential source of information and socialization, but not one that the industrialists could endorse wholeheartedly.59 SESI also attempted to compensate somewhat for the gross inadequacies of state-sponsored health care and recreational facilities, and even, to a lesser extent, for the scarcity of adequate low-income housing. As for the government, Vargas did issue certain decrees to goad the industrialists into action on such matters as providing cafeterias in factories with more than five hundred workers. But both Vargas and his successors generally acted to encourage independent efforts by the industrialists. This forms an interesting contrast with the administration of Juan Perón in Argentina, which zealously promoted government-sponsored services and vocational training programs to ensure greater worker loyalty and identification with the Peronist state.60

It is difficult to determine whether Simonsen and company took on these functions, or a symbolic performance of them in many cases, because the state manifested no intention of providing similar services on an adequate scale, or whether they did so because they felt themselves more capable and reliable than the government and wished to garner the political credit for such “reforms.” Given the enthusiasm with which FIESP’s leaders tackled the project, it seems probable that they felt themselves, or their organization, to be the proper vehicle to promote these social services. Thus, even during the presidency of Eurico Dutra (1946-51), when the industrialists gained direct control over the social policies of the central government (Dutra’s choice for his second minister of labor was the FIESP president, Morvan Dias de Figueiredo), the leading industrialist spokesmen preferred autonomous ventures to state-directed activities.61

Direct control over a wide array of educational and social services for urban workers was also attractive as a means to construct a new image of the industrialists as progressive and enlightened, and ready to create an alliance with labor in the name of national development and social peace. It meant that activities that might be “manipulated” by unfriendly politicians or independent labor leaders, if sponsored by the state, would consistently reflect the views and objectives of the industrial bourgeoisie. For example, soon after its founding SESI moved to defuse the potentially explosive issue of industrial accidents by forming a team of consultants whose mission was to recommend solutions for workplace hazards, and by circulating a bimonthly publication devoted to the problem of occupational safety.62 In a more political vein, during the 1950s SESI began using its extensive recreational facilities to organize parades and “workers’ Olympics” for May Day, with the explicit intention of transforming a traditional commemoration of worker militancy into a celebration of “social peace.”63 Thus, minimizing the role of the state in certain areas meant greater ideological and organizational control for the industrialists, and meant that they, and not some potentially unreliable politicians, would get the credit for such ventures.64

It is not that industrialists shunned electoral politics. Simonsen ran for federal senator in 1947 and used SESI employees to help recruit working-class support for his campaign.65 But even the most supportive administration might prove untrustworthy due to the pressures involved in electoral politics and all of its ramifications. In the case of Dutra, that pressure came ultimately from the United States and its postwar economic policies, which caused a contraction in Brazil’s industrial sector. In the case of Dutra’s successor, the now popularly elected Getúlio Vargas, the pressure came from Vargas’s working-class supporters and their demands for real improvements in their standard of living. Apparently, the industrial leadership felt that it was worth an extra payroll tax of 2 or 3 percent to allow SENAI and SESI independence from such pressures. Moreover, the fact that SENAI and SESI had been created by federal law, and that the contributions from industrialists were therefore compulsory, resolved the issue of how to recruit support from the industrialist rank- and-file who had been less than forthcoming when funding was solicited on a voluntary basis.66 In short, SENAI and SESI had been organized to combine the best elements of both worlds: the state’s capacity for coercion and the private sector’s preference for autonomy.

The process which spawned SENAI and SESI clearly raises some questions about the strong state-weak bourgeoisie model that has informed so much of the literature on the industrialists’ role in Brazilian development. At the very least, it demonstrates that the industrialists, or more correctly the industrialist leadership, played a much more active role in the formation of labor policy than previously assumed, and did not simply abdicate intellectual and social functions to the state. It also raises some questions about the strength of the state that emerged after 1930. Certainly, the Vargas regime created a much more interventionist state than had existed under the Old Republic (1889-1930), but its relations with the industrialists indicate that there were serious limits to its capacity to create and control services that could be crucial to the state’s image and power.67 The limited resources, both political and fiscal, of the Vargas regime and its immediate successors meant that the industrialists, not the state, would be the major source of industrial training and social services, however inadequate, for several decades. In strictly financial terms, this may have been beneficial to the state, but it meant losing potential means to solidify worker support and identification with a supposedly new state dedicated to national development.

