productivity and labour politics jeh

8/14/2019 Productivity and Labour Politics JEH

1/35


2/35

94 Colistete

compromise with management in which workers could gain from eco-

nomic growth, real wage increases, and welfare in exchange for partici-pating in a strong productivity drive at the firm level.2

Assessing the overall evidence from country studies, Charles Maiermaintained that the United States, France, West Germany, and Japandespite their particular social and institutional conditionsoffered ahistory of capitalist convergence towards a corporatist settlement in-volving labor and management, which gave rise to a high level of eco-nomic growth and relative stability in industrialized countries during thepostwar years. He then wondered to what degree Latin American so-cieties could also be assimilated into this overall corporatist model.3

Perhaps surprisingly, Maiers question actually is highly pertinent tothe postwar history of Latin American countries, not the least becausemany of them underwent fast economic growth and rapid industrializa-tion at the same time Europe and Japan were experiencing their goldenage. Even more importantly, however, the question may help to shednew light on the causes of the highly unstable social, political, and insti-tutional conditions that prevailed in Latin America after the SecondWorld War. Was this chronic instability caused only by the weakness ofdemocratic institutions in Latin America? Or was it due to an ingrainedauthoritarian and interventionist stance by Latin American elites? Even

though one might be tempted to answer yes to these questions, any con-clusion should be placed in the context of the prewar European back-ground of far-reaching political instability, repression, and destructionon a scale that had never been seen before. As Mark Mazower remindsus, Europe was far from committed to democracy by the 1940s and fas-cism was the most European ideology of those that marked the DarkContinent in the twentieth century.4 If this is true, what happened inLatin American societies to prevent them overcoming their own dra-matic past of authoritarianism, repression, and inequality during thegolden age of postSecond World War economic growth? This is an is-

sue that has long intrigued economic and social historians of LatinAmerica.5 By exploring the notion of a social compact for growth inLatin American context, we can formulate new hypotheses about thepolitical economy of growth and income distribution in the region.

2 Maier, Postwar Social Contract, p. 148.3 Ibid.4 Mazower,Dark Continent.5 See, for example, Abel and Lewis, eds., Welfare; Bethell and Roxborough, Introduction;

Bulmer-Thomas, Economic History; Cardoso and Helwege, Latin Americas Economy; Coats-worth and Taylor, eds., Latin America; Furtado, Economic Development; Haber, ed., PoliticalInstitutions; Maddison et al.,Political Economy; and Thorp,Progress.


3/35

Productivity, Wages, and Labor Politics 95

One crucial reason for the failure of democratic stabilization in

postwar Brazils political economy lies with the relations between theemerging trade union movement and employers. Brazil was the larg-

est industrial economy in Latin America by the 1950s, and this articleargues that labor politics in this country after the Second World War

diverged sharply from the overall pattern observed in advanced in-dustrial societies. Although left-wing trade union leaders took over

the key positions in the Brazilian labor movement, demanding sub-stantial improvements in workers social standing, they were not seen

as reliable partners by industrialists and other elite groups in Brazil.As a result, industrial relations in Brazil became highly confronta-

tional and antagonistic, undermining any prospects of a social con-tract that could favor sustained economic growth along with social

reform.The article is set in Brazils main industrial center at the timethe

state of So Paulo. So Paulo not only had the largest share of indus-

trial output in Brazil (47.8 percent in 1949 and 55.7 percent in 1959,according to the Industrial Censuses), it was also home to the most

active and influential employers organization, the Federation of In-dustries of the State of So Paulo ( Federao das Indstrias doEstado de So Paulo or FIESP), as well as trade unions. The analysis

covers the governments and phases of labor relations in Brazil fromthe end of the Second World War to the end of the JuscelinoKubitscheck government (January 1961). As there are no official sta-tistics on manufacturing industry in 1960 and 1961, the article ex-

tends its quantitative data to 1962, which is used as a reference yearfor interpolation. Initially, the article provides an overview of how

wages and productivity evolved in postwar Brazil and then brieflyoutlines the features of Brazils political economy, steps which are

important to set the context and establish the main issues to be exam-

ined in the following sections. The distinct parts of the main argu-

ment are presented in sequence, by looking at industrial relations inSo Paulo and Brazil after 1945. The aim is to show which were theparticular political solutions to the labor question in the 1940s and

the 1950s and the behavior of wages and productivity. The articlestresses the links between domestic politics and the international

Cold War juncture, and it examines the attempts made by interna-tional trade union organizations to support moderate trade union

leaders who could establish a settlement with industrialists and gov-ernments in Brazil. In the end, we can see how historical develop-

ments in industrial relations had a wide-ranging impact on Brazilian

postwar economic and social history.


4/35

96 Colistete

WAGES AND PRODUCTIVITY: AN OVERVIEW

The key empirical evidence for the social compact for growthhypothesis is that real wages grew in tandem with productivity dur-

ing the crucial years of recovery and expansion of European andJapanese economies after 19451947.6 By taking part in a productiv-

ity drive at the firm level and limiting their wage demands to the

growth of productivity, labor obtained steady increases in real wagesand welfare, which formed the basis for a corporatist accommodation

between labor and management. Indeed, manufacturing wages and productivity in the advanced industrial countries grew by 3 percent

per annum between the mid-1950s and mid-1960s. Fast growth, fullemployment, high productivity, and rising wages were thus key fea-

tures of the postSecond World War settlement in European andJapanese economiesuntil inflation struck in the early 1970s and

wages surged past productivity.7

Wages and productivity evolved differently in postwar Brazil. Dataon these variables are limited and somewhat tenuous. I have attemptedto construct wage and labor productivity series by using interpolatedand backward-extrapolated estimates from available statistics.8 A sum-mary of the methodological procedures is presented in the Appendix. In

order to make clear the limitation of the series, data obtained by interpo-lation and backward extrapolation appear in brackets.Overall trends are clear even before we take a closer look at the data.

Thus Figure 1 shows that, for most of the 19451962 period, labor pro-ductivity exceeded real wages in Brazils manufacturing industry. Testsof structural stability, using a regression model with dummy variables,show that there was a change in the level and the slope of Brazils labor productivity series, indicating the existence of a structural break in1954an outcome not replicated by wages.9 Such a result is revealing because the 1950s were a period of high economic growth in Brazil,

with an average Gross Domestic Product growth of 7.4 percent peryear.10 Besides, developmentalist or populist economic policiesare usually assumed to have had a redistributive impact in postwar

6 Maier, Politics of Productivity; Maier, Postwar Social Contract. Other approaches thathave stressed the role of productivity, wage restraint, and high investment include Eichengreen,Institutions; Cameron and Wallace, Macroeconomic Performance; and Soskice, Wage De-termination.

7 Armstrong, Glyn, and Harrison, Capitalism, pp. 12021; and Eichengreen, Institutions.8 See Tables 1 and 2 in the section Labor Politics in the Early Postwar Years.9 See the Appendix for the tests.10 Average rate of growth between 1950 and 1960, with data from Zerkowski and Veloso,

Seis Dcadas.


5/35


020406080

100120140160180

1945 1947 1949 1951 1953 1955 1957 1959 1961

WagesLabor Productivity

FIGURE 1WAGES AND LABOR PRODUCTIVITY IN BRAZIL, 19451962

Note: Wages and labor productivity are taken from Table 1.Sources: See the Appendix.

Latin America.11 It is hard to avoid the conclusion that, because of the jump in productivity and the relatively slow growth of wages, firmswere experiencing an increase in their profits at the expense of workerswages after 1954 (Figure 1).

The trend is similar in So Paulo. Real wages in So Paulo laggedbehind labor productivity from 1956, as can be seen in Figure 2. As inBrazil as a whole, a regression model with both intercept and slopedummy variables shows that So Paulos labor productivity series is notstable across time, indicating the presence of a structural break in 1954(see the Appendix). It seems that, even at the heart of the industrialtransformation in Brazil, workers were not able to keep pace with ra-tionalization and improvements in production organization by firms,which led to significant increases in labor productivity during the sec-ond half of the 1950s and early 1960s.

The behavior of wages and productivity could conceivably reflect theinitial phase of the Kuznets curve, as inequality increased during thefirst stages of economic development. However, it seems more likelythat specific, institutionally determined factorssuch as inflationary

11 Developmentalist economic policies aimed at rapid industrialization, with foreign and do-mestic private capital as active partners of the state (Bielschowsky, Pensamento EconmicoBrasileiro; and Sikkink,Ideas). Populist economic policies were originally defined by historiansas policies implemented by charismatic politicians who, based on multiclass political coalitions,sought to establish a direct relation with the people by pursuing nationalism, state intervention,rapid industrial growth, and ameliorative distributive measures in Latin America, between the1920s and 1960s. See Conniff, ed., Latin American Populism. For a view of the different ap-proaches to and new uses of the concept of populism, see Knight, Populism.


6/35

98 Colistete

020406080

100120140160180

1945 1947 1949 1951 1953 1955 1957 1959 1961

Wages

Labor Productivity

FIGURE 2WAGES AND LABOR PRODUCTIVITY IN SO PAULO, 19451962

Note: Wages and labor productivity are taken from Table 2.Sources: See the Appendix.

macroeconomic policies, minimum-wage policies, strategies by firms,and workers bargaining powerwere the main causes shaping the dis-tribution of wages and profits in Brazil, not only in the 1950s but also in

the 1960s.12 This is a hypothesis that still remains to be investigatedempirically and is beyond the scope of the present article. My concernhere is a different issue, one related to labor politics, wages, and pro-ductivitynamely, why did Brazils political economy fail to give birthto a social compact that could help design proper economic and socialpolicies in order to promote growth along with increasing welfare? In particular, the focus will be on the relations between employers andworkers that may have prevented industrial wages from catching upwith productivity in a favorable economic situation that was marked byhigh economic growth and employment.

