{ SIX }

Slave Societies: Extreme Inequality

In Part One of this book we analyzed the transformation of ternary societies into ownership societies, focusing on European trajectories. In so doing, we overlooked not only the case of non-European trifunctional societies but also the fact that between 1500 and 1960 or so, European countries established systems of colonial domination throughout the world. These systems profoundly affected not only the development of Europe but also that of the entire globe. In Part Two, we will study slave and colonial societies and the way in which the transformation of non-European trifunctional societies (notably India, where ancient status distinctions remain unusually visible to this day) was altered by their encounter with proprietarian European colonial powers. These processes and trajectories are crucial for understanding the present structure of global inequality both within and between countries.

This chapter begins by looking at what is without a doubt the most extreme type of inequality regime: slave society. Slave societies existed long before European colonialism, and the history of how they grew, were justified, and disappeared raises fundamental questions for any general history of inequality regimes. In particular, we will discover that the ways in which slavery was abolished in the modern era—in the United Kingdom in 1833, France in 1848, the United States in 1865, and Brazil in 1888—as well as the various forms of financial compensation offered to slaveowners (but not to slaves) tell us a great deal about the quasi-sacralization of private property in the nineteenth century, out of which came the modern world we know today. In the United States, moreover, the question of slavery and racial inequality has had a lasting impact on both the structure of inequality and the political party system. In subsequent chapters we will study postslavery colonial societies in the context of what might be called the “second colonial era” (1850–1960), dwelling first on the case of Africa and then on India and other countries (notably China, Japan, and Iran) to see how their inegalitarian trajectories were altered by colonialism.

Societies with Slaves; Slave Societies

Slavery was present in the most ancient societies of which written traces survive, specifically in the Near East in the second and first millennia BCE, in Pharaonic Egypt and Mesopotamia. The Babylonian Code of Hammurabi, which dates from about 1750 BCE, details the rights of slaveowners. Theft of a slave was punishable by death, and a barber who cut the lock of hair by which slaves were identified at the time could have his hand cut off. In the Old Testament, which dates from the first millennium BCE, vanquished peoples were regularly enslaved by their conquerors, and parents sold their children into slavery when they could not pay their debts. Traces of slavery survive from well before the explicit emergence of the trifunctional schema, which sought to organize society around three classes (clergy, warriors, and workers, with a laboring class that was unified and free, at least in theory); this was formalized around the year 1000 in Europe and as early as the second century BCE in India. In practice, slave and trifunctional logics long coexisted in certain societies because the process of unifying the status of workers, which in theory implied not only the end of slavery but also the end of serfdom and other forms of forced labor, was a complex one that lasted for centuries in Europe, India, and other civilizations.1

It is useful to begin by recalling Moses Finley’s distinction between, on the one hand, “societies with slaves,” in which slaves existed but played a relatively minor role and represented only a small fraction of the population (usually only a few percent), and on the other hand, “slave societies,” in which slaves occupied a central place in the structure of production and power and property relations and accounted for a significant share of the population (on the order of several dozen percent). Slaves were found in nearly all societies before the nineteenth century. These were “societies with slaves” in Finley’s sense, generally with fairly small slave populations. For Finley, there were very few true slave societies: Athens and Rome in antiquity and then Brazil, the southern United States, and the West Indies in the eighteenth and nineteenth centuries. In these cases, slaves may have represented from 30 to 50 percent of the total population (or even more in the West Indies).²

Subsequent research has shown that slave societies, while relatively rare, were quite a bit more common than Finley imagined. In antiquity one finds substantial concentrations of slaves throughout the Mediterranean and Near East, in Carthage and Israel as well as numerous Greek and Roman cities, with important variations depending on the political-ideological, economic, monetary, and commercial context.³ Between the fifteenth and nineteenth centuries, we find many examples of non-Western slave societies, such as the Kingdom of the Kongo (comprising parts of Angola, Gabon, and present-day Congo), the Sokoto Caliphate (in the northern part of what is now Nigeria), and the Kingdom of Aceh (on the island of Sumatra in today’s Indonesia), where slaves are estimated to have accounted for 20–50 percent of the population. The Sokoto Caliphate, considered the largest African state at the end of the nineteenth century (with a population of more than 6 million, of whom about 2 million were slaves), is a particularly important case, because slavery and other forms of forced labor continued there until it was incorporated into the British Empire at the beginning of the twentieth century.⁴ There were very likely other slave societies that have yet to be discovered and still others that have not left sufficient traces to be studied in detail.⁵ As for the African slave trade, it has been estimated that it involved some 20 million enslaved persons between 1500 and 1900 (two-thirds of whom were shipped across the Atlantic to the West Indies and the Americas and one-third across the Sahara to the Red Sea and Indian Ocean). The trade was organized both by states and by European, Arab, and African traders. Such numbers represent a significant demographic drain on sub-Saharan Africa, given the limited population of the continent in this period.⁶

The other limitation of Finley’s classification is that in practice there exist many forms of slavery and forced labor. What we see in history is a continuum of labor statuses ranging from absolute servitude to complete “freedom,” an infinite variety of situations defined by the actual rights of individuals, which are always a specific sociohistorical construct. In the most extreme “industrial” forms of slavery, such as we find in the Atlantic trade, slaves had virtually no rights. Pure labor power, they were treated as movable property (chattel slavery). Slaves then had no personal identity (not even an officially recognized name); no right to private life, family, or marriage; no property rights; and of course no mobility rights. Their mortality rate was extremely high (roughly one-fifth died in crossing the Atlantic and almost another fifth in the year that followed), and they were continually replaced by new slaves from Africa. Under the Black Code of 1685, promulgated by Louis XIV to regulate slavery in the French West Indies and in part to limit abuses there, slaves could own nothing; their meager personal effects belonged to their owners.

By contrast, under serfdom, serfs certainly had no mobility rights, since they were required to work the lord’s land and could not leave to work elsewhere. But they did have a personal identity: some signed parish registers, and they generally enjoyed the right to marry (though in some cases this required approval by the lord) as well as in principle the right to own property, generally of small value (and again with the master’s approval). In practice, however, the boundary between slavery and serfdom was never clear and could vary quite a bit depending on the context and the owner.⁷ By a gradual process that began in the final decades of the eighteenth century and accelerated after the abolition of the Atlantic trade in 1807 (which took several more decades to take full effect), plantations in the West Indies, United States, and Brazil began to rely on the natural increase of the Negro population. In the United States, this second phase of slavery proved more profitable than the first, and the number of slaves increased from 1 million in 1800 to 4 million in 1860. In some cases, fear of slave revolts led to harsher treatment of slaves: for instance, Virginia, the Carolinas, and Louisiana adopted laws in the period 1820–1840 that mandated heavy sentences for anyone who taught a slave to read. Nevertheless, the mere fact that forms of private and family life developed in this period made the situation of slaves in the United States, West Indies, and Brazil quite different from that of slaves in the era of continual replenishment of the labor force by new arrivals from overseas. It is by no means certain that the condition of serfs in medieval Europe was much better than that of slaves in the New World.

In the current state of research, it would appear that the 4 million slaves exploited in the southern United States on the eve of the Civil War (1861–1865) constituted the largest concentration of slaves that ever existed. Our knowledge of ancient slave societies is quite limited, however, as are the sources available for the study of slave systems other than the Euro-American trans-Atlantic systems of the eighteenth and nineteenth centuries. The most common estimates of ancient slavery suggest that about 1 million slaves (compared with a free population of about 1 million) worked in the region of Rome in the first century, and from 150,000 to 200,000 slaves worked in the region of Athens in the fifth century BCE (compared to 200,000 free citizens). These estimates do not cover all of Roman Italy or ancient Greece, however, and should be regarded as suggestive orders of magnitude and nothing more.⁸

More importantly, the meaning of servile status varied so widely that such purely quantitative comparisons make only limited sense. In the Sokoto Caliphate in the nineteenth century, some slaves held high positions in the bureaucracy and army.⁹ In Egypt from the thirteenth to the sixteenth centuries, the Mamluks were freed slaves who rose to occupy high military posts and ultimately seized control of the state. Slave soldiers played an important role in the Ottoman Empire until the eighteenth or nineteenth centuries, as did female domestic and sex slaves.¹⁰ In ancient Greece, some slaves (a small minority, to be sure) served as high public officials, often in positions calling for high skills such as the certification and archiving of judicial documents, verification of coinage, and inventorying of temple properties—tasks requiring expertise that it was deemed best to remove from the political arena and assign to individuals without civil rights and therefore no claim to higher office.¹¹ We find no trace of such subtle distinctions in Atlantic slavery. Slaves were assigned to work on plantations, and the virtually absolute separation of the black slave population from the white free population was unusually strict, unlike in most other slave societies.

The United Kingdom: The Abolition Compensation of 1833–1843

Our next task will be to review the various abolitions of Atlantic and Euro-American slavery in the nineteenth century. This will give us a better understanding of the various arguments advanced to justify or condemn slavery as well as the variety of possible postslavery trajectories. The UK case is particularly interesting because, like the British transition from trifunctional to proprietarian logic, it was extremely gradual.

Parliament passed the Slavery Abolition Act in 1833, and between then and 1843 it was gradually put into effect, with complete indemnification of slaveowners. No funds were appropriated to compensate slaves for the damages they or their ancestors had suffered, whether serious physical harm or mere loss of wages for centuries of unpaid labor. Indeed, slaves were never compensated, not under this abolition law or any other. To the contrary, as we will discover, former slaves, once emancipated, were obliged to sign relatively rigid and undercompensated long-term labor contracts, which left most of them in semi-forced labor for long periods after their official liberation. By contrast, in the British case slaveowners were entitled to full compensation for their loss of property.

Concretely, the British government agreed to pay slaveholders an indemnity roughly equal to the market value of their stock of slaves. Fairly sophisticated payment schedules were established in function of each slave’s age, sex, and productivity so as to offer the fairest and most precise compensation possible. Some 20 million pounds sterling, or 5 percent of the UK’s national income at the time, was paid to some 4,000 slaveowners. If the British government had decided in 2018 to spend a similar proportion of national income, it would have had to disburse 120 billion euros, or an average of 30 million euros for each of 4,000 slaveowners. Clearly, these were very wealthy people, many of whom owned hundreds of slaves and in some cases several thousand. The expenditure was financed by a corresponding increase of public debt, which was repaid by British taxpayers; in practice this meant mostly modest or average families, in view of the highly regressive tax system in force at the time (based primarily on indirect taxes on consumption and trade, like most tax systems before the twentieth century). To get an idea of orders of magnitude, note that total public spending on schools and other instruction (at all levels) was less than 0.5 percent of annual national income in the United Kingdom in the nineteenth century. Compensation to slaveowners thus amounted to more than ten years’ worth of educational spending.12 The comparison is all the more striking when one realizes that underinvestment in education is generally considered one of the major causes of Britain’s decline in the twentieth century.¹³

It so happens that the parliamentary archives chronicling these decisions, which at the time seemed perfectly reasonable and justified (at least in the eyes of the minority of property-owning citizens who wielded political power), have recently been the subject of extensive study, which has culminated in the publication of two books and a comprehensive online database.¹⁴ Among the descendants of the slaveholders who were generously indemnified in the 1830s was a cousin of former prime minister David Cameron. Some voices demanded that the state be reimbursed for the sums paid out—sums that formed the basis of many a family fortune still intact today, with slave assets having long since been replaced by real estate and financial holdings. Nothing came of those demands, however.

The Slavery Abolition Act of 1833 emancipated roughly 800,000 slaves, mostly (some 700,000 in all) in the British West Indies (Jamaica, Trinidad and Tobago, Barbados, the Bahamas, and British Guiana), together with a smaller number in the Cape Colony in South Africa and the island of Mauritius in the Indian Ocean. The population in these territories consisted mostly of slaves, but compared with the population of the United Kingdom in the 1830s (roughly 24 million), the number of emancipated slaves represented only about 3 percent of the total metropolitan population. Otherwise, without the large number of British taxpayers relative to the number of emancipated slaves, it would have been impossible to bear the high cost of completely indemnifying slaveholders. As we will see, things looked very different in the United States: the amount of the compensation that would have been required all but ruled out a financial solution.