Many of the recent debates about the nature of the post-1930 state in Brazil have examined the notion that the Vargas regime inaugurated an estado de compromisso. As noted above, this concept has been developed by several distinguished Brazilian social scientists who have argued that no single social group or class was able to establish political hegemony after 1930; thus, various competing factions ceded control to a strong, autonomous state that did not rule in the name of a specific class or political group, but gained legitimacy by mobilizing urban popular support.68 Recently, a series of studies inspired by a critique of the “strong state” have taken the estado de compromisso idea to task, arguing that the authoritarian state that emerged from the crisis of the 1920s, with its emphasis on industrialization and national development, represented the triumph of the industrial bourgeoisie and its economic and social project.69

The documentation generated by FIESP during the 1930s and ’40s generally supports the latter view. The power and influence of the FIESP leadership, even during the early 1930s, was quite impressive, and the years of the Estado Nôvo clearly mark a victory for the industrialists’ conception of the role of the state in conflicts between capital and labor. Yet, without resurrecting the estado de compromisso model, or the strong state-weak bourgeoisie dichotomy, we need to acknowledge the nature and limits of the industrial bourgeoisie’s hegemonic position during these years. First, just as it is a mistake to assume that one can generalize about the working class as a whole from the activity and consciousness of the labor leadership, we cannot simply generalize about the industrial bourgeoisie from the pronouncements and publications of its leadership. FIESP officials constantly had to battle the apathy of the organization’s membership, as well as the members’ reluctance to endorse any revision of labor relations that might involve obligations or concessions on the employers’ part. Thus, it was its very relationship with the state that made the industrial bourgeoisie “stronger,” since the leadership was able to use legal mechanisms to enlist the rank and file in its efforts, and to create a more cohesive and homogeneous bourgeois class.

Even more important, we should not overestimate the political power or confidence of the industrialists. They viewed even the friendliest politicians as susceptible to making concessions to workers, concessions that could undermine both the economic and social objectives of the industrial bourgeoisie. The result was a considerable ambivalence toward the state among the leaders of FIESP and a genuine fear that the state could become an instrument to undermine the industrialists’ hegemony and to promote the interests of labor.70 It is this ambivalence, and this threatening aspect of the state in its political incarnation, that is missing in the critiques of the estado de compromisso. Only by acknowledging the complexities of the relationship between the state and the bourgeoisie, and the latter’s genuine insecurity about its position even in the limited democratic system of postwar Brazil, can we understand the industrialists’ strong commitment to maintaining control over activities elsewhere relegated to the state, and weak commitment to an open political system.


Among the distinguished works that make this argument, see Boris Fausto, A revoiução de 1930, 8th ed. (São Paulo, 1982), 12-50; Warren Dean, The Industrialization of São Paulo, 1880-1945 (Austin, 1969), 151-206; and Hobart A. Spalding, Jr., Organized Labor in Latin America (New York, 1977), 178-185. I will not explore the role of the labor movement in this process or the ultimate impact of Vargas’s policies on that movement. It is implicit in my argument that the new labor legislation was largely a response to pressures and demands from the workers themselves. It should also be noted that projects for “social control” rarely produce the precise results intended by their formulators.


The best source on the industrialists’ role in the 1932 revolt is the circulars of the Centro/Federação das Indústrias do Estado de São Paulo (hereafter CIESP) in the Biblioteca Roberto Simonsen, São Paulo. See also Clovis de Oliveira, A indústria e o movimento constitucionalista de 1932 (São Paulo, 1956) and Dean, Industrialization, 181-196.


Two influential works arguing for a strong state from a conservative perspective are Francisco Campos, O estado nacional, sua estrutura, seu conteúdo ideológico (Rio de Janeiro, 1941) and Francisco José de Oliveira Vianna, Instituições políticas brasileiras, 2 vols. (Rio de Janeiro, 1974). From a leftist perspective, see Fernando Henrique Cardoso, Empresário industrial e desenvolvimiento econômico no Brasil (São Paulo, 1964); Luciano Martins, Industrialização, burguesia nacional e desenvolvimiento (Rio de Janeiro, 1968); and Florestan Fernandes, A revolução burguesa no Brasil (Rio de Janeiro, 1974). For a critical discussion of this shared assumption of the left and right, see Marilena Chauí, “Apontamentos para uma crítica da Ação Integralista Brasileira,” in Ideologia e mobilização popular, Chauí and M. S. Carvalho Franco, eds. (Rio de Janeiro, 1978), 19-30.


This argument can be found in the stimulating essay by Eli Diniz, “Empresário, estado e nacionalismo: Ideologia e atuação politica nos anos trinta,” in Empresariado nacional e estado no Brasil, Diniz and Renato Raúl Boschi, eds. (Rio de Janeiro, 1978), 45-107. Two other important works portray the industrialists as mainly trying to exercise a “veto power,” with limited success, at least until the Estado Nôvo. See Ángela Maria de Castro Gomes, Burguesia e traballio: Política e legislação social no Brasil, 1917-1937 (Rio de Janeiro, 1979) and Marisa Sáenz Lerne, A ideologia dos industriais brasileiros, 1919-1945 (Petrópolis, 1978), 125-158.


