1
W. E. B. Du Bois called “the problem of the color line” the pivotal question for the twentieth century.1 In South Africa, the United States, and elsewhere, the century has brought both reinforcements of the color line and militant struggles against it. Momentous conflict has confounded expectations that racial domination and mobilization, supposedly archaic residues of another age, would simply wither away. Far from displacing race as an issue, industrialization, class conflict, rising nationalism, and state consolidation have actually spurred racially defined contention. As a result, struggle centers increasingly on state action or inaction. The end of Du Bois’s fateful century challenges us to rethink the links between racial conflict and nation-state building and development.
The initial driving force of this study is a major puzzle of widespread concern: What accounts for twentieth-century race relations? This question can be broken down into more manageable parts, though each impinges on the others. How, and to what extent, were such relations prefigured by earlier history? How did political and economic processes during this century shape divergent racial orders? And how did challenges from below emerge, interact with, and force changes in racial domination from above? My approach to these issues is comparative, rejecting U. S. exceptionalism without diminishing the distinctiveness of that or other cases.
Discrimination according to race predates this century and became pervasive where Europeans sought to justify their rule. But modern, official policy did not always build upon prejudice to encode racial domination, or did so in varying forms of exclusion from citizenship rights. Similar early practices of discrimination did not consistently produce the same legal racial orders and outcomes. And whether the state was so deployed to enforce racial domination shaped all other aspects of race relations. Where authorities legally reinforced discrimination, the unintended result was major social conflict between those who ruled via exclusion and those demanding inclusion in the polity. In the absence of official racial domination, such conflict was muted.
States made race: amid pervasive discrimination, official actions enforced racial distinctions or did not, with profound consequences. This state-centered argument builds upon and incorporates previous explanations of modern race relations. But to argue simply that states can and do impose official racial domination – that states enact such state policies – is tautological. The key is to explain why states so act. Specific state policy is historically embedded, reflects ideology, is constrained by dominant political and economic claims, and responds to protests. Under what circumstances do each of these particular factors become salient, and with what results?
Political and economic elites have consistently been eager to ensure the stability of their societies by building institutions of coercion and coordination. By consolidating state power within varying structures, they sought to preserve their power, to bolster their legitimacy, and to provide the conditions for economic growth. But such stability was often undermined by major violent conflict, such as the British–Afrikaner ethnic conflict in South Africa and the North–South regional conflict in the United States, dividing prominent loyalties. To diminish these conflicts, elites acted strongly to strike bargains, selling out blacks and reinforcing prior racial distinctions and ideology in order to unify whites. The state instantiated “white nationalism,” with the torque of this enforced racial identity proving powerful enough to integrate populations otherwise at war and engaged in ongoing competition. “To bind up the nation’s wounds” among whites, blacks were bound down, and the wound of race was left to fester. Polities were shaped by this dynamic, with varying forms of official racial domination in South Africa and the American South persisting until black opposition replaced intrawhite conflict as the most pressing challenge to the nation-state. Brazil, by contrast, with relatively little intrawhite conflict, was less pressured to reconcile whites via racial domination, and the state instead embraced the ideology of “racial democracy,” even amid continued discrimination.
Analysis of race making may then tell us something more generalizable about the processes and effects of nation-state building through either exclusion or inclusion. Not only have such institutional rules consolidated particular social cleavages, but manipulation of cleavages such as race or ethnicity has also shaped how dominant institutions and loyalty to them were built. Selective exclusion was not tangential to nation-state building, as liberals argue, but was instead central to how social order was maintained.2 If constructions of race and of nation-states have been so linked as prominent developments of this century, focusing on the former will also reveal fundamental aspects of nation-state building not otherwise apparent.
Bounding the State and Institutionalizing Identities
Racial domination rests on a distinction of peoples according to categories of physical difference. Until at least the mid twentieth century, people of European ancestry generally assumed that such racial demarcations reflected natural white superiority. The color line was seen as having been drawn by God or biology. Slavery, proscriptions against miscegenation, colonialism, imperialism, manifest destiny, racially exclusive forms of citizenship or nationalism, and exploitation were all justified by whites as preordained in nature.3 Such racism became “a mode of thought endemic in Western civilization,”4 buttressed by eugenics, social Darwinism, and explicitly racist theories such as those of Count Gobineau.5 Primordialism serving the interests of whites made the domination of darker peoples seem inescapable.
Assumptions of natural racial distinctions, as notably enforced by the Nazis, have been refuted. After the Second World War, the scientific community reached a consensus that race had no scientific basis.6 Categorical distinctions of race were recognized to be artificial and inconsistent with the continuum of skin color evident in the world’s increasingly mixed population. If biology did not draw the color line, then race had no basis in fact. Marx and Engels had earlier argued that “all fixed, fast-frozen relations, with their train of ancient and venerable prejudices and opinions, are swept away” by class conflict.7 Following Weber and Durkheim, liberals also assumed that as mere myth such a distinction would fade with development or assimilation.8 They were wrong.
Vicious discrimination has remained, even if particular forms of racial domination varied. “Modernity has not sealed off the pathway to this bottomless depth of sorrow and chaos,” but rather reinforced aspersions of human difference.9 Describing race as having no basis in science did not dispel its power as myth. Prejudice has continued to serve as a criterion of stratification, embedded in economic relations, political institutions, and ideology.10 If the basis of such distinctions was not naturally ascriptive, then it had to be socially constructed. But seeing race as situational only begins the analysis of why and how such lines are drawn. Specifying where this process is centered can help in understanding it.