On the Proprietarian Justification for Compensating Slaveholders

It is important to insist on the fact that the policy of indemnifying slaveowners seemed self-evidently reasonable to British elites at the time. If one confiscated slave property without compensation, why wouldn’t one confiscate the property of those who had owned slaves in the past but exchanged them for other assets? Wouldn’t all existing claims to property then be in danger? These are the same proprietarian arguments we encountered previously in other contexts, in connection for instance with corvées during the French Revolution and absentee landlords in Ireland in the late nineteenth and early twentieth centuries.15

Think, too, of the novels of Jane Austen I discussed in the previous chapter. In Mansfield Park, it so happens that Sir Thomas owns plantations in Antigua while Henry Crawford does not, but these facts have no particular moral connotation given the extent to which different kinds of assets and different forms of wealth (land, government bonds, buildings, financial investments, plantations, and so on) seem to be interchangeable, as long as they yield the expected annual income. By what right should Parliament be allowed to ruin one of these gentlemen and not the other? Indeed, it was not easy to see an “ideal” solution as long as one refused to question the logic of proprietarianism. Of course, it might have been deemed just to demand more of those who had enriched themselves through slaveownership, not only by depriving them of their “property” but also by compensating the slaves, for example, by transferring to them ownership of the parcels on which they had worked for so long without remuneration. But to finance the indemnity, it might also have been justifiable to tax all property owners on a sliding scale according to their wealth. This would have made it possible to share the burden with the many people who had owned slaves in the past and, more generally, all who had enriched themselves by conducting business with slaveholders, for instance, by buying the cotton and sugar they produced, which played a central role in the economy of the day. But it was precisely this general questioning of property, which would have become almost inevitable once one raised the question of compensating slaves (or simply accepted noncompensation of slaveowners), that nineteenth-century elites wished to avoid.

The necessity of compensating slaveowners was obvious not only to the political and economic elites of the time but also to many thinkers and intellectuals. We come back to the distinction between the “radical” and “moderate” Enlightenment that we encountered in the discussion of the French Revolution.¹⁶ Although some “radicals” such as Condorcet defended the idea of abolition without compensation,¹⁷ most “liberals” and “moderates” considered compensation of owners to be a self-evident and uncontroversial preliminary to any discussion. Among them was Alexis de Tocqueville, who stood out in French debates on abolition in the 1840s for compensation proposals that he believed to be ingenious (and they were, for slaveowners, as we will see later). To be sure, moral arguments about equal human dignity did play a role in abolitionist debates. But as long as those arguments failed to provide a comprehensive vision of how society and the economy were organized and a precise plan describing how abolition would fit into the proprietarian order, they failed to elicit much support.

In the eighteenth and nineteenth centuries, numerous Christian abolitionists tried to explain that Christian doctrine itself demanded an immediate end to slavery and that it was the advent of Christianity that had made ending ancient slavery possible. Unfortunately, this argument was incorrect. Any number of bishoprics in Christian Europe owned slaves until at least the sixth or seventh century, and this hastened conversions and abetted Islam’s penetration into Spain in the eighth century.¹⁸ Not until the year 1000 did slavery end in Western Europe, and it took several more centuries for serfdom to disappear, while in Orthodox Russia it lingered until the end of the nineteenth century. In these debates, many historians and scholars of the antiquity, notably in the German school, opposed the arguments of Christian abolitionists on the ground that it was slavery that allowed the other classes of society to engage in the higher artistic and political pursuits that made ancient civilizations, especially Greece and Rome, great. To oppose slavery was therefore tantamount to opposing civilization and settling for egalitarian mediocrity. Some even sought to prove that slavery and civilization were intimately related by arguing that humanity had achieved its highest population level in antiquity, which was no truer than the assertions of the Christian abolitionists but at least seemed plausible, given the intellectual climate of the period: from the Renaissance to the nineteenth century, the Middle Ages were seen as dark ages.¹⁹

It is also interesting to note that debates on abolition, which were particularly spirited in the United Kingdom and France between 1750 and 1850, made free use of figures and statistics thought to reveal the comparative merits of servile and free labor.²⁰ Abolitionists such as Pierre Samuel Du Pont de Nemours (1771) and André-Daniel Laffon de Ladebat (whose calculations in 1788 were more sophisticated) estimated that free workers were so much more productive than slaves that planters should have been able to earn greater profits by emancipating their slaves and transporting to the West Indies some of the cheap labor that could be found in abundance in rural France and elsewhere in Europe. Slaveowners were not persuaded by these scientific calculations (which in fact were not very credible). Indeed, they estimated that servile labor was just as productive as free labor if not more so given the harshness of the work and the need for corporal punishment. Slaveowners in many countries also insisted that since free labor was more costly but no more productive than slave labor, switching would straightaway make it impossible to compete with rivals in other colonial empires. No one would buy their sugar, cotton, or tobacco, and the nation’s output would plummet along with its greatness if somehow the anti-economic and antipatriotic fantasies of the abolitionists were put into practice.

In the end, there is no evidence that the end of the Atlantic slave trade in 1807 damaged the profitability of plantations. Those who had lived off the trade did have to find other employment, but planters soon realized that it could be less costly to rely on the natural increase of the slave population. The decision to end the slave trade was in any case taken first by Britain, followed by the United States and France in 1808–1810, and then by other European powers at the Congress of Vienna in 1815, at a time when new breeding practices had already become widespread and proven their efficacy. If Britain’s landowning and industrial elites agreed to support abolition in 1833, it was probably in part because they believed that at that moment wage labor would allow for economic growth just as profitable as slave labor (and of course it may have been tempting to take revenge against the Americans for their independent ways and economic backwardness)—provided, of course, that slaveowners were fully compensated for their losses, as in Britain, since it was highly unlikely that free labor’s greater efficiency would have sufficed to compensate the slaveholders, notwithstanding the abolitionists’ protests to the contrary. The abolition of slavery imposed a cost on slaveowners, and in the United Kingdom the public choice was for British taxpayers to bear that cost, thus illustrating both the political power of the slaveholders and the grip of proprietarian ideology.

France: The Double Abolition of 1794–1848

The abolition of slavery in the French colonies was unusual in that it took place in two stages. The first abolition was decided by the Convention in 1794 following a slave revolt in Saint-Domingue (Haiti), but slavery was subsequently reinstated under Napoleon. Ultimately, abolition was definitively adopted in 1848, following the fall of the monarchy and advent of the Second Republic. The French case reminds us of what was no doubt the primary reason for the abolition of slavery: not the magnanimity of Euro-American abolitionists or the pecuniary calculations of slaveowners but the rebellions staged by slaves themselves and the fear of further unrest. The crucial role of slave rebellion is obvious in the abolition of 1794, the first major abolition of modern times, which was a direct consequence of the fact that Haitian slaves had already freed themselves by force of arms and were preparing to declare their country’s independence.

It is also quite clear in the case of the British Slavery Abolition Act of 1833, which came less than two years after the Christmas Rebellion of 1831 in Jamaica—a revolt whose bloody echoes in the British press made a deep impression on public opinion, reinforcing the abolitionist position in the debates of 1832–1833 and persuading slaveowners that it would be wiser to accept generous financial compensation than to take the risk that their plantations in Jamaica and Barbados might someday meet the same fate as those of Haiti. The Christmas Rebellion, which ended with mass executions, itself followed another uprising in British Guiana in 1815 and the Guadeloupe revolt of 1802, which ended with the execution or deportation of roughly 10,000 slaves, some 10 percent of the population—an event that led the French authorities to temporarily reinstate the slave trade in the 1810s in order to repopulate the island and get the sugar plantations going again.21

It is important to bear in mind that the largest concentration of slaves in the Euro-American world on the eve of the French Revolution was found in France’s island colonies. In the 1780s, French plantations in the West Indies and Indian Ocean were home to 700,000 slaves (or 3 percent of the population of metropolitan France at the time, which was about 28 million), compared with 600,000 in British possessions and 500,000 on plantations in the southern United States (which had just won its independence from Britain). In the French West Indies, the major concentrations of slaves were found in Martinique, Guadeloupe, and above all Saint-Domingue, which alone was home to 450,000 slaves. Renamed Haiti (from an old Amerindian name) when independence was proclaimed in 1804, Saint-Domingue at the end of the eighteenth century was the jewel of French colonies, the most prosperous and profitable of all, thanks to its production of sugar, coffee, and cotton. Occupying the western part of the island of Hispaniola, where Columbus had landed in 1492, it had been a French colony since 1626; the eastern part of the island belonged to Spain (and later became the Dominican Republic), as did the large nearby island of Cuba (where slavery would continue until 1886).

In the Indian Ocean, the two French slave isles were Île-de-France (the larger of the two in the eighteenth century; it was occupied by the English in 1810 and became a British possession under the name Mauritius after the defeat of Napoleon in 1815) and the Île Bourbon, which was renamed Réunion during the Revolution and remained French in 1815. Plantations on those two islands housed nearly 100,000 slaves in the 1780s, compared with 600,000 in the French West Indies, 450,000 of them in Saint-Domingue alone.

FIG. 6.1.  Atlantic slave societies, eighteenth and nineteenth centuries

Interpretation: Slaves represented roughly a third of the population of the southern United States from 1800 to 1860. The slave share fell from nearly 50 percent to less than 20 percent between 1750 and 1880. It surpassed 80 percent in the British and French West Indies in the period 1780–1830 and rose as high as 90 percent in Saint-Domingue (Haiti) in 1790. Sources and series: piketty.pse.ens.fr/ideology.

Note, moreover, that these were veritable slave islands: slaves accounted for 90 percent of the population of Saint-Domingue in the late 1780s (or even 95 percent, if one counts metis, mulattos, and free men of color). We find comparable levels in the rest of the British and French West Indies in the period 1780–1830: 84 percent in Jamaica, 80 percent in Barbados, 85 percent in Martinique, and 86 percent in Guadeloupe. These were the most extreme levels ever observed in the history of Atlantic slave societies and, more generally, in the global history of slave societies (Fig. 6.1). For comparison, slaves represented 30 to 50 percent of the population of the southern United States and Brazil in the same period, and available sources suggest comparable proportions in ancient Athens and Rome. The British and French West Indies of the eighteenth and early nineteenth centuries are the best documented historical examples of societies in which nearly the entire population consisted of slaves.

It is quite obvious that when the proportion of slaves reaches 80 or 90 percent, the risk of rebellion is very high, no matter how fierce the repressive apparatus. The case of Haiti was particularly extreme in that the slave population grew at a very rapid rate and the number of slaves was significantly greater than on the other islands. Around 1700, the total population of the island was about 30,000, more than half of whom were slaves. In the early 1750s, Haiti was home to 120,000 slaves (77 percent of the total population), 25,000 whites (19 percent), and 5,000 metis and free men of color (4 percent). At the end of the 1780s, the colony comprised more than 470,000 slaves (90 percent of the total population); 28,000 whites (5 percent); and 25,000 metis, mulattos, and free people of color (5 percent; Fig. 6.2).

FIG. 6.2.  A slave island in expansion: Saint-Domingue, 1700–1790

Interpretation: The total population of Saint-Domingue (Haiti) increased from barely 50,000 in 1700–1710 (of which 56 percent were slaves, 3 percent free people of color and mixed race, and 41 percent white) to more than 500,000 in 1790 (of which 90 percent were slaves, 5 percent free people of color and mixed race, and 5 percent white. Sources and series: piketty.pse.ens.fr/ideology.

On the eve of 1789, roughly 40,000 Africans were arriving every year in Port-au-Prince and Cap-Français to replace deceased slaves and replenish the slave supply, which was then growing at an extremely rapid rate. The system was in a phase of accelerated expansion when the French Revolution broke out. In 1789–1790 free blacks began demanding the right to vote and to participate in assemblies. This seemed logical in view of the grand proclamations about equal rights emanating from Paris, but their demands were rejected. The great slave uprising began in August 1791 after a meeting at Bois-Caïman in the Northern Plain; among the participants were thousands of marrons, or fugitive slaves, who for decades had used the mountain’s islands as a refuge. Despite military reinforcements dispatched from France, the insurgents quickly gained the upper hand and seized control of the plantations while the planters fled the country. The commissioners sent from Paris had no choice but to declare the emancipation of all slaves in August 1793, a decision that the Convention extended to all the colonies in February 1794, setting the revolutionary government apart from previous regimes (even if the decision was in reality imposed by the revolts). Yet the decision barely had time to take effect before the slaveowners persuaded Napoleon to restore slavery in 1802 on all the slave islands except Haiti, which declared its independence in 1804. It was not until 1825 that Charles X recognized Haiti’s independence and 1848 that abolition was extended to other territories, including Martinique, Guadeloupe, and Réunion.

Haiti: When Slave Property Becomes Public Debt

The Haitian case is emblematic, not only because it was the first abolition of the modern era following a victorious slave revolt and the first independence secured by a black population from a European power but also because the episode ended with a gigantic public debt that did much to undermine the development of Haiti over the next two centuries. If France finally agreed to recognize Haitian independence in 1825 and to end its threat to invade the island with French troops, it was only because Charles X extracted from the Haitian government a promise to pay 150 million gold francs to compensate slaveowners for the loss of their property. The government in Port-au-Prince really had no choice, given France’s obvious military superiority, the embargo imposed by the French fleet pending a settlement, and the real risk of an occupation of the island.

