7

“Order and Progress”

INCLUSIVE NATION-STATE BUILDING IN BRAZIL

In 1822, Pedro I was enthroned as emperor of newly independent Brazil. The backdrop curtain for the coronation, reproduced in a print by Regis DeBret, symbolized Brazil’s image of itself at this transitional moment. At the center is the Portuguese prince, ruler of the empire. Grouped around his throne paying homage are representations of the Brazilian nation, including black and white soldiers, workers, and black slaves. It is an image of state founding remarkable for its projection of centralized power personified by Pedro, inclusive unity, and popular loyalty. A comparable image of state founding in the United States or South Africa would surely have represented the culmination of conflict, whether the American War of Independence or Civil War, or South Africa’s Boer War. And surely the comparable American or South African representation would not have included blacks as symbolic members of the nation represented by that state. Nor would South African or U.S. imagery have placed at its center a descendant of the previous colonial power. But Brazil’s foundation as an independent empire had not been the result of conflict, but of a peaceful transition that retained historical links to the colonial power of Portugal.

A second image from Brazilian history carries similar symbolism. In 1889, the son of Pedro I lost his Brazilian throne with the declaration of a republic. Before Pedro II sailed back to his ancestral homeland of Portugal, Brazil threw a party for him on the Ilha das Cobras in Rio’s majestic Guanabara Bay. By all accounts, this was a festive occasion, including white and some mulatto elites. What is remarkable is that such an inclusive and cordial festivity occurred at all. The transition from empire to republic, almost simultaneous with abolition, had been achieved with little rancor. A celebration was in order. It is simply impossible to imagine a comparable party being thrown jointly by Americans and their former colonial rulers to celebrate the forced withdrawal of the British, with Southerners to celebrate the Union’s victory and abolition after the Civil War, or by Afrikaners and British to celebrate the founding of South Africa after the Boer War. Brazil was not born of comparably bloody conflict, but rather emerged from a relatively gradual and peaceful transition. Instead of gunpowder smoke and human carnage, Brazil emerged from slavery and empire in a hail of flowers and dancing.

Unity and Discrimination

Brazil’s peaceful transitions from colony to empire and then to republic were made possible by, and indicate, the capacity of a state able to finesse change. In part, this relatively unified state was a legacy of the traditions of Portuguese rule. Colonialism had established more centralized rule and greater bureaucratic capability than the decentralized capitalism imported by Britain and Holland to eastern North America and South Africa. The earlier withdrawal of the Dutch and the arrival in Brazil of the Portuguese king fleeing Napoleon had further diminished internal conflict that might have disrupted state coherence. This tradition was then reinforced during the empire. As a result, the Brazilian state remained strong enough to pursue its ideal of “peace and harmony . . . to avoid confrontation.” The preservation of authority and relative absence of violent conflict encouraged little mobilization. What regional conflict did emerge consistently ended in compromise.¹

The Brazilian state’s early tendency to conciliate should not be seen as a sign of weakness, but as a signal of a state able and confident enough to bargain away conflict in order to preserve its authority.² Patrimonial rule was never seriously challenged, as is reflected in the small number of deaths due to internal conflict. The state and its related elites remained largely hegemonic, even if official authority did not fully penetrate into vast areas of territory.

The state’s encouragement of an integrating social order was shaped and reinforced by cultural norms and discourse inherited from Portuguese colonialism. Informed by Catholicism, the Portuguese had constructed an inclusive hierarchy. Unlike in South Africa and the United States, no absolute social or racial dichotomy was enforced. This inclusive hierarchy would be adjusted after abolition, with blacks still at the bottom and mulattoes as intermediaries. By thus maintaining the image of a social and racial continuum, Brazilian culture and policy diminished the prospect of antagonism.³ This argument was contradicted by discrimination and would be contested among elites and later by Afro-Brazilians (see Chapter 10). But the image of social inclusion remained a strong element of culture, consistent with state policies of incorporation.

The use of compromise to preserve stability guided Brazil in ending slavery, a process so contentious elsewhere. Abolition in Brazil did not disrupt the social hierarchy in which blacks, slave and free, were already incorporated at the bottom. The end of slavery could be achieved gradually, provoking remarkably little of the kind of opposition that elsewhere inflamed racist ideology and practice. Slavery had been a national institution in Brazil, but a regional one in the United States and a more ethnic one among Afrikaners in South Africa. In 1819 “no region had less than 27 percent slave out of its total population. . . . [S]laves continued to be distributed throughout the Empire at a remarkably uniform rate.”⁴ Similarly, “abolition sentiment and agitation was not limited to any one section of Brazil,” with pre-abolition manumission already widespread.⁵ When the poorer Northeast eventually did become more committed to abolition and Southerners called for immigration to replace slaves, abolition could be managed by the state with little tension or social disruption, which might otherwise have curtailed emancipation or challenged the state itself. Brazil escaped anything comparable to the American Civil War, and achieved emancipation without civil strife.⁶ Indeed, Brazilians were eager to avoid the sort of conflict they had seen erupt in the United States.

Abolition nonetheless did have a significant political impact in Brazil. Former slaveholders angry at the crown for abolition threw their support behind the growing republican movement, already strong among an underpaid and dissatisfied military, advocates of greater religious freedom, and industrialists.⁷ Abolition and republic reinforced each other: the military declared the republic a year after abolition. The 1888 prediction that with abolition “your Highness has redeemed a race but lost a crown” proved correct.⁸ In some respects, “1889 did not mark a significant break in the Brazilian historical process.”⁹ The republic was ushered in by the same elites who had prospered under the empire, and with so little violence or rancor as to make possible a pleasant farewell party for the emperor.

Despite the lack of conflict in its founding, republicanism did fundamentally change the form of government. Without the centralizing power and symbolism of the emperor, the republic adopted an American style of federalism, breaking up unitary authority into more autonomous states.¹⁰ Brazil moved in a direction opposite to the greater centralization that characterized the U.S. Reconstruction period and South Africa after the Boer War. In part, the Brazilian ability to shift toward state decentralization demonstrated the greater confidence of Brazilian elites that devolution of power would not bring the sort of conflict already avoided. Despite a major peasant revolt in 1897, decentralization would last until 1930, when the tradition of a more strongly centralized state was reasserted.¹¹

Peaceful abolition and republican federalism provided little pressure or capacity for central state intervention to assist freed slaves. Neglect of those slaves earlier manumitted set the pattern. Brazilian abolitionists had rarely raised the issue of how freed blacks would be treated, and once abolition was achieved, most blacks simply continued to work as before, but now for low wages.¹² The long-established social and racial hierarchy remained intact, in contrast to the U.S. South’s social order resting on a dichotomy of slave and free shaken by war and abolition. Brazil’s “racial order remained almost identical to that which existed under the caste system; and the negro never found in the white man any real support.”¹³ With no Civil War ushering in abolition, there was also no Reconstruction. Proposed land redistribution was aborted, and in general, the issue of black advancement was eclipsed by the rising tide of republican fever.¹⁴ With no expectation of state intervention to help blacks and little pressure to impose formal segregation already maintained informally, power could be and was diffused for a time. Blacks themselves emerged “from slavery materially and psychologically ravaged,” lacking any means to advance themselves or to compete, isolated in rural areas or in the newly emerging urban slums, or favelas.¹⁵