Domination has been officially encoded in racial terms, suggesting that the state plays a role in constructing and enforcing the institutional boundaries of race. And as social movements have developed against racial domination, such protest reflects the terms of that state rule. The state sets the focal point for analysis, providing clues about the interconnected processes of race making from above and below. Following Weber, if public policy by the state makes explicit the terms of domination or resistance,11 then it should be possible to so specify these processes. And briefly adopting the shorthand of analyzing the state focuses the question of what states do, before turning to the question of why.12
States are compulsory and continuous associations claiming control of society within a territory. They differ over time and place in their capacity to rule, in their autonomous power, and in their particular forms. But all states seek to contain challenges or instability threatening order and growth. To do so, ruling elites structure relations between states and civil society, shaping norms and reinforcing social and political identities accordingly.13 The central identity encouraged by states is that of the nation, defined as the popular loyalty of a population held together in being obliged to serve and be served by the state.14 By encouraging allegiance to the nation, states enhance their claimed monopoly of legitimacy.15 Accordingly, state elites and others have been inspired by the goal of the nation-state, a coincidence of institutional rule and allegiance to it that would diminish internal conflict. As John Stuart Mill argued, “the boundaries of governments should coincide in the main with those of nationalities” or “fellow-feelings.”16
Benedict Anderson has suggested that the nation emerges as an “imagined community.”17 His argument discusses preconditions for the emergence of such nationalist solidarity. Economic development, literacy, mass communication, education, and urbanization make possible the emergence of such expansive loyalty, but this implies a more spontaneous and inclusive process of nation building than was often the case. Official boundaries purposefully defined and enforced who was imagined as part of the nation and who was not.18 Such imaginings were often not benign. Where states were formed before a nation emerged, the explicit efforts of the state to limit and encourage selective nationalism are particularly evident. Specified exclusion has provided a crucial referent demarcating those included.19 In such instances, group formation, identities, ideas, and “social categories . .. are shaped, manifested and entrenched through the state.”20 Inclusion then solidifies loyalty among those officially incorporated. While such nationalism may erode certain group loyalties by their inclusion, it also reinforces solidarity among others purposefully excluded.21
Citizenship is a key institutional mechanism for establishing boundaries of inclusion or exclusion in the nation-state.22 It selectively allocates distinct civil, political, and economic rights, reinforcing a sense of commonality and loyalty among those included.23 But by specifying to whom citizenship applies, states also define those outside the community of citizens, who then live within the state as objects of domination. Even in formal democracies, some are not included nor have their interests served. Such imposed exclusion inadvertently may serve as a unifying issue, mobilizing the excluded group to seek inclusion in the polity as a central popular aspiration.24 Gradual expansion of citizenship is then gained through protracted contestation.25 The goal of gaining citizenship rights, which were not originally universal, thus has often served as a frame for mobilization.26 Extending a provocative analogy, groups use their voice to overcome their enduring and forced exit from the polity.27
The process of defining the nation with rules of citizenship is of obvious relevance for how racial categories are established and reinforced. Coercive powers have been used to define citizenship according to race – states bind the nation they claim to represent by institutionalizing identities of racial inclusion and exclusion. The extension of citizenship rights has been blocked by constructing racial boundaries. Barbara Fields has suggested that such official racial domination began at a particular historical moment, when the modern nation-state emerged.28
States then play a central role in imposing the terms of official domination, with unintended consequences. Official exclusion, as by race, legitimates these categories as a form of social identity, building upon and reshaping historical and cultural solidarities. In the short run, such exclusion benefits those included and hurts others. But in the longer run, institutionalized exclusion may further consolidate subordinate identity and encourages self-interested mobilization and protest. State-sanctioned racial categories impose real costs on their subjects but also offer oppressed populations both legal grounds for redress and bases for political mobilization. The reverse reasoning would also hold. Where and when a state does not impose categorical exclusion, potential subordinates are not so directly disadvantaged. But in the longer run, the preferable absence of such exclusion also may deprive informally subordinated peoples of an unavoidable demarcation unifying them as a group. Ironically, the lack of official racial domination then becomes an obstacle to effective collective action forcing redress of real inequality. In either case, state rule and solidarity from below continually interact.
Arguing that states have the capacity to establish racial boundaries of citizenship, provoking resistance accordingly, does not explain why a state does or does not enforce a racial cleavage. Such an explanation must be cognizant that the state is itself not unitary, but reflects divisions of the society in which it is imbedded, and is itself often torn by competing global, national, and local pressures. For instance, does the state construct and use race because of international pressures, to preserve itself, to ensure revenues, to meet specific interests, to avoid class conflict, or to reflect popular racist beliefs? To proceed further, it is useful to turn to comparative historical analysis, looking both inside the state and outside to its social context. How autonomous was the state? How consequential were different state structures? The analytic category of “the state,” though suggestive, must be broken down to explain the motives for particular forms of nation building and for specific official policies and outcomes.
Comparing and Excavating Historical Foundations
South Africa, the United States, and Brazil are the three most prominent cases in which European settlers dominated indigenous populations or slaves of African origin. In all three of these major regional powers, modern social and economic indicators demonstrate significant disparities between black and white. The legacies of slavery and of conquest established a common pattern of discrimination.29 Indeed, there has been a rich tradition of comparison between two or three of these “big three” cases, the histories of which have often interacted.30
The common history of racial dynamics, so understood, makes the comparison of South Africa, the United States, and Brazil obvious. Their differences make comparison fruitful. Given earlier economic development in the United States and other significant variations of history, culture, state structures, and demography, the similarities in South African and U.S. rhetoric and practices of racial domination are striking, particularly in regard to segregation policies between the world wars.31 There has also been remarkable comparability in the ideology and strategy of resistance challenging racial domination.32 Yet a dichotomy of racial domination and conflict in South Africa and the United States, contrasted with Brazil’s so-called racial democracy and relatively mild black protest, is too simple. Despite the commonality of early discrimination and maintained inequality, post abolition state policies encoded differing racial orders, with varying results. South African apartheid was more centrally imposed and pervasive than was Jim Crow segregation and exclusion in the more decentralized post-Reconstruction United States. And if South African race relations bark most loudly and the United States a bit more softly, why was Brazil the dog that didn’t bark? These divergent outcomes present a useful puzzle.