BRAZILS POLITICAL ECONOMY IN THE POSTWAR YEARS

Brazils political economy from the 1940s to the early 1960s wasdeeply affected by two major developments in earlier decades. First, thedemands of the growing industrial working class for economic benefits,

12 This is what Albert Fishlow, John Wells, Rodolfo Hoffman, and others argued in a well-known debate over the increasing income concentration in Brazil between 1960 and 1970.These authors held that income concentration was mostly explained by political and institutionalfactors related to the military government from 1964not by a structural increase in incomeconcentration in the first stages of economic development (as other authors, such as Carlos Lan-goni, maintained). See the contributions in Tolipan and Tinelli, eds., Controvrsia.


7/35


social rights, and political participation called into question the tradi-

tional relations between state and society in Brazil. The existence of thisworking-class challenge was an important aspect of the political eventsin the 1930s, which started with the seizure of power by a coalition offorces led by Getlio Vargasthe Revolution of 1930. At first, the newgovernment was responsive to working-class demands and made an at-tempt to acknowledge the rights of emerging social forces. Between1932 and 1937, new social legislation enacted by the government ad-dressed many of the issues labor organizations had been fighting forsince at least the 1910s.13

Despite government action in the area of social rights, the limitations

of the reformist zeal of the political forces that took power in the Revo-lution of 1930 were made clear within a few years. As trade unions andpolitical groups independent from the government pushed forward theircampaign for extended economic and social benefits, the newly foundedinstitutional order, established by the Constitution of 1934, began tocrumble. In January 1935 the government decreed a National SecurityLaw, which heralded a period of harsh repression of labor organiza-tions. In November 1937 Vargas and his supporters shut down Congressand consolidated their authoritarian rule, giving birth to the EstadoNovo (New State) regime, which would last until October 1945. A new

era of relations between the state and the workforce was then initiated,in which the labor problem became a matter of national security.14

Second, the 1930s also saw the establishment of the framework thatwould regulate industrial relations over the next decades. By 1939 suc-cessive government decrees had laid down a state-corporatist structurein which both employer and labor representations were defined as con-sultative organs attached to the state. In the end, the new legislation en-forced a three-tier hierarchy in which trade unions (in cities), federa-tions (in states), and confederations (at the national level) wouldrepresent the interest of employers and employees in a given trade. The

Ministry of Labor, Industry, and Commerce (set up in 1931) held the power of authorizing, supervising, and controlling the official unions,which in turn were forbidden to declare or publicize political and reli-

gious ideas. At the same time, in 1932 the government began to set up alabor court system to mediate in the legal disputes between labor and

management.15

13 Gomes,Burguesia, ch. 6; and Vianna,Liberalismo, pp. 14149, 18197. For the changes inthe state and attitudes by industrialists during and after the Revolution of 1930, see Diniz, Em-presrio; and Weinstein,For Social Peace.

14 Gomes,Inveno, ch. 4; and Vianna,Liberalismo, pp. 119206.15 Decrees 19,770, 19 March 1931; 24,694, 12 July 1934, and 1,403, 5 July 1939 (the latter

was reviewed by Decree 2,533, 29 June 1940). On this legislation: Simo, Sindicato, ch. 4.


8/35

100 Colistete

During the Estado Novo, the state-corporatist structure of indus-

trial relations was further strengthened by complementary legislation.In addition to Decree 1,403 of 1939, which consolidated the interven-

tion in trade union issues initiated in 1931, in July 1940 a mandatoryunion tax (imposto sindical) on all employees provided official trade

unions, federations, and confederations with stable funding for theirregular work as appendages of the state. In May 1939 a new Labor

Justice system was set up with a three-level structure comprising lo-cal councils (Juntas de Conciliao e Julgamento), regional tribunals

(Tribunais Regionais do Trabalho), and a top federal body responsi- ble for last-instance decisions (Tribunal Superior do Trabalho). Fi-

nally, in November 1943, the Consolidated Labor Laws (Consolida-o das Leis do Trabalho or CLT) brought together and expanded the

social legislation enacted during the previous years. The Consoli-dated Labor Laws were the final step in the establishment of an au-

thoritarian corporatist system that would regulate labor-management

conflicts in the following decades in Brazil, notwithstanding the endof theEstado Novo regime in late 1945.

16

This authoritarian corporatist system of industrial relations was aresponse to the social conflicts that marked the industrialization

process in Brazil, particularly from the 1910s onwards. During the

early 1930s, independent left-wing activists resisted attempts by theMinistry of Labor to co-opt the whole labor movement into the stateagencies. Even when militants decided to join the official trade un-ions (as occurred in 1933/34), they sought to preserve their inde-

pendence in the face of the governments position.17

It was only withthe crackdown on the labor movement from 1935 that the govern-

ment and industrialists could regain the initiative in the sphere of in-dustrial relations in So Paulo and other industrial centers in Brazil.

For their part, industrialists in So Paulo welcomed the approach to

labor and the speeches on social harmony between classes formulated

by theEstado Novo ideologues. As illustration, in 1943 the Federa-tion of Industries of the State of So Paulo urged its affiliated firmsto provide transport for their workforce to take part in the official

commemoration of the Labor Day, because it promis[ed] to be anexpressive demonstration of the spirit of social harmony which, for-

tunately, govern[ed] the relations between labor and management.18

16 Rodrigues, Sindicato, ch. 2.17 Gomes,Inveno, pp. 15052.18 Circular FIESP, no. 68/43, 28 April 1943. The Federation of Industries of the State of So

Paulo was created in 1931 immediately after the already mentioned Decree 19,770, which set upthe official syndical structure. The prior organization, Centre of Industries of the State of SoPaulo (Centro das Indstrias do Estado de So Paulo or CIESP), was maintained along with the


9/35


Such an optimistic assessment would suddenly change in early 1945,

when a strike wave hit industrial centers in Brazil, as will be shown.With regard to the economy, the overthrow of Vargas and the end of

Estado Novo in October 1945 did not affect the optimism that followedthe end of the Second World War. There was hope that the world econ-omys return to normality would ensure acknowledgement of Brazilsparticipation in the war and material achievements, so that the countrywould no longer be relegated to a marginal position in internationaleconomic and political affairs. At the same time, a smooth transitionfrom wartime was expected, owing to the substantial trade surplus withEuropean partners during the war.19

The story of how such an optimistic economic outlook was frustratedin the immediate postwar years has been told by economic historians.20

As early as 1946 the liberal exchange policy that had been implementedby the newly-elected Eurico Gaspar Dutra government (January 1946January 1951) was hit by capital flight and soaring imports, with neithercompensatory investments from nor exports to the dollar area. As happened to countries in Latin America and other regions, the reservesamassed during the war were shown to be worthless in the face of thedollar shortage. In 1947 the worsening exchange crisis forced the gov-ernment to do a U-turn on its economic policy, with a rapid escalation

of exchange restrictions and import controls, coupled with strict appli-cation of expenditure-reducing policies (the latter until mid-1949). Be-sides, Latin America was not included in the large-scale U.S. aid forpostwar recovery, the Marshall Plandespite lobbying by its govern-ments and elites. This was more than a transitory setback, and it helpedmake controlling imports a fixture in Brazils (and Latin Americas)economic policy during the following decades.21

Such a context of crisis, restrictive practices, and lack of foreign re-sources strengthened the position of those who resisted arguments in fa-vor of markets and who favored a strategy of industrial diversification

led by the state.22

In a society still largely dependent upon primary pro-duction and exports, the aim of economic modernization through rapidindustrial growth attracted strong support among business groups, themiddle class, workers, and the state bureaucracy. In fact, the social basis

Federation of Industries, although the latter became the official representative of industrialists inSo Paulo. See Leme, Ideologia, ch. 1.

19 For Brazils economic relations in the Second World War, see Abreu,Brasil.20 Vianna, Poltica Econmica Externa; and Saretta, Poltica Econmica.21 Thorp,Progress, ch. 5; and Rabe, Elusive Conference. On the U.S. policy towards Brazil

in the postwar years, see Weiss, Cold Warriors.22 For the views of, and debates between, liberals and interventionists in Brazil, see Biel-

schowsky,Pensamento Econmico.