It is important to measure the significance of that sum of 150 million gold francs, which was fixed in 1825. Following lengthy negotiations, the figure was based on the profitability of the plantations and the value of slaves prior to the Haitian revolution. It represented 2 percent of French national income at the time or the equivalent of 40 billion euros in today’s money.22 The amount is therefore comparable to the sum paid to British slaveowners following the Slavery Abolition Act, taking account of the fact that the number of slaves “emancipated” in Haiti was half the number of British slaves freed in 1833. More significant, however, is the ratio of the debt to the resources at Haiti’s disposal at the time. Recent research has shown that the sum of 150 million gold francs represented more than 300 percent of Haiti’s national income in 1825—in other words, three years of production. The treaty also provided that the entire amount should be paid within five years to the Caisse des Dépôts et Consignation (a public banking institution created during the revolution and still in existence today), where it would be paid out to the despoiled slaveowners (which was done), while the Haitian government was required to refinance the loan from the Caisse with new loans from private French banks so as to spread the payments out over time (which was also done). It is crucial to recognize the magnitude of the sums involved. With refinancing at an annual interest of 5 percent, typical for the time—not even counting the juicy commissions that the bankers did not fail to add on in the course of numerous partial defaults and renegotiations over the subsequent decades—this meant that Haiti was obliged to repay the equivalent of 15 percent of its national product every year, indefinitely, simply to pay the interest on the debt without even beginning to pay down the principal.

Of course, former French slaveowners had no difficulty showing that the island had been far more profitable during the era of slavery. In fact, on the basis of estimates that it is possible to make today, roughly 70 percent of Saint-Domingue’s output from 1750 to 1780 was realized as profit to French planters and slaveholders (who represented just over 5 percent of the island’s population)—a particularly extreme and well-documented example of egregious colonial extraction.²³ Of course, it was difficult to require a theoretically sovereign country to continue to pay 15 percent of its output indefinitely to its former owners merely because it no longer wished to live in slavery. Meanwhile, the island’s economy had suffered greatly from the aftermath of the revolution, the embargo, and the fact that much of its sugar production had been relocated to Cuba, which remained a slave society and where many planters had sought refuge during the insurrection, in certain cases taking some of their slaves with them. Haiti’s insertion into the regional economy was complicated, moreover, by the fact that the United States, worried by the Haitian precedent and little disposed to sympathy for slave rebellions, refused to recognize or deal with the country until 1864.

Though subject to multiple and often chaotic renegotiations, the Haitian debt was largely repaid. In particular, Haiti ran very significant trade surpluses throughout the nineteenth and into the early twentieth centuries. After the earthquake of 1842 and the subsequent fire in Port-au-Prince, France agreed to a moratorium on interest payments from 1843 to 1849. But the payments resumed thereafter, and recent research shows that French creditors managed to extract an average of 5 percent of Haiti’s national income from 1849 to 1915, with substantial variation depending on the period and the political state of the country: the island’s trade surplus often amounted to 10 percent of national income but sometimes fell to zero or slightly below, with an average of about 5 percent over this period. This is a significant average payment to sustain for such a long period of time. It was nevertheless less than the amount implied by the agreement of 1825, which led French banks to complain regularly that Haiti was a delinquent borrower. With the support of the French government, the banks ultimately decided to cede the rest of their loans to the United States, which occupied Haiti from 1915 to 1934 to restore order and protect American financial interests. The 1825 debt was not definitively repaid and officially wiped from the books until the early 1950s. For more than a century, from 1825 to 1950, the price that France insisted Haiti pay for its freedom had one main consequence: namely, that the island’s economic and political development was subordinated to the question of the indemnity, which was sometimes violently denounced and at other times accepted with resignation, according to the ebb and flow of endless political and ideological cycles.²⁴

This episode is fundamental. It illustrates how the logic of slavery and colonialism was related to the logic of proprietarianism. It also shows how deeply ambivalent the French Revolution was regarding questions of inequality and property. At bottom, the slaves of Haiti took the Revolution’s message of emancipation more seriously than anyone else, including the French, and it cost them dearly. These events also remind us of the close and persistent relation between slavery and debt. In antiquity, slavery for debt was quite common; we find traces of it in the Bible as well as on Mesopotamian and Egyptian steles, which depict endless cycles of debt accumulation and enslavement, sometimes punctuated by periods during which debts were canceled and slaves freed in order to restore social peace.²⁵ In English, the importance of the historical link between slavery and debt is illustrated by the term “bondage,” which refers to the relations of dependency that characterize the servile or slave condition. From the thirteenth century on, “bond” also refers to the legal and financial ties between creditor and debtor as well as to the ties of dependency between landlord and peasant. The legal systems that took hold in the nineteenth century abolished slavery, and at the same time, they ended imprisonment for debt and, above all, intergenerational transmission of debt. There is, however, one form of debt that can still be transmitted across generations, allowing potentially unlimited financial burdens to weigh on progeny, who must pay for the sins of their parents: namely, public debt, like that which postslavery Haiti was obliged to repay from 1825 to 1950. We find many similar cases of colonial debt in the nineteenth and twentieth centuries, to say nothing of the growing public debt that many countries have incurred in recent decades.²⁶

Abolition of 1848: Compensation, Disciplinary Workshops, and Indentured Workers

Let us turn now to the abolition of 1848. Following the passage of the British Slavery Abolition Act of 1833 and its implementation in the period 1833–1843, the abolition debate became ubiquitous in France. There were still 250,000 slaves in the French colonies, especially Martinique, Guadeloupe, and Réunion, whereas those of Jamaica and Mauritius had been set free, arousing fears of new revolts. Nevertheless, debate once again hit a snag over the question of compensation. For slaveowners and their supporters, it was inconceivable that they should be deprived of their property without a fair indemnity. But the idea that the full burden should be borne by public treasury, and therefore the taxpayers, who had already been called on to finance the “émigré billion” in 1825, did not seem quite right.27 Shouldn’t the slaves, who after all would be the primary beneficiaries of the measure, also pay? Alexandre Moreau de Jonnès, a dedicated abolitionist well known for his statistics on slaves and masters in the colonies, which he had compiled using census data and administrative surveys from the early seventeenth century on, proposed in 1842 that slaves should reimburse the entire amount of the indemnity by performing “special work projects” (travaux spéciaux) without pay for as long as necessary. He also insisted that this would be a way of teaching slaves the meaning of work.²⁸ Some commentators pointed out that this transitional reimbursement period might well last quite some time, which would be tantamount to not emancipating the slaves at all: it would merely transform the servile condition into a condition of perpetual debt, just as the former corvées had been transformed into debt during the Revolution.

Tocqueville thought he had found the perfect combination when he proposed in 1843 that half the indemnity be paid to slaveholders in the form of government annuities (hence by increasing the public debt, to be repaid by the taxpayers) and the other half by the slaves themselves, who would work for the state for ten years at low wages, allowing the wage differential to be used to reimburse their former owners. In that way, he argued, the solution would be “fair to all participating parties,” since the former slaveowners would, after ten years, be obliged to pay “the increased price of labor” due to emancipation.²⁹ Taxpayers, slaves, and slaveowners would thus all be made to pay their fair share. A parliamentary committee chaired by Victor de Broglie came up with a similar solution. No one involved in these debates—which admittedly took place in fora dominated by property owners (since just over 2 percent of adult males were eligible to vote for the Chamber of Deputies between 1830 and 1848, and they had to choose their representatives from among the 0.3 percent of wealthiest individuals)—seems to have given serious consideration to the idea that it was the slaves who ought to be indemnified for centuries of unpaid labor. This would have allowed them to become owners of a portion of the land on which they had worked as slaves, and they might then have been able to work for themselves, as Irish peasants did under the agrarian reforms of the late nineteenth and early twentieth centuries (admittedly with generous state compensation to the landlords, at least up to the time of independence).

In any case, the debate went nowhere until the mid-1840s because slaveowners rejected emancipation and threatened to stop it, with armed force if necessary. Only after the fall of the monarchy and the proclamation of the Second Republic in 1848 was Victor Schoelcher’s committee able to secure passage of an abolition bill, which provided compensation for slaveholders somewhat less generous than the British act of 1833 under a cost-sharing arrangement ultimately similar to the one proposed by Tocqueville. Slaveowners received an indemnity calculated on a basis half as large as previously envisioned (which was nevertheless quite substantial).³⁰ In addition to indemnifying slaveholders, the abolition decrees promulgated on April 27, 1848, included articles “punishing vagabondage and begging while calling for disciplinary workshops in the colonies,” the purpose of which was to ensure that planters would have an adequate supply of cheap labor. In other words, under the Schoelcher emancipation, not only were slaves not indemnified or offered access to landownership, but in addition, slaveowners were paid and a regime of semi-forced labor was established, which kept former slaves under the control of planters and allied state authorities. In Réunion, the prefect immediately explained how the new regime would work: former slaves would be required to sign long-term work contracts either as plantation workers or domestic employees or else be arrested for vagabondage and sent to the disciplinary workshops envisioned by the law promulgated in Paris.³¹

To understand the context of the time, it is important to note that laws of this type, in which the state de facto served employers and landlords by imposing strict discipline on labor and keeping wages as low as possible, were common everywhere; they simply caught a second wind in the colonies after the abolition of slavery. Specifically, since many emancipated slaves refused to work for their former masters, British and French authorities developed new systems that allowed workers to be shipped in from elsewhere. In the case of Réunion and Mauritius, the additional labor came from India, for example. The French called these imported laborers engagés, and the British, “indentured workers.” Engagement meant that the Indian workers brought in to replace the slaves were required to reimburse the cost of transportation borne by their employers; this reimbursement extended over a lengthy period, say ten years, and was taken out of their wages. If their job performance was unsatisfactory or, worse, if they were accused of some disciplinary infraction, the reimbursement period could be extended for another ten years or more. Surviving court documents from Mauritius and Réunion show clearly that, since the courts were strongly biased in favor of employers, this system led to exploitation and injustice not identical to slavery but not far removed from it either. The sources also show how employers and courts in a sense negotiated the transformation of the labor discipline regime. Owners slowly agreed to abandon the methods of corporal punishment that had been in wide use under slavery, but only on condition that the authorities help them by imposing financial and legal sanctions that had the same effect.³²

It also bears emphasizing that this type of legal regime, which was very hard on workers (and on the poor generally), was also quite widespread in European labor markets. In 1885, Sweden still had a law on the books allowing anyone without either a job or sufficient property to live on to be arrested and sentenced to a term of forced labor.³³ We find similar laws throughout Europe, notably in the United Kingdom and France, but Swedish law was particularly harsh and remained in force for an unusually long time, which is consistent with what we have seen of Sweden’s exacerbated proprietarianism in the late nineteenth century.³⁴ As it happens, this regime was about to be radically transformed in a number of European countries, including Sweden, in the late nineteenth and early twentieth centuries as unions were authorized, workers obtained the right to strike and engage in collective bargaining, and so on. In the colonies—and not just the former slave islands—the transition took longer: in Chapter 7 we will see that perfectly legal forms of corvée and forced labor persisted into the twentieth century in the French colonial empire, especially in the interwar years and virtually up to the time of decolonization.

Forced Labor, Proprietarian Sacralization, and the Question of Reparations

Several lessons emerge from these episodes. First, there are many gradations of labor between forced and free, and it is important to look closely at the details of the relevant legal system (the point being that they are not merely details). This is true in particular regarding immigrant workers today, whose right to negotiate wages and working conditions is often quite limited, whether in the petro-monarchies of the Persian Gulf or in Europe and elsewhere in the world (particularly for undocumented workers). Indeed, labor law in general calls for close attention. Second, these debates attest to the power of the quasi-sacralized private property regime that dominated the nineteenth century. Had conflicts and events taken a different course, other decisions might have resulted. But those that were taken demonstrate the power of the proprietarian schema.

Schoelcher, who is remembered as a leading abolitionist, said he was embarrassed by the compensation paid to slaveholders but insisted that it was impossible to proceed in any other way once slavery was enshrined in a legal setting. The Romantic poet Lamartine, also an abolitionist, forcefully voiced the same argument in the Chamber of Deputies: it was absolutely necessary, he said, to grant “an indemnity to the colonists for the portion of their legally owned property in slaves, which is to be confiscated. We will never do anything else. Only revolutions confiscate without compensation. Legislators do not act that way: they change, they transform, but they never ruin. They always respect acquired rights, no matter what their origin.”35 No clearer statement of the case is imaginable: the refusal to distinguish among different types of acquired rights to property was the basis of the belief that slaveowners should be compensated (and not slaves). These episodes are fundamental. For one thing, they enable us to set in perspective the reemergence of certain forms of quasi-sacralization of property in the twenty-first century (regarding, in particular, integral repayment of public debt, no matter what its amount or duration, as well as the argument that the private wealth of billionaires is fully legitimate and sacrosanct, regardless of magnitude or origin). For another, they shed new light on the persistence of ethno-racial inequalities in the modern world, as well as the complex but unavoidable issue of reparations.