Studies which develop the concept of the estado de compromisso include Fernandes, A revolução burguesa; João Manuel Cardoso de Mello, O capitalismo tardio (São Paulo, 1982); and Sônia Draibe, Rumos e metamorfoses: Estado e industrialização no Brasil, 1930-1960 (Rio de Janeiro, 1985). For an interesting critique of the estado de compromisso concept, see Maria José Trevisan, 50 anos em 5: A FIESP e o desenvolvimentismo (Petrópolis, 1986), 15-41.


The strongest argument to this effect can be found in Edgar Salvadori de Decca, 1930: O silêncio dos vencidos (São Paulo, 1981), 135-182. Two other studies that reject the notion of a weak bourgeoisie are Paulo Sérgio Pinheiro and Michael M. Hall, A classe operária no Brasil, 2 vols. (São Paulo, 1979–81), II, 9-14 and Emília Viotti da Costa, “Brazilian Workers Rediscovered,” International Labor and Working Class History, 22 (Fall 1982), 28-38. Also, aside from the question of the industrialists’ role, it is clear that several key pieces of labor legislation preceded Vargas’s seizure of power. See Castro Gomes, Burguesia e trabalho, 85-107.


A collection of Simonsen’s early essays, Á margem da profissão (São Paulo, 1932), presents several key speeches and writings from this period, including “O trabalho moderno” (1919) and “Orientação industrial brasileira” (1928), the speech he gave at the founding ceremonies of CIESP. Among the many laudatory biographies of the industrial engineer Roberto Mange, written by his disciples, see Ítalo Bologna, Roberto Mange, um pioneiro (São Paulo, 1972). For biographical information on Arruda Pereira (Simonsen’s collaborator in business and politics) and Ferraz, see “Figuras exponenciais do SESI,” SESI Jornal (São Paulo), Mar. 31, 1948, p. 4 and Aug. 31, 1948, p. 3. On Souza Noschese, see Boletim Informativo da FIESP, 681 (Oct. 4, 1962), 592. Simonsen, Arruda Pereira, and Ferraz, in particular, shared several characteristics: they were engineers turned businessmen, had done some studies in the United States, and were active members of Rotary Club International. John French has an illuminating discussion of Simonsen’s emerging project for social control, which French calls “welfare capitalism,” in “Industrial Workers and the Origin of Populist Politics in the ABC Region of Greater São Paulo, Brazil 1900-1950” (Ph.D. diss., Yale University, 1985), 198-200.


The 1926 “Código de Menores” was particularly objectionable to employers, since it set a minimum age of 14 and a maximum work day of six hours. Its effective application would have eliminated child labor from industry. Ironically, the child labor law later passed under Vargas was much less stringent and, hence, much more acceptable to the industrialists. See Circulares do CIESP (in Biblioteca Roberto Simonsen [São Paulo]), 32 (Apr. 2, 1929), 42 (May 15, 1929), 45 (May 29, 1929), 47 (June 4, 1929), 53 (June 14, 1929), 55 (June 21, 1929), 61 (July 30, 1929), 71 (Sept. 30, 1929), 80 (Jan. 30, 1930), 206 (Oct. 26, 1931) and Sindicato dos Industriais de Fiação e Tecelagem (São Paulo), “Relalório do exercício 1931–1932,” 15–17 (in Arquivo do CIFT). As early as 1921, São Paulo’s Centro dos Industriáis de Fiação e Tecelagem (CIFT) sent a request to the federal government to initiate a carteira profissional that would be required of all industrial workers and would facilitate the process of social control. Pinheiro and Hall, A classe operária, II, 197-201; CIFT (São Paulo), “Relatório do exercício 1928-1929,” 9.


Octávio Pupo Nogueira, A indústria em face das lets do trabalho (São Paulo, 1935), 3. This author also argued against the Lei das Férias of 1926, arguing that only intellectual work was fatiguing, and that workers with free time on their hands would be a threat to the social order. A indústria, 52–70.


At least one major figure in CIESP, the textile manufacturer Jorge Street, went beyond Simonsen and his coterie in the 1920s by advocating recognition of labor unions and collective bargaining. Street’s 1919 article to this effect has been reprinted in Edgard Carone, ed., O pensamento industrial no Brasil (1880-1945) (Rio de Janeiro, 1977), 398-403. For an excellent discussion of Street’s position see French, “Industrial Workers,” 192-194.


Circular do CIESP (Jan. 21, 192.9). This and another document, signed by the lawyers Plínio Barreto and Antônio Mendonça, cite the Versailles Treaty as establishing a role for government in the regulation of industrial relations. This position, however, did not prevent CIESP’s lawyers from mounting arguments against specific pieces of labor legislation.