Not only were blacks abandoned to their fate, but their prospects were diminished by official efforts to further encourage European immigration. The end of slavery and rising industrialization had produced great demand for free labor, which the Brazilian elite preferred to fill with imported whites rather than by training and advancing blacks.¹⁶ Widespread immigration was directly encouraged, with private funds until 1889 and with public funds thereafter used to pay for the passage from Europe.¹⁷ State support for immigration was deemed necessary to overcome reluctance among Europeans to come to Brazil; it was explicitly described as a form of compensation for former slaveholders who had lost their captive labor pool.¹⁸ From 1870 until 1963, over five million immigrants came to Brazil. Eighty percent of the 3.7 million arriving before 1930 settled in the industrial center of the Southeast around São Paulo. By 1920, 50 percent of industrial workers were foreign-born.¹⁹ Those who came brought with them European ideas of white racial superiority, though there is some debate about how influential these immigrants’ views were in Brazil.²⁰ While European immigration was encouraged, resulting advantage was enjoyed without formal, domestic racial domination. And such inequality cannot be explained as being caused by white labor’s interests. The preference for importing white labor was established when Brazil still had few white workers, with mulattoes used to fulfill intermediate economic functions elsewhere filled by poorer whites.²¹

Racial prejudice in Brazil was no doubt influenced by European ideas, but it was already well established before the bulk of immigrants arrived. Fears of earlier slave revolts and racially based resistance to dependence on black workers had motivated state support for white immigration as an alternative source of labor. In this sense, Brazil’s immigration policies mirrored U.S. policies encouraging black return to Africa. Rather than exporting blacks, Brazil favored importing whites, but the purpose was the same – to dilute the black population. In 1890, Brazil banned the immigration of blacks, a ban again debated and ultimately reinforced in the 1920s and 1930s.²² As late as 1945, Brazilian legislation stipulated a preference for immigration that would “develop in the ethnic composition of the population the more desirable characteristics of European ancestry.”²³ This continued preference for white immigrants was defended, as it had been at the turn of the century, as providing Brazil with a steady “injection of civilization” as a means to “purify the race.”²⁴

The ban on black and encouragement of European immigration were complementary components of a general policy toward “whitening,” shifting the population’s color balance toward lighter shades. Encouragement of miscegenation, already pervasive, further reinforced this process. And the image of class mobility for mulattoes encouraged accommodation. Thus, whitening further served to diminish nonwhite solidarity and the prospects for conflict the elites feared. Black solidarity and potential mobilization were diluted; as long as it was possible to become white, there seemed little reason to insist on defining oneself as black, with all of the associated negative stereotypes.

Whitening led to a real increase in the proportion of Brazil’s population identifying itself that way. So-called whites rose from 44 percent in 1890 to 62 percent by 1950, with a corresponding fall in the proportion of the population identified as mulatto.²⁵ Black numbers were relatively diminished by white immigration, intermarriage, low reproduction, and social rewards encouraging people to redefine themselves as lighter.

Whitening as a conscious policy was both explicit and implicit, and a consequence of Brazil’s early racial discrimination. The image of Brazilian racial tolerance is directly confronted by the stated post-abolition preference for white immigrants and for a diminution of the Afro-Brazilian presence. Denigration of slaves had clearly shifted post-abolition to a greater explicit focus on racial images, as it had in the United States.²⁶ At the same time, Brazil’s particular history of miscegenation forced a subtle but significant reinterpretation of European eugenics. “They borrowed racist theory from Europe and then discarded two of that theory’s principal assumptions – the innateness of racial difference and the degeneracy of mixed bloods.”²⁷ Brazilians embraced their miscegenation, assuming that such mixing would produce ever lighter generations because of the greater strength of white genes.²⁸ Meanwhile, they encouraged upwardly mobile mulattoes to align themselves and even define themselves with whites, and blacks to define themselves as mulatto.²⁹ The resulting later increase in mulatto categories, a “browning” more than a whitening of the population, was embraced as consistent with overall lightening.³⁰ Whitening was thus a reflection of Brazil’s racial hegemony, where whites were explicitly preferred, with some flexibility about including mulattoes in that or intermediate categories.³¹ Racial prejudice was clearly inscribed into the Brazilian social order and official policies, reflecting white popular ideology and culture.

In light of the U.S. and South African experiences, as well as its own history, Brazil’s post-abolition experience of racial prejudice and discrimination is not surprising. What remains remarkable is the constrained and distinct application of these principles. Whitening could be and was achieved without the establishment of biracial categories or official segregation. Indeed, such forced segregation might have discouraged miscegenation, which Brazilians embraced as means to dilute the black presence established by slavery. Instead, integration of a sort was encouraged, not to advance blacks but to absorb and dilute them culturally as a distinct group. White racial domination was still the goal, albeit unstated. But the means of achieving this goal appeared to be the opposite of U.S. and South African policies.

The Persistent Myth of Racial Democracy

Post-abolition Brazil inherited an entrenched social hierarchy and racial prejudice. But instead of enforcing official racial domination, Brazil projected an image of an inclusive nation-state and racial democracy. Compared to South Africa or the United States, Brazil entered the modern age with a more consolidated state apparatus then decentralized in the republic. There had been no major intrawhite conflict that would be resolved through official racism and policies encouraging white unity. Instead of provoking conflict by means of such exclusion, Brazil sought stability and maintenance of its social hierarchy through incorporation of all, consistent with its cultural predilections. Earlier national consolidation required no legal racial crutch for its preservation, but rather was maintained by avoiding policies of segregation, amid debates and challenges.

Post-abolition Brazil’s distinctive lack of formal segregation was understood at the time as a significant departure from contemporaneous U.S. official practice. No less a figure than Theodore Roosevelt visited Brazil and in 1914 wrote of Brazil’s “tendency to absorb the Negro . . . [and] to draw no line against the Negro.” Roosevelt went on to quote approvingly an anonymous Brazilian’s comments:

“You of the United States are keeping blacks as an entirely separate element and you are not treating them in a way that fosters their self respect. They will remain a menacing element in your civilization. . . . [W]e Brazilians have chosen [an alternative that] will in the long run, from the national standpoint, prove less disadvantageous and dangerous than the one you of the United States have chosen.”³²

Roosevelt’s quote suggests two elements that help to explain Brazil’s alternative racial order. First is the notion of blacks as a potential “menace” to social order. Brazil had directly experienced this threat during its long history of slave revolts, more extensive than those in the United States. With a state and culture fixated on the idea of unity and stability, it is not surprising that they chose to suppress such conflict. They sought to diminish the threat of future unrest by providing a seeming means of inclusion and by avoiding formal segregation, which would likely encourage animosity and protest for redress. By contrast, the post–Civil War United States had experienced a less comparable threat from blacks, a relatively much smaller proportion of the population. U.S. whites were therefore less fearful of the prospect of black revolts and used official segregation to head off any such threat.

Roosevelt’s quote from his Brazilian informant also suggests that the American and Brazilian models of race relations were alternatives among which both countries were somehow free to choose. But because in the United States a violent regional conflict was healed, in part, by Northern appeasement of Southern demands for racial domination, the country was not then free to choose the alternative advocated by Theodore Roosevelt. Social order, shaken by war, had to be restored. Abandoning the ideology and practice of racial domination might very well have reopened wounds that had already once torn apart the Union. The imperative for legal racial domination had been so established in America, as it had been in South Africa.

Brazil experienced much less violent regional and no ethnic conflict requiring resolution of the kind elsewhere achieved through the projection of white racial unity. Describing racial domination in the United States and South Africa as the result of such pressure for reconciliation is consistent with Brazil’s having less comparable conflict to be reconciled and enacting no official racial domination for this purpose. In Brazil, a consolidated nation-state had long been in place, with civil wars elsewhere in the region reinforcing Brazil’s efforts to remain distinguished for its “white,” or peaceful, revolutions.³³ This result had significant turning points, and emerged gradually.