In 1904, when Haiti celebrated the hundredth anniversary of its independence, the government of the Third Republic refused to send an official delegation. French officials were in fact quite dissatisfied with the rate at which Haiti had been paying down its 1825 debt and felt that it was out of the question to indulge such a delinquent borrower, particularly at a time when the colonial empire, then in a phase of rapid expansion, frequently needed to be disciplined with coercive debt strategies. In 2004, when Haiti celebrated the bicentennial of its independence in a very different political context, the government of the Fifth Republic came to the same conclusion but for different reasons. The French president refused to attend the ceremony because it was feared (not without reason) that Haitian president Aristide would seize the opportunity to demand that France compensate Haiti for the odious debt that the small island republic had been obliged to repay for more than century (the value of which Aristide put at 20 billion in 2003 US dollars)—a demand that the French government had no intention of entertaining on any grounds whatsoever. In 2015, the French president, on a visit to Haiti in the wake of the 2010 earthquake and the lengthy reconstruction operations that followed, reiterated this position. To be sure, France owed Haiti a sort of “moral” debt, but it was out of the question even to consider any kind of financial or monetary reparations.

It is not my place to resolve this complex issue here or to say what exact form French compensation to Haiti ought to take (especially when there is nothing to prevent us from thinking about more ambitious forms of transnational justice or intergenerational reparations; I will come back to this later).³⁶ Nevertheless, I must point out the extreme weakness of the arguments raised by those who refused to reopen the Haitian case while defending other forms of reparation. In particular, the argument that all this is ancient history cannot withstand scrutiny. Haiti reimbursed its French and American creditors from 1825 to 1950, that is, until the middle of the twentieth century. But compensation is still being paid today for expropriations and injustices that took place in the first half of the twentieth century. Think, for example, of the confiscation of Jewish property by the Nazis and allied regimes (including the Vichy government in France) during World War II. It took far too long to establish lawful restoration procedures for these injustices, but eventually it was done, and repayment continues to this day. Think, too, of current reparations for expropriations by Communist regimes in Eastern Europe after World War II, or of the law passed in the United States in 1988 granting $20,000 to Japanese Americans interned during the war.³⁷ By refusing any discussion of the debt Haiti was forced to pay back to France because it no longer wished to be enslaved, even though the payments made from 1825 to 1950 are well documented and wholly uncontested, one inevitably runs the risk of giving the impression that some crimes are more deserving of punishment than others.

Since the early 2000s, several French organizations have been calling for an exercise in national transparency regarding the compensation to former slaveowners paid by the Caisse des Dépôts in connection with the indemnity of 1825 as well as the compensation paid under the law of 1848.³⁸ Neither case has been examined in detail, unlike the British compensation of slaveowners (which admittedly was investigated only recently). It is possible that the relevant French archives are not as well preserved as Britain’s parliamentary archives. That should not prevent a thorough examination of the issues, nor should it prevent France from paying substantial reparations to Haiti or, for that matter, from paying for appropriate educational materials and museum exhibits (there is no museum of slavery worthy of the name in France, not even in Bordeaux or Nantes, ports that owe their prosperity to the slave trade). The cost of the latter would be ridiculously small compared to the cost of reparations to Haiti, but the pedagogical benefit would be huge.

On May 10, 2001, the French National Assembly, acting at the behest of Christiane Taubira (a representative from French Guiana), passed a law “tending toward the recognition of the slave trade and slavery as a crime against humanity.” But the government and majority at the time took care to excise Article 5, which set forth the principle of reparations and would have established a commission to look into the issues; it would never see the light of day.³⁹ Apart from the question of financial reparations to Haiti, another large-scale compensation also backed by Taubira seems difficult to avoid: the question of agrarian reform in Réunion, Martinique, Guadeloupe, and Guiana, the purpose of which would be to allow the descendants of former slaves to have access to parcels of land in places where most of the land and financial assets remain in the hands of the white population, often descended from the families of planters who benefited from the indemnities of 1848. In 2015, Taubira, by then minister of justice, sought unsuccessfully to remind the French president of the importance of the Haitian debt issue and of agrarian reform in France’s overseas départements.

Yet to judge by the indemnification of Japanese Americans, which American leaders resisted for decades, or that of French Jews whose property was confiscated during the war and who had to wait until the early 2000s for a committee to be named to look into their grievances, it is quite possible that agitation around these outstanding slavery-related issues will someday succeed and lead to reparations that seem unthinkable today. On the other hand, the case of the Japanese Americans, who received compensation that continues to be denied to the descendants of former African American slaves and to the Mexican Americans who were deported in veritable anti-foreigner pogroms during the Depression (especially in California), reminds us that racial and cultural biases (along with the legal, financial, and political resources available to those seeking indemnities) sometimes play a role in determining who gets what.⁴⁰

United States: Abolition by War, 1860–1865

We turn now to the case of the United States, which is particularly important for our study given the preeminent role that the United States, self-proclaimed leader of the “free” world since 1945, plays in the global interstate system. It is also the only case of abolition precipitated by a violent civil war, in a country where legal racial discrimination persisted until the 1960s and ethno-racial inequalities (or inequalities perceived and represented as ethno-racial) continue to play a structuring role today in the economy, society, and politics. The countries of Europe, which long regarded America’s singular history with astonishment, continue to wonder how the Democratic Party, which was the party of slavery at the time of the Civil War (1860–1865), became the party of the New Deal in the 1930s, of civil rights in the 1960s, and finally of Barack Obama in the period 2008–2016, changing imperceptibly and without major discontinuity. Europeans would nevertheless do well to follow the US trajectory in detail because it is not totally unrelated to the structure of inequality, political conflict, and debates over immigration that has emerged in postcolonial European societies over the past several decades and whose long-term evolution raises many similar questions.

To begin, it is important to note that the system of slavery that existed in the United States in the second slave era (1800–1860) enjoyed an extremely prosperous existence. The number of slaves increased sharply from 1 million in 1800 to 4 million in 1860, or five times the number of slaves on the French and British slave islands at their peak. Although it is true that the slave trade persisted in clandestine fashion until 1820 or so, the fact remains that the dizzying growth in the number of slaves was achieved mainly through natural increase, thanks to a certain improvement in living conditions and the development among the enslaved of forms of private and family life unknown in the eighteenth century; in some cases, this went together with forms of religious education and expansion of literacy, a slow and subterranean process, which despite repressive southern laws to stop it helped arm black abolitionists for the struggle ahead. At the time, however, nothing augured the end of the system. The population of the southern states was 2.6 million in 1800: 1.7 million whites (66 percent) and 0.9 million blacks (34 percent). By 1860 the population had increased nearly fivefold to more than 12 million: 8 million whites (67 percent) and 4 million blacks (33 percent; Table 6.1). In other words, the system was experiencing rapid but relatively balanced growth, and nothing portended impending doom.

In some states, to be sure, the population was as much as 50–60 percent black, but nowhere did the black share of the population attain the levels seen in the West Indies (80–90 percent). Between the 1790s and the 1850s, land use in the United States became increasingly specialized. While the proportion of slaves remained constant in Virginia at around 40 percent throughout this period, in South Carolina it rose gradually from 42 percent in 1800 to 57–58 percent in the 1850s; it also rose in Georgia and North Carolina. In Mississippi and Alabama, newly admitted to the Union from 1817 to 1819, the proportion of slaves increased significantly between the 1820 census and that of 1860, rising to 55 percent in Mississippi, almost as high as in South Carolina. Meanwhile, states close to the Mason-Dixon line separating North from South saw their proportion of slaves stagnate, as in Kentucky (at around 20 percent), or sharply decline, as in Delaware (which went from 15 percent in 1790 to less than 5 in 1860). In New Jersey and New York, where slaves accounted for less than 5 percent of the population in the 1790 census, slavery was gradually abolished after 1804, and no slaves remained in official census figures after 1830 (Fig. 6.3).

TABLE 6.1
The structure of the slave and free population in the United States, 1800–1860
		Total (thousands)		Black slaves		Free blacks		Whites		Total (%)		Black slaves		Free blacks		Whites
Total United States, 1800		5,210		880		110		4,220		100%		17%		2%		81%
Northern states		2,630		40		80		2,510		100%		2%		3%		95%
Southern states		2,580		840		30		1,710		100%		33%		1%		66%
Total United States, 1860		31,180		3,950		490		26,740		100%		13%		2%		85%
Northern states		18,940		0		340		18,600		100%		0%		2%		98%
Southern states		12,240		3,950		150		8,140		100%		32%		1%		67%
Interpretation: The number of slaves in the United States quadrupled between 1800 and 1860 (from 880,000 to nearly 3.95 million) while the slave share of the population in the South remained fairly stable (at about one-third). The slave share of total population declined (owing to the even more rapid growth of the population in the North). Note: All slave states as of 1860 are classified as Southern: Alabama, Arkansas, North and South Carolina, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, Missouri, Tennessee, Texas, and Virginia. Sources and series: piketty.pse.ens.fr/ideology.

FIG. 6.3.  Proportion of slaves in the United States, 1790–1860

Interpretation: The proportion of slaves in the population increased or remained at a high level in the principal slave states of the South between 1790 and 1860 (35–55 percent in 1850–1860 and as high as 57–58 percent in South Carolina), whereas slavery disappeared in the Northern states. Sources and series: piketty.pse.ens.fr/ideology.

It bears emphasizing that these figures are very well known in the US case because a census of both free and slave populations was conducted every ten years from 1790 on. The census was particularly important because, under the terms of the famous “Three-Fifths Compromise,” the number of slaves played a key role in determining the number of seats assigned to each state in the House of Representatives and therefore the number of members of the Electoral College, which chooses the president: each slave counted for three-fifths of a free person. Beyond that, it is important to recall the importance of slaveownership in the birth of the Republic. Virginia was by far the most populous state (with a total population of 750,000, including slaves, in the first census of 1790, which was equal to the combined population of the two most populous northern states, Pennsylvania and Massachusetts). Virginia furnished the country with four of its first five presidents (Washington, Jefferson, Madison, and Monroe, all slaveowners), the only exception being John Adams of Massachusetts. Of the fifteen presidents who served prior to the election of the Republican Abraham Lincoln in 1860, no fewer than eleven were slaveowners.

The slave system in the southern United States was also of decisive importance for the production of cotton, without which the textile industry could not have developed in the North, and which was also crucial for industrial development in Britain and Europe. It is important to keep in mind the unprecedented scale of the Euro-American slave system in the period 1750–1860 (Fig. 6.4), which was truly the crucial period in Europe’s rise to industrial dominance. Until the 1780s, the West Indies, and especially Saint-Domingue, had been the principal producer of cotton. After the collapse of Saint-Domingue’s slave plantations in the 1790s, the torch was passed to the southern states of the United States, which achieved new heights in the number of slaves and cotton production capacity in the period 1800–1860: the slave population was multiplied by four and cotton output by ten, thanks to improved techniques and intensified production. In the 1850s, on the eve of the American Civil War, 75 percent of the cotton imported by European textile factories came from the southern United States. As Sven Beckert has recently shown, it was this “empire of cotton,” intimately associated with slave plantations, that was the heart of the Industrial Revolution and more generally of the economic domination of Europe and the United States. In the eighteenth and early nineteenth centuries, the British and French were still uncertain what they might sell to the rest of the world, so much so that they were prepared to launch the Opium Wars of 1839–1842 and 1856–1860 to capture the China drug trade, but the transcontinental organization of the empire of cotton enabled them to establish their control over global textile production, radically increasing its scale and ultimately flooding the planet’s textile markets during the second half of the nineteenth century.⁴¹

FIG. 6.4.  The rise and fall of Euro-American slavery, 1700–1890

Interpretation: The total number of slaves on Euro-American plantations in the Atlantic region reached 6 million in 1860 (4 million of whom were in the southern United States, 1.6 million in Brazil, and 0.4 million in Cuba). Slavery in the French and British West Indies (to which I have added Mauritius, Réunion, and the Cape Colony) reached its zenith in 1780–1790 (1.3 million), then declined following the revolt in Saint-Domingue (Haiti) and the abolitions of 1833 and 1848. Sources and series: piketty.pse.ens.fr/ideology.