De Decca, 1930: O silêncio, 135-182. Taylorism, standardization, and assembly-line production became subjects of discussion and debate in Brazil surprisingly early. According to the standard sketches of Simonsen’s career, he first applied Taylorist principles in a Santos construction firm in the 1910s, and Mange was credited with introducing rationalized training methods in the 1920s. “Figuras exponenciais do SESI,” SESI Jornal, Mar. 31, 1948, p. 3; Bologna, Roberto Mange, 1-4. In a talk given in 1920, writer and educator Afrânio Peixoto discussed Taylorism (“A educação nacional: Aspectos femininos,” repr. in A. Peixoto, Ensinar a ensinar [São Paulo, 1937], 62). In 1929, CIESP (together with the government of São Paulo) sponsored a series of lectures on Taylorism and closely related topics. Circular do CIESP, 66 (Aug. 26, 1929). Two years earlier, the newspaper of São Paulo’s graphic workers had denounced “rationalization” as the “war cry of the bourgeoisie.” O Trabalhador Gráfico (São Paulo), Aug. 1927, p. 7.


Simonsen, “O traballio moderno,” 1, 11, 12, 35. The Marxism quote is from Simonsen, As finanças e a indústria (São Paulo, 1931), 43, cited in de Decca, 1930: O silêncio, 165. For a stimulating discussion of the movement to “purify the workplace,” see Margareth Rago, Do cabaré ao lar (Rio de Janeiro, 1985), 32–47.


Circular do CIESP (FIESP), 114 (Nov. 7, 1930), 117 (Jan. 3, 1931), 130 (Apr. 18, 1931), 246 (Feb. 17, 1932), 260 (Apr. 1, 1932). Even after federal forces defeated the Paulistas, FIESP requested that the implementation of the Lei das Férias be postponed due to the disorganization of local industry. Circular, 334 (Oct. 10, 1932), 335 (Oct. 15, 1932), 341 (Oct. 28, 1932).It should be noted that in May 1931, in response to Vargas’s decree-laws creating sindicatos and state-level federations, CIESP also became known as the Federação das Indústrias do Estado de São Paulo (FIESP).


The most detailed statement of the FIESP leadership’s position on the various legislative proposals can be found in a letter (probably authored by Simonsen) sent by FIESP to Minister of Labor Lindolfo Collor. Circular da FIESP, 228 (Dec. 17, 1931). Despite a variety of criticisms and objections, the letter ended by declaring that the members of FIESP “are unanimous in warmly applauding the urgent effort being made by Your Excellency with the intention of bestowing on our nation social legislation whose high, noble, and patriotic goal is to provide aid and protection to the worker.”


On the Paulista industrialists and Vargas’s early economic policies, see Dean, Industrialization, 181-206. As for Vargas’s relations with the labor movement, Pupo Nogueira (as always, articulating the most conservative response) claimed, in the early 1930s, that most industrialists originally favored the creation of a ministry of labor, but soon saw that it was only interested in courting the masses. A indústria em face das leis, 43.


Pupo Nogueira, A indústria em face das leis, 34-35.


The law to be enforced was the decree requiring two-thirds of all factory employees to be Brazilian citizens. Circulares da FIESP, 430 (July 19, 1933). 451 (Oct. 10, 1933).


French, in emphasizing the continued differences between Vargas and the FIESP leadership, argues that scholars have wrongly concluded that “the Estado Nôvo’s social and labor policies were … supported by Brazilian employers.” But his work exaggerates the industrial leadership’s opposition to the Vargas administration’s unionization drive and compulsory union tax, and confuses points of dispute with lack of support. “Industrial Workers,” 214, 229-231.


The brutal repression of the left following the failed uprising of the Aliança Nacional Libertadora (1935) undoubtedly contributed to industrialist confidence in the Vargas regime, but it is also clear that the FIESP leadership actively supported most of the measures affecting industrial relations instituted during this period. Sáenz Leme, A ideologia dos industriais, 148-158. As for Vargas’s attitude toward industrialization, I agree with those who argue that his government’s commitment to national economic reorganization with an emphasis on industrialization preceded the Estado Nôvo. An excellent discussion of this can be found in Joan L. Bak, “Cartels, Cooperatives, and Corporatism: Getúlio Vargas in Rio Grande do Sul on the Eve of Brazil’s 1930 Revolution,” HAHR, 63:2 (May 1983), 255-275. However, this commitment became much more apparent to the industrialists in the mid-1930s. Draibe, Rumos e metamorfoses, 100-119; Diniz, “Empresariado,” 74-105.


Circular da FIESP, 41/39 (May 31, 1939).