10/35

102 Colistete

for such a development had already been established in previous dec-

ades. The industrialist leadership, for example, had become an influen-tial voice in all debates and decisions regarding economic and socialpolicy already in the early twentieth century.23 By 1945 the Federationof Industries of the State of So Paulo was the most powerful employerinterest group in Brazil, with a history of close collaboration with theEstado Novo regime and access to the spheres of policy making. Theworking class, in turn, had expanded in an impressive way, with largeconcentrations in the main industrial cities, so that its interests could nolonger be ruled outas was shown by the corporatist initiatives by gov-ernments in the 1930s mentioned before.24 The middle classes were also

captured and influenced by the thrust of industrial growth, with the ex-pansion of new professions, services, and urban centers. To a large ex-tent, therefore, economic policy aimed at rapid industrialization seemsto have been a response to powerful interests that had been in gestationduring the previous decades. The strong appeal of industrial growth wasultimately conveyed by the developmentalist theses of the EconomicCommission for Latin America (ECLA), set up in 1948 and highly in-fluential among Brazilian industrialists throughout the 1950salthoughsuch an influence was not always significant among other businesselites in Latin America, such as the Argentinean.25

Following a world-wide trend in the postwar years, the achievementof rapid industrial growth in Brazil was closely associated with plan-ning, state intervention, and protectionism.26 Apart from direct invest-ments in basic industries (such as oil and steel), and in line with otherhistorical experiences of late industrialization, the Brazilian state alsomobilized financial resources, ensured protection against foreign competition, set up a range of incentives, and established the sectors in-tended by industrial policy. The automotive industry is an illustrativeexample of state intervention in postwar Brazil. The second Vargasgovernment (January 1951August 1954) established the initial policy

framework for building up the motor vehicle and parts manufacturingindustry in Brazil. After a brief pause, caused by Vargass suicide inAugust 1954 and a transition period, the program was reauthorized and

23 Luz,Luta; Stein,Brazilian Cotton Manufacture; and Dean,Industrialization.24 The population of the city of So Paulo jumped from 579,033 to 1,326,261 inhabitants be-

tween 1920 and 1940. According to the Demographic Censuses, the workers employed in themanufacturing industry made up 17.3 percent (1920) and 17.2 percent (1940) of the total popu-lation and were the largest occupational group in the city of So Paulo. Data from IBGE,Anurio Estatstico 1936, p. 56; and IBGE, Censo Demogrfico 1960.

25 On Brazil, Leopoldi, Industrial Associations, pp. 13840. The reception of ECLAs thesesamong business sectors in Argentina is examined by Sikkink, Influence.

26 On the world-wide trend, Cox,Production, ch. 7; and for Latin America and Brazil, Thorp,Progress, ch. 5.


11/35


extended by the Kubitscheck government (January 1956January 1961),

as part of its comprehensive Targets Plan (Plano de Metas). On bothoccasions, features that seem to have characterized successful economicintervention by the state in late-industrializing societies were also atwork in Brazil: an active technical bureaucracy, a close relationshipwith business, and clear, long-term goals.27 The result was that both thesecond half of the 1950s and the automotive industry became symbolsof the developmentalism in Brazilan economic ideology aimed atpromoting fast industrial growth led by the state, with foreign and do-mestic private capital as active partners.28

A noticeable outcome of the developmentalist era in Brazil were the

high rates of output growth and economic diversification, which madethe country one of the main newly industrializing economies at thetime.29 Growth in Gross Domestic Product reached 7.5 percent per yearbetween 1945 and 1960, whereas average industrial output achieved 9.5 percent per year growth in the same period.30 Apparently, economicgrowth could have set a solid basis for accommodation between the in-dustrial working classes and employers, under the auspices of an inter-ventionist state. However, evidence on wages and productivity shows arather different picture: the gap between industrial wages and profits in-creased in the heyday of developmentalism and, as we shall see, a con-

frontational pattern of labor relations continued to develop in Brazil.

LABOR POLITICS IN THE EARLY POSTWAR YEARS

By early 1945 Brazil was experiencing all the uncertainties and socialtensions of a world that had gone through years of war, material de-struction, genocide, and social disruption. Although the country had notbeen a location of combat during the Second World War, public opinionclosely followed the events, mobilized to take part in the conflict, andhelped force a government sympathetic to the Axis to send the army to

Europe to join the Allies. The institutional and political situation, there-fore, was awkward. On the one hand, the defeat of fascist regimes inEurope fueled the expectations of a new era of democracy, popular par-ticipation, social justice, and economic development. On the other hand,the country was still under the rule of the Estado Novo dictatorship.Vargass resignation and the end ofEstado Novo in October 1945 did

27 Shapiro,Engines, ch. 2. See Evans,Embedded Autonomy, for a treatment of ideal types andhistorical patterns of the relations between business, states, and developmental policies, includ-ing Brazil.

28 Sikkink,Ideas; and Bielschowsky,Pensamento Econmico.29 Maddison et al.,Political Economy, p. 43.30 Calculated from Zerkowski and Veloso, Seis Dcadas.


12/35

104 Colistete

not solve the impasse, because a coalition of conservative groups led by

Dutra, a former Ministry of War during theEstado Novo regime, endedup winning the presidential elections in December 1945. Thus, the res-toration of institutional legality headed by conservative groups closelylinked to the early dictatorship only reasserted the uncertainties and so-cial tensions in Brazilian society. In particular, this situation raised seri-ous doubts about the ability of new political and social groups to replacethe authoritarian elites that traditionally held power in Brazil. As re-gards the outcomes, there was skepticism whether the benefits of eco-nomic development could be extended beyond the powerful landed andentrepreneurial classes.31

The particular political and social alignments established in the imme-diate postwar years, characterized by continuity with the authoritarian

past, deeply marked labor relations in Brazil in the decades to come. La-bor issues, as we know, continued to be regulated by the highly central-

ized and interventionist system of industrial relations inherited from the Estado Novo regime, which prevented trade unions from being autono-mous and independent from the state. Nonetheless, labor politics in post-

war Brazil shows a complex history of working-class organization andconflicts, which went far beyond the corporatist system of industrial rela-

tions. A major example were the workers initiatives and organization on

the shop floor, which, in contrast to the standard stories in Brazilian laborhistory, paralleled the trade unions and even clashed with them at times.

32

Relations between industrialists, workers, and governments were thuscharacterized by conflict, instability, and repression, as the history of the

governments and the industrialists dealings with workers clearly shows.

A repressive stance by governments and industrialists was alreadyclear in the first months of the new Dutra government, elected in De-cember 1945. Early in the year, there had been strikes in hundreds offactories in So Paulo which were led by workers apparently withoutlinks with the official trade unions. This first wave of strikes was fol-

lowed by new demonstrations and industrial actions in the followingmonths. It seems that the strikes and the assertive stance by the rank-and-file caused a deep impression upon the new government and the in-dustrialists. By the end of 1945 So Paulo industrialists were already

31 Maranho, Sindicato; and Bethell, Brazil.32 The standard view of Brazils labor history in the 1940s through the early 1960s has been

endorsed by major works on Latin American and Brazilian industrial working class, such asRoxborough, Urban Working Class, pp. 34057; Roxborough, Labor Control, pp. 26061;and Humphrey,Fazendo o Milagre, ch. 1. Nonetheless, historical studies heavily based upon primary sources have painted a rather different picture. See Costa, Em Busca da Memria;Colistete, Labour Relations; Fontes, Trabalhadores; French, Brazilian Workers ABC; Negro,Ford Willys; and Wolfe, Working Women, Working Men.


13/35


denouncing what they saw as the importation of exotic redistributive

ideologies, which were being put into practice by social democratic par-ties in Europe (particularly in the United Kingdom, according to the in-dustrialists) in the wake of wartime destruction. In Brazil, the Federa-tion of Industries of the State of So Paulo pointed out that it seemedthat the workers want[ed] . . . to replace management, and divide whatappears to them superfluous. This is the social justice preached by theirattitudes. The madness is so great that it causes dismay to comment.33

Alarmed by labor demands, in early 1946 the industrialists hailed astrike law passed by the Dutra government, which in practice imposed aban on industrial action. According to the Federation of Industries of the

State of So Paulo, labor attitudes at the time were a call for the violentdissolution of our traditional political and social institutions, whichmake up our secular style of political organization.34

From the outset the Dutra government was committed to the industri-alists and soon showed its willingness to curb the emerging labormovement. In addition to the strike law of March 1946, the governmentdecreed a ban on political meetings and demonstrations in cities (includ-ing So Paulo) affected by strikes in February 1946, held back long-overdue trade union elections, imprisoned trade unionists accused of be-ing leftists, and stormed newly founded organizations backed by the

communists (such as the United Movement of Workers or MUT).35 Thefinal crackdown, however, was still being prepared. In September 1946,Federation of Industries of the State of So Paulos director, MorvanDias Figueiredo, was appointed to the Ministry of Labor, Industry, andCommerce and signaled that he was only waiting for the state electionsto act.36 Meanwhile, there were informed reports that the inner circlesof the Dutra government had long been preparing to outlaw the Com-munist Party.37

Eighteen months of hesitation came to an end on 7 May 1947, whenthe Federal Supreme Court did outlaw the Communist Party. This deci-

33Revista Industrial de So Paulo [hereafterRISP], no. 13 (1945): 64.34RISP, no. 15 (1946): 21. For the antistrike legislation: Decree no. 9,070, 15 March 1946,

Revista do Trabalho (MayJune 1947).35 United Kingdom, Public Records Office [hereafter PRO], Foreign Office [hereafter FO]

371/61204, D. Gainer, Annual Report on Brazil 1946, 13 January 1947, p. 11; United States,National Archives and Records Administration [hereafter NARA], E. Rowell, Monthly LaborReview (1 April to 1 May 1946), Record Group [hereafter RG] 84, 850.4, pp. 1617, and E.Rowell, Monthly Labor Review (1 May to 1 June 1946), RG 84, 850.4, p. 8.

36 United Kingdom, PRO FO 371/61204, C. German, Trade Union Developments in Brazil, 9January 1947, p. 2.

37 Since March 1946 at least, the U.S. government had been informed about plans to ban theCommunist Party. See for example United States, NARA, E. Hoover to F. Lyon, Federal Bureauof Investigation, 8 March 1946, RG 59, 832.00B/3846.