Including Brazil is pivotal, reminding us that legal racial domination was not inevitable. While South Africa and the United States used earlier discrimination as the base for constructing continued racial domination, the connection between past and present in Brazil is more problematic. There, early prejudice and large-scale slavery seem oddly inconsistent with the later absence of legal segregation. This apparent disconnect between the foundation of historical discrimination and modern policies might be reconciled by rejecting the rosy Brazilian self-image of post-abolition racial democracy. But while discrimination and inequality are evident in modern Brazil, the lack of any official policy comparable to either apartheid or Jim Crow remains remarkable. Alternatively, Brazil’s past may have been particularly tolerant, prefiguring its present, as argued by Gilberto Freyre.33 But Brazil had imported an African population widely denigrated as “forever constituting one of the causes of our inferiority as a people.”34 If the historical record contradicts the thesis of early tolerance, Brazil’s later racial order was not so preordained. Instead, the past was reconceived into a benign image, unlike elsewhere. Belief in racial tolerance was thereby reinforced, the legacy of inequality was camouflaged, and quiescence among Afro-Brazilians was encouraged. Brazil’s distinctiveness did not then rest upon less historical discrimination, but rather in the country’s purposeful denial of that legacy. What accounts for this divergence?
The most obvious difference of conditions across the three cases here compared is demographic. There is an indigenous African majority in South Africa, a roughly equal mix of African and European ancestry in Brazil, and an African-descendant minority in the United States (though some Southern states had close to a majority black population by the mid-nineteenth century).35 This demographic difference was consequential, but does not coincide with the continuum of outcomes. A black majority in South Africa and a black minority in the United States were subject to more or less strict forms of post-abolition legal exclusion by race, while in the intermediate case, Afro-Brazilians were not.
Differences of early labor coercion also do not correspond with later outcomes. To meet labor needs not met by indigenous populations, the Southern United States and Brazil engaged in massive importation of slaves from Africa, becoming the two largest slaveholder societies in modern history.36 But out of these legacies of extensive bondage, the United States later enacted legal segregation and Brazil did not. Tannenbaum argued that Brazil’s post-abolition racial democracy was the result of a more humanitarian form of slavery, influenced by Catholic doctrines of natural equality, and the historical record does suggest that Brazilian slavery permitted some greater rights for slaves and a higher level of manumission than elsewhere.37 But the overriding fact remains that Brazilian slavery was one of the most extensive forms in history, particularly brutal and deadly. Even the Church itself owned slaves and was long unable to force abolition.38 Though Brazilian slavery (and Catholic influences over it) certainly took different forms than elsewhere, to call it “humanitarian” is absurd.
Slavery established heinous patterns of discrimination in both Brazil and the United States, and so differences in their common systems of extensive slavery cannot directly explain their later divergence of racial orders. Indeed, if a softening of slavery in itself predetermined later tolerance, then South Africa’s greater dependence on free indigenous labor and British-imposed abolition as early as 1833 should have produced the most tolerant racial order of the three cases. But South Africa, with its more limited legacy of slavery, enacted the most pernicious form of institutionalized racism.
Analysts have also suggested that Brazil’s post-abolition racial democracy was a bequest of early patterns of Portuguese colonialism. But was such colonialism more racially tolerant than that imposed by the British or Dutch in South Africa and the United States? Again, the historical record suggests not. Portuguese colonialism was distinctive in the greater extent and longevity of direct, centralized state involvement, but the Portuguese state used its power to maintain slavery, to impose its own color bar, and to establish discrimination against blacks in Brazil as in its African colonies.39 Such early state practices cannot then be credited for having imported early racial tolerance.
A final historical explanation for Brazil’s relative tolerance is that the high level of miscegenation made it impossible to draw strict racial boundaries. The mobility of mulattoes purportedly made strict segregation and discrimination unworkable. According to Carl Degler, “the mulatto escape hatch” of advancement vitiated any efforts at imposing a harsh racial order.40 High levels of miscegenation practiced by outnumbered and disproportionately male Portuguese are a historical fact. And free mulattoes did serve intermediary functions of control over slaves. But few mulattoes advanced very far and most were and remained subject to discrimination comparable to that suffered by blacks.41 The image of mulatto mobility is largely a myth.
Comparison further suggests that the existence of a large number of mulattoes is not in itself sufficient to produce or explain later tolerance. In both South Africa and the United States there was a substantial intermediate population, but this did not prevent authorities from eventually grouping these mixed peoples with blacks as common victims of racial domination. The higher proportion of mulattoes in Brazil would have made comparable segregation more difficult, but not impossible. That Brazil did not attempt to go down this road cannot be explained by the history of miscegenation, but rather by how this fact was interpreted and how categories were constructed.
None of this is to deny the importance of early historical legacies, but rather to demonstrate that the past did not directly prefigure later outcomes of legal racial orders. Instead, nationalists would shape historical interpretations and omissions to legitimate the present, with such uses of history and culture subject to contestation.42
Constructing Racial Domination
Legacies of prior relations, slavery, and ideology laid a foundation for later and varying social constructions, specifying race as a salient issue. But as Brazil most notably demonstrates, the past may be affirmed or reinvented, according to particular biases, as history is brought forward. The modern house of race in each country was built on similar foundations of prejudice, but constructed according to varying circumstances. Like a Roman church built on top of an older basilica covered by the debris of time, the modern structures of racial order settled on top of the older constructions of discrimination. But they were built according to modern specifications, with architects and their critics disputing renovations. The result was a conjuncture of old and new. Why and how was the past carried forward or denied?