The circulares of FIESP during the period of Falcão’s ministry (1937-41) contain virtually no critical comments about the minister of labor. Falcão’s right-wing Catholic condemnation of “the Jewish spirit of voracious capitalism” as the cause of social unrest, however, may have alienated some industrial leaders, such as Horácio Lafer, who was Jewish. Marcondes Filho, on his part, was a Paulista by birth, and apparently had some close friends in São Paulo’s business circles. His rhetoric and policies also strongly emphasized class collaboration. Fundação Getúlio Vargas (hereafter FGV), CPDOC, Arquivo Falcão, 41.00.00/2, Discurso ao Congresso de Direito Social sobre a Questão Social” (1941); Arquivo Marcondes Filho, 44.11.07, “Discurso na Inauguração da V Feira Nacional das Indústrias—São Paulo” 1944), 44-12 08, “Primeiro Congresso Brasileiro de Indústria” (1944), 44.02.01, letter from Wallace Simonsen (1944), and 45.11.05, letter from Morvan Dias de Figueiredo, president of FIESP (1945). The latter thanks Marcondes Filho “for the services provided to Paulista industry.”


Circulares da FIESP, 327 (Sept. 5, 1932), 333 (Oct. 7, 1932), 343 (Nov. 3, 1932), 473 (Dec. 26, 1933), 514 (May 3, 1934), 518 (May 28, 1934), 551 (Aug. 20, 1934), 567 (Sept. 28, 1934). Circular 567 returned to the subject of an insurance cooperative despite the “complete indifference” shown by Paulista industrialists to the idea.


Circular da FIESP, 393 (Mar. 3, 1933). For a discussion of the results of this study, see IDORT, Atas da 10a reunião do Conselho Consultivo, May 31, 1937.


The Escola Livre had many ties with FIESP and IDOET. Not only was Simonsen one of its founders, but Mange and other active members of IDORT served on its faculty. On the founding of the Escola Livre, see Charles O’Neil, “Educational Innovation and Politics in São Paulo: 1933-34,” Luso-Brazilian Review, 8:1 (Summer 1971), 56-68. On studies of working-class living conditions by students at the Escola Livre and similar institutions, see Maria A. Guzzo Decca, A vida fora das fábricas: Cotidiano operário em São Paulo, 1920-1934 (Rio de Janeiro, 1987), 50-57.


Mensagem enviado ao congresso legislativo de São Paulo pelo governador dr. Altino Arantes (São Paulo, 1917), 10-11; Mensagem do Gal. Waldomiro Castilho de Lima como Interventor Federal do Estado de São Paulo (São Paulo, 1933), 281; Horácio Silveira, O ensino technico-profissional e doméstico em São Paulo (São Paulo, 1935), 13-15; Relatório dos trabalhos da Escola Profissional Masculina da Capital (São Paulo, 1924), 8-9. The director of the Escola Profissional condemned informally trained workers as incapable of “generating new formulas” for industry, but also admitted that his school’s own teaching staff had poor preparation, and that there was a very high drop-out rate among its students. For Mange’s criticisms of traditional apprenticeship methods see Mange, Relatório do Serviço de Ensino e Seleção Profissional, Estrada de Ferro Sorocabana, ano de 1931 (São Paulo, 1932), 10-18, and A formação de técnicos para a indústria (São Paulo, 1940), 4; Hamilton Galli/Centro Roberto Mange, Origem e evolução do ensino profissional ferroviário no Brasil (São Paulo, 1967), 1-3; and Paulo Ernesto Tolle, Retrospecto, realizações e problemas do Departamento Regional do SENAI, de São Paulo (Rio de Janeiro, 1972), 3-4.


O Trabalhador Gráfico, Jan. 1906, pp. 1-3, Sept. 26, 1920, p. 4. An article on p. 1 of the Mar. 7, 1926 issue of O Trabalhador Gráfico, by a “young worker,” demanded “technical instruction paid for by the bosses”: an article in the Aug. 1927 issue (p. 2) lamented “the conditions under which [apprenticeship] is currently done, without method and without following any criteria.”


Aprígio de Almeida Gonzaga, Relatório dos Trabalhos da Escola Profissionai Masculina (São Paulo, 1920), 23.


On the industrialists’ objections to the Lei dos Dois Terços, see Circulares da FIESP, 762 (Aug. 25, 1936) and 788 (Nov. 26, 1936).


Mange, Relatórios do Serviço de Ensino e Seleção Profissional, Estrada de Ferro Sorocabana, 1930-1933 (São Paulo, 1934). The Revista IDORT, in its Jan. 1932 issue, published a detailed account of Mange’s 1931 “experiment,” which the Centro Ferroviário (as it was known after 1934) reprinted in pamphlet form as Ensino profissional racional no curso de ferroviários de Sorocaba (São Paulo, 1936). On Mange’s center and its relation to labor conflicts, see Liliana R. Petrilli Segnini, Ferrovia e ferroviários (São Paulo, 1982), 63-82.