14/35

106 Colistete

sion was likely to have been influenced by the Cold War confrontation,

sparked off by President Trumans speech in the U.S. Congress inMarch 1947. A British diplomat even pointed out that the new TrumanDoctrine had been seen in Brazil as a green light to anti-communistmeasures.38 The Cold War also reasserted a deep-rooted mistrustamong industrialists and other conservative sectors (such as the militaryand the Catholic church) of open labor politics in Brazil. Along with theban on the Communist Party, an additional decree set out intervention inall trade union organizations allegedly under the influence of the com-munists. As the British Labor Attach noted at the time, [o]pen inter-vention on this scale . . . seems quite unprecedented even under the Dic-

tatorship.39 The catch-all repression affected all labor groups that hadnot been closely linked with theEstado Novo and the Ministry of Labor.In fact, the attack against the newly reorganized labor in Brazil was partof an overall trend in most Latin American countries at the time.40

Repression also reached the workplace. The changing political situa-tion apparently encouraged firms that resisted negotiating wage in-creases directly with workers to embark on a series of repressive meas-ures on the shop floor. For example, the largest metalworkingcompanies in the state of So Paulo confronted walkouts after February1946 and took advantage of the strike law decreed by Dutra to fire

workers on strike. As early as the end of March, in two industrial citiesof Greater So Paulo, Santo Andr and So Bernardo do Campo, about1,400 workers were dismissed by metalworking companies. Firms alsoblacklisted workers who had taken the lead in the walkouts and beenfired so that they would not be able to find other jobs.41 At the sametime, the industrial companies sought to assert authority in the work-place by bringing in the police. In certain cases, management infiltratedpolicemen to identify and help to dismiss strike leaders. In others, thepolice was called in to quell meetings and to dissuade workers from tak-ing their demands to management.42

By August 1947 well-informed sources reported that more than 170trade unions all over Brazil had fallen under the governments new law,with their officials being replaced by appointees of the Ministry of La-

38 United Kingdom, PRO FO 371/61204, G. Young, Outlawing of the Communist Party inBrazil, 4 May 1947, p. 2. For the Truman Doctrine, see LaFeber,America, Russia, and the ColdWar, pp. 4958.

39 Decree no. 23,046, 7 May 1947. United Kingdom, PRO FO 371/61205, C. German, LabourReport no. 23, 18 June 1947, p. 3.

40 Bethell and Roxborough, Introduction.41Hoje, 8 March 1946, p. 7, and 20 March 1946, p. 8.42Hoje, 4 June 1946, p. 5.


15/35


bor.43

In So Paulo, trade unions involved in labor demands during the

previous months had their officials ousted, leftist affiliates expelled, andtheir operations closely watched by the Ministry of Labor and the politi-

cal police (Department for Political and Social Order or DOPS). At thesame time, unofficial organizations set up after 1945 were persecuted and

declared illegal. The most noticeable case was the Confederation of Bra-zilian Workers (Confederao dos Trabalhadores do Brasilor CTB), es-

tablished by over 1,500 delegates at a national congress in Rio de Janeiroin September 1946. The Confederation of Brazilian Workers was en-

dorsed by about 800 trade unions and arose as the most representative na-tion-wide labor organization during the immediate postSecond World

War period.44 Even so, the May 1947 Decree explicitly mentioned theConfederation of Brazilian Workers as provoking interference with out-

put, order and discipline and decreed that all trade unions affiliated orcontributing to it would be subject to intervention.45

The Confederation of Brazilian Workers was also important becauseits creation represented a fundamental schism in the trade union move-ment, which would characterize labor politics in Brazil throughout the1950s. Indeed, the national congress that set up the Confederation ofBrazilian Workers had originally been a result of a compromise be-tween left-wing and independent trade unionists, on the one hand, and

those directly sponsored by the Ministry of Labor, on the other. Duringthe congress, however, the group close to the government walked outwhen confronted by an overwhelming majority of independent and left-wing delegates.46 As a response to the Confederation of BrazilianWorkers, the Ministry of Labor first announced the creation of a neworganization, the National Confederation of Labor (or CNT). Con-fronted with legal difficulties and negative reactions in Congress, how-ever, the Dutra government decided to recognize a national confedera-tion already established by the Consolidated Labor Laws. Thus, on 28October 1946 President Dutra signed Decree Law 21,798 creating the

National Confederation of Industrial Workers (Confederao Nacional

43 This number implied that, of the estimated 460,000 workers affiliated to trade unions inBrazil, about 300,000 had been affected by the interventions. See United Kingdom, PRO FO371/61205, C. German, Labour Report no. 26, 16 August 1947, p. 3.

44 Henry, Developments, p. 440.45 United Kingdom, PRO FO 371/61205, C. German, Labour Report no. 23, 18 June 1947,

appendix 1, p. 2.46 Henry, Developments, pp. 43940; United States, NARA, E. Rowell, Monthly Labor

Review (1 August to 1 September 1946), RG 84, 850.4, p. 6, and E. Rowell, Monthly LaborReview (1 September to 1 October 1946), 16 December 1946, RG 84, 850.4, pp. 34. For in-dependent trade unionists I mean the trade union leaders who did not follow the Ministry ofLabors orientation in labor issues and were not associated to the Communist Party.


16/35

108 Colistete

dos Trabalhadores da Indstria or CNTI).47 During its short existence,

the Confederation of Brazilian Workers became by far the most influen-tial organization among trade unions in So Paulo and Brazil as awhole. The National Confederation of Industrial Workers was officiallyestablished by the Ministry of Labor only in June 1947, just after theban on the Confederation of Brazilian Workers.48

The major reason for the split in the labor movement was that theConfederation of Brazilian Workers retained the allegiance of inde-pendents and left-wingers, who had progressively taken the lead in themajor trade unions since 1945. While there is no firm evidence as to theextent of communist influence, one source estimated that nine out of the

15 officials of the provisional executive committee chosen in the Con-federation of Brazilian Workers congress were linked to the CommunistParty.49 The first elected general secretary of the Confederation of Bra-zilian Workers, for instance, was Roberto Morena, the most importantcommunist trade unionist in Brazil during the postwar period. The splitin Brazils organized labor movement in fact anticipated the similar1949 schism in the World Federation of Trade Unions (WFTU), whichwould lead to the formation of the International Confederation of FreeTrade Unions (ICFTU) and of two opposed, antagonistic internationallabor organizations during the Cold War. In contrast to the European

experience, however, the split did not provoke a division of the left orthe isolation of the communists in the Brazilian labor movement, a dif-ference whose implications would only become clear in the future.50

One of the chief aims of the Dutra governments labor policy was toestablish a new cadre of trusted officials in labor unions who couldspeak to the government and to industrialists. The key figure in this pol-icy was Figueiredo, the Federation of Industries of the State of SoPaulos representative in charge of the Ministry of Labor. When repres-sion hit all active trade unions in beginning 1947, Figueiredo sought toreplace suspected union officials with unionists (ministerialistas) affili-

ated with the Estado Novo apparatus. In fact, the aim of finding loyal

47 United States, NARA, E. Rowell, Monthly Labor Review (1 October to 1 November 1946),RG 84, 850.4, p. 10.

48 Henry, Developments, p. 440; United Kingdom, PRO Ministry of Labour and NationalService [hereafter LAB] 13/498, C. German, Labour Report no. 15, 3 January 1947, pp. 12;United Kingdom, PRO FO 371/61205, C. German, Labour Report no. 24, 26 June 1947.

49 United States, NARA, E. Rowell, Brazilian Labor Congress, 9 December 1946, RG 59,832.504/12-946, enclosure. Rowell also mentioned estimations of the actual number of commu-nist delegates in the Confederation of Brazilian Workers congress: 200 out of 1,500. See UnitedStates, NARA, E. Rowell, Monthly Labor Review (September 1 to October 1, 1946), RG 84,850.4, p. 3.

50 On WFTU split and ICFTU composition, see Carew, Schism and Conflict. The ICFTUwas founded in December 1949.


17/35


labor leaders had been under way for some time. For example, during

the strikes of 1946 the Federation of Industries of the State of So Paulowould only negotiate with workers affiliated with the Estado Novostructure. The evidence suggests, however, that industrialists and thegovernment were not successful in consolidating a politically reliablelabor leadership. Early in 1948, Figueiredo had already recognized thatappointed trade unionists enjoyed little support among the rank and fileand that they were not a viable alternative to the left or to independentunionists.51 The only alternative at the time, Figueiredo implied, was tosuspend all free elections for trade unions and stick to a hard-line policyon labor.

Still, the history of labor relations in Brazil between 1945 and 1950was not only a story of repression.

52Other things were going on, as

Tables 1 and 2 make clear. The tables show the evolution of average

real wages, labor productivity, and the unit labor costs W/LP, that is,

the wage costs per unit of output, which gives the wage share of work-

ers in output. Initially, average real wages in Brazil and the state of

So Paulos manufacturing industry fell sharply in 1947: 11.0 and

11.4 percent, respectively. It seems therefore that Dutras economic

and wage policy caused a major adjustment in labor markets from the

end of 1946, with a significant impact on real wages. Dutras policies

had lasting effects, as real wages in Brazil and So Paulos manufac-turing industry only regained their 1946 level in 1949 (Tables 1 and 2,

column 1).