Comparison with South Africa and the United States highlights the relatively inclusive nature of Brazil after abolition, and the absence of official racial domination. State-led nationalism was purposefully inclusive. Brazil established no systematic discrimination, and accordingly Jim Crow laws were not enacted.³⁴ Informal racial discrimination was evident, but official racial domination was not. This was not due to lack of capacity by the Brazilian state, which emerged from a past of centralized power able to manage relatively successfully major transitions with little violence and to support immigration and project nationalist ideology.

Despite republican decentralization, the Brazilian state did act, to some extent by omission, to enforce its particular internal racial order. Racial categories were encoded in the census but not used as the basis of official discrimination. Instead, on occasions when the census was carried out after 1890, it generally encouraged the categorization of as many people as possible in intermediate categories of mixed ancestry, demonstrating that such mixing was leading to increased whitening. Though census categories were official constructs, results were deemed as scientific evidence of Brazil’s racial continuum, lightening, and lack of a firm racial divide. Brazilians shared this assumption.³⁵ State enumeration and popular culture combined to reinforce an image of complexity, countering any sense of biracial distinctions on which segregation might have been constructed.

State actions reinforced a racial order, if not legal domination, but the tensions and conflicts over policy should not be ignored.³⁶ Black immigrants were banned, though this much-debated official discrimination applied only to outsiders. Official forms of segregation applied to Afro-Brazilians within the country were also debated and then purposefully rejected as heightening the potential for internal conflict.³⁷ Vagrancy was outlawed and illiterates were not enfranchised, but while these policies adversely affected blacks they were not complemented by racially encoded segregation. Between 1891 and 1907, proposals to establish a formal color bar were debated in the Brazilian parliament, but were not enacted.³⁸

The relative preservation of state power coincided with preservation of the social hierarchy, reducing the imperative for legal buttress. Abolition had occurred so gradually that the racial order was not shaken. Blacks were left unable to compete with whites, even in the absence of official racial domination. Slavery had ended with so little disruption that no color line had to be drawn to protect whites.³⁹ This left the Brazilian state free to distance itself formally from any official policies of racial domination and instead to claim the higher moral ground by outlawing such practices. The absence of Jim Crow and the existence of laws against such discrimination consistently convinced American visitors, including many prominent blacks, that Brazil was a “racial paradise.”⁴⁰

To firmly establish the image of Brazil as such a racial paradise required not only official rejection of present discrimination, but also the purposeful reinterpretation of Brazil’s past. For those interested in avoiding racial conflict, it was not enough to establish new laws banning discrimination. The history of discrimination was denied as well, so as to avoid any use of such history to foment antagonism or mobilization. Such prevalent historical reinterpretation is most evident in post-abolition descriptions of slavery, which consistently focused on paternalistic tolerance. There was little intellectual dissent and sufficient earlier rhetoric of historical tolerance to embrace. The actually harsh treatment of the majority of slaves was purposefally brushed under the rug of modern images and “the myth of the benign master.”⁴¹ Images of Catholic- or Portuguese-inspired racial tolerance and ease of mixing were celebrated along similar lines, reinforcing cultural assumptions of inclusion.

As further proof of the historical legacy of tolerance, white Brazilians trumpeted the nation’s embracing of African culture. African-influenced dance and music, the candomblé religion and other cultural forms “sprang back” after 1888, were meshed with European traditions in a variety of synchronism, and then were re-Africanized.⁴² But projecting the survival of African culture as a signal of tolerance ignored its roots in the long-maintained slave trade.⁴³ African culture, together with benign images of slavery and miscegenation, became grist for the mill of revisionism.

The absence of official discrimination and complementary reinterpretations of history emerged as the basic building blocks for a state-supported ideological projection of Brazil as a racial democracy. This term was popularized by Gilberto Freyre, whose interpretations of Brazilian history thus provided the foundation on which other scholars built. According to Carlos Hasenbalg, “Freyre created the most formidable ideological weapon against blacks.”⁴⁴ He and his followers created an image of history and the present that denied any subordination against which blacks might resist. Inequality remained largely unchallenged by blacks, encouraged to see no barriers to their advancement.

Racial democracy was both a product of and a crucial reinforcement for post-abolition republican nationalism. This was the ideological glue that held together the Brazilian nation. Like the selectively applied liberal creed in the United States, racial democracy was projected as the underpinning for national unity and social peace. These were the paramount concerns of the Brazilian elite who purposefully projected an image of harmony to dampen earlier tensions and to avoid confrontation within a single inclusive culture.⁴⁵ Brazilians embraced the view that they had had no problem with race so that there would be no such problem.⁴⁶ This ideology of assimilation was embedded in a nationalist celebration of racial tolerance and unity.⁴⁷ The Brazilian nation projected itself as analogous to the romanticized image of the big house of the slaveholder, in which everyone had their place in a paternalistic and inclusive hierarchy.

Racial democracy would become a source of Brazilian national pride and criticism of the tolerant image became tantamount to treason.⁴⁸ Even numbers that contradicted the image of racial democracy were condemned to oblivion. Freyre himself described “racial statistics as irrelevant.”⁴⁹ Even before Freyre had refined the official view, racial specifications were simply omitted from the Brazilian census of 1900 and 1920, with no census at all conducted in 1910 and 1930.⁵⁰ Later censuses camouflaged the racial divide with categories of mixed “color.” Studies of racial discrimination, already hampered by the lack of racially categorized statistics, were later outlawed.⁵¹ Such extreme measures to preserve the image intact suggest that racial democracy resonated with a central core of Brazilian tradition.

The convergence of Brazilian nationalism and the image of racial democracy had the almost magical result of projecting an established social and racial order without conflict. All had their place in a “natural plan,” reinforcing social distance.⁵² That order clearly reinforced inequalities, but these were generally understood to reflect economic class rather than race, despite the superimposition of black and poor.⁵³ Undereducated blacks were unable to compete for jobs, and as illiterates were not permitted to vote. But their deprivation and exclusion were not associated with any explicit rules of racial domination.⁵⁴ Blacks were formally free to educate themselves, to advance economically, and to become absorbed. They embraced such supposed opportunity, believing in their own mobility despite evidence to the contrary.

Not seeing any official coercion or barriers to their absorption, Afro-Brazilians were encouraged to accept their inferior condition as a “pathology of normality.”⁵⁵ Racial democracy deprived Afro-Brazilians of any legally explicit cause for their subordination against which they might mobilize. Even contemporary Afro-Brazilian activists have been forced to conclude that purported absorption long worked to discourage blacks from identifying themselves in terms of race, thereby avoiding racial conflict. As we shall see in Chapter 10, until fairly recently Afro-Brazilians did “not even pose a threat to the system.”⁵⁶ Blacks found themselves in a racial paradise that was indeed heavenly for those elites whose status was seamlessly maintained, but purgatory for the supposed beneficiaries of racial democracy.

Brazil’s commitment to an inclusive, nonracial nationalism and to conflict avoidance is underlined by its persistence. Social order was maintained throughout the twentieth century’s varied and significant political transitions. The Brazilian polity shifted from republic to dictatorship to democracy to military rule and to more decentralized democracy, but each shift was accomplished relatively peacefully. Official racial domination was consistently avoided, while the effective racial order was left largely unchallenged. The state remained intact, even as it failed to settle firmly on a political expression or form, while the nation itself remained relatively unified. The lack of a founding conflict comparable to the American Civil War or South African Boer War would seem to have allowed a legacy of unity strong enough to withstand institutional rearrangements. And as a central part of that entrenched unity, the image of racial democracy was preserved and continued to deflect potential racial conflict.