Documents relating to the functioning of the interministerial commission (Labor and Education) can be found in FGV, CPDOC, Arquivo Gustavo Capanema, 34.11.28 (Organização do ensino profissional), 35.10.18/2 (Organização do ensino industrial), 38.04.30 (Aprendizagem industrial). The passage cited can be found in the minutes of the sixteenth meeting of the committee, on July 25, 1939 (38.04.30, doc. III–1).


FGV, CPDOC, Arquivo G. Capanema, 35.12.00, doc. I–10, “O ensino profissional na Alemanha” (1938). Fuchs also extolled the strict separation of the sexes in German vocational education, which produced “feminine women and real men.” On the apprenticeship program in Nazi Germany, see John Gillingham, “The ‘Deproletarianization’ of German Society: Vocational Training in the Third Reich,” Journal of Social History, 19:3 (Spring 1986), 423-432.


FGV, CPDOC, Arquivo G. Capanema. 41.09.13. doc. I–15.


Mange, A formação de técnicos, 8-9; FGV, CPDOC, Arquivo G. Capanema, 38.04.30, doc. IIa-1, “A aprendizagem nos estabelecimentos industriais,” by Joaquim Faria Góes Filho (1939); doc. IIb-1, Ata da 9a reunião, June 30, 1939.


FGV, CPDOC, Arquivo G. Capanema, 38.04.30, doc. IIa-1.


Ibid., doc. IIa-1, part VI, p. 2.


Ibid., docs. IIb–1 and IV-2.


Ibid., doc. IIa–1, letter from Simonsen to Gustavo Capanema, June 7, 1940; 41.09.13, doc. I-1, report from Euvaldo Lodi to Capanema, Sept. 13, 1941. Lodi, as president of the Centro Industrial Brasileiro, attended the Aug. 25, 1939 meeting of the commission and presented the industrialists’ position for three solid hours. Philippe Schmitter notes that “Simonsen had ingeniously turned the impending obligation into a political asset …” (Interest Conflict and Political Change in Brazil [Stanford, 1971], 184).


Capanema drew up an alternative proposal, according to which the government would contribute a certain amount of money annually and retain more control, but to no avail. Ibid., 38.04.30, doc. IIb–1, letter from Capanema to Vargas.


Ibid., 41.09.13, doc. I-6, letter to Capanema from Fuchs, July 21, 1942. Aside from Fuchs’s letter, an unsigned 1942 memorandum, included among the SENAI documentation in the Arquivo Capanema, argues for more direct government control of the entity through a Comissão Nacional de Educação Professional to be composed of educators, not industrialists. In addition, it critiques the “regionalist spirit” of SENAI: according to the service’s regulations, most of the funds were to be used in the region where they were collected even though, according to this anonymous author, it was the poorest states that needed such programs the most. Ibid., 41.09.13, doc. I–2.


On industrialization and education in the United States, see Samuel Bowles and Herbert Gintis, Schooling in Capitalist America (New York, 1976) esp. 191-195.


See, e.g., SENAI-São Paulo (hereafter SENAI-SP), Comemorações do jubileu de prata (1942-1967) (São Paulo, 1968).


Mange, Missão do Serviço Nacional de Aprendizagem Industrial (São Paulo, 1943), 6; SENAI-SP, Relatório do Departamento Regional de São Paulo, 1942-43 (São Paulo, 1944), 20; “Cursos Ordinários,” Informativo SENAI (São Paulo), 5:54 (July 1950), 3; Circulares da FIESP, 75/44 (June 6, 1944), 109/44 (Aug. 7, 1944).


Mange, Missão do SENAI, 2-3; Planejamento e administração unificada da aprendizagem no Brasil (São Paulo, 1949), 6-8; SENAI-SP, Relatório (1944), 9; SENAI-SP, O SENAI completa 20 anos (São Paulo, 1962), 3.


For further discussion of SENAI in a more contemporary context, see Ralph Edfelt, “Occupational Education and Training: The Role of Large Private Industry in Brazil” and Cláudio de Moura Castro, “Academic Education versus Technical Education: Which is More General?,” in Educational Alternatives in Latin America, Thomas J. La Belle, ed. (Los Angeles, 1975), 384-413: 434-461.


SENAI-SP, Relatório (1945), 10.


Mange, Planejamento, 14–16.


Informativo SENAI, 6:60 (Jan. 1951), 2. An early annual report described “domestic education for female students, to prepare them for the “services of the home.” It noted that the SENAI schools found time to teach the young women cooking and sewing by cutting back on their classes in math and Portuguese. SENAI-SP, Relatório (1945), 29-30.