Meanwhile, the internal balance of political and ideological power in the United States also changed radically between 1800 and 1860. In 1800, the population of the United States was roughly 5.2 million, almost equally divided between the southern slave states (with a population of 2.6 million, including slaves) and the northern nonslave states (also with a population of 2.6 million). Many of the northern states had only recently abolished slavery following the lead of Massachusetts in 1783 (although strict racial discrimination continued there until the Civil War, especially in the schools, much as it would continue in the South until the 1960s). By 1860, the picture looked quite different: although the population of the South nearly quintupled (from 2.6 to more than 12 million), that of the North had grown sevenfold (from 2.5 to nearly 19 million). Thus nonslave states now represented more than 60 percent of the total population and more than two-thirds of the free population (Table 6.1). The North had also become considerably more diversified since it now consisted of two distinct parts with different economic bases and different political and ideological attitudes: on the one hand, the Northeast, which included the metropolises of New York and Boston and the industrial and financial fortunes of New England; and on the other hand, the Midwest, represented by both the small farmers of the new Western frontier states and the great meat and grain distribution networks that flourished around Chicago, the region from which Lincoln sprang. In other words, although the slave South with its cotton plantations was growing rapidly, it belonged to a political space that was growing even more rapidly, whose economic and political-ideological models were based on free labor. The West and frontier territories remembered their coming of age before being admitted to statehood themselves, under the “colonial” tutelage of the federal government and the original states: hard-won land was often confiscated by the central government for the benefit of powerful interests.

Bear in mind, however, that the North initially had no intention of demanding immediate abolition of slavery in the South (much less racial equality). The central issue was the status of the new territories to the west. Lincoln and the Republicans wanted them to be free because that was the development model they knew, and they could see the West’s full potential as part of an integrated continental and global economy. “The great interior region … already has above 10,000,000 people, and will have 50,000,000 within fifty years if not prevented by any political folly or mistake,” Lincoln declared to Congress in 1862, adding that this prosperity called for a united nation because this vast interior region “has no seacoast—touches no ocean anywhere. As part of one nation, its people now find, and may forever find, their way to Europe by New York, to South America and Africa by New Orleans, and to Asia by San Francisco; but separate our common country into two nations, as designed by the present rebellion, and every man of this great interior region is thereby cut off from some one or more of these outlets, not perhaps by a physical barrier, but by embarrassing and onerous trade regulations.”⁴² By contrast, southerners feared that if free states were allowed to develop in the West, the slave states would end up a minority in the United States, unable to defend their distinct way of life (a judgment that was not entirely wrong). Slaves began to flee in growing numbers, and even though the Fugitive Slave Act, passed by Congress in 1850, significantly reinforced previous laws, compelling authorities in the free states to assist slave hunters in tracking down their presumed property and providing stiff prison sentences for anyone convicted of aiding fugitive slaves, the southern states felt that they needed a solid political coalition to defend their economic model over the long run.⁴³

Lincoln was elected in November 1860 on a promise to refuse to extend slavery to the new states of the West. In late 1860 and early 1861, he repeatedly stated that he asked for nothing more than unequivocal acceptance of the fact that the new states would be free, along with the beginning of an extremely gradual process of emancipation in the South, with compensation for slaveholders—a process which, had it been accepted, might have prolonged slavery until 1880 or 1900, if not longer. But southerners, like the white minorities in South Africa and Algeria in the twentieth century, refused to give in to a majority they judged to be distant and alien to their world; they chose secession instead. South Carolina voted to secede from the Union in December 1860, and by February 1861 it had already been joined by six other states, forming the Confederate States of America. Lincoln still held out hope for dialogue, but in April 1861, shortly after the inauguration of the new president, the Confederates seized Fort Sumter in the harbor of Charleston, South Carolina, capturing the federal troops stationed there, which left Lincoln no choice but to go to war or accept the partition of the country.

Four years and more than 600,000 dead later (that is, more dead than in all other conflicts in which the United States has been involved, including the two world wars, Korea, Vietnam, and Iraq), the war was over: the Confederate armies surrendered in May 1865. In view of the damage done by the southern forces, compensating former slaveowners was unthinkable. To enlist black support for the Union armies, Lincoln persuaded Congress to pass the Thirteenth Amendment, emancipating the slaves, in April 1864 (without any compensation to either slaveholders or slaves); this was ratified by all the states, including the southern states occupied by the armies of the North, in December 1865. It was made clear that the amendment carried no implication concerning political, social, or economic rights for freed slaves. Early in 1865, Union military authorities had indeed hinted to emancipated slaves that they would receive “forty acres and a mule” when the war was over; had this program been adopted nationwide, it would have amounted to a large-scale agrarian redistribution. No law to compensate slaves was adopted by Congress, and the “forty acres and a mule” slogan became a symbol of Yankee deception and hypocrisy.⁴⁴

On the Impossibility of Gradual Abolition and Compensation in the United States

Could gradual abolition with compensation of slaveowners, such as Lincoln proposed to the South in 1860–1861, have worked in the United States? Given the sums at stake, it seems unlikely without a very large (and highly improbable) transfer of funds from the North to southern slaveholders, or else a very long transition period, extending to the very end of the nineteenth century or the early decades of the twentieth. Without the war or slave revolts (hard to imagine because the slave population was a smaller proportion of the population than in the West Indies),45 the most probable outcome would have been continuation of the slave system. With powerful interests at stake and the slave regime prospering and expanding rapidly in 1860, the South was not ready to accept a peaceful end to slavery.

To gain a better idea of the sums involved, recall that the compensation paid by the British in 1833 cost taxpayers roughly 5 percent of GDP, which is a lot, even though the number of slaves was smaller (about 3 percent of the British population at the time) and British GDP per capita was extremely high for the era. Slaves were then very valuable assets, and the market price of a slave was generally about ten to twelve years of an equivalent free worker’s wages. What does this work out today in today’s terms? Assume a slave does work for which a free worker would be paid 30,000 euros (2,500 euros a month, or roughly the average wage in France and Western Europe today), and assume that this labor brings in at least that much revenue for the slave’s employer. Then the selling price of that slave would be between 300,000 and 360,000 euros. It is easy to see that in a society where slaves represented virtually the entire work force, their market value could reach astronomical levels, potentially as high as seven or eight years of annual production (700–800 percent of national income).⁴⁶ Recall that France saddled Haiti with a debt equivalent to three years of Haitian national income in 1825 yet remained convinced that it was making sacrifices compared to what slaves in Saint-Domingue actually yielded in profit.

In the case of the American South, where slaves represented about a third of the population, there exist numerous sources that tell us how the price of slaves varied with age, sex, and productivity. Recent research has shown that in 1860, the market value of slaves exceeded 250 percent of the annual income of the southern states and came close to 100 percent of the annual income of all the states.⁴⁷ If compensation had been paid, it would have been necessary to increase the public debt, and taxpayers would have been saddled with interest and principal payments for decades.

To sum up, in order to free the slaves without despoiling their owners, the country as a whole would have had to bear the financial burden. The former slaveowners would have become bondholders, to whom US taxpayers (including former slaves) would have owed a substantial debt. This is exactly what happened in the United Kingdom and France (with the special case of Haiti), except that in the United States the sums at stake were considerably larger given the scope of the slave system. Recall that annual public expenditure on education, at all levels of government, did not exceed 1 percent of national income in any country in the nineteenth century. A federal debt of 100 percent of national income would therefore have represented more than a century of investment in education, to say nothing of the fact that interest on that debt alone (roughly 5 percent of national income) would have consumed five times the amount of tax revenue spent on all primary schools, high schools, and colleges and universities in the country. Note, moreover, that the debt contracted during the Civil War—the first major federal debt in US history, stemming from the mobilization, upkeep, and arming of more than 2 million Union soldiers for five years—amounted to $2.3 billion in 1865, or roughly 30 percent of US national income, which at the time seemed a gigantic amount; repayment of that debt was the source of complex political conflicts in the decades to come. It would have taken three or four times the cost of the war itself to compensate former slaveowners at market prices. It is reasonable to think that the people involved were no fools: when Lincoln proposed abolition with compensation in 1860–1861, everyone knew that true compensation was impossible: one side or the other would have found the amounts unacceptable. The real question was therefore whether to put the problem off until later or to accept an immediate freeze on extension of slavery to the new states in the West. Southern slaveowners rejected the latter option.

It is interesting, moreover, to note that both Jefferson and Madison tried to estimate the cost of compensation in the 1810s; both discovered that it would have been enormous (on the order of one year’s national income at the time). Both also submitted proposals for coming up with such a sum. It could have been done, they argued, by selling a third to a half of all land in the public domain, particularly new land in the West.⁴⁸ This would have meant giving vast estates in the new territories to the former slaveholders, estates that would have replaced the small family farms of the settlers then moving into those territories, which would have provoked significant social and political tensions. Proposals of this sort were entertained from time to time between 1820 and 1860, but it was difficult to imagine circumstances under which a majority coalition prepared to run the risk could have been assembled at the federal level without radically altering the political system.

On the Proprietarian and Social Justification of Slavery

The abolition of slavery posed difficult ideological problems to nineteenth-century proprietarian societies, which feared that abolition without compensation of slaveowners would ultimately undermine the whole proprietarian order and system of private property. In the US case, this fear was aggravated by the magnitude of the compensation that would have been required; had it been attempted, it might have provoked other kinds of tension, so in the end it became difficult to see any way out of the country’s predicament.

Beyond these proprietarian concerns, the conflict over slavery in the United States had very deep political and ideological underpinnings, which stemmed from quite distinct models of development and visions of the future. The southern rural slaveholder position was forcefully articulated by John Calhoun, who served as vice president of the United States from 1825 to 1832 in addition to stints as secretary of war, secretary of state, and long service as a senator from South Carolina, a post he held until his death in 1850. As leader of the slave power in the Senate, Calhoun repeatedly described “slavery as a positive good” rather than the “necessary evil” acknowledged by other defenders of the system, whom he deemed pusillanimous. Calhoun’s principal argument rested on the values of paternalism and solidarity that he saw as essential to the slave system. For instance, according to the Democratic senator, the ill and elderly were much better treated on southern plantations than in the urban industrial centers of the North, the United Kingdom, and Europe where workers who were no longer able to work were left to die in the streets or in wretched poorhouses.

According to Calhoun, that would never happen on a plantation, where the old and sick remained members of the community and were treated with dignity and respect until the day they died.49 For Calhoun, plantation owners like himself embodied an ideal of agrarian republicanism and local community. By contrast, the industrialists and financiers of the North were hypocrites who pretended to worry about the fate of the slaves but whose real objective was to turn them into proletarians to be exploited like the rest, only to be discarded once they could no longer work. No doubt Calhoun’s speeches failed to sway dedicated abolitionists, who were familiar with accounts of corporal punishment and mutilation inflicted on plantation slaves and had heard the tales of fugitive slaves like Frederick Douglass. But for many other Americans at the time, the idea that southern planters took at least as much interest in their slaves as northern capitalists did in their workers seemed plausible (and in some cases, no doubt, the claim was not totally false).

Calhoun’s rural republican ideal had points in common with Thomas Jefferson’s ideal of a democracy of yeoman farmers but with one essential difference: Jefferson saw slavery as an evil he did not know how to eliminate. “I tremble for my country when I reflect that God is just, and that his justice cannot sleep forever,” worried the man who wrote the Declaration of Independence and who nevertheless could not imagine the possibility of a peaceful emancipation. “We have a wolf by the ears, and we can neither hold him, nor safely let him go. Justice is in one scale, and self-preservation in the other.” For Jefferson, who was speaking at the time in the 1820 congressional debate about extending slavery to Missouri (which he supported, as he supported the right of Missouri settlers to refuse to admit free blacks to the new state), emancipation could be envisioned only if it was accompanied not only by just compensation for the slaveowners but also by immediate expatriation of all former slaves.⁵⁰

Such fears of inevitable vengeance by freed slaves, or merely of the impossibility of cohabitation, were widespread among slaveowners. This explains the creation of the American Colonization Society (ACS) in 1816. Its mission, ardently supported by Jefferson, Madison, Monroe, and many other slaveowners, was precisely to deport emancipated slaves to Africa. This was in a sense an extreme form of the segregation of blacks and whites practiced in the South from 1865 to 1965. If the two groups were to be separated, why not put an ocean between them? This project was a resounding failure. Between 1816 and 1867, the ACS relocated fewer than 13,000 emancipated African-Americans to Liberia, less than 0.5 percent of the total number of slaves (which was nevertheless enough to seriously perturb the subsequent development of Liberia, which has remained divided between “Americos” and natives to this day).⁵¹ Whatever Jefferson may have thought, emancipation could only have taken place on American soil, and steps would have needed to be taken to ensure good relations between whites and blacks afterward, for instance, by seeing to it that former slaves and their children would have access to schools and political rights. Unfortunately, this was not the path that was chosen, no doubt because former slaveholders were convinced that peaceful cohabitation with their former slaves was impossible.