As earlier in South Africa and the United States, Brazil embraced state consolidation as a prerequisite for economic development. Decentralized republican rule was abandoned in favor of a resurgent state centralization after the 1930s, a process led by Getúlio Vargas, who served as chief of a provisional government from 1930 to 1934, Congressionally elected president until 1937, dictator until 1945, and again elected president from 1951 to 1954. Faced with minor armed revolts in 1935, Vargas “used the rebellion as a vehicle to enlarge federal power” under a state of siege. He forced the subordination of the states, their governors and militias, and even large landowners weakened by the Depression.⁵⁷ Industrialists, the middle class, and the national military generally supported Vargas’s centralization, agreeing on the need for state economic intervention and having tired of the costly rivalries that had grown under republican federalism.⁵⁸ Appealing to the Brazilian tradition of hierarchy, Vargas convinced many that his efforts to reinforce state and national unity were more important than a divisive “democracy of parties.”⁵⁹ Support for this position is suggested by the 1937 dictatorship having been established without a shot fired.⁶⁰

Vargas engineered a massive expansion of the centralized state, comparable to that achieved earlier in South Africa and in the United States under the contemporaneous New Deal. The number of public employees rose from 30,000 in 1912 to 65,000 in 1938, and government receipts doubled in the first eight years of Vargas’s rule.⁶¹ Various social services were established and by 1951 the state had even moved to take over some industrial activities with the formation of the Petrobrás oil company. Resulting opportunities for massive patronage further reinforced popular support for the regime.⁶²

Vargas modernized the Portuguese colonial and Brazilian tradition of a strong state extending its power into society to reinforce the established hierarchy. His Estado Novo was the quintessential corporatist regime of state-ordered interest representation.⁶³ Vargas licensed mandatory trade unions, ensuring their members social benefits such as a minimum wage in return for guarantees against strikes. By 1944, half a million had joined these unions, which continued to operate under the same rules for the next four decades.⁶⁴ The FUNRURAL organization established a similar pattern for rural workers.⁶⁵ By 1945, Vargas allowed and established political parties, though even these remained influenced by his vision.⁶⁶ It is difficult to imagine a more complete convergence of the nation and state, with the state orchestrating popular organization and reinforcing nationalist loyalty through corporatist institutional design.

Brazil under Vargas consolidated the nation-state to an even greater extent than the United States and arguably South Africa without resort to racial exclusion for white unity. Instead, Vargas reached out to Afro-Brazilians and was rewarded with their loyalty. He encouraged the image of racial democracy. And his policies gave credence to the ideology of tolerance. Vargas ended support for white immigrants who competed with blacks. He encouraged expressions of African culture such as samba, candomblé, and spiritistic umbanda, in which he reportedly participated. In 1934, he supported the first national Afro-Brazilian Congress. He signed the 1951 Arinas anti-discrimination law, and partially opened the civil service.⁶⁷ At the same time, Vargas sought state control over black mobilization and religious and cultural practices, consistent with his corporatist philosophy. But “Vargas never spoke [for] racism, allowed candomblé . . . created social security . . . so he got black support.”⁶⁸ This support was only partially diminished when Vargas closed the largest black organization, the Frente Negra, when all such popular organizations were banned in 1937.

Vargas’s support for and from blacks and the highpoint of racial democracy rhetoric coincided with a dramatic resurgence of state centralization. The state provided for all, of any color, according to corporatist categories that did not include race.

After a decade of increasingly left-leaning democracy that followed Vargas’s death, a 1964 coup brought the military to power for two decades of repressive rule. Adding national security to the traditional emphasis on national unity, the junta crushed divisive popular mobilization and put restraints on what had become an increasingly militant trade union movement.⁶⁹ At the same time, the military asserted an economic form of nationalism by breaking with the International Monetary Fund and appropriating foreign investments.⁷⁰ State-led investments produced a major economic boom. Phenomenal growth funded a further expansion of the state and a dramatic increase in salary for civil servants and the military.⁷¹ The middle class particularly benefited, but the poor, particularly the black poor, suffered as inequality increased. From 1960 to 1990 the top decile enjoyed an 8.1 percent increase in income, while the bottom half saw their income decrease by 3.2 percent.⁷² In short, the military state served the interests of the same elite that had long dominated Brazil. Such rule was maintained with repression but relatively little mass confrontation, consistent with the general pattern of Brazil’s historical avoidance of conflict.⁷³

Economic growth was built in part on the profits from paying black labor less than whites, as earlier in South Africa and the United States. But because this wage differential was not enforced by law it did not provoke major racial conflict, which might have disrupted growth. Recent analysis reveals pervasive economic inequalities according to race, for instance with black mean income in 1960 less than half of whites, and with mulattoes much closer to the lower black level. In the relatively developed area of Rio de Janeiro, the 1960 average monthly income for blacks was Cr$5,440, for mulattoes Cr$6,492, and almost double that, Cr$11,601, for whites.⁷⁴ A leading Afro-Brazilian activist, João Jorge Santos Rodrigues, argues, with some exaggeration, that “Brazil has no black middle class.”⁷⁵ Such racial inequality was widely attributed to historical legacies and informal discrimination. Inequality purportedly reflected the lower skills of blacks as a class rather than as an officially categorized race.⁷⁶ For instance, 53 percent of blacks were illiterate in 1950, compared with 26 percent of whites; educational attainment was correspondingly unequal and occupations were distributed accordingly.⁷⁷ The disproportionate number of blacks living in the favela slums, where education and job opportunities have been limited, reinforced this inequality.⁷⁸ While dominant white economic interests have been served by depressed black wages, no official racial order was constructed to enforce these interests. Instead, class differences were projected to camouflage racism.

Economic growth under military rule came with increased economic competition and class tensions, and fed growing state centralization, but still without resort to official racial domination. Instead, the ideology of racial democracy was further reinforced as a component of national unity and as a means of avoiding racial conflict.⁷⁹ Candombé was sanctioned and even participated in by military officials.⁸⁰ But such symbolic gestures were no longer as effective at winning support in light of the declining fortunes of poor blacks. As a result, the junta preserved the image of racial democracy not so much by positive actions but with repression against its critics. In 1969, the National Security Council outlawed studies of racial discrimination as subversive. In 1970 race categories were again omitted from the census. Scholars studying race (including the later president, Fernando Henrique Cardoso) were exiled or at least forced to retire from universities. Even a television debate on race was canceled by the authorities.⁸¹ The very concept of race was effectively banned as an enemy of national security.

The irony of the authoritarian regime using repression to reinforce the image of racial democracy and curtail criticism of it was lost on the generals. But it suggests that the image of racial democracy was more valued for its use in bolstering national unity than as a commitment to real efforts at equal representation and avoiding discrimination. The scholarship and statistical studies squashed by the regime would have demonstrated this contradiction and undermined the projected image. Instead, authoritarianism had the effect of reinforcing white rule under the cover of racial tolerance.