The import of inherited conditions and ideas in shaping state policy may be most evident at key turning points, when lengthy processes are crystallized. In the aftermath of abolition and state consolidation during the late nineteenth or early twentieth century, South Africa, the United States, and Brazil all faced extended moments of relative indeterminacy and an unhappy repertoire of alternative possible configurations of race relations. With the formation of South Africa after Britain’s Boer War victory, with abolition and the U.S. North’s Civil War victory, and abolition and republic in Brazil, the issue of whether and how to include free blacks in the polity could not be avoided. Focusing on such pivotal historical moments may help to explain why forms of racial domination were enacted or not.43 At the same time, such turning points can only be understood in the broader historical context in which they developed.44
The dilemma of whether and how to incorporate an African descendant population had long divided whites, undermining whites’ potential loyalty to the state. The more contentious the issue had been, the greater the degree of resolution required. Ruling elites had to manage conflict and design race relations anew, to resolve or avoid division among whites that could threaten central rule, stability, and development. Without a solution, instability would disrupt political consolidation of the nation-state and economic growth. Particularly in South Africa and the United States, ruling elites faced problems of white disunity exacerbated by disputes over the future treatment of blacks.45 Their overriding interest in order and security required building a unifying racial order, to which ideals of equality, freedom, and justice were subordinated.46 Their solutions built upon ideologies of racism that had emerged to defend exploitation or slavery against liberal reform factions.
Was there something similar about transitional moments in South Africa and the United States, but different for Brazil, that can account for the pattern of outcomes? One possible answer stands out immediately. Abolition and/or state consolidation in South Africa and the United States emerged out of the bloodiest conflicts in the histories of those countries, while Brazil suffered nothing remotely comparable to the Boer War or Civil War.47 This divergence fits the general pattern of outcomes. But why would imperial conquest or internal wars and their aftereffects have produced apartheid and Jim Crow? Indeed, just the opposite might seem logical. It was the more liberal abolitionist British and Northerners who won those wars, yet later policies toward blacks were closer to those advocated by the defeated Afrikaners and Southerners. Something more than who won was at work. Furthermore, why would the lack of major internal war result in the absence of official racial domination in Brazil? Again this result is counterintuitive. The more peaceful process of abolition and state consolidation in Brazil left the earlier racial order and social hierarchy relatively undisturbed, yet in Brazil earlier racial discrimination was not codified in later state institutions.
Looking just at pivotal moments cannot fully account for the divergence in racial orders. Outcomes even ran counter to the victors’ possible advantage. If the British had aligned themselves with and protected black South Africans, they could have forged a majority coalition to outvote Afrikaners. With the hubris of victory, the U.S. North assumed they could indeed forge such a coalition with blacks and impose reforms under Reconstruction, but this project was abandoned. Instead, the British and the North sought a coalition with the Afrikaners and the white South, respectively, because those defeated whites had earned a reputation in the prior conflict as being capable of violent disruption. Afrikaners and Southerners posed a viable threat of continuing disruption after the Boer War and during Reconstruction. Liberal interest in blacks was overshadowed by the stronger imperative to unify whites within the nation-state, in part because blacks themselves, divided by ethnicity in South Africa and outnumbered in the United States, were not seen as a comparably violent threat.48 Earlier images of racial inferiority, shared by whites across ethnic or regional lines, reinforced the idea that blacks were not capable of organized and united disruption, and so could and should be subordinated. The strategic and ideological imperatives for a white coalition then set the terms of official racial domination. During the “moments of madness”49 capping cataclysmic and violent change, official policies established the trajectory of racial order for close to a century.
South African and U.S. nation-state consolidation followed a similar logic. “Liberal” and industrializing forces defeated agrarian powers tied to coercive labor practices, with the victors then facing the challenge of reconciling divided loyalties reinforced by conflict. Ethnic tensions and sectionalism had to be diminished, even if they could not be altogether vanquished.50 Afrikaners and Southerners demanding that citizenship not be extended to blacks were appeased. Intra-white conflict was diminished by projecting white racial unity and domination. Blacks were sacrificed on the altar of white unity. In this sense, in the aftermath of the Civil and Boer Wars, further open conflict was contained by inscribing inequalities on which whites generally agreed into social institutions.51 As Stinchcombe concludes, “the tragedy of social life is that every extension of solidarity . . . to nation, presents also the opportunity of organizing hatred at a larger scale.”52 Institutionalizing common prejudice against blacks reinforced white nationalism.53 Both countries followed a “golden law .. . that every white bargain must be sealed by an African sacrifice.”54The pragmatic goal of conflict resolution and nation-state building consistently eclipsed liberal calls for racial justice.
Why did intrawhite conflict trigger the particular use of racial cleavage for domination and nation building? In South Africa and the United States, issues of race and the treatment of Africans were embedded in emerging ethnic or regional conflict, reinforcing existent racist ideologies. Issues of race exacerbating intrawhite conflict were then turned into an instrument for encouraging unity. And as elites victorious in war were more concerned with winning over their white rivals than blacks, constructing a racial order of white supremacy could and did diminish intrawhite conflict. Racial domination was embraced or allowed because it was potentially powerful enough to gradually integrate white populations in violent conflict with each other. Continuing ethnic or section conflicts would have exacerbated disruption. Neither capital nor labor could be so excluded because prosperity required the inclusion of both business and workers. And national unity could not be built on the exclusion of foreigners, who were needed as immigrant labor.