“Reconstruction” and the Birth of Social Nativism in the United States

These debates about the justification of slavery must be taken seriously because they had a fundamental impact on what came later, not only in terms of persisting racial inequality and discrimination in the United States but also, more generally, regarding the specific structure of political, ideological, and electoral conflict in the United States since the nineteenth century. Foreign observers—and sometimes natives as well—are often astonished that the Democratic Party, which in 1860 defended slavery against Lincoln’s Republican Party, often with arguments close to those of Calhoun and Jefferson (both eminent Democrats), subsequently became the party of Franklin D. Roosevelt and the New Deal and, in the 1960s, the party of John F. Kennedy, Lyndon B. Johnson, the Civil Rights Act, and the War on Poverty, before becoming the party of Bill Clinton and Barack Obama (1992–2000, 2008–2016). We will come back to this in Part Four, when we compare the evolution of socioeconomic structures and political cleavages in the United States and Europe in the twentieth and early twenty-first centuries, along with other large democratic countries such as India and Brazil. And we will see then that this peculiar political-ideological trajectory is in fact rich in instruction and implications for the entire world.

At this stage, note simply that it was by small adjustments and without major discontinuity that the Democratic Party ceased to be Jeffersonian and Calhounian to become Rooseveltian and Johnsonian (and ultimately Clintonian and Obamian). In particular, it was by denouncing what they perceived as the hypocrisy and selfishness of the Republican industrial and financial elites of the Northeast, rather as Calhoun had done in the 1830s, that the Democrats were able to regain power at the federal level in the 1870s and establish the basis of the coalition that would bring them success in the era of the New Deal. From 1820 to 1860, conflict at the ballot box usually pitted Democrats, who were especially well established in the South (as they were throughout the period 1790–1960), against the Whigs, who replaced the Federalists in the 1830s before themselves being replaced by the Republicans in the 1850s and who usually scored their best results in the Northeast. Until 1860, when the Republicans adopted a platform advocating the extension of “free labor” to the West (along with gradual abolition of slavery in the South), the two camps had carefully avoided confrontation over the slavery question, which had been temporarily closed with the Missouri Compromise of 1820 (under the terms of which Missouri was admitted to the United States as a slave state at the same time as the free state of Maine). Constant tension remained, however, especially around the issue of fugitive slaves. In the South, candidates of both parties vied to defend slavery, with each camp accusing the other of tolerating northern abolitionists. In practice, within each southern state, the Democrats drew their main support from white voters in rural counties where plantations were dominant (so that it was difficult to imagine a future without slavery), while the Whigs drew the educated urban vote.⁵²

During Reconstruction, which lasted from 1865 to 1880 or so, the Democrats were quite assiduous in denouncing the financial and industrial elites of the Northeast, who they claimed pulled the strings of the Republican Party for the sole purpose of defending their interests and increasing their profits.⁵³ They focused their accusations on one issue in particular: repayment of the war debt, in relation to the monetary system with its dual gold and silver standards (bimetallism). Briefly, Democrats alleged that Boston and New York bankers were concerned solely with collecting comfortable interest on the sums they had lent to pay for the war, whereas the country needed a loose money policy to expand credit to small farmers and manufacturers and finance modest pensions for veterans, even if it meant tolerating moderate inflation and privileging paper money (the so-called greenbacks) and silver dollars over the gold standard to which the bankers wanted to return immediately. The other major issue was the customs tariff: like the Federalists and Whigs before them, the Republicans wanted to impose high tariffs on imported textiles and manufactured goods from the United Kingdom and Europe to protect industry in the Northeast and ensure a flow of cash into the federal treasury (partly to repay the debt and partly to finance infrastructure they deemed useful for industrial development).⁵⁴ The Democrats, traditionally protective of states’ rights and wary of expanding the federal government, had a field day denouncing the selfishness of New England elites, who they said were always eager to take money from people’s pockets to feather their own nests, whereas the West and South needed free trade to expand the market for their agricultural produce.

The Democrats also took up the cause of new immigrants from Europe, mainly Irish and Italian, whom Protestant Republican elites viewed with a wary eye and sought to deny the right to vote by delaying the grant of American citizenship and imposing educational requirements on suffrage. It was partly for this reason, moreover, that northerners allowed southern whites to regain control of their states and deny former slaves the right to vote. At bottom, many Republicans believed that blacks were not ready for citizenship; hence they had no interest in fighting to give them the vote, especially since they wanted to go on denying that right to newly arrived immigrants in the Northeast (at a time when Democrats in New York and Boston were trying to naturalize Irish and Italian immigrants as fast as they could to swell the ranks of their supporters). The Fourteenth Amendment, adopted in 1868 to replace the three-fifths rule, provided that seats in the House of Representatives would henceforth be apportioned on the basis of population, but if the right of adult males to vote was “in any way abridged … the basis of representation … shall be reduced.” This provision might have provided an efficient way to exert pressure on the southern states, but it was never enforced, because the states of the Northeast realized that they had a great deal to lose in view of their own interest in limiting the right to vote.⁵⁵ This was clearly an important fork in the road.

Finally, the Fifteenth Amendment, adopted in 1870, forbade (in theory) any racial discrimination regarding the right to vote, but its application was left entirely to the states. Segregationist Democrats were on the way to regaining control of the southern states in a climate of extreme violence marked by numerous lynchings and attacks on former slaves who attempted to assert their new rights and show themselves in public. At times the situation verged on insurrection, as in Louisiana in 1873, when there were two rival governors (one a Democrat, the other a Republican elected with black votes). In view of the determination and organization of the segregationists, who had always held power in the South, it would have taken a very strong will on the part of the North to impose racial equality, and that will simply did not exist. Most northerners blamed the war on a small minority of extremists among the large plantation owners and felt that it was time to leave the rest of the South in peace to manage its own affairs and deal with inequality as it saw fit. Once southerners regained control of their state governments, police, constitutions, and courts and, above all, once the last federal troops departed in 1877 (the date that marks the official end of Reconstruction), southern Democrats were free to put in place the segregationist regime that for nearly a century would allow them to deny blacks the right to vote and exclude them from white schools and public facilities.⁵⁶ A specially tailored labor law that made it possible to keep plantation wages low was also introduced,⁵⁷ and growing numbers of blacks who had briefly nursed the hope of full freedom and of some day being able to work their own land began to consider the possibility of a “great migration” to the North.⁵⁸

Such was the new Democratic platform: intransigent defense of segregation in the South, loose money and restructuring of the war debt, opposition to tariffs on manufactured goods, and support for white immigration in the North. More generally, Democrats opposed what they saw as the financial and industrial aristocracy of the Northeast, which had waged the Civil War and freed the slaves only to increase its profits and defend its interests. It was on this complex mix of issues that the Democrats won a majority in Congress in 1874 and won the presidential election of 1884 (having already won more votes, but not the presidency, in 1876, only a little more than ten years after the end of the Civil War). Alternation between parties is normal in a democracy, and these Democratic victories were in part a consequence of the voters’ natural fatigue with the Republicans, who had also been tarnished by various financial scandals, as often happens to parties in power. Nevertheless, it is interesting to try to understand the coalition of ideas and aspirations that allowed this alternation to take place so soon after the war, as this coalition would exert great influence on what came later.

Succinctly put, the political ideology that the Democratic Party developed during Reconstruction partook of what one might call “social nativism,” or, in this instance perhaps, “social racialism,” because blacks were just as much natives of the United States as whites (and more so than the Irish and Italians), even if slaveowners would have been glad to deport them to Africa. One might also speak of “social differentialism” to denote political ideologies that promote a measure of social equality but only within a segment of the population—among whites, say, or people considered to be true “natives” of the territory in question (with the understanding that what is at stake has more to do with the supposed legitimacy of different groups with a claim to occupy the land than with their actual native status), as opposed to blacks or others considered to be outside the community (like non-European immigrants in Europe today). In this instance, the “social” dimension of social nativism was just as real as the “nativism”: Democrats succeeded in convincing white voters from the lower and middle classes that they were more apt to defend their interests and advance their prospects than the Republicans.

Later in this book we will see how this social-nativist Democratic coalition from the era of Reconstruction contributed to an ambitious program of inequality reduction in the United States, especially with the creation of federal income and estate taxes in the 1910s and the New Deal in the 1930s before finally jettisoning its nativism with the turn to civil rights in the 1960s. We will also study the common features and above all the profound differences between the trajectory of the Democratic Party in the United States in the period 1860–1960 and the development of social nativism in the early twenty-first century, especially in Europe and the United States (but now under the auspices of the Republican Party).⁵⁹

Brazil: Imperial and Mixed-Race Abolition, 1888

We turn now to the case of Brazil. Although less studied than the British, French, and American cases, the abolition of slavery in Brazil in 1888 is also highly instructive. In contrast to the American South, where the number of slaves jumped from 1 million to 4 million between 1800 and 1860, Brazil did not experience spectacular growth of its slave population in the nineteenth century. The country was already home to 1.5 million slaves in 1800, and their number increased only slightly between then and abolition in 1888 (Fig. 6.4). Despite increasingly urgent complaints from the British, Brazilian slave traders continued to do business throughout much of the nineteenth century, at least until 1860, but on a steadily diminishing scale. The important point is that the trade did not allow for growth as rapid as that achieved through natural increase in the United States. Racial mixing and gradual emancipation were also much more widely practiced in Brazil, which helped limit growth of the slave population. In the 2010 Brazil census, 48 percent of the population declared itself to be “white,” 43 percent “mixed race,” 8 percent “black,” and 1 percent “Asian” or “indigenous.” In fact, the available research suggests that, however people may describe themselves, more than 90 percent of Brazilians today are of mixed origins, European African and/or European Amerindian, including many who describe themselves as “white.” All signs are that racial mixing was already extremely advanced in Brazil by the end of the nineteenth century while it remains quite marginal to this day in the United States.⁶⁰ However, racial mixing does not prevent social distance, discrimination, or inequality (which remains exceptionally high in Brazil today).

The relative stability of the number of slaves (1–1.5 million) in a rapidly growing population in the period 1750–1850 is reflected in the decreasing proportion of slaves, which fell from 50 percent in 1750 to 15–20 percent in 1880—still a high number (Fig. 6.1). Note, too, that the proportion remained above 30 percent in some regions. Historically, the largest concentrations of slaves were found in the sugar plantations of the Nordeste, particularly around Bahia. During the eighteenth century some slaves were moved south (especially to Minas Gerais) following the development of gold and diamond mines, which were soon exhausted; more slaves were then moved south with the development of coffee plantations in the regions of Rio de Janeiro and São Paulo in the nineteenth century. In 1850, the population of Rio was 250,000, of whom 110,000 were slaves (44 percent), a slightly higher proportion than in Salvador de Bahia (33 percent).

In 1807–1808, when the court of Lisbon abandoned the Portuguese capital under threat from Napoleon’s troops and moved to Rio de Janeiro, the population of Brazil was around 3 million (half of whom were slaves), roughly the same as the population of Portugal. An event unique in the annals of European colonialism then ensued: in 1822, the heir to the Portuguese throne—after renouncing his Portuguese title to the great consternation of his court—became emperor of Brazil under the name Pedro I, the first head of the newly independent state. The decades that followed were marked by numerous slave rebellions in a country that had already seen many autonomous communities founded by fugitive slaves, starting with the quilombo dos Palmares in the seventeenth century, a veritable black republic that survived in a mountainous region for more than a century before succumbing to troops dispatched to put an end to this subversive experiment.⁶¹ A first law mandating emancipation of slaves at age 60 was passed in 1865 after lengthy debate. In 1867, Emperor Pedro II delivered a long speech in which he raised the issue of slavery, provoking an outcry in the Chamber of Deputies and the Senate, then dominated by wealthy property owners and elected by less than 1 percent of the population, with many slaveholders among them.