On the other hand, existing data show that there was a significant in-crease in industrial wages during the remaining years of the Dutra gov-ernment: real wages grew at an annual compound rate of 10.0 percent inBrazil and 11.3 percent in So Paulo, between 1947 and 1950 (as calcu-lated from Tables 1 and 2, column 1).53 Such an outcome suggests thatalongside the hard-line labor policy of both government and companies,there were wage increases above the current levels of inflation between

1947 and 1950. This result is particularly relevant given that there wasno modification in minimum-wage levels during these years. A possiblereason for such increasing wages is the existence of a social subsis-tence wage, as argued by John Wells. Another possibility is that spe-cific labor policies implemented by firms and growing pressures byshop-floor organizations, fueled by expanding industrial output from

51 United Kingdom, PRO FO 371/68167, C. German, Labour Report no. 37, 26 January 1948,p. 2.

52 Quantitative and qualitative data on welfare and detailed description of measures aimed toelicit workers effort are provided in Colistete,Labour Relations.

53 The method used for calculating these and the following annual compound rates of growthis shown in the Appendix.


18/35

110 Colistete

TABLE 1

WAGES, LABOR PRODUCTIVITY, AND UNIT LABOR COSTS IN MANUFACTURINGINDUSTRY, BRAZIL, 19451962(indexes, 1952 = 100)

Year

Wages(W)(1)

Labor Productivity(LP)(2)

Unit labor Costs(W/LP)

(3)

1945 [64] [74] [0.87]1946 [73] [74] [0.98]1947 [65] [85] [0.76]1948 [67] [88] [0.76]1949 78 91 0.861950 [85] [94] [0.90]

1951 [92] [97] [0.95]1952 100 100 1.001953 91 101 0.901954 102 106 0.971955 104 109 0.951956 110 118 0.931957 113 126 0.901958 116 143 0.811959 110 145 0.761960 [114] [148] [0.77]1961 [117] [151] [0.78]1962 121 153 0.79

Notes: Indexes in column 1 refer to average real wages. Figures in brackets were obtained by in-

terpolation or backward extrapolation. See the Appendix for details.Sources: See the Appendix.

1948, pushed up real wages despite widespread repression at trade un-ion level. Both hypotheses, however, still have to be tested.54

The indexes in Tables 1 and 2 suggest that real wage growth washigher than labor productivity growth between 1945 and 1946, whichmay help explain industrialist opposition to workers demands in suchyears and the use of repression in an attempt to discipline labor markets.Anyhow, there was a recovery of labor productivity in the following

years. In Brazil, the unit labor costs fell from 0.98 in 1946 to 0.76 in1947 and 1948, though it was followed by an increase to 0.86 and 0.90in 1949 and 1950, respectively. As for So Paulo, the same indexdropped from 1.23 in 1946 to 0.90 and 0.89 in 1947 and 1948, followed by an increase to 1.02 and 1.07 in 1949 and 1950, respectively. Thelikely implication in both cases is that firms took advantage of Dutraslabor policies to rationalize their production processes and to drive theirworkers to higher levels of effort. Such policies appear to have even al-lowed firms to partially accommodate workers demands in the form ofhigher real wages, whereas in So Paulo real wages even overcame

54 Wells, Industrial Accumulation, p. 301; and Colistete,Labour Relations, ch. 2.


19/35


TABLE 2

WAGES, LABOR PRODUCTIVITY, AND UNIT LABOR COSTS IN MANUFACTURINGINDUSTRY, SO PAULO, 19451962(indexes, 1952 = 100)

Year

Wages(W)(1)

Labor Productivity(LP)(2)

Unit Labor Costs(W/LP)

(3)

1945 [80] [73] [1.09]1946 [88] [71] [1.23]1947 [78] [87] [0.90]1948 [81] [91] [0.89]1949 96 94 1.021950 106 99 1.07

1951 [103] [100] [1.03]1952 100 100 1.001953 101 102 0.991954 112 107 1.051955 112 107 1.051956 114 116 0.981957 116 125 0.921958 120 141 0.861959 108 146 0.741960 [110] [147] [0.75]1961 [112] [148] [0.75]1962 113 149 0.76

Notes: Indexes in column 1 refer to average real wages. Figures in brackets were obtained by in-

terpolation or backward extrapolation. See the Appendix for details.Sources: See the Appendix.

productivity in the end of the period. Partial accommodation of laborsclaims, however, took place amidst a political and institutional situationthat was harmful for workers organization. Confrontation already char-acterized labor relations in Brazil by the end of the 1940s.

THE RECOVERY OF TRADE UNIONS

Trade unions experienced a slow recovery from the setbacks during the

Dutra government. Although Vargas had been elected in the end of 1950 inthe wake of a broad appeal to labor, during the first year his government

took a careful line regarding the trade unions under intervention. On 1 May

1951 the Minister of Labor, Danton Coelho, officially announced his inten-

tion of holding free elections for trade union offices during the following

months. In particular, Coelho promised to abolish the ideology certifi-

cate (certificado de ideologia), a personal document issued by the politi-

cal police and aimed to identify left-wing candidates and prevent them

from running for trade union office. Despite the announced liberalizing

measures, the ideology certificate was officially revoked only in September

1952; in the meantime, the convocation of trade union elections went


20/35

112 Colistete

ahead slowly throughout 1951 and 1952.55

Together with a cautious ap-

proach to organized labor, the Ministry of Labor launched an affiliationcampaign for new members of trade unions in order to strengthen the posi-

tion of progovernment trade unionists in the labor movement. The aim was

clearly to confront the expected growth in the number of leftists and inde-

pendents in the forthcoming trade union elections.56

The replacement of Danton Coelho by Segadas Viana at the Ministryof Labor in September 1951 reinforced the governments ambiguous at-titude towards organized laboran attitude that combined speeches onsyndical freedom and frequent interventions in trade union issues.57 Itwas only with the appointment of Joo Goulart to the Ministry of Labor

in June 1953 that an aggressive labor policy was implemented by theVargas government. The new Minister led a large section of trade un-ionists from the Brazilian Labor Party (Partido Trabalhista Brasileiroor PTB), and he sought to establish his group as a powerful force in theBrazilian labor movement. Goularts strategy was to build up a workingrelationship with the communists and other left-wingers while at thesame time presenting himself as an alternative to the left.58 AlthoughGoulart would be removed from the Ministry of Labor during a seriouspolitical crisis in February 1954, he retained a strong influence over theBrazilian Labor Partys labor leaders and also over the government bu-

reaucracy during the remainder of the Vargas administration and be-yond. Goularts influence over labor was a major asset in his successfulbid for the Vice Presidency between 1956 and 1961.59

The result of the gradual slackening of the governments grip ontrade unions, more clearly after late 1951, was a resurgence of organ-ized labor in the Brazilian political scene. In So Paulo, the main ex- pression of this was the setting up of the Inter-Syndical Unity Pact

55 United Kingdom, PRO FO 371/103253, L. Mitchell, Labour Report no. 69, 25 February1953, p. 7; and United States, NARA, H. Hammond, Quarterly Labor Report Third Quarter1951, 17 October 1951, RG 59, 832.06/10-1751, p. 14.

56

United States, NARA, H. Hammond, Minister of Labors Report on Status of Syndicates, 2March 1951, RG 59, 832.062/3-251, p. 2.57 United Kingdom, PRO FO 371/103253, by L. Mitchell, Labour Report no. 75, 29 October

1953, p. 11.58 United States, NARA, I. Salert, Labor Report Second Quarter 1953, 20 August 1953, RG

59, 832.06/8-2053, I. Salert, Labor Report Third Quarter 1953, 18 November 1953, RG 59,832.06/11-1853, pp. 25, and I. Salert, Labor Last Quarter, 1953, 4 February 1954, RG 59,832.06/2-454, pp. 12.

59 On the removal of Goulart, United Kingdom, PRO LAB 13/503, by L. Mitchell, LabourReport no. 79, 9 March 1954. For Goularts influence over labor, United States, NARA, J.Fishburn, Annual Labor Report for 1958, 12 March 1959, RG 59, 832.06/3-1259, pp. 13. Run-ning and winning again as Vice-President in the 1960 elections, Goulart achieved BrazilsPresidency after the resignation of the newly elected president, Jnio Quadros, less than sevenmonths after being nominated. President Goulart was overthrown by the military coup of April1964.


21/35


(Pacto de Unidade Inter-Sindicalor PUI) by the largest and more mili-

tant trade unions in the state. The Inter-Syndical Unity Pact was a resultof a huge industrial action, which brought Greater So Paulo to a stand-still during March and April 1953. In the following years, the Inter-Syndical Unity Pact turned out to be the most representative and influ-ential labor organization in So Paulo, despite the fact that it was nevergranted official recognition by successive governments. As with theConfederation of Brazilian Workers in 1946/47, the Inter-SyndicalUnity Pact formed a broad coalition of independents and left-wing ac-tivists who held key positions in So Paulos trade unions.60

Still, the realization of trade union elections, and the creation of

strong parallel organizations such as the Inter-Syndical Unity Pact, didnot put an end to the threats looming over the newly reorganized labormovement. In particular, following Vargass suicide in 1954, the shortCaf Filho government (August 1954November 1955) took an uncom-promising attitude towards organized labor by interfering in trade unionissues. The Minister of Labor, Napoleo Guimares, barred electedtrade unionists accused of being leftists from taking office (for example,at the So Paulo Trade Union of Textile Workers); acted to quell indus-trial actions (for example, in the tramway industry in Rio de Janeiro);and enacted a decree banning all inter-union committees (including the

Inter-Syndical Unity Pact) set up by organized labor in the previousmonths in Brazil.61 Only with the end of the Caf Filho tenure, and theappointment of Nelson Omegna as Minister of Labor, could the tradeunion movement work again without systematic intervention by thegovernment.