There remains a danger of exaggerating the similarity between the cases of South Africa and the United States. South Africa’s newly forged state authority had either to exclude or include blacks, and to unify whites it pursued exclusion imposed from the center. The potential threat from the African majority impelled such strong intervention, which was matched by increasingly strong central state capacity, reinforced by the strength of the British Empire and encoded in a nonfederal political system of concentrated power.55 Unlike South Africa, in the United States reforms were attempted, but the central state was too weak to persist in Reconstruction in the South, or even to impose a uniform racial order against the black minority. Much as the federalist compromise had allowed for state building a century earlier, a decentralized state of limited capacity was held together by allowing for local variation. Rather than exert its authority, the central state acted by withdrawal, consolidating its authority as best it could by avoiding further conflict. African-Americans were abandoned anew, with Lincoln’s earlier rhetoric of national reconciliation applied only to whites. A generation after the Civil War, Jim Crow was gradually imposed as law in the South and informally maintained in the North, serving both to diminish regional conflict and to unify white Southerners. Segregation was enforced by mob rule, to which the federal government turned a blind eye. Guarantees of equality were unenforced, whereas in South Africa no such guarantees existed and the state acted with force and impunity.
The pattern of racial domination used to diminish intrawhite conflict in South Africa and the United States should also not be misunderstood as having been reified at one point in time. What appears neatly functional after the fact was instead messy, refined by the feedback of ongoing political party competition among whites maneuvering to find varying solutions to real problems.56 Even the identifications of racial, ethnic, or regional actors remained fluid, shaped by ongoing conflict and policy. Increasingly consolidated as a group, Afrikaner nationalists were not easily appeased. Throughout much of the first half of this century their party representatives sought to trump the racial segregation policies of the English-speakers. But the overriding policy of black exclusion helped to contain English–Afrikaner conflict within a single polity. This ongoing competition finally led to an Afrikaner-controlled state in 1948, with segregation reinforced under apartheid. And in the United States, Southern distinctiveness and antagonism against the North were used to reinforce the Democratic Party there after Reconstruction. Regional antagonism and competition did not disappear, but again were successfully contained within a single polity, albeit a loosely controlled one. The original “deal” of a white coalition was made and remade, with outcomes varying to limit ongoing conflict. Indeed, such continued tensions kept alive the intrawhite dynamic refining racial domination in an ongoing process. A racial strategy aimed at preserving nation-state unity was repeatedly and purposefully invoked by selective democratic competition, with both some success and unforeseen consequences.
The other major or potential conflict that continued to shape racial orders was economic. Functionalist arguments have suggested that Jim Crow and apartheid were implemented to protect native or immigrant whites from black competition amid industrialization and urbanization, slower to develop in Brazil.57 But in the United States, Jim Crow was enforced more fully in the less industrialized South, and racial domination was not encoded in Brazil amid urbanization and later job competition in the more industrialized Southeast. South African and U.S. segregation did serve the interests of capital in supplying cheap black labor, and the interests of privileged white labor by ensuring reserved jobs or higher wages.58 But to reduce racial orders to an expression of such white class interests ignores major divisions within and between capital and labor, and the extent to which states maintained racial policies despite economic costs and pressures for reform.59 Capital’s demands were not always met. Instead, racial orders emerged to diminish or avoid inter- and intraclass conflict, all the more potentially disruptive in overlapping with “ranked” ethnic or regional tensions.60 To hold together the nation-state, preserving stability needed for growth, whites were unified across class by race in South Africa and the United States. Economic interests were subordinated to white racial unity, with this class compromise made explicit and enforced by state policy varying in response to ongoing class tensions. Race trumped class.61
Brazilian nation-state making and race making followed an altogether different linked trajectory. There was no competing European fragment akin to the Afrikaners in South Africa, so that the Portuguese were unchallenged in establishing central colonial authority.62 In general, slavery was countrywide, producing little regional conflict over abolition, nor the elaboration of a programmatic racist ideology to defend slavery against such challenges.63 A relative lack of early conflict was further preserved by Brazil’s slow economic growth, contrasted with faster development producing more class conflict earlier among whites in South Africa and the United States. Brazilian unity had been preserved by compromise, and the continued rule of Portugal and Portuguese-descendant emperors ensured the persistence of an established order and religious-social hierarchy, with slaves at its bottom.64 Brazil then modernized and peacefully transformed itself from colony to empire and to republic, and from slaveholding to free.65 A “prefabricated” central state was in place when the winds of modernity hit. Abolition did raise the question of whether and how to incorporate blacks, but in the relative absence of major intrawhite conflict, there was little impetus to unify whites through racial exclusion. Whites were already relatively united; the nation was already bound up, at least among elites.
Brazilian elites found that they could maintain their long-established social order of white privilege without enforcing racial domination. Having experienced larger slave revolts, the Brazilian ruling class was more fearful of blacks than their U.S. or South African counterparts. They were eager to submerge potential racial conflict. Policies were debated and discrimination was evident in preferences for European immigrants and bars on Africans entering the country.66 Racial categories were even encoded, but Brazil did not use these categories as a basis for domination. Inherited and continued racial inequality was denied or explained as reflecting unavoidable but fluid class distinctions. Some Afro-Brazilians were able to advance themselves, encouraging accommodation. Further miscegenation was promoted to “whiten” and unite the population. Seeking “the best of both worlds, such racial democracy maintains the structure of white privilege and non-white subordination, [and] avoids the constitution of race into a principle of collective identity and political action.”67 That prejudice was not legally instantiated had significant consequences. With no official segregation requiring strong state intervention, the republic decentralized power with little conflict or dissent.68
This analysis suggests a schematic pattern. For all three countries, racial discrimination was historically embedded, distinguishing blacks as blacks. But elites sharing this prejudice acted differently in response to the particular conflicts they faced or feared. In the United States and South Africa, nineteenth-century builders of the nation-state faced the threat of a potential triadic tension between two white groups and blacks. A coalition among whites was encouraged by enforcing an ideology and practice of racial domination, shifting the conflict to a more manageable dyadic form of white over black. Intrawhite conflict required reconciliation, and racial domination was imposed to unify white nationalism and allow for state centralization. In Brazil, with no comparably violent ethnic and less regional conflict standing in the way of unity, no reconciliation by exclusion was pursued. A more inclusive nationalism developed in place of a biracial divide. The earlier consolidated Brazilian state required no dyadic racial crutch to preserve the polity even as the form of rule repeatedly shifted. Thus, racial domination was constructed selectively and strategically as a tool for pursuing the goal of the nation-state, a twentieth-century preoccupation.