If the persistence of elite power and national unity under military rule is notable, then the same persistence of the social order during the transition toward a federal democracy is even more remarkable.⁸² The junta itself began the process of liberalization and by continuing to control this process was able to ensure protection of the military’s own interests in the transition. Military representation was secured in the first post-junta cabinet of 1988.⁸³ The transition process itself was long and cautiously pursued by increments of gradual reform elaborated through “conversations among gentlemen and generals.” The result was an elite pact, much as analysts have described the negotiated transition in South Africa after 1990.⁸⁴ And also as in South Africa, this negotiated transition was made possible by, and effectively ensured, the maintenance of economic privilege.⁸⁵ But unlike South Africa, the Brazilian military was thus able to negotiate a transition with a minimum of violence.⁸⁶ Military rule and some central authority were abandoned, but elite power and wealth were preserved. A proliferation of parties further diluted opposition.

This brief description of the Brazilian transition toward democracy in the 1980s runs the risk of understating the extent of popular mobilization and conflict that was evident, albeit contained. Pressure for an end to authoritarian rule emerged via massive strikes in 1978 and 1979. By then numbering more than ten million, these unionized workers were angered by economic decline. Civic organizations, women’s groups, churches, and Afro-Brazilian associations further added to this pressure throughout the protracted transition.⁸⁷ However, such social mobilization generally followed at least some initial state reforms and remained relatively divided in the face of elite efforts to manage the process from above.⁸⁸ More fundamentally, analysts have consistently suggested that civil society mobilization remained isolated and weaker than the Brazilian centralized state.⁸⁹ The legacies of “clientalism, co-optation, and corporatism” could not preclude all social mobilization, but the Brazilian state had the strength to contain such protest sufficiently to avoid disruption of an elite-led process of transition.⁹⁰

It is particularly striking that Brazil did not suffer more conflict around issues of race amid growing economic competition and continued inequality. Earlier comparisons with South Africa and the United States suggested that Brazil’s relatively slow growth and the persistence of a precapitalist social hierarchy constrained potential black-white competition that otherwise might have provoked more explicit racial domination and conflict.⁹¹ But Brazil did not remain the economic laggard implied by such analysis.⁹² Even in the Southeast, where industrialization and immigration produced high levels of competition by the 1930s, legal discrimination was not used, though “social practices were clearly discriminatory.”⁹³ More recently, the proportion of the population engaged in manufacturing has doubled, rising from 12.9 percent in 1960 to 24.4 percent in 1980. At the same time, GDP per capita increased by 60 percent, while income inequality and competition grew markedly.⁹⁴ Still, expectations that economic development would replace the old hierarchy with competition-inspired formal racial segregation and conflict have not proven correct. “During the twentieth century Brazil has moved from its old paternalistic type of race relations toward the competitive model,”⁹⁵ yet Brazil has not imposed official racial domination. Even if a Jim Crow equivalent would have been difficult to impose in the 1930s or during later growth, some movement in this direction would have been possible. But the post-abolition legacy of tolerance was maintained. White capital and white labor advanced without formal segregation or major racial conflict. As a result, social hierarchy and order were preserved.

Economic (and social) discrimination is pervasive in Brazil, but without formal racial domination. The absence of legal segregation was connected to the absence of any major early intrawhite conflict that might have required rapprochement via white unity. The Brazilian state was less challenged by regional or ethnic conflict than South Africa and the United States. Even Brazilian trade union activism was long contained by corporatism.⁹⁶ Indeed, the relative lack of regional and ethnic conflict, which elsewhere was superimposed with class, reduced overall antagonism. And racial democracy, like corporatism more generally, conveyed an image of everyone’s interests being served. Employers faced no formal color bar that might impede their flexible use of the cheapest available labor. White workers benefited from relatively higher wages. Black workers who saw no official racial impediment could believe that they might advance, even though they generally did not. A flexible and inclusive racial order purportedly served the interests of all and therefore encouraged unity. Until recently, all Brazilian sectors have avoided the issue of race. Racial democracy thus helped unite the nation, forging a class compromise as in South Africa and the United States, but a compromise that did not take the form of racial domination, which elsewhere provoked conflict.

What emerges most starkly from this historical overview is a picture of the resilient Brazilian state containing and avoiding conflict. This state has consistently been able to maintain national unity of a sort and to protect the interests of a well-established elite and hierarchy. The forms of rule have varied significantly, from colony, empire, republic, and military dictatorship to more federal democracy. But these transitions, all managed from above, produced remarkably little change in the fundamental social order. The “Order and Progress” inscribed on the Brazilian flag have been maintained. Social movements and conflicts have emerged, but have been largely tamed by absorption into the social and institutional framework dominated by the state and elites. The state remained the expression of a remarkably unified nation, using its power to shape national loyalty and obedience. Brazil emerges as a quintessential example of hierarchical rule and state-led nationalism.

As a significant portion of the population, Afro-Brazilians were inescapably absorbed into the impressive project of Brazilian nation building orchestrated by the state. After abolition, the Brazilian elite was concerned about the prospect of racial conflict, which might have disrupted national unity and state building. To avoid such conflict, the elite consciously avoided imposing antagonistic racial exclusion. It instead gradually refined an image of racial democracy as a fundamental component of Brazilian nationalism, consistent with cultural assumptions of inclusion. Blacks were embraced, at least rhetorically, to avoid division. African culture was celebrated, though more substantive political and economic rights did not follow. Laws were enacted penalizing color discrimination in public places, not as an acknowledgment that there was such discrimination but ostensibly to prevent it.⁹⁷ Constitutional assurances that “racism constitutes a crime” were later enacted, though no enabling laws were passed and few prosecutions of such a crime were pursued.⁹⁸ The state proclaimed itself the representative of all Brazilians, though few blacks were employed as civil servants.⁹⁹ Many blacks were long deprived of the franchise, but on the basis of illiteracy, and not race per se. The murder of street children and police brutality suggests significant violence against blacks continuing with impunity, but not as official policy.

In official terms, the rule was racial democracy. In practice, the state was often far from democratic in general terms. Laws were selectively enforced. Policies toward blacks were more notable for the lack of legal discrimination than for any positive assertions of black political, economic, or social rights.

The two themes of nation-state unity and racial democracy are interconnected. National unity was proclaimed by the Brazilian elite to contain and avoid disruption to its continued privilege and its efforts at state building. Racial democracy was projected as a central component of that national unity, in order to avoid the sort of race conflict Brazilians had experienced with the slave revolts and watched in dismay unfold in the United States. Having accomplished abolition without a conflict comparable to the Civil War, the elite was eager to maintain the benefits of unity by projecting an image of inclusiveness. And as no regional or ethnic conflict comparable to that in the United States or South Africa had impelled the need to unify whites through racial exclusion, this strategy was consciously rejected. This relative absence of violent intrawhite struggle and the persistence of a unified nation-state made it possible for the upper classes to rule without legally racializing their hegemony. Instead, the image of unity and inclusiveness preserved and protected the social order, more than it was an outcome of that order. Afro-Brazilians did not benefit from this ideological project. As we shall see, they were largely deflected from protest against social discrimination. The lack of any official rules for such discrimination deprived them of a target for mobilization that might have forced redress of inequality. The elite did benefit; it preserved its place at the top of the hierarchy.

Racial democracy was an ingenious form of social engineering, largely through omission rather than commission. The Brazilian state ruled without official racial domination. White interests were protected. To reinforce the social order, racially distinct legal treatments were not encoded and racial conflict was largely avoided. The Brazilian nation-state emerged relatively unscathed from a century in which racial conflict tore at the fabric of the United States and South Africa.