Under the Kubitscheck government, trade unions became more activeand influential. Overall, the Ministry of Labor refrained from exercisinga tight control over trade union issues, and sought to bolster the moder-ate and Brazilian Labor Party wing of the labor movementagainst theleftists and independents who were at the head of the most powerful of-

ficial trade unions and parallel organizations. In So Paulo, this policyallowed for a substantial expansion of the Inter-Syndical Unity Pact,which by mid-1957 claimed to have about 104 affiliated trade union or-ganizationsincluding the largest ones in the state of So Paulo.62 Inaddition to dealing with traditional labor issues, the Inter-Syndical

60 For example, Notcias de Hoje [hereafter NH], 9 May 1954, front page and p. 6, 15 May1954, pp. 6 and 8, and 2 September 1954, p. 2.

61 United States, NARA, I. Salert, Visit to Labor, Industry and Commerce Minister Alencas-tro Guimares, 21 September 1954, RG 59, 832.06/9-2154, and I. Salert, Quarterly Labor Re-view Third Quarter 1954, 8 October 1954, RG 59; United Kingdom, PRO LAB 13/503, by L.Mitchell, Labour Report no. 81, 14 October 1954, pp. 12.

62NH, 1 May 1957, p. 2.


22/35

114 Colistete

Unity Pact and individual trade unions embarked on campaigns for the

revision of minimum wage levels, social security reform, and even onmacroeconomic issues such as inflation and industrial policy.63 In thelate 1950s the Inter-Syndical Unity Pact was replaced by the Trade Un-ion Council (Conselho Sindical dos Trabalhadores or CST) as the ma-jor interunion organization in the state of So Paulo. Launched in Sep-tember 1959, the Trade Union Council was apparently an attempt byBrazilian Labor Party unionists and followers of Jnio Quadros (Gover-nor of So Paulo and elected President in 1960) to regain the initiativein the labor movement in So Paulo. However, the left and independ-ents decided to participate in the venture and soon took over the major

positions in the new trade union organization.64The years 19561961 also witnessed growing ideological disputes

within the labor movement. On the domestic scene, the hard-line tradeunionists traditionally close to government remained in control of statefederations and national confederationssuch as the National Confed-eration of Industrial Workers. The influence of this group over the rankand file and intermediate labor leaders was small, so that its major ac-tivity was restricted to high-level articulation and contacts (with gov-ernment officers, the U.S. embassy, and the ICFTU) aiming to containthe growth of the left in the Brazilian labor movement.65 On the other

hand, there was fierce competition among Brazilian Labor Partys un-ionists, independents, left-wingers, and other groups for hegemony inthe trade unions. In So Paulo, significant examples of this were, first,the formation of the Trade Union Renewal Movement (MovimentoRenovador Sindical) from a split in the Communist Party in 1957; sec-ond, the growth of the Catholic militancy; and, third, the emergence ofgroups associated to Jnio Quadros and Joo Goulart.66 Notwithstand-

63 United States, NARA, B. Sowel, Official Report on Labor, Year 1958, 16 March 1959, RG59, 839.06/3-1659, pp. 23; and United Kingdom, PRO LAB 13/1339, R. Morris, Labour Re-port no. 16/59, 10 June 1959.

64

United Kingdom, PRO LAB 13/1339, R. Morris, Labour Report no. 33/59, 10 October1959, p. 1, and R. Morris, Labour Report no. 40/59, 12 November 1959, pp. 12; United States,NARA, J. Fishburn, Brazilian Labor 1959, 7 December 1959, RG 59 832.06/12-759, pp. 67,W. Cochran, City of So Paulo Trade Union Convention, 18 April 1960, RG 59, 832.062/4-1860, and R. Burton, Goulart Appears to Have Headed Off Split in So Paulo Labor Movement,31 August 1960, RG 59, 832.062/8-3160. For a strong criticism, by a left-wing group, of thedissolution of the Inter-Syndical Unity Pact and participation in the Trade Union Council, seeAo Socialista (October 1959), p. 3. The group was theLiga Socialista Independente.

65 For instance, United States, NARA, J. Fishburn, Deterioration in Leadership and Anti-Communist Unity of the Three Older Labor Confederations, 30 June 1960, RG 59, 832.062/6-3060.

66 On the Movimento Renovador Sindical, see Harding, Political History, pp. 33043. Forthe Catholic militants, see Martins,Igreja, chs. 12. On the Brazilian Labor Party and Goulartssupporters, Benevides, PTB, chs. 56; and on Jnio Quadross supporters: United States,NARA, J. Fishburn, Brazilian Labor 1959, 7 December 1959, RG 59, 832.06/12-759, pp. 67.


23/35


ing increasing disputes, the Communist Party remained the single most

powerful political force in the labor movement in So Paulo and Brazilduring the second half of the 1950s. It was made clear in regional andnational trade union conferences in 1960, when the communists wereable to decisively influence the sorts of issues discussed and the resolu-tions adopted.67 Similarly, in 1958 the U.S. Labor Attach judged thatthe communists had significant influence or control over one-third ofthe 250 trade unions and over six out of 15 federations in the state ofSo Pauloincluding the largest in terms of number of workers, theSo Paulo textile and metal workers trade unions.68 Once again, thegovernment and the industrialists had failed to create a representative

wing of labor leaders who could successfully compete with the leftistsand independents.

What happened to wages and productivity during the recovery oftrade unions? As we know, Vargas and his followers sought to forge acloser relationship with workers between 1951 and 1954. However, realindustrial wages were unstable at the time, contrary to what one wouldexpect when the government was seeking closer ties to labor. In Brazilsmanufacturing industry, an estimated real wage increase of 8.7 percentper year between 1950 and 1952 was followed by a drop of 9.0 percentin 1953, and then by an increase of 12.1 percent in 1954 (Table 1, col-

umn 1). The record of industrial wages in So Paulo was worse than inBrazil as a whole. Wages there dropped 2.8 percent between 1950 and1952 and increased only 1 percent in 1953. Not until 1954 was there anappreciable improvement in average wage levels (10.9 percent between1953 and 1954) (Table 2, column 1). The struggle for real wage in-creases continued thereafter, although there were substantial yearlyvariations, especially in So Paulo. In Brazils manufacturing industry,industrial wages grew 2.8 percent per year between 1956 and 1958, fellby 5.2 percent in 1959, and increased 3.6 percent in 1960. At the end ofKubitscheck presidency, average manufacturing wages were only 3.6

percent above their level at the beginning of the governmentor 10percent if 1962 is taken as a more reliable reference year (Table 1, col-umn 1). So Paulos industrial wages did worse. After an increase of 3.0percent per year between 1956 and 1958, manufacturing wages fell by10 percent in 1959 and then rose slightlyonly 1.9 percent in 1960.Overall So Paulos manufacturing wages fell by 3.5 percent during the

67 United States, NARA, W. Cochran, City of So Paulo Trade Union Convention, 18 April1960, RG 59, 832.062/4-1860, and J. Fishburn, Third National Labor Congress, 4 August 1960,RG 59, 832.062/8-460.

68 United States, NARA, B. Sowell, Briefing Paper on Labor in So Paulo, 27 October 1958,RG 59, 832.06/10-2758, p. 16.


24/35

116 Colistete

Kubitscheck government, or 0.9 percent if 1962 is chosen for compari-

son (Table 2, column 1).Labor productivity growth was relatively slow in the early 1950s

2.7 percent per year in Brazil and 0.8 percent per year in So Paulo be-tween 1950 and 1953. As a result, unit labor costs oscillated without aclear trend in the period: in Brazil, labor productivity grew more slowlythan wages between 1950 and 1952 but regained the lost ground in1953; in So Paulo, labor productivity did better in 1953 but was belowwage levels in 1950/51 (Tables 1 and 2, columns 2). Yet, in a clear de- parture from its earlier performance, labor productivity exhibited astrong upward trend in the 19541962 period. Tables 1 and 2 indicate

that labor productivity in Brazils manufacturing industry rose 5.2 per-cent per year between 1954 and 1962, whereas in So Paulo the in-crease was of 5.0 percent per year in the same period (columns 2). Con-sequently, unit labor costs (columns 3) experienced a noticeable declinebetween 1954 and 1959: from 0.97 to 0.76 (in Brazil) and from 1.05 to0.74 (in So Paulo), with just a slight increase between 1960 and 1962(to 0.79 in Brazil and 0.76 in So Paulo). This was an unprecedentedoutcome during the postwar years and it suggests that organized laborcould not keep its share of the industrial output from diminishing at atime when labor productivity was soaring.

The odds of a corporatist accommodation between labor and man-agement, which could turn the upsurge in productivity into an equiva-lent increase in real wages and welfare, were slim during the 1950s andearly 1960s. Industrialists and the most influential sections of the tradeunion movement had very different views. As a result, there were fre-quent disputes between industrialists and the trade union movementover setting wages. The political divide also prevented labor organiza-tions from participating directly or indirectly in the formulation of eco-nomic and social policy. Particularly during the second half of the1950s, the trade union movement was increasingly concerned with more

general economic issues, which might have become a basis for collabo-ration with the government and even industrialists. Nonetheless, indus-trialists showed little disposition to negotiate wages and other issuesthat interested organized labor. For example, So Paulo industrialistssystematically rejected all attempts in the 1950s to shorten the workweek, grant maternity leaves, limit night work, or provide compensationfor accidents.69 The crucial issue for Brazilian industrialists seems tohave been that they were still highly suspicious of the leadership of themajor trade unions and interunion organization.