Faced with a new surge of slave revolts and threats of dissolution, Brazil’s Parliament finally agreed in 1871 to pass a so-called free womb law declaring that children born to enslaved mothers would be emancipated, thus leading gradually to complete abolition. Owners of the mothers of the beneficiaries of this law, known as “ingenues,” were obliged to raise them until the age of 6 in order to qualify for a state indemnity, paid in annual rents (juros) of 6 percent; alternatively, they could keep the young blacks until the age of 21, forcing them to work without pay, in exchange for a smaller indemnity. Meanwhile, debate on outright abolition continued. From 1880 on, the tension in the country was palpable, so much so that many travelers in the Rio and São Paolo provinces in 1883–1884 believed that revolution was imminent. In 1887 the army declared that it could no longer cope with slave revolts and would no longer arrest fugitive slaves. It was in this context that Parliament enacted general abolition in May 1888, shortly before the fall of the imperial regime in 1889, after it was abandoned by the landed aristocracy whose interests it had been unable to defend. The fall of the regime led to the adoption of the first republican constitution in 1891.⁶²

Slavery was ended, but Brazil had not seen the end of the extreme inequality that flowed from it. The constitution of 1891 eliminated the wealth qualification for voting but took care to deny the vote to the illiterate, a provision extended by the constitutions of 1934 and 1946. This immediately excluded about 70 percent of the adult population from the polls in the 1890s; the excluded still represented more than 50 percent of the population in 1950 and roughly 20 percent in 1980. In practice, it was not only former slaves but the poor in general who were banished from political life for a century, from the 1890s to the 1980s. For comparison, India did not hesitate to introduce true universal suffrage in 1947 despite vast social and status differences inherited from the past and despite the country’s poverty. Note, too, that if the European countries that extended the suffrage to all men in the late nineteenth and early twentieth centuries had made the right to vote conditional on literacy, a substantial proportion of citizens (particularly in rural districts and among the elderly) would have been excluded. In practice, moreover, literacy requirements often end up granting inordinate power to local officials in charge of registering voters. Similar requirements were used to prevent blacks from voting in the southern United States until the 1960s.

Beyond the slavery question and access to the vote and education, relations between workers and employers remained extremely harsh in Brazil throughout the twentieth century, particularly between landowners on the one hand and agricultural workers and landless peasants on the other. Abundant evidence attests to the extreme violence of social relations in the sugar-producing regions of the Nordeste, where landlords relied on police and state officials to quell strikes, restrain wages, and exploit agricultural labor without limit, especially after the military coup of 1964.⁶³ Not until the end of the military dictatorship in 1985 and the promulgation of the constitution of 1988 was the right to vote finally extended to everyone, regardless of education. The first election by universal suffrage took place in 1989. In Part Four I will return to the evolution of political conflict in Brazil during the first decades of universal suffrage.⁶⁴ At this stage, I will simply insist on a conclusion we have encountered before: namely, that it is impossible to understand the structure of inequality today without taking into account the heavy inegalitarian legacy of slavery and colonialism.

Russia: The Abolition of Serfdom with a Weak State, 1861

We turn finally to the abolition of serfdom in Russia, decided by Tsar Alexander II in 1861. Besides the fact that this major turning point in Russian and European history coincides exactly with the American Civil War, it is interesting to note that the debates surrounding it raised issues comparable with the issue of compensation to slaveowners but with specificities linked to the weakness of the Russian imperial state. Note, too, that the form of serfdom practiced in Russia in the eighteenth and nineteenth centuries was generally considered to be quite harsh. In particular, serfs were not allowed to leave their estates or have access to the courts. Until 1848, serfs were in theory not allowed to own land or buildings. Yet there was in practice quite a wide range of situations across the huge expanse of Russian territory. On the eve of abolition, it has been estimated that European Russia was home to more than 22 million serfs, or nearly 40 percent of the population of Russia west of the Urals, dispersed over a vast landscape. Many worked on immense estates, some of which employed thousands of serfs. Rights and living conditions varied with the region and owner. In some cases, serfs rose to occupy positions in which they helped administer estates and were able to accumulate property.65

The emancipation of the serfs in 1861, triggered in part by Russia’s defeat in the Crimean War (1853–1856), involved many different processes—making it impossible to analyze here. In particular, the abolition of serfdom was followed by agrarian reform, which ultimately gave rise to various forms of communal property, whose effects on agricultural growth have generally been deemed to be much less positive than emancipation itself.⁶⁶ One important aspect of the Russian Emancipation Act of 1861 was that it included a complex mechanism for indemnifying the owners of serfs for their loss of property, in some ways comparable to the compensation of slaveowners in the British, French, and Brazilian cases (1833, 1848, and 1888 respectively). The general principle was that, to gain access to communal lands, former serfs were required to pay reimbursements to the state and to their former owners for a period of forty-nine years. In principle, then, these payments would have continued until 1910. The terms of the law were renegotiated many times, however, and most of the payments ended in the 1880s.

Broadly speaking, it is important to note that the process was fairly chaotic and not carefully monitored by the central government, whose administrative and judicial capacity was limited. In particular, there was no imperial cadastre so that it was difficult to allocate or guarantee new land access rights. Tax collection, recruitment of soldiers, and the lower echelons of the court system were largely delegated to the nobility and local elites, as was often the case in trifunctional societies in which the formation of the central state had not progressed very far. Hence the ability of the imperial government to transform power relations in the Russian countryside was relatively limited. The mobility of peasants continued to be restricted, officially under community control, to be sure, but in practice all signs are that former serf owners continued to play a preponderant role.

In the eyes of many historians, the emancipation acts of 1861 even led in many cases to reinforced landlord control over the peasantry, for nothing was really done to develop an independent justice system or professional imperial bureaucracy, which would have required a significant increase in the yield of the tax system.⁶⁷ The fragile fiscal and financial organization of the Russian central state also explains in part why the imperial government required former serfs to pay landlords for forty-nine years to secure their redemption, rather than envisioning a monetary indemnity financed by public debt and therefore by taxpayers, as in the United Kingdom and France for the abolition of slavery. A new wave of agrarian reforms was attempted in Russia in 1906, with limited effect. Finally, in April 1916, in the midst of World War I, the imperial government opted for a fiscal reform much more ambitious than anything previously attempted, including a progressive tax on total income rather similar to the one adopted in France in July 1914.⁶⁸

Clearly, it was too late. The Bolshevik Revolution broke out in October 1917 before much headway had been made with this reform; it is impossible to know whether the imperial Russian state could have carried it out successfully. The failed experiment with abolition of serfdom in Russia reminds us of a crucial fact: the transformation of trifunctional and slave societies into ownership societies requires the formation of a centralized state capable of guaranteeing property rights; exercising a monopoly of legitimate violence; and establishing a relatively autonomous legal, fiscal, and justice system—otherwise local elites will continue to wield power and maintain subaltern classes in a state of dependence. In Russia, the transition was made directly to something new: a communist society of the soviet type.

---

1.     1.  See Chap. 2 for the European case and Chap. 8 for India.

2.     2.  See M. Finley, Ancient Slavery and Modern Ideology (Penguin, 1980).

3.     3.  See D. M. Lewis, Greek Slave Systems in Their Eastern Mediterranean Context, c.800–146 BC (Oxford University Press, 2018). See also J. Zurbach, “La formation des cités grecques. Statuts, classes et systèmes fonciers,” Annales. Histoire, sciences sociales, 2013.

4.     4.  See P. Lovejoy and J. Hogendorn, Slow Death for Slavery: The Course of Abolition in Northern Nigeria, 1897–1936 (Cambridge University Press, 1993); P. Lovejoy, Jihad in West Africa During the Age of Revolutions (Ohio University Press, 2016).

5.     5.  There are also many intermediate cases in which slaves represented a fraction of the population that was neither tiny nor dominant: for instance, from 10 to 15 percent in Portugal and Morocco in the late fifteenth and sixteenth centuries. See the online appendix (piketty.pse.ens.fr/ideology).

6.     6.  The population of sub-Saharan Africa has been estimated at 40 million in 1500 and 60 million in 1820. Many researchers have measured the extremely negative long-term effects on the regions that lost the most population. See the online appendix.

7.     7.  The terms used to denote the various forms of forced labor have ambiguous origins. The words esclave and slave come from raids on Slavic populations in the fifth to eighth centuries, and the exploitation that followed was described as servage (serfdom).

8.     8.  See, for example, Lewis, Greek Slave Systems, as well as W. Scheidel, “Human Mobility in Roman Italy: The Slave Population,” Journal of Roman Studies, 2005.

9.     9.  See P. Lovejoy, Jihad in West Africa. Lovejoy also insists on the fact that the large slave populations in Sokoto in the nineteenth century (1.5–2 million at the end of the century, and nearly 4 million if one includes West Africa) should be compared with the rapid growth in the United States. In both cases, the growth of the slave population was fueled by the end of the Atlantic trade, which Sokoto’s Muslim leaders pressured Britain to implement in the late eighteenth and early nineteenth centuries.

10.   10.  See M. Zilfi, Women and Slavery in the Late Ottoman Empire (Cambridge University Press, 2010).

11.   11.  See P. Isnard, La démocratie contre les experts. L’esclavage public en Grèce ancienne (Seuil, 2015). There were barely 2,000 such public slaves, however, out of a total of 200,000.

12.   12.  See the online appendix for an analysis of the amounts at state. In terms of 2018 euros, the 120 billion paid in indemnities amounted to an average of 150,000 euros for each of 800,000 slaves, or a payment of about 30 million to an average slaveholder with 200 slaves. See below the discussion of the price of slaves (in terms of average income at the time) in the US context.

13.   13.  See Chap. 11.

14.   14.  See N. Draper, The Price of Emancipation: Slave-Ownership, Compensation and British Society at the End of Slavery (Cambridge University Press, 2010); C. Hall, N. Draper, K. McClelland, K. Donington, and R. Lang, Legacies of British Slave-Ownership: Colonial Slavery and the Formation of Victorian Britain (Cambridge University Press, 2014). The Legacies of British Slave-Ownership (LBS) database can be consulted at http://www.ucl.ac.uk/lbs/.

15.   15.  See Chaps. 3–5.

16.   16.  Chap. 3.

17.   17.  In Réflexions sur l’esclavage des nègres (1781), Condorcet even proposed that slavemasters pay compensation in the form of a pension to former slaves.

18.   18.  This negative experienced encouraged Christian kingdoms in northern Spain to reduce their dependence on slavery from the eighth or ninth century on. See R. Blackburn, The Making of New World Slavery. From the Baroque to the Modern (Verso, 1997), pp. 39–40.

19.   19.  See Finley, Ancient Slavery and Modern Ideology, chap. 1.

20.   20.  On these debates, see the impressive book by C. Oudin-Bastide and P. Steiner, Calcul et Morale. Coûts de l’esclavage et valeur de l’émancipation (18^e–19^e siècles) (Albin Michel, 2015).

21.   21.  On the strange period of on-again/off-again abolition in Guadeloupe prior to the official restoration of slavery in 1802, see F. Régent, Esclavage, métissage et liberté. La Révolution française en Guadeloupe 1789–1802 (Grasset, 2004). On the context of abolition in Britain in 1833, see esp. the books of N. Draper and C. Hall (LBS) cited above in note 14.

22.   22.  Taking equivalent percentages of GDP or national income at different points in time seems to me the best way to compare sums over the course of history. It is tantamount to indexing amounts to nominal growth of the economy, which leads to results between indexing by price level and indexing by nominal average yield on invested capital (which is distinctly higher than nominal long-run growth). See the online appendix for a more detailed discussion of these issues.

23.   23.  See the online appendix and the estimates of S. Henochsberg, Public Debt and Slavery: The Case of Haiti (1760–1815) (Paris School of Economics, 2016). The equivalent of roughly 55 percent of domestic output (or, more precisely, value added) was exported for the benefit of the owners while 15 percent was consumed or accumulated locally by the planters.

24.   24.  These devastating cycles began as early as 1804, when power was seized by Jean-Jacques Dessalines, who put in place a hyperauthoritarian, monarchical, anti-white, isolationist regime following the surrender in 1803 of the French expeditionary force (whose mission was to exterminate all the insurgents) and the 1802 arrest of Toussaint Louverture (who vigorously defended a continuing white presence and the possible of peaceful partnership with the metropole and integration into the internal economy). The subsequent history of the island has been marked by similar cycles of denunciation and resignation.

25.   25.  See, for example, D. Graeber, Debt: The First 5000 Years (Melville House, 2011), pp. 81–84. See also A. Testard, L’esclave, la dette et le pouvoir (Errance, 2001).

26.   26.  Think, for example, of the debts that Greece and other southern European countries owe to Germany, France, and other northern European countries, or of the recent debts that many African and Asian countries have incurred toward China, or of the Argentine debt to a consortium of international creditors. We will have more to say about the differences and similarities among these cases as well as the debt that France imposed on Germany with the Treaty of Versailles. See esp. Chaps. 10 and 12.

27.   27.  See Chap. 4. The “émigré billion,” intended to compensate the nobility for rent and property lost between 1789 and 1815, affected a far larger number of noble landowners and represented roughly 15 percent of annual GDP in 1825; the 300 million francs envisaged for abolition represented about 2 percent of annual GDP in 1840.