Comparative Racial Domination

AN OVERVIEW

European descendants in South Africa, the United States, and Brazil established early legacies of racial discrimination against African descendants, despite dramatic differences in population mix and mixing. Slavery in all three countries reinforced pervasive images of inequality. By the time slavery ended, such discrimination was often buttressed by pseudo-science and encoded in varying forms of legislation, evident even in Brazil, where later interpretations of Portuguese rule, Catholicism, slavery, and miscegenation were purposefully shaped to deny this inheritance and to project an image of racial tolerance. But Brazil’s self-image as an exception to the discourse, culture, and practice of racial discrimination was more creative than historical.

Historical legacies of discrimination were not automatically or thoughtlessly brought forward into the modern era. All three countries came to junctures at which the preexisting social order and relations between blacks and whites had to be consciously reconfigured. Of course, the resolution of race relations would not be fully decided at any single moment, and was instead elaborated over decades of conflict, political and economic competition, and varying policies. But the establishment of the post-abolition, formally unified nation-state in all three countries did crystallize the issue of race relations, forcing decisions that set the course of those relations for generations.

In South Africa and the United States, the conflict-ridden process of nation-state consolidation set the terms of official racial domination. Ethnic conflict between Afrikaners and English-speakers in what became South Africa, and regional conflict between South and North in the United States, came to a head in the dramatic violence of the Boer and Civil wars. In both, the relatively liberal industrializing power won, only to face the challenge of how to reconcile with its former adversaries, incorporating them as productive citizens rather than allowing their separate and destabilizing self-determination. Further conflict had to be contained to avoid instability undermining the primary fruit of victory, which was control over the newly unified or reunified state as an organ of economic development. In South Africa, the British concluded that abandoning commitments to enfranchising and advancing blacks was the surest way to appease the Afrikaners and to consolidate union – deletion of a commitment to the native franchise in the 1902 peace treaty symbolized this deal. Continued Afrikaner nationalism reinforced English efforts at appeasement, leading finally to Afrikaner rule and apartheid. In the United States, the North was gradually forced to a similar conclusion about how to appease its former adversary. The North’s efforts to impose reforms under Reconstruction brought sufficient Southern antagonism to persuade the North to abandon the effort. Sectional tensions were then diminished by allowing the South to impose Jim Crow, with blacks again paying the price of white reconciliation.

For all their dramatic differences, these patterns linking nation-state consolidation and formal racial domination were similar in the United States and South Africa. In both, the nation-state was divided and policies of racial domination were designed to diminish such division. Earlier racial discrimination and the racist ideology developed to defend it were encoded or allowed by state action to reinforce institutional power and white loyalty to it. Legal exclusion of blacks helped to gradually unify the core constituency of whites.

Brazil was “exceptional,” not in its lack of racial discrimination, but in failing to establish official racial domination. Brazil suffered nothing comparable to the Boer or Civil War, and was not as riven by overarching antagonisms. Its colonial legacy of a strong state and established hierarchy produced less pressure for regional and none for ethnic reconciliation encouraged elsewhere by white racist ideology and black exclusion. With less pressure for intrawhite reconciliation, no racial domination was constructed. Instead, the Brazilian elite feared regional division and conflict with blacks, present in large numbers and previously prone to revolt. To head off this threat, the elite purposefully projected a form of nationalism that included blacks under an official racial democracy, despite ongoing discrimination. The established hierarchy remained in place.

This schematic description of the different challenges to nation-state consolidation and the resulting official racial domination or racial democracy is artificially neat. Such analysis projects fixed subjects – Afrikaners, British, North and South – whose conscious unity was not simply given or ascribed. Each was divided internally by political and economic differences. But the violent process of nation-state building in South Africa and the United States solidified antagonistic ethnic and regional identities. And as these conflicting identities were reinforced, their competing interests repeatedly had to be reconciled to maintain the stability needed for further nation-state building. Continued tension exacerbated antagonism, solidifying the units to be reconciled in repeated iterations and official responses to the social conflicts in which states were embedded. It was precisely because the nation-states were not united that policies to encourage such consolidation were enacted and refined. Specific policies varied according to the individual characteristics of this messy, ongoing dynamic. National unity in Brazil was also achieved in an ongoing process, but was not riven by comparable levels of violent conflict. The predominant solidarity was a unified nation, with conflicts over the form of rule but with each regime encouraging dependence on the state.

It must be acknowledged that identifying the historical process of identity constructions and efforts at reconciliation does not do justice to the complexity of the comparative experience. For instance, the post–Boer War South African state imposed segregation and exclusion relatively quickly and centrally, while such policies in the post–Civil War United States were imposed after a lag, locally and unevenly. In large part, this difference was due to varying state structures, the form and level of central state power, and types of challenges to it.

South Africa quickly became a strong state, backed by the British Empire and by agreement among English-speakers and Afrikaners regarding the benefits of centralized power. These ethnic groups, solidified within themselves by their prior conflict, continued to compete for control of the state. But they did not dispute the power of or the need for the central prize over which they wrestled. Nor did they disagree over the need to maintain and reinforce centrally imposed racial domination. The perceived threat from the African majority impelled such a consensus, as compared with the lesser threat posed by the African-American minority.

The contrasts with the United States are marked. The Civil War did not produce a consensus for centralized power, but instead reinforced the dispute with advocates of states’ rights. The balance of federal power had been and remained the central issue of political competition. After the North’s imposition of Reconstruction intensified continuing sectional animosity, racial domination was imposed locally. The weaker central authority in the United States gradually and grudgingly ceded the issue of black rights to the states, restoring the balance of federal authority and states’ rights as a compromise necessary for national unity. The weakness of central power forced its own solution of federal withdrawal. While the North, like Brazil, discriminated informally rather than imposing legal Jim Crow, it was a party to and allowed for Jim Crow in the South.

Brazil again differs, with central state authority preserved until the republic, and a balance of central and regional power established thereafter. Political competition remained largely under the control of an elite, vying over whether the state should be democratic or dictatorial. The power of that state was not as disputed as in the United States and the lack of official racial domination was countrywide.

In all three cases, race making was tied to the imperative of nation-state building and challenges to it. Elites sought consolidation of state institutions and popular loyalty to them to ensure stability and development, using race selectively as a means to that end. Having fought in part – or so they alleged – for blacks’ rights and won, both the Northerners and the South African English abandoned this commitment in order to pursue in peace the more pressing goal of intra-white reconciliation. Blacks were expendable, useful as scapegoats for core white unity, a choice that crucially coincided with existing prejudice and racist ideology. The Brazilian state and social hierarchy faced no challenges comparable to those posed by the Afrikaners or South, and so nation-state consolidation could proceed without racial domination but with informal discrimination.

The role of war in nation-state building was also shaped by and in turn shaped racial dynamics. Earlier in Europe, state consolidation had been achieved through warfare. And as formative states had united and called upon the populace to fight, they had extended citizenship rights accordingly. In the United States, South Africa, and Brazil, war making again led to extensions of the franchise, albeit to whites only, though blacks had expected that their participation in the military would also be rewarded by better treatment and inclusion. But after helping the victorious British in the Boer War, the North in the Civil War, and Brazil against the Dutch in 1630–54, blacks in each case subsequently returned to subordination. By contrast, in later external wars, blacks fighting for Brazil against Paraguay in the 1860s did help bring abolition, and blacks fighting for the United States in the Second World War helped inspire some reforms of segregation. Thus, while external war generally reinforces national loyalty and encourages inclusion, internal war more often exacerbates subnational loyalties that may be healed through specific exclusion.