69Boletim Informativo da FIESP, no. 98 (1951): 31, no. 156 (1952): 326, no. 223 (1954): 2632, and no. 452 (1958): 70102.


25/35


TRADE UNIONS AND THE COLD WAR: THE ICFTU IN BRAZIL

The conflicts among diverging orientations in the trade union move-

ment were aggravated by the Cold War. In particular, international laborpolitics started to play a part in disputes and conflicts within the trade un-

ion movement in Brazil during the second half of the 1950s. TheICFTUs inter-American section, Organizacin Regional Interamericana

de Trabajadores (or ORIT), had been set up in January 1951 at a confer-

ence in Mexico to co-ordinate the anticommunist trade unions in theWestern Hemisphere. A few years later, in April 1953, an ICFTU/ORIT

office was established in Rio de Janeiro with the express aim of fightingthe left-wing influence in Brazils trade unions. The office was located at

the National Confederation of Industrial Workers headquarters and em-ployed one full-time trade unionist, who was in charge of propaganda andcontacts between the ICFTU/ORIT and Brazilian labor unions. A Consul-

tative Council was initially composed of organizations that had been inliaison with ICFTU since its foundation in 1949the national confedera-

tions of industrial and commercial workers plus five federations.70

Despite its ambitious beginnings, the ICFTU/ORIT office in Brazilwas beset with quarrels and internal disputes, so that it was ineffective

throughout the rest of the 1950s. The major obstacle to the operation ofthe office seems to have been the cautious support provided by thesmall number of organizations that made up the Consultative Council.For instance, the long-standing president of the National Confederationof Industrial Workers, Deocleciano Cavalcanti, on various occasionsshowed his reticence in following the strict anticommunist line pre-scribed by the ICFTU/ORIT, and even in being publicly associated withthese international organizations.71 As a high-level observer noted in1957, with the exception of Pequeno [president of the National Federa-tion of Urban Workers] and his friends, the others seem[ed] to be afraid

of showing too much sympathy with ORIT and the ICFTU.72

70 On the foundation of the Rio Office: Netherlands, International Institute of Social History[hereafter IISH], ICFTU Archives, Regional Fund Committee, 4th meeting, Brussels, 911 Feb-ruary 1953, p. 4; South America, General Correspondence, 5319, T. Gomez to J. Oldenbroek,18 March 1953; Brazil, General Correspondence, 5366, Arturo Juregui to J. Oldenbroek, 13September 1952; Brazil, General Correspondence, 5366, T. Gomez to J. Oldenbroek, 17 August1953. For the offices affiliates: Netherlands, IISH, ICFTU Archives, Brazil, Correspondencewith/on the Rio Office, 5373, J. Arajo to J. Oldenbroek, 16 August 1954.

71 For example, Netherlands, IISH, ICFTU Archives, Brazil, General Correspondence, 5368,D. Angelo to C. Millard, 1 December 1958; Memorandum (Confidential), 4 January 1958; R.Otero to H. de Horne, 20 January 1958.

72 Netherlands, IISH, ICFTU Archives, Brazil, General Correspondence, 5367, Herman to C.Millard, 17 December 1957.


26/35

118 Colistete

The cautious approach that the leading unionists from the Brazilian

Labor Party took to the ICFTU/ORIT seems to have resulted from thegreat influence that leftists and independents had in the postwar trade

union movement in Brazil. In the highly competitive and radicalized

contest for the key positions in the Brazilian labor movement during

the 1950s, labor leaders could not risk being linked to an organization

widely regarded as controlled by the U.S. government. Moreover, the

ICFTUs activities were not tied to an aid program that could foster

their influence among Brazilian labor, as happened in Europe during

the implementation of the Marshall Plan.73

In such a political and eco-

nomic environment, the impact of the ICFTU/ORIT office could only

be indirect and limited. Although the office could reinforce Cold Warlabor politics by providing an international link to the anticommunist

forces operating in Brazil, it could hardly be successful in establishing

the sort of solid anticommunist trade union block it had originally

planned.

The ICFTUs limited impact had important consequences for la-bor-management relations in the context of the highly polarized poli-

tics of the Cold War. Despite strenuous efforts, the government, in-dustrialists, and the ICFTU were not able to consolidate a cadre of

moderate labor leaders able to hold key positions in the larger trade

unions and, in particular, among the rank and file and intermediatelabor leaders. Nor was there a split in the labor movement that iso-lated the communists and leftists. It is true that the left and independ-ents showed great eagerness to collaborate with the government in a

program that included fast industrial growth and economic develop-ment. The majority of the labor leaders also tended to support price

controls, anticyclical economic policies and protection of national in-dustry. However, in all these cases the leadership and the rank and

file did not put aside the demand for better wages, workplace im-

provements and social rights. Instead, trade unions and the Inter-

Syndical Unity Pact advanced a view that economic developmentshould incorporate and share its benefits with the working class, bymeans of increasing real wages and welfare. A leftist leadership and

a trade union movement that pursued economic gains, welfare, andsocial reforms were hardly the sort of partners that industrialists

sought. A confrontational pattern of labor relations, which had firstemerged in the 1940s, became a regular feature of Brazils political

economy during the 1950s and the early 1960s.

73 Carew,Labour, chs. 78 and 10.


27/35


CONCLUSIONS

The tensions and conflicts generated by the industrialization processin Brazil were directly reflected in the sensitive area of labor relationsduring the postSecond World War period. Notwithstanding the highlycentralized system of industrial relations inherited from the EstadoNovo, state intervention failed to inhibit the organization and politicalexpression of an increasingly numerous and diversified working class.The result was the taking over of a large section of official trade unionsby left-wing and independent groups and the proliferation of organiza-tions parallel to the official corporatist structure after 1945. Compared

to other historical experiences, even the powerful anticommunist forcesorganized world-wide by the ICFTU had little hold over Brazilian laborpolitics during the 1950s. As a result, Brazil never experienced a corpo-ratist accommodation between labor and managers based on sharedgoals of economic growth and real wage increases tied to productivity.Instead it witnessed a widening gap between wages and productivity,and the workers share in output fell. Prospects of eliciting labor co-operation at the firm level and in the society at large under such circum-stances could be hardly achieved.

Two questions remain. First, why were left-wing groups so promi-

nent in Brazilian trade unions? The reason partly seems to lie in the re-pressive stance taken by both governments and employers in the imme-diate postwar years. Repression alienated even moderate labor leadersand made workers in general very distrustful of the cadre of officialtrade unionists cultivated by governments and industrialists. The leftand independents rapidly filled the gap and consolidated their positionsin the labor movement, despite the hostile political environment. Thecommunists in particular were able to establish a relatively solid foun-dation among the most militant sections of labor leaders and the rankand file. When the political situation changed, left-wing groups and in-

dependents occupied the main positions in the official trade union struc-ture (and set up parallel organizations), soon becoming the most activeand reliable representatives of organized labor in So Paulo and Brazil.

Second, why did industrialists tend to be so reticent about negotiatingwages and welfare with organized workers? It seems that the uncom-promising stance taken by industrialists was greatly reinforced by theDutra governments full commitment to Cold War politics. The briefpolitical liberalization between 1945 and early 1947 was marked by in-dustrial leaders calling for order and containment of the wave of indus-trial actions. The opportunity to curb the re-emerging labor movement

came with the Truman Doctrine of early 1947. Industrialists gave their


28/35

120 Colistete

full support to the wave of repression in trade unions and played their

part at the firm level by firing workers and tightening factory discipline.Therefore, rather than accepting a new role for the working class in theBrazilian political economy, and supporting open politics and negotiat-ing economic benefits and social rights, the industrialists backed an adjustment to the new postwar juncture that relied upon repression andlarge-scale intervention in labor issues.

A possible reason for such development goes beyond a deeply in-grained bent for authoritarianism by Brazilian industrialists. Productiv-ity growth, we know, was possibly slow in the early postwar years, andthat may help explain the hard-line stance by industrialists. Moreover,

long-standing industrial practices, based on low wages, poor conditions,and limited training of the labor force, tended to undermine the searchfor incremental improvements in industrial processes and organizationalstructures that could lead to steady increases in productivity. Industrialcompanies in So Paulo and Brazil may have thus seen the containmentof workers demand for better wages, conditions, and welfare as thebest way to assure their own survival as economic entities. But if such ahypothesis seems plausible for the early postwar years, the same is nottrue for the second half of the 1950s, when productivity soared in manu-facturing industry. At that point it was peculiar developments in the

trade union movement that kept industrialists from negotiating wagesand welfare with organized labor. They simply did not consider the left-ists and independents who had taken over the major trade unions inBrazil to be reliable partners.

The anticommunist policy promoted world-wide by the U.S. govern-ment found in Brazilian elites and governments its most dedicated sup- portersto the extent that the largest country in Latin America soonbecame a safe area in terms of communist threat. In particular, the Dutragovernments attempts to evict the communists from public life showedthat the U.S. government could direct its energies to other regions less

successful in their struggle against left-wingers. This situation helped toproduce an unexpected outcome, at least for local elitesthe exclusionof Brazil (and Latin America) from the major inflows of U.S. capitalduring the 1940s and 1950s. Another consequencethough less recog-nized todaywas that independents and left-wingers established a firmfoothold in Brazils trade union movement. Only at the end of the 1950sdid the U.S. government become aware of the ground it had lost to theleft-wing forces in Brazilian trade unions and of the dangerous situationcreated for its strategic interests in Latin America.