This divergence of South Africa and the United States from Brazil follows a pattern similar to one found in Europe, as described by Rogers Brubaker. Divisions within the potential German nation were purposefully overcome by state policies of ethnically exclusive citizenship aimed at unifying those included. By contrast, an earlier unified French state adopted formal rules of “civic,” inclusive citizenship. Already achieved French unity did not require reinforcement with official “ethnic” exclusion, though informal discrimination is increasingly evident.69 In South Africa, the United States, and Brazil, colonialism, slavery, and geography had left a substantial and historically differentiated black population. “Bounded citizenship”70 according to race was constructed in South Africa and to a lesser extent in the United States to bolster nation building, much as German unity was reinforced by excluding ethnic groups defined as not German. In the relative absence of regional or any ethnic division, race was not soused in Brazil, akin to the French example of long avoiding official ethnic exclusion.
For all their variation, racial orders helped to achieve stability and development. In South Africa and the United States, ethnic or regional and class conflict were contained by encouraging white racial unity. In South Africa after 1948, Afrikaner political power over a strong state gradually satisfied and enriched its constituency, dampening animosity with the English-speakers. In the United States, allowing for Jim Crow served to help reconcile North and South and to unify whites as whites, thereby reducing the divisive pressures that had once and could again tear apart the Union. Continued regional and party competition reinforced the post-Reconstruction deal of white unity encouraged by racial domination. The central state subsequently grew in capacity as the South was gradually Americanized.71 With the greater resolution of intrawhite conflict in the United States, as in South Africa, economic growth and structural change proceeded. In Brazil, racial democracy in its own way helped to preserve the polity and growth, with class and racial conflict contained within a corporatist order. But if these racial orders did gradually diminish internal conflict, how then do we account for the abandonment of apartheid and Jim Crow, and rising challenges to Brazil’s racial order?
Challenging the Racial Order
Whites had the political, economic, and military power to take the lead in defining themselves as a race and in establishing racial order. But this domination also consolidated mobilization by blacks, building upon their own historical solidarity of shared culture, experience, and fate.72 Using race to reinforce national unity created new conflicts, challenging that unity.73 Refinements of racial domination toward greater repression or reform and the eventual abandonment of official segregation all came in response to real or perceived challenges posed by blacks. To complete an analysis of these dynamics of race making, “a history from above . . . is thus no less essential than a history from below,”74 with either incomplete on its own. Previously concerned with explaining why states do or do not impose racial domination, we must reverse the terms of analysis to explore the consequences of state policy. How did enactment of racial domination or the ideology of racial democracy contribute to resistance that challenged those orders? In other words, if state institutions were constructed to encourage exclusive or inclusive national identities, how did these same institutions shape identities75 and action among blacks, and with what results?
Analysis of pressures from below should avoid any artificially neat state determinism. Intrablack distinctions of color or origin remained, but emergent racial identity among blacks was evident in political, religious, and other associations before racial domination was codified. And such identity has remained after the legal demise of apartheid or Jim Crow, and even in the absence of such policies in Brazil. Nor have protests always been tightly linked to state policy. Activists have applied varying strategies, learning the strengths and weaknesses of each and refining their efforts accordingly.76 Often more than one form of protest was pursued at a time, and often the same people participated in different movements. And blacks have often pursued their own interests by migrating, as from the U.S. South to the North. Indeed, racial identity, movement, and protest cannot be understood simply as a response to actions from above. Blacks have pursued their own ideas and actions, often forcing a reconfiguration of rule imposed upon them.
To the extent that those dominated retain their own agency, their actions must be explained in terms of structural and cultural forces.77 What specifically accounts for the formation of a self-consciously identified racial group capable of protest? Even when race is so reinforced as a salient identity, why does it only sometimes lead to mobilization? And where or when such mobilization emerges, why do its ideology, strategy, and goals vary? When and if protest does emerge in varying forms, what accounts for its timing and effectiveness? If race making from above and below are interconnected processes, then any explanation of changing rules of official domination should also help explain actions by those so dominated. Such patterns would be most evident in South Africa and the United States, where post-abolition race policy and repression have been explicit and protests massive. Yet even in Brazil, where racial identity has grown and the reality of racial democracy has been challenged, more limited forms of protest may accord with similar patterns.
Attempts to answer why and how mobilization against racial domination develops have remained incomplete. Black protest (with or without white allies) has been explained as a result of relative deprivation, though deprivation has not consistently provoked mobilization.78 If relatively constant deprivation cannot explain outbursts and variations in the form of protest, then such outcomes may be dependent on the availability of resources, including allies, financing, and internal organization.79 But again, such resources have been available without protest resulting, and when it does result, its form varies.80 Finally, mobilization has been explained as the response to political opportunities, which encourage people to pursue specific means of collective action.81 Both reforms and repression may provide such opportunities for protest, but neither have always been acted upon.82
Identity formation is a prerequisite for mobilization. In the absence of a self-conscious group, there is no collectivity that can interpret and act upon its situation. The fact that relative deprivation can inspire mobilization does not explain why such deprivation is perceived and acted upon by a group defined by race. That resources and political opportunities are seized as a basis of mobilization also does not explain why such openings are interpreted and acted upon according to race.