Democracy itself played a role in these processes, though this form of rule did not produce the same policies in different contexts. Selective inclusion in democracy continued to reinforce and reflect racial orders. Racially specified democracy in South Africa and the United States enabled white voters to ensure their dominance over outvoted or nonvoting blacks. Party competition in South Africa from 1910 to 1948 preserved the polity while leading to further black disenfranchisement and segregation to bolster white unity. Post-Reconstruction party competition in the United States had similar results. In the South, even some whites were disenfranchised to preserve state and local Democratic Party power and the image of white unity. Democrats and Republicans then vied for national power, again preserving the polity amid deepening racial domination. Limited democracy in republican Brazil and from 1945 to 1964 did not produce the same outcomes: blacks were disenfranchised by subterfuge and no official racial domination was enacted. Thus, democracy served as a kind of transmission belt for reinforcing the distinct directions of racial policy.

And what was the role of economic forces in shaping divergent legal racial orders? Earlier comparative studies argued that in the United States and South Africa, industrial expansion brought competition for jobs between blacks and whites, with whites then acting to protect themselves with legal color bars.¹ By contrast, goes this explanation, Brazil’s relatively slower economic development brought little competition between blacks and whites, even amid European immigration, so that whites perceived less need to impose a harsh racial order to protect themselves. Paternalistic race relations were preserved.

There are numerous problems with such an economistic account. Racial domination was evident before industrialization and heightened black–white competition. And the legacy of slavery and past discrimination itself contained the threat whites felt from black advancement. Though competition was often stronger between whites, particularly with rising immigration, blacks became the primary target of discrimination.² Regional and demographic differences also intruded. Both the more industrial regions of the U.S. North and Brazilian Southeast had fewer blacks than the less developed regions in which blacks were more concentrated, the U.S. South and Brazilian Northeast. But amid informal discrimination, rising industrial development and competition did not produce fully fledged Jim Crow or apartheid anywhere in Brazil or in the U.S. North, even when blacks later migrated there. In the U.S. South, as late as 1930 more than 40 percent of black males were employed in agriculture.³ Yet it was in the less industrialized South that Jim Crow came of age, with the acquiescence of the North. And Jim Crow was most strictly enforced in the deep South, as in Mississippi, where industrial competition was at its lowest. If industrial competition causes segregation, then Jim Crow should have been born a Yankee in the early twentieth century, not a rural Southerner in the late nineteenth.

Racial domination was perhaps then impelled by capital’s efforts to maintain a split labor market, reducing the cost of labor. Dominant economic interests benefited from rules ensuring a lower wage scale for blacks and from the racial division of workers, who then found it more difficult to unify in demanding higher wages. According to this argument, business reinforced racial antagonism, pressed for more rigid forms of racial domination, and reaped the fruits of official segregation.⁴ Florestan Fernandes and others in the São Paulo School in Brazil similarly argued that capital there discriminated against blacks to reduce wages, but without official edict.⁵

Explaining racial domination as a tool of capitalist exploitation over labor recognizes the central role of intrawhite class competition, but such analysis falsely assumes that capital alone determines state policy. Not only were capital’s demands often rebuffed, but capital itself was often divided by sector. For instance in South Africa, agrarian capital consistently resented the efforts of mining interests to pull labor off the farms and into the pool of cheap, forced migrant labor.⁶ And English manufacturers’ calls for reform of the color bar as an impediment to economic growth were resisted by Afrikaner employers. While Marxists pointed to the mining sector and Afrikaner capital to blame business for apartheid and liberals hailed manufacturers’ calls for reform, neither argument about the role of capital could be generalized, for capital remained divided.⁷ In the United States, business was uncertain of its interests and also divided, often according to region. Businessmen in the North advocated concessions to the South to avoid the instability of Civil War but then later attacked President Andrew Johnson’s accommodationist policies.⁸ Northern capital was wary of the instability provoked by Congressional Reconstruction, but these concerns were diluted by expectations of profit from military rule of the South, and were subordinated to ideology and party political imperatives for reform until Reconstruction was abandoned to restore stability. Primary and agricultural producers did benefit from cheap black labor. But capital’s interests were not consistently served, for the exclusion of blacks was often and in many ways costly, particularly for manufacturers. Fractions of capital were forced to “adjust to the requirements of a racially stratified labor force” ensuring higher wages for whites.⁹

A related argument suggests that white labor enforced segregation to ensure its advantage over black labor. Certainly in South Africa, and arguably in the United States, white workers used their popular strength and votes (more numerous than white business) to deny blacks equal justice. They forged a racially bounded working class and supported parties that were committed to preserving whites’ economic position above their black competitors.¹⁰ As C. Vann Woodward concluded, Jim Crow was enforced by the Democratic Party “to please the crackers,”¹¹ much as South African segregation can be described as having been enforced by the National Party to serve poor Afrikaners.

Reducing racial domination to specific white working-class interests overstates the power of such pressure. White workers rarely determined state policy on their own; nor were their demands always consistent, for as more sophisticated Marxists acknowledge, working-class solidarity cannot be assumed as given.¹² For instance, U.S. white labor was divided over Jim Crow by region. And even within each region, labor’s interests and advocacy varied. Many Southern white workers supported populism, while others supported segregation. Even Northern unions shifted in their policies toward blacks. White labor also suffered under Jim Crow, with segregation dividing the working class as a whole, making it possible for capital to pay lower wages and to use racially divided workers against each other. White workers were often willing to pay the price of racial domination by accepting division by race and a lower rate of economic growth.¹³ The fact that there were costs to such policies suggests that narrow working-class interests did not consistently produce racial policies.

Official racial orders were shaped by, but cannot then be reduced to, specific class interests. Capital and white labor were often in conflict, for instance in South Africa’s 1922 Rand revolt. Indeed, had racial domination been determined by one class’s interests, then class conflict and related party competition would have disrupted the racial order. No interests were sufficiently dominant or uniform to produce the persistent outcome of racial domination. And classes were often internally divided into ethnic, regional, or sectoral factions with distinct interests and predilections and party representatives, undermining a strict class dichotomy. That whites were united neither within nor between classes argues against any effort to reduce racial domination to such a narrow interpretation.

Comparative analysis then suggests some refinement of the conclusion that “racial domination . . . is essentially a class phenomenon.”¹⁴ Class interests were significant, but this alone does not tell us very much. The interrelation between racial order and class interests is more complex than can be described by reducing the former to the latter. Instead, racial domination has been constructed with or without serving distinct economic interests, and not because of them. And racial orders varied over time, mixing informal economic discrimination, political exclusion, and legal segregation.

Specific class interests were not determinant, but economic interests were relevant to the construction of racial domination. What was determinant was an overriding white interest in peace and stability as essential for economic development. Growth could proceed only if competition did not provoke open conflict, thereby threatening the market. Karl Polanyi has argued that this overriding economic interest in avoiding conflict accounts for “the long peace” engineered by the great powers during the nineteenth-century economic and trade expansion.¹⁵ A similar process took place within countries, where growth depended on avoiding domestic unrest as well. Much as the balance of power provided for peace among the major European powers, consolidation of the nation-state within countries was impelled by the necessity of avoiding internal conflict that otherwise would have disrupted the economy, if not destroyed it. In South Africa and the United States, racial domination was formally constructed and reinforced to avoid such further disruptive conflict between regions, ethnic groups, or classes. Such policies did constrain economic development, but less than open violent conflict would have. In Brazil, racial democracy served not so much to heal previous conflict as to avoid its further emergence. The imperative of stability as a precondition for growth provided an overarching economic incentive for conflict control or avoidance through either racial domination or racial democracy.