Mange, Missão do SENAI, 1.


Euvaldo Lodi, Positivos os indícios de que a criação do SENAI foi um ato acertado da indústria (São Paulo, 1949), 5.


A. C. Pacheco e Silva, “A fádiga industrial,” Boletim SENAI (Rio de Janeiro), Nov. 1946, P. 13. This author also claimed that women would find such labor less monotonous than men, offering as evidence the fact that women can knit for hours without feeling fatigued.


Joaquim Faria Góes Filho, the member of Vargas’s commission on vocational training who authored the document arguing against universal apprenticeship (and a future national director of SENAI), recommended that unskilled and semiskilled workers be provided with courses in literacy, hygiene and nutrition, civic and moral education, and general technical knowledge. FGV, CPDOC, Arquivo G. Capanema, 38.04.30, doc. IIa–1.


Simonsen, Discurso pronunciado a 25 dejulho de 1946 em São Paulo, na instalação do l° Conselho Consultivo do SESI (São Paulo, 1946), 14.


Largely at the urging of Simonsen and other representatives of FIESP, Federal Decree-Law no. 7.249 authorized factories with over three hundred workers to set up discount food stores that could sell basic items at no more than 10 percent over the wholesale price, but were exempt from paying taxes. FGV, CPDOC, Arquivo Marcondes F°, 45.10.20, doc. 1; Circulares da FIESP, 19/45 (Jan.19, 1945), 24/45 (Jan 30. 1945) The selection of locations for subsequent posts, outside the workplace, was determined not only by fear of growing Communist influence among urban workers, but also by specific political goals, such as gaining popular support for Simonsen’s campaign for the federal senate in 1947.On competition with Communists, see Arquivo Geral do SESI-São Paulo (hereafter AGS-SP), Atas do Conselho Regional do SESI, July 10, 1947 and Robert J. Alexander, Labor Relations in Argentina, Brazil and Chile (New York, 1962), 108. On Simonsen’s campaign, see French, “Workers and the Rise of Adhemarista Populism in São Paulo, Brazil 1945-47,” HAHR, 68:1 (Feb. 1988), 11–12.


The best summary of SESI’s early activities can be found in SESI—18 anos, a series of pamphlets published by SESI-SP from 1964 to 1966 describing the alleged accomplishments of the various divisions and subdivisions, including “Abastecimento,” “Alimentação,” “Assistência Social,” “Higiene e Segurança Industrial,” “Educação Fundamental,” “Melhoria da Saúde,” and “Orientação Social.”


Interview with Hélvio Pinheiro Lima, São Paulo, July 14, 1986. The interviewee, now head of SESI-SP’s public relations department, began working for SESI as a social educator in 1946.


A social educator would visit a factory, evaluate its social services, and recommend a percentage for exemption. In some cases, the company resisted making any payments to SESI, arguing that its own facilities allowed it to dispense with SESI services. See, for example, the correspondence between SESI and the Companhia Nitro-Química Brasileira in AGS-SP, processo 34/616, Aug. 4, 1947, Aug. 26, 1947, Sept. 29, 1947, and Sept. 17, 1949. Several SESI employees claimed that factory visits were often perfunctory, with exemptions granted according to the owner’s influence, but I did find a number of cases where exemptions were denied or set at a much lower level than requested by the employer. See, e.g., the correspondence between Pirelli S.A. and SESI, processo 53/618, Mar. 6 and 16, 1948.


For an interesting discussion of the way in which SENAI and SESI funds increased the power of the state-level industrialist federations, see Schmitter, Interest Conflict, 182–186.


SESI’s attitude toward the labor unions was ambivalent, to say the least. From the outset, SESI-SP carried out a vigorous campaign to encourage workers to join unions, believing that a larger membership would dilute the influence of Communist activists. In addition, SESI subsidized activities and services of unions that were regarded as “friendly.” At the same time, SESI subsidized Catholic “Círculos Operários” as an alternative to the unions, and its directors became particularly enraged when the owner of the Companhia Taubaté Industrial decided to grant direct subsidies to his workers’ union rather than use SESI services. AGS-SP, Atas da Diretoria, July 12, 1947, Jan. 23, 1948; processos 79 (Correspondência com Sindicatos), processo 125/188, Mar. 28, 1950.


Despite the typically negative tone, I think Robert Alexander was correct when, in comparing social activities of industrialists in Brazil and Argentina, he contended that “Perón was not anxious to have the employers provide special services for their workers—he wanted the workers to feel that all their blessings flowed from him and from the trade-union movement which he controlled.” Labor Relations, 18.


Circular da FIESP, 44/48 (Apr. 24, 1948). This circular includes an account of Pres. Dutra’s recent visit to the headquarters of the CNI during which he specifically praised the work of SENAI and SESI.