It is not sufficient to argue that certain social factors polarize racial identity, for this essentializes race as preexisting and cannot account for different ways in which race has been socially constructed, interpreted, and used. Analysis must explain why social distinctions and conflict come to be projected in terms of physical differences or purported race in the first place, rather than alternative cleavages. Social movement theorists seeking to explain black protest have generally failed to make this shift from using race to analyze events rather than making race the object of analysis.83 Instead, like theorists of racial domination, they have tended to assume as a given the salient identity around which mobilization (or domination) builds. Activists have also tended to essentialize their race. The question remains, Why does race become salient at all, with blacks singled out? Why and how any such identity becomes politicized must be specified.84
The defining identity of those provoked to mobilization may or may not be shaped by official policies of exclusion and enforced deprivation. A state-led process of identity formation is not necessarily applicable to all forms of protest. Some new social movements have been demarcated precisely by their autonomous process of identity consolidation.85 But state action to impose race has built upon and reinforced historical and cultural foundations of black solidarity, helping to provoke collective response.
The Faustian bargain of officially encouraging white unity at the expense of blacks had a later cost. Racial domination had the unintended consequence of consolidating and legitimating subordinated racial identity into a potential basis for resistance. Extending Karl Marx’s argument about class solidarity, a group must self-consciously exist “in itself” before it can act “for itself.”86 Rules enforced according to physical differences helped reinforce such group self-consciousness among blacks, who then interpreted and responded to structural conditions accordingly. Blacks emerging from history with greater solidarity as a group were then provoked to action for themselves as that group. Internal divisions often remain unresolved, but as domination and conflict proceed, those experiences may submerge conflicts within the subordinated group. As state power was consolidated and increasingly penetrated civil society, it also deepened imposed identities and expanded the potential for group mobilization.87
In the United States and South Africa, white racial identity was enforced by state institutions and apparatuses diminishing intrawhite conflict.88 But the consolidation of segregation after the Boer War and American Reconstruction also provoked further efforts at blackracial unity, such as those spearheaded by the African National Con- gress (ANC) or by Du Bois. Dominant institutions thus reinforced and shaped dominated identity, which then asserted itself and pressed for change in those institutions. By contrast, Brazil’s absence of an official race order provided no explicit target of state policy against which racial identity could be consolidated.89
Racial identity can be expressed or channeled into varying strategies of mobilization. Subordinated blacks may seek to enter white society to achieve voice and material benefits, but may also withdraw to avoid prejudice and develop self-worth.90 Protest can then be aimed at integration and/or separatism, both of which can and have been pursued with more or less militancy. These alternatives are not mutually exclusive and have been pursued simultaneously by different people, often interacting, reinforcing and enriching each other.91 Both approaches are “always going on . . . [but] there is always a dominant form.”92
The timing of protest and preferences for integration or separatism, militancy or accommodation, are connected to shifts in state policy presenting distinct resources and opportunities. State building not only creates opportunities for collective action, but the form of such action also reflects the form of the state.93 No less a figure than Nelson Mandela subscribes to this reasoning, arguing that
A freedom fighter learns the hard way that it is the oppressor who defines the nature of the struggle, and the oppressed is often left no recourse but to use methods that mirror those of the oppressor. At a certain point, one can only fight fire with fire. . . . [For instance,] If the oppressor uses violence, the oppressed have no alternative but to respond violently.94
If this reasoning is correct, we should see the timing and forms of mobilization influenced by shifts between greater state repression or reform.
Interaction between state policy and black protest can be specified, though unifying whites through heightened racial domination may not immediately produce mass protest. During the first half of this century in South Africa and the United States, blacks generally engaged in migration or limited peaceful petitioning, remaining hopeful for redress and eager to escape or avoid repression. (The Garvey movement was a notable exception, advocating return to Africa rather than redress.) Throughout, identity was being consolidated and organizations were being built, though relatively stable state policy and repression provided little opportunity or resources for protest. Even in the 1960s, South Africa was able to impose relative quiet.
Transitions and crises in state rule provided openings for mobilization. When whites had unified enough to allow for relaxing segregation, reforms tended to encourage more moderate, often integrationist forms of protest.95 The ANC Youth League and later the United Democratic Front in South Africa, the U.S. civil rights movement, and the Frente Negra in Brazil emerged during such periods of relative reform. These protest responses to reform strengthened white hard-liners who wanted to reverse reforms. Resulting black disillusionment then encouraged more militant, often separatist or nationalist movements, such as the Pan Africanist Congress and Black Consciousness in South Africa, Black Power in the United States, and the Movimento Negro Unificado in Brazil.96 When whites turn away from blacks, blacks tend to follow suit.
Finally, state structure and policy also shape the aim of protest. In South Africa, no constitutional guarantees of equal rights existed for blacks to appeal to, and central authority was stronger earlier and had been used to reinforce racial domination. Rather than look for reform from the center, protest eventually focused on forcing an end to the state as it existed. Such protest built to the equivalent of internal war; racial domination could only be ended through a fundamental political transition.
The situation in the United States was dramatically different. After Reconstruction, constitutional amendments had guaranteed equal rights, suggesting that later protest might force federal authorities to fulfill those abandoned obligations. The civil rights movement in the South purposefully attempted to engage federal intervention against Jim Crow.97 By contrast, discrimination in the North was de facto rather than de jure, with federal intervention both less likely and less effective than in the South, where laws could be and were changed. Northern mobilization was thus less successful at gaining federal intervention, provoking greater frustration, militancy, and violence. But amid such regional distinctions, most African-Americans supported greater state centralization, expecting that such authority would then enforce constitutional guarantees locally resisted.