Specific racial orders were constructed and refined by a process of class compromise rather than the serving of one class. As described by political scientist Adam Przeworski, for such a compromise to become hegemonic in establishing an explicit social order, a minimum of objective conditions of development must have been reached.¹⁶ With such development, the pressure to satisfy and coordinate competing interests grows, for instance with capital demanding a minimum of profits.¹⁷ The way in which such compromise is achieved is uncertain and variable, rather than implying a static solution, much as racial orders have been reinforced and reformed. The specific terms of compromise vary according to shifting pressures. According to Gramsci, such compromise repeatedly emerges

as a continuous process of formation and superseding of unstable equilibria (on the juridical plane) between the interests of the fundamental group and those of subordinate groups – equilibria in which the interests of the dominant group prevail, but only up to a certain point, i.e. stopping short of narrowly corporate economic interests.¹⁸

The pressure for continual compromise remains, particularly where the alternative is open conflict. “The decision to compromise depends, in the end, on a comparison of the best compromise that can be obtained with the consequence of no compromise.”¹⁹ In South Africa and the United States, racial domination emerged out of such a class compromise to avoid further conflict. The Rand revolt was not repeated and U.S. Southern populism was contained.

This argument helps to combine political and economic analyses of racial domination. Racial orderings are constructed to diminish conflict and to allow for economic growth, in a mutually reinforcing way. Racial domination and its refinement diminished overlapping and ongoing intrawhite regional or ethnic and class conflict dividing the nation and threatening growth. Indeed, race was a long-standing “central ideological underpinning” of each of the developing nation-states discussed here, though subsequent social orders differed.²⁰ Each such construct was shaped by economic interests, for ideology or state action and policy do not emerge in a vacuum. And such particular interests and their impact on policy vary. But the overriding interest served by racial orders was not class-specific, but more general: to preserve the overall state and its economy, in which industrialization and related social changes could and did proceed.

This argument fits with a more general analysis of the role of the state in constructing class compromise. Development heightens class conflict at the same time that it provides resources to build state capacity. For such conflict to be diminished, for the polity to be preserved and consolidated, the state must act.²¹ The policies of the state then make explicit the terms of the class compromise it engineered, in historically specific forms.²² The varying institutions of segregation in South Africa and the United States and the ideology of racial democracy in Brazil were constructed in a political process. Conflict remained, forcing variations in such outcomes and underlining the difficulties of achieving stability. Economic interests and related demands had to be tamed and met. When this was not sufficient, state force was used, in wars or to put down revolts or strikes.²³

The state was the instrument for conflict management, using racial order as a tool. Economic development exacerbated the potential for conflict among whites, which the state then had to manage to ensure revenues. And with such stability, growth provided the state with resources to so manage society. Capital, white labor, and fractions thereof all lent

legitimacy to the racial order and race domination. Each calls on the state to take control of the subordinate worker, to draw racial lines. . . . The well developed racial state, very much the product of capitalist development and the period of intensification, does not wither away with its feckless business, farm and perhaps trade-union supporters.²⁴

In other words, state-imposed racial domination helped to prevent class warfare, meeting a general interest in preserving a functioning economy.²⁵ A class reductionist argument, for instance in assuming the policy determination of capitalists, ignores both that states do make policy and that such policy often contradicts dominant class interests.²⁶ The actual process of state action again differed according to state structures.

The state’s own interest was not autonomous but shaped by varying popular interests that had to be satisfied enough to preserve peace and maintain growth. States themselves were not unitary. Rather than being insulated from class and related interests, “the state itself was penetrated and structured by the very interests whose conflicts it seeks to stabilize and control.”²⁷ In South Africa, capital and white workers were served by racial domination, which helped to contain intra- and interwhite class and ethnic conflict. With a more porous state in the United States, capital and white workers were similarly served by segregation, taking different forms in North and South. In Brazil, racial democracy and corporatism more generally long contained significant class conflict, not as aggravated by regional or any ethnic tensions. Solutions to conflict via differing racial orders were thus determined by overlapping political and economic challenges needing to be addressed. That the state could provide such solutions served as well to reinforce its own power. But such outcomes that appeared functional were refined in response to varying pressures. Particularly where tensions between ethnic groups or regions reinforced class divisions, conflicts were fitfully addressed by purposeful and varying state action and policies of racial domination aimed at preserving the nation-state and its economy. Despite economic costs of segregation, growth could and did proceed, bringing structural changes that would contribute to further refinements of the racial order.

Ongoing party and economic tensions thus kept alive and reengaged the same dynamic that determined racial order, providing an intervening explanation for refinements of that order. Under political and economic pressures, state elites continually remade racial domination or racial democracy. In South Africa and the United States, state action convinced whites in different parties, classes and class factions that their interests were being met. In Brazil, some blacks even believed they were being included. In all three cases, the nation-state and its economy were thus preserved by purposeful state action and responses to social pressures.

This analysis suggests not only the process that led to the elaboration of racial domination or racial democracy, but also part of the complex process by which apartheid or Jim Crow later came to be abandoned. In South Africa, Afrikaner Nationalists came to power first under Hertzog and again after 1948 under the National Party. They used their control of the strong state to bolster Afrikaner cultural, social, and economic standing until greater parity with the English was achieved. With ethnic and class antagonism diminished, the strategic imperative for racial domination to encourage white unity faded. Nation-state consolidation, at least for whites, had been largely accomplished. Similarly in the United States, localized racial domination via Jim Crow appeased the South, which over the course of a century was reincorporated into the nation-state and its economy. With diminished sectional and class antagonism, white unity no longer needed to be reinforced by racial domination. In this sense, in both countries racial domination served the strategic purpose for which I argue it had been enacted.

Institutionalized racial domination did not end simply because it was no longer needed; practices were not so functionally enforced or abandoned. As institutionalized reflections of racist ideology, apartheid and Jim Crow had a life of their own, beyond their strategic purpose to encourage white unity. As I will discuss in Part Three, such racial domination also had the unintended consequence of provoking greater mobilization, agency, and protest by subordinated blacks, bringing a new and increasing threat to the nation-state that could only be defused by ending official racial domination. The nation-state in both South Africa and the United States had been strongly enough consolidated to abandon its racial crutch. A more muted form of black protest emerged in Brazil, where there was no official racial domination to challenge or demolish.

Racial orders as class compromise also provided the conditions for nation-state preservation and for growth to proceed. But unforeseen consequences would gradually undermine racial orders and impel new forms of class and political compromise. Growth brought significant changes in systems of production and society, with industrialization requiring more skilled and urbanized labor, and consumption by blacks. Migration expanded to areas of opportunity. These changes were made possible by racial orders, though the costs of segregation also grew. Economic development also consolidated and empowered blacks, setting the conditions for blacks to challenge established patterns of racial domination.

The significant implication of this analysis is its emphasis on the political determinance of racial domination. Earlier historical legacies, discourse, and culture were important, but their later influence was not direct but rather depended upon subsequent interpretations. Demographic factors were also important, but again how such physical differences were interpreted in varying forms over time was shaped by political processes. The imperative for nation-state consolidation produced explicit and varying forms of official domination where ethnic or regional and class combatants sought resolution of their conflict by projecting white unity and black exclusion. Where nation-state consolidation faced fewer such obstacles of regional or ethnic conflict, Brazilian ruling classes did not encode racial domination to unify whites. They instead projected racial democracy to unify an increasingly mixed populace, heading off internal conflict. Racial domination was constructed and imposed (or not) according to these contexts, framed by the goal of nation-state consolidation, until black demands and contestation grew to challenge this consolidation.