SESI’s Serviço de Higiene e Segurança Industrial published CIPA Jornal from 1950 to 1975. “CIPA” stands for Comissão Interna para Prevenção de Acidentes, a form of committee created by Vargas’s labor legislation in the early 1930s, and required in all factories with more than one hundred employees. Since these committees included employer representatives as well as workers, they were usually of limited utility in improving safety conditions in the workplace.


On the May Day “Jôgos Operários,” see SESI Jornal, May 31, 1948, p. 8, and French, “Workers and the Rise of Adhemarista Populism,” 33, n. 87. For a positive account of the SESI initiative, see O Metalúrgico (São Paulo), May 1947, p. 1. The latter was the official newspaper of the São Paulo city metalurgical workers’ union, which had just been subjected to government intervention. For a much more negative account, see O Trabalhador Gráfico, Apr.-May 1954, pp. 1-2, which described the Jôgos Operários as a “legacy of fascism.”


It is hardly surprising that the literature of SESI and SENAI constantly emphasized that their services were the result of the industrialists’ generosity and initiative. But it must have been especially pleasing to the industrial leadership when Everardo Dias, former socialist activist and graphic worker, praised SENAI (and its commercial counterpart, SENAC) in his História das lutas sociais no Brasil (São Paulo, 1962). According to Dias, such courses “should have been the culminating idea, thought, effort of the labor government of Getúlio Vargas, and instead became the idea and work of an enlightened, advanced, progressive, understanding, and truly nationalist bourgeoisie …” (212-213).


There are reports from the social educators who participated in Simonsen’s campaign in AGS-SP, processo 3/114, Jan. 13, 1947. According to French, although Simonsen managed to get himself elected federal senator, he did very poorly in the industrial suburbs of São Paulo. “Industrial Workers,” 448-449.


A letter sent by FIESP to the military governor of São Paulo in Nov. 1932 outlined an ambitious worker assistance plan, similar to SESI, as an alternative to the Let das Férias. Though it was to be funded and managed by the industrialists, the letter argued that such a social program “should emanate from the federal government through its Ministry of Labor.” Text of letter included in Circular da FIESP, 341 (Nov. 1, 1932). There also was a brief campaign by FIESP in early 1946 to create a “Fundo de Assistência ao Trabalhador” on a voluntary basis. Though the directorate claimed that there was an enthusiastic response from its membership, the Fundo was quietly merged with SESI when the latter was founded in July 1946. Circulares da FIESP, 54/46 (Mar. 13, 1946), 75/46 (Apr. 23, 1946), 83/46 (May 2, 1946), 120/46 (July 6, 1946).


Objections to the industrialists’ direct role in worker education and training, first articulated in the early 1940s with the creation of SENAI, have persisted into the 1980s. During the debates in the Constituent Assembly that produced Brazil’s new constitution, delegates raised the issue of SENAI, arguing that such a key educational service should be subject to direct control by the state. The industrialists, however, mounted a vigorous and successful campaign to maintain control of SENAI as well as of schools operated by SESI. The directors of SENAI and SESI have also argued that these entities have been effective precisely because they are subordinated to the CNI and, on the regional level, FIESP, rather than to the inefficient government bureaucracy. J. F. Góes Filho, O SENAI: Traços do seu passade e perspectivas emergentes (Rio de Janeiro, 1981), 15.


The most important discussions of the estado de compromisso and the role of worker mobilization can be found in the classic essays by Francisco Weffort, “Estado e massas no Brasil” and “O populismo na política brasileira,” in O populismo na política brasileira (Rio de Janeiro, 1978), 45-78.


This position is articulated most clearly in Chauí and Carvalho Franco, Ideologia e mobilização popular., de Decca, 1930: O silêncio; and Trevisan, 50 anos em 5.


A clear example of this is the unsuccessful campaign by Morvan Dias de Figueiredo, Armando de Arruda Pereira, Antônio Devisate, and other key figures in FIESP during the 1940s and 50s to cancel or severely limit the workers’ right to strike. Circular da FIESP, 36/46 (Feb. 6, 1946); Boletim Informativo (F/CIESP), 247 (June 28, 1954), 11-12, 319 (Nov. 14, 1.955), 13-17, 522 (Oct. 7, 1959), 100-101. The latter article, an editorial, referred to a bill in the federal senate to guarantee the right to strike as an “incentive to anarchy.”

Author notes


The author wishes to thank Warren Dean and Erich Goode for their suggestions, and is especially grateful to John French for his many helpful comments and criticisms. The research for this article was made possible by fellowships from the National Endowment for the Humanities, the Council for International Exchange of Scholars, and the Fulbright-Hays Post-Doctoral Training Program.