As a crucial part of this dynamic, analysis of mobilization helps to complete analysis of race making in accounting for the pressures that forced an end to legal racial domination where it had been enacted. Unlike in Brazil, this was the aim of those social movements provoked by apartheid and Jim Crow. But such protest had been long-standing in South Africa and the United States, even while apartheid and Jim Crow were maintained. State strength and coercion remained. What changed?
The interaction of racial domination and mobilization eventually produced a new dynamic leading to greater inclusion. Preservation of the nation-state no longer depended upon racial domination by whites, who had become more unified on this basis. And the preservation of the nation-state, stability, and growth came to be challenged by black agency and protest consolidated by subordination. To head off this new challenge, apartheid and Jim Crow were abandoned. Such reform also reflected international conditions and economic structural changes, which had altered the quality and nature of interactions. But the undermining of racial domination brought about by industrialization, migration, urbanization, and reliance on black labor and consumption was also made possible by prior racial orders preserving polities and economies.
This dynamic, bringing together the two parts of my analysis, suggests considerable analytical leverage, but the reality was much less neat. The end of legal racial domination was protracted. And the legal demise of apartheid and Jim Crow has not ended discrimination or the salience of race. Identities have persisted even after enforcing institutions were reformed. The dynamics of race making remain.
What Is at Stake?
That states constructed legal racial orders prompts the questions of why, how, and with what consequences. As David Laitin has argued, such a “pattern of political cleavages . . . are best explained by a focus on the actions of a hegemonic state. . . . [Even] the political organizations that make demands on the state are themselves partly a function of state actions.”98 But left unspecified, this argument does not account for particular cleavages and their consequences. In general terms, how did particular situations shape race making? Or more directly, why did state-building elites respond to varying pressures and seek to consolidate their support by using race to achieve a specified nationalism? 99
Race was not chosen randomly as a bulwark for nationalism. Issues of race had been long imbedded in practices of slavery, exploitation, and discrimination. These practices were justified with images of white racial superiority, which were widely believed. This ideological foundation could be and was built upon. Elites’ use of race to solidify white support for the nation-state could piggyback on this powerful “symbolic repertoire.”100 Indeed, elites generally shared this belief structure with the white popular sector. State structures, boundaries of democratic inclusion, and citizenship rules were shaped by such essentialist beliefs.101 These outcomes were also rational within the historical context in which they occurred. The same issue of race that previously had exacerbated division among whites in South Africa and the United States could now be used to heal that division. But neither such rationality nor ideology can fully account for the persistence of these relationships.102 Indeed, racial domination was maintained even when it was costly and challenged.
In Part One, this book takes up the question of how the enactment or not of legal racial domination during the twentieth century was shaped by earlier experiences. More specifically, how did colonialism, slavery, and miscegenation influence what would become apartheid in South Africa, Jim Crow in the United States, and racial democracy in Brazil? If these legacies cannot directly explain why Brazil did not encode legal racial domination, or the varying forms of such policies in South Africa and the United States, then the past is not so constraining. Cultural and historical inheritances play a role in policy formulation, but state actions may also shape how such legacies are interpreted to legitimate that action. At stake is why and how aspects of history are brought forward.
Part Two addresses the circumstances that shaped the particular racial orders that were enacted and how the past was used accordingly. Why were certain policies encoded? Why was race used in particular ways? What accounts for changes over time? To answer these questions, we must trace the interests and actions of real actors. It is not sufficient to talk abstractly about the state, as if this were a unitary or intentional actor establishing unchanging policy. We know the policy outcomes, but to account for these we must explore the actual and ongoing pressures, deliberations, debates, and consequences. At stake is why institutional rules codify particular identities and relations.
Part Three takes a different perspective, focusing on how state policy shaped and was shaped by black protest. Such mobilization, or the threat of it, played a central role in motivating racial domination and then in undermining it. But to make this argument, I challenge the prevailing tendency to take for granted the identity around which protest forms. I argue instead that it was shifting state policy that reinforced prior black solidarity, providing the basis and the target against which protest emerged. At stake is how codified identity shapes mobilization and forces institutional change.
Finally, in the Conclusion, the threads of the argument are brought together. If the prior analysis holds, it should then also help to explain why racial domination was officially abandoned and what the implications are of continued informal racial discrimination. In the process, I argue that the impetus for resolving conflict by means of nation-state building was integral to race making and to the unmaking of legal racial domination. But I also suggest the reverse, that within the three cases examined here issues of race have been integral to nation-state building. Nationhood was institutionalized on the basis of race; the political production of race and the political production of nationhood were linked. Even changing structures of the state, whether more or less centralized or corporatist, were shaped by the deals elites made about race in pursuit of nation-state consolidation.
There is also a more far-reaching potential payoff. In contrast to the image of the nation-state as an inclusive “imagined community,” a distinctive and divisive route to nation-state building is suggested. State-imposed exclusion of a specified internal group, used to reinforce the allegiance and unity of a core constituency, may be a more pervasive pattern. Indeed, nation-states have often been based on such exclusion, not only according to race, but also ethnicity, class, and other cleavages. Even in supposedly inclusive nations and democracies, assimilation often rested on informal distinctions and hierarchy. Such selective nation-state building would then explain why particular cleavages become institutionalized. And such manipulation of cleavages may be central to how and in what form the nation-state was built and later challenged. If so, the dynamics explored here may be suggestive of a more widespread historical pattern of nation-state building through exclusion, reinforcing cleavages and thereby exacerbating conflicts that continue to bedevil us.