28.   28.  “Emancipation should be gradual and partial, not simultaneous and en masse, because otherwise it might turn into a subversive revolution, as in Haiti. The masters of slaves must be compensated by an indemnity, which should be equivalent, insofar as possible, to the value of the property of which they have been deprived. This indemnity cannot be borne by the Metropole, because it constitutes a capital of 300 million francs, the mere interest on which would soon overload the public debt of France.… It is obvious that since sacrifices must be made to this end, the slaves, who will derive immense benefits from it, should naturally and necessarily make them. As they are being admitted to the class of citizens, it will be useful to instruct them, through salutary practice, that a common law holds that every man ameliorate his position by hard and intelligent labor.” A. Moreau de Jonnès, Recherches statistiques sur l’esclavage colonial et les moyens de le supprimer (1842), pp. 252–253.

29.   29.  See Oudin-Bastide and Steiner, Calcul et Morale, pp. 122–123.

30.   30.  In the debates of the 1840s the proposed indemnity averaged around 1,300 francs per slave (leading to a total estimate of 300 million francs), whereas the reference value in the 1848 bill was 600 francs (or four to six years of waves for an equivalent free worker). See the online appendix.

31.   31.  In 1843, Tocqueville proposed that former slaves be deprived of property rights for a long period of time, from ten to twenty years, to give them time to acquire a taste for work and effort; this lesson might be lost if they were to discover the comforts of property too quickly (and “unnaturally”). This proposal was ultimately omitted from the 1848 law, however. See Oudin-Bastide and Steiner, Calcul et morale, pp. 202–203. On the background of the 1848 decrees, see also N. Schmidt, La France a-t-elle aboli l’esclavage? Guadeloupe, Martinique, Guyane 1830–1935 (Perrin, 2009).

32.   32.  See esp. A. Stanziani, “Beyond Colonialism: Servants, Wage Earners and Indentured Migrants in Rural France and on Reunion Island (c. 1750–1900),” Labor History, 2013; A. Stanziani, Sailors, Slaves, and Immigrants: Bondage in the Indian Ocean World 1750–1914 (Palgrave Macmillan, 2014); and A. Stanziani, Labor on the Fringes of Empire: Voice, Exit and the Law (Palgrave, 2018). See also R. Allen, “Slaves, Convicts, Abolitionism and the Global Origins of the Post-Emancipation Indentured Labor System,” Slavery and Abolition, 2014.

33.   33.  See E. Bengtsson, “The Swedish Sonderweg in Question: Democratization and Inequality in Comparative Perspective, c. 1750–11920,” Past and Present, 2018, p. 10.

34.   34.  In the United Kingdom, the “Master and Servant Law” remained on the books until 1875. See S. Naidu and N. Yuchtman, “Coercive Contract Enforcement: Law and the Labor Market in 19th Century Industrial Britain,” American Economic Review, 2013. In France, the livret ouvrier (work permit), which was toughened in 1854 and abolished in 1890, allowed former employers to warn future employers about troublesome workers and thus seriously harm the prospects of workers deemed to be troublesome. See R. Castel, Les métamorphoses de la question sociale (Fayard, 1995), pp. 414–415.

35.   35.  See Chamber of Deputies, sessions of April 22, 1835, and May 25, 1836.

36.   36.  That is, forms of transnational justice based on equal rights, independent of place of birth, or distant origins. See Chap. 17.

37.   37.  The American indemnity was limited to persons still living in 1988 (roughly 60,000 of the 120,000 Japanese Americans interned between 1942 and 1946), for a cost of $1.2 billion.

38.   38.  See esp. L. G. Tin, Esclavage et réparations. Comment faire face aux crimes de l’histoire, (Stock, 2013). The author is also president of the Conseil représentatif des associations noires (CRAN).

39.   39.  Article 5 read as follows: “A committee of qualified persons is charged with determine the damages suffered and examining the conditions of reparations due as a result of this crime. The competence and mission of this committee will be set by decree of the Council of State.”

40.   40.  Estimates of the number of Mexican Americans expelled in the period 1929–1936 range from 1 to 1.5 million (of whom some 60 percent were born in the United States). Deportations were often organized with the support of local and federal authorities. Some recent estimates put the number of deported as high as 1.8 million (most of whom never returned). See A. Wagner, “America’s Forgotten History of Illegal Deportations,” The Atlantic, 2017.

41.   41.  See S. Beckert, Empire of Cotton. A Global History (Knopf, 2014). In Chap. 9 I will return to the role of slavery and colonial rule in the Industrial Revolution.

42.   42.  Quoted in N. Barreyre, L’or et la liberté (Editions de l’EHESS, 2014), p. 27; published in English as Gold and Freedom: The Political Economy of Reconstruction, trans. A. Goldhammer (University of Virginia Press, 2015), p. 17.

43.   43.  The Fugitive Slave Act led to an increase of 15–30 percent in the price of slaves in the border states compared with states farther south, which suggests that the risk of flight was deemed to be serious by slave dealers. See C. Lennon, “Slave Escape, Prices and the Fugitive Act of 1850,” Journal of Law and Economics, 2016. Kidnapping of free blacks in the North was common and inspired the film Twelve Years a Slave, dir. S. McQueen (Fox Searchlight, 2013).

44.   44.  Indeed, the film director Spike Lee ironically chose it as the name of his production company: “Forty Acres and a Mule.”

45.   45.  The proportion of slaves was as high as 75 percent in some counties, such as Nottoway County in Virginia, not far from Southampton County where Nat Turner led a rebellion in 1831, recently dramatized in The Birth of a Nation, dir. N. Parker (Fox Searchlight, 2016).

46.   46.  Assuming that wages account for 60–70 percent of GDP, which itself depends on many factors and in particular on the legal regime defining “free” labor. With an average yield on capital on the order of 5 percent, the price of slaves could in principle approach twenty years of wages, but allowing for risk and the cost of slave upkeep (food and clothing) explains why the apparent yield was closer to 8–10 percent. See the online appendix.

47.   47.  In the South, the market value of all slaves exceeded the value of all other private property (land, buildings, and equipment). See T. Piketty, Capital in the Twenty-First Century, trans. A. Goldhammer (Harvard University Press, 2014), figs. 4.10–4.11 and the online appendix.

48.   48.  See W. Shade, Democratizing the Old Dominion: Virginia and the Second Party System, 1824–1861 (University Press of Virginia, 1996), pp. 191–193. On the amounts involved, see the online appendix.

49.   49.  “I may say with truth, that in few countries so much is left to the share of the laborer, and so little exacted from him, or where there is more kind attention paid to him in sickness or infirmities of age. Compare his condition with the tenants of the poor houses in the more civilized portions of Europe—look at the sick, and the old and infirm slave, on one hand, in the midst of his family and friends, under the kind superintending care of his master and mistress, and compare it with the forlorn and wretched condition of the pauper in the poorhouse.” Speech delivered by John Calhoun on February 6, 1837, in the Senate.

50.   50.  “The cessation of that kind of property, for so it is misnamed, is a bagatelle which would not cost me a second thought, if, in that way, a general emancipation and expatriation could be effected; and, gradually, and with due sacrifices, I think it might be. But as it is, we have a wolf by the ears, and we can neither hold him, nor safely let him go. Justice is in one scale, and self-preservation in the other.” Thomas Jefferson to John Holmes from Monticello (April 22, 1820), The Writings of Thomas Jefferson, vol.15 (1903), pp. 248–250. See also B. Shaw, “A Wolf by the Ears: M. Finley’s Ancient Slavery and Modern Ideology in Historical Context,” in M. I. Moses, Ancient Slavery and Modern Ideology, ed. B. Shaw (Markus Weiner, 1998).

51.   51.  Note, too, that many slaveholders supported the ACS’s idea of deporting free blacks, whose growing numbers and alleged propensity to foment revolts worried them—all while maintaining slavery. See Shade, Democratizing the Old Dominion, pp. 194–195. The Liberian constitution of 1846, adopted with guidance from the ACS, reserved political power and the right to vote for the Americos, who held the post of president exclusively until 1980.

52.   52.  See Shade, Democratizing the Old Dominion. See also R. McCormick, The Second Party System: Party Formation in the Jacksonian Era (Norton, 1966). The “first party system” featured Democratic-Republicans (renamed Democrats in 1828) against the Federalists. After the presidential election of 1796, won by John Adams (a Boston Federalist), the Federalists suffered heavier and heavier losses and were replaced in the 1830s by the Whigs, named for the British liberal party of that name. This gave rise to the “second party system,” Democrats versus Whigs. The third system began in 1860, when Lincoln’s Republicans took on the Democrats. The principal point of stability in the period 1790–1960 is that the Democrats (and their predecessors, the Democratic-Republicans) always scored their best results in the South, while the Federalist-Whig-Republicans did best in the Northeast. A useful source for mapping presidential elections from 1792 to 2016 is the American Presidency Project. An analysis widely accepted by American political scientists is that the third-party system was transformed into a fourth in 1896–1900 with the arrival of the “populist” movement and demands for redistribution, and then a fifth in 1932 with the arrival of the Roosevelt coalition, and a sixth after 1960 and the civil rights movement. (Some see a seventh since the election of Donald Trump.) See, for example, S. Maisel and M. Brewer, Parties and Elections in America (Rowman, 2011). On the evolution of the US party system, see Chaps. 14 and 15.

53.   53.  On the structure of political conflict in Reconstruction, see N. Barreyre’s excellent book, L’or et la liberté. Une histoire spatiale des Etats-Unis après la guerre de sécession (Éditions de l’EHESS, 2014); published in English as Gold and Freedom: The Political Economy of Reconstruction, trans. A. Goldhammer (University of Virginia Press, 2015).

54.   54.  On the ideas and strategies of Boston financial elites (called “Brahmins” in the political vocabulary of the time), see the illuminating book by N. Maggor, Brahmin Capitalism: Frontiers of Wealth and Populism in America’s First Gilded Age (Harvard University Press, 2017). Some Bostonians tried to invest in southern plantations but soon realized that the “darkies” had no intention of working for nothing (and nursed the “chimerical” hope of owning their own land). Many then reoriented their priorities for investing capital accumulated in northeastern textiles by turning to the West (where they had to deal with pioneers also seeking to protect themselves by, for instance, writing public regulation of water and railroads into their state constitutions).

55.   55.  See Barreyre, L’or et la liberté, pp. 175–176.

56.   56.  In the 1870s, many blacks had the right to vote (and voted massively Republican) in southern states; some of these (such as Louisiana and South Carolina) had up to 40 percent black representation in their legislatures. Then segregationist laws and rigged educational tests were introduced, decreasing black participation in southern elections from 61 percent to 2 percent between 1885 and 1908. See S. Levitsky and D. Ziblatt, How Democracies Die (Penguin, 2018), pp. 89–91.

57.   57.  It was illegal, for example, to offer plantation workers higher wages to lure them away from their jobs; violators faced high fines. See S. Naidu, “Recruitment Restrictions and Labor Markets: Evidence from Postbellum U.S. South,” Journal of Labor Economics, 2010.

58.   58.  This process was very gradual. The proportion of African Americans living in the South decreased slowly from 92 percent in 1860 to 85 percent in 1920 before dropping rapidly to 68 percent in 1950 and 53 percent in 1970, where it then stabilized (with a slight upward tendency since 2000).

59.   59.  See Chaps. 10–11 and Chaps. 14–16.

60.   60.  See the online appendix.

61.   61.  See, for example, B. Bennassar and R. Marin, Histoire du Brésil (Pluriel, 2014), pp. 102–108.

62.   62.  See Bennassar and Marin, Histoire du Brésil, pp. 369–370.

63.   63.  This is true especially for the Nordeste, especially in the Pernambuco region, whose democratically elected governor—a man who had tried to develop cooperatives, launch ambitious literacy programs, and enforce some minimal respect for work rules—was violently overthrown by putschists after the coup. See F. Juliao, Cambao (le joug). La face cachée du Brésil (Maspero, 1968); R. Linhart, Le sucre et la faim. Enquête dans les régions sucrières du nord-est brésilien (Editions de minuit, 1981).

64.   64.  See Chap. 16.

65.   65.  See esp. T. Dennison, “Contract Enforcement in Russian Serf Society, 1750–1860,” Economic History Review, 2013.

66.   66.  See A. Markevitch and E. Zhuravskaya, “The Economic Effects of the Abolition of Serfdom: Evidence from the Russian Empire,” American Economic Review, 2018.

67.   67.  See T. Dennison, “The Institutional Framework of Serfdom in Russia: The View from 1861,” in S. Cavacioocchi, Serfdom and Slavery in the European Economy, 11th–18th Centuries (Firenze University Press, 2014). See also N. Moon, The Abolition of Serfdom in Russia, 1762–1907 (Routledge, 2001).

68.   68.  See N. Platonova, “L’introduction de l’impôt sur le revenu en Russie impériale: la genèse et l’élaboration d’une réforme inachevée,” Revue historique de droit français et étranger, 2015.