Compared to the postwar history of Western European countries and

Japan, the different outcomes of labor politics in Brazil could not be


29/35


more striking. Whereas in Europe and Japan, noncommunist trade union

leaders paved the way for a social democratic contract for growth, inBrazil the agenda for social reform was headed by the communists andother left-wing groups. Whereas in Western Europe and Japan man-agement counted on moderate labor leaders for a productivity drive inexchange for economic and social benefits, in postwar Brazil, employ-ers held an antilabor policyboth on the shop floor and in society atlargeeschewing compromise with demands for real wage increasesand welfare by the leftist labor militancy. The divided world of the ColdWar made even more difficult any settlement conducive to a compactfor growth in Brazil, which had the difficult task of bringing together

management and communist labor leaders.The irony is that where the U.S. anticommunist foreign policy found

its most willing followers, the social basis for sustained growth and so-cial reform turned out to be more and more unlikely. In Europe, in turn,U.S. policy-makers had to deal with consolidated social democraticgroups in order to prevent the hegemony of the pro-Soviet trade unionmilitancy. The adoption of a crude anticommunist policy gave way tobargaining large-scale resources and assistance epitomized by the Mar-shall Plan. Meanwhile, in Brazil and Latin America, eager cold-warriorssought to implement by all means available the anticommunist line

hinted at by the Truman Doctrine. Does this mean that the social com-pact for growth could only become a historical reality because of thelimits posed to the U.S. anticommunist thrust in Europe and Japan?

In any case, for Brazil the effects of postwar labor politics were clear.Even when conditions were favorable for economic growth and democ-racy, relations between industrialists and labor remained essentiallyhostile and antagonistic during most of the 1950s and early 1960s. Re-sults were felt a few years later, when the acute political and social con-flicts reached their climax with the military coup that, for a long time,radically altered the prospects for democracy and social reform in

Brazil.


30/35

122 Colistete

Appendix: SourcesYearly data for Tables 1 and 2 were obtained as follows:

WAGES

19451948: Extrapolated from the level of real average manufacturing wages in1949 by using the rates of growth of the industrial wages compiled by the Instituto deAposentadoria e Penses dos Industririos (IAPI), Brazil and state of So Paulo: Bra-zil, Instituto Brasileiro de Geografia e Estatstica [hereafter IBGE], Anurio, 1941 to1945, 1947, 1948, 1951. IAPI figures refer to wages paid in July.

1950 (So Paulo), 19521958: Registro Industrial, Brazil: Brazil, IBGE, Anurio,1952, 1955, 1956, 1957, 1959, 1960, 1965. Registro Industrial, state of So Paulo:Brazil, IBGE,Anurio, 1952; Brazil, IBGE,Produo Industrial de So Paulo, 1952,1953, 1954, 1955; Brazil, IBGE, Produo Industrial Brasileira, 1956, 1957, 1958.Mean wages of monthly average production workers.

1949 and 1959: Censo Industrial, Brazil and state of So Paulo: Brazil, IBGE,Anurio, 1955, 1963. Mean wages of monthly average production workers.

1950 (Brazil), 1951, 1960, 1961: estimated by interpolation using the formula ofannual compound rate of growth (see below).

Nominal average wages were converted to cruzeiros at 1952 rate by using the fol-lowing price indexes: Table 1-Brazil: the consumer price index ( ndice Geral dePreos ao Consumidor) of the Ministry of Labor between 1948 and 1962, weighted bythe share of national manufacturing output from the states (fourteen) included in thenational survey; for 19451947, national wages were deflated by the consumer priceindex of So Paulo Prefecture. Table 2-So Paulo: So Paulos consumer price indexof the Ministry of Labor between 1948 and 1962; for 19451947, So Paulos wageswere deflated by the consumer price index of So Paulo Prefecture. Source: Brazil,IBGE,Estatsticas Histricas.

LABOR PRODUCTIVITY

1950 (So Paulo), 19521958, 1949 and 1959: number of workers, same sources aswages above. The labor productivity data were obtained by dividing the valor datransformao industrial(a measure similar to the industrial value added: see Brazil,IBGE, Estatsticas Histricas, p. 370) by the monthly average number of productionworkers.

1950 (Brazil), 1951, 1960, 1961: estimated by interpolation using the formula ofannual compound rate of growth (see below).

19451948: number of workers, same sources and procedures as wages above.Given the lack of data similar to industrial value added for such years, measures wereestimated by backward extrapolation of the annual compound rate of growth (by fit-ting a log-linear trend) of the real valor da transformao industrialbetween 1949-1952 (Brazil) and 1949-1953 (So Paulo). The annual compound rates of growth ob-tained were 3.8 percent (Brazil) and 5.1 percent (So Paulo). The years 1949-1952(Brazil) and 1949-1953 (So Paulo) were chosen as the basis for extrapolation becauseof their relative stability before a structural break identified in Brazil and So Paulosvalor de transformao industrialseries (1953 and 1954, respectively) by a dummyvariable test for structural stability identical to that presented below for the labor pro-

ductivity series.


31/35


Nominal valor da transformao industrialwas converted to cruzeiros at 1952 rate

by using the ndice de Preos por Atacado - Disponibilidade Interna, produced byFundao Getlio Vargas. Source: Brazil, IBGE,Estatsticas Histricas.

CALCULATING THE ANNUAL COMPOUND RATE OF GROWTH

When all values of a series were available, the annual compound rate of growth wascalculated, first, by fitting a log-linear trend (with time as the explanatory variable) onall years of the relevant period and using the indexes presented in Tables 1 and 2; sec-ond, by taking the exponential of the regression coefficient, subtracting 1 and multi- plying by 100. For interpolation of missing values, the following procedure wasadopted: consider that the annual compound rate of growth (i) is implicitly defined by

t

t iXX

10010

where

Xt= value ofXat time t

Given X0 and Xt from the previous formula, the annual compound rate of growth be-tween 0 and tmay be obtained as follows

1001

/1

0

tt

X

Xi

TESTING STRUCTURAL STABILITY IN LABOR PRODUCTIVITY SERIES

The following model was estimated for both Brazil and So Paulo

LP = 1 + 2Di +B1Xi +B2(DiXi) =i (1)

whereLP = the 1952 index of labor productivity,Di = 0 for observations before 1954(Brazil and So Paulo) and = 1 for observations from 1954 (Brazil and So Paulo),Xi= time, andi = error term.

Regressions for the subperiods are

E(LP /Di = 0,Xi) = 1 + 1Xi (2)andE(LP /Di = 1,Xi) = (1 + 2) + (1 + 2)Xi (3)

The results based on models 1, 2, and 3 are as follows

Brazil LP= 71.005 29.618Di + 3.642Xi + 2.916DiXit= (22.941) (3.518) (6.621) (3.749)

R2 = 0.979F= 221.883

Model for subperiod 19451953 (Di = 0): LP = 71.005 + 3.642XiModel for subperiod 19541962 (Di = 1): LP = (71.005 29.618) + (3.642 + 2.916)Xi

= 41.387 + 6.558Xi


32/35

124 Colistete

So Paulo LP = 71.362 27.484Di + 3.867Xi + 2.413DiXi

t= (18.337) (2.596) (5.591) (2.467)R2 = 0.964F = 126.207

Model for subperiod 19451953 (Di = 0): LP = 71.362 3.867XiModel for subperiod 19541962 (Di = 1): LP = (71.362 27.484) + (3.867 + 2.413)Xi

= 43.787 + 6.280Xi

As the differential intercept and the differential slope coefficients (2 and 2, respec-tively) are statistically significant (as indicated by tstatistics) in both Brazil and SoPaulos basic models, there is strong evidence that the labor productivity series are notcontinuous through timein other words, that there is a structural break in the series.The models for subperiods show the different intercept and slope coefficients for each

subperiod in which the Brazil and So Paulos series were split.

REFERENCES

Abel, Cristopher, and Colin Lewis, eds. Welfare, Poverty and Development in LatinAmerica. Houndmills: Macmillan, 1993.

Ao Socialista, October 1959. So Paulo.Abreu, Marcelo. Brasil e a Economia Mundial, 19301945. Rio de Janeiro: Civiliza-

o Brasileira, 1999.Arquivo Edgar Leuenroth. Universidade Estadual de Campinas. So Paulo.Arquivo do Sindicato das Indstrias de Fiao e Tecelagem em Geral do Estado de

So Paulo, So Paulo.

Armstrong, Paul, Andrew Glyn, and John Harrison. Capitalism since 1945. Oxford:Blackwell, 1991.

Benevides, Maria Vitria. O PTB e o Trabalhismo. Partido e Sindicato em So Paulo,19451964. So Paulo: Brasiliense, 1989.

Bethell, Leslie. Brazil. In Latin America between the Second World War and theCold War, 19441948, edited by Leslie Bethell and Ian Roxborough, 3365.Cambridge: Cambridge University Press, 1992.

Bethell, Leslie, and Roxborough, Ian. Introduction: The Postwar Conjuncture inLatin America: Democracy, Labor, and the Left. In Latin America between theSecond World War and the Cold War, 19441948, edited by Leslie Bethell andIan Roxborough, 132. Cambridge: Cambridge University Press, 1992.

Biblioteca Roberto Simonsen. Federao das Indstrias do Estado de So Paulo. So

Paulo.Bielschowsky,

productivity and labour politics jeh

Documents