Even today, roughly 150 years since its demise as a legal institution, American slavery remains a focus of controversy, albeit of a rather different character than that which preceded its extinction. In the mid-1800s Southern planters passionately defended the morality of their “peculiar institution” against equally impassioned denunciations by (mostly Northern) abolitionists. Planters insisted that the system of slave labor was good and right not only for themselves and their communities but also for the slaves. The abolitionists, however, long ago won the day both on the field of battle and on the field of public discourse, and few if any will take up the challenge of a moral defense of slavery today. That particular controversy, which for so long troubled the American conscience, is over. The controversies that swirl around slavery today are of a more historical and technical nature. Why did the slave system of labor, for example, embed itself so deeply in the American South while gradually disappearing in virtually every other New World colony? Was American slavery comparatively humane, at least relative to other New World systems of slavery? Did American slaves live in materially worse conditions than free industrial workers in the North? Did slaves benefit substantially from the value that they produced? Was the American slave-labor system inefficient in comparison with “free” agricultural labor? Was American slavery moribund by the eve of the Civil War, or was it dynamic and expanding? In short, setting aside the issue of its morality, was slavery at least economically rational? Much of the controversy that has surrounded these questions—and the consequent impulse to reexamine the evidence relating to them—was instigated by the publication of Robert Fogel and Stanley Engerman’s Time on the Cross: The Economics of American Negro Slavery in 1974. In this groundbreaking text, the authors attacked the two contending approaches that had dominated scholarship on American slavery up to that period. The first approach was exemplified by the well-known American historian Ulrich B. Phillips, particularly in his classic American Negro Slavery (1918). Phillips’s study focused almost exclusively on white slaveowners and their organization and management of the slave system of labor. Regarding the slaves themselves, he infamously remarked that whites helped the slaves to become as good as they were. Phillips’s approach can be summarized as a view of slavery through the owners’ eyes, and his work falls into a category that is sometimes called the literature of slave domination. No doubt this approach was not merely ideologically satisfactory from the perspective of white Southerners but also convenient from the perspective of historians, as slaveowners left a much wider array of artifacts and records from which a picture of a time and place could be reconstructed than did illiterate slaves with few possessions. Working from such material does seem likely, however, to produce a very one-sided account of the American experience of slavery. The alternate approach to which Fogel and Engerman equally responded was, in a sense, the mirror opposite of the first approach. It could be called the literature of slave victimization, and it is probably best represented in Stanley M. Elkins’s Slavery: A Problem in American Institutional and Intellectual Life (1959). Instead of seeing slavery through the owners’ eyes, this approach attempted to reconstruct something of the slaves’ own experience of slavery. Elkins noted that, in contrast to Brazil and the Caribbean Islands, the American South experienced no large and sustained slave revolts, and he tried to explain this absence by arguing that the American slave-labor system in the South, especially the Deep South, was, from the slave’s perspective, harsher than the comparable systems in the Caribbean and in South America. Elkins argued that Southern slavery was similar to Nazi concentration camps in the sense of constituting a “total experience” that utterly isolated the individual slave, cutting off all meaningful social relations except with the paternal figure of the owner as “the good father.” The individual slave was thus rendered psychologically defenseless and then systematically transformed into what Elkins termed a “Sambo”—a docile and childlike creature who identified with the very master who was the source of his emasculation. In a way, Elkins’s book and the enormous lit- erature that it spawned re-created the same picture of “the slave as the object of the Master’s will” that lay at the heart of Phillips’s book but now revealed from a different perspective. Fogel and Engerman responded to the literatures of domination and victimization by stressing two main themes. On the one hand, their analysis emphasized the resilience of African Americans trapped within the slave system and the spaces of autonomy and distinctive culture they were able to carve out and preserve for themselves within the system. On the other hand, they argued that slavery itself was not based simply on irrational or archaic domination or victimization but rather on an economically rational and highly effective system of (exploitative) production. They employed new econometric techniques to engineer a careful reassessment of the extensive raw statistical data on slavery, and they argued that their analysis revealed that, contrary to received wisdom, the American slave was not lazy but worked as hard as the white laborer. Although critics of slavery argued that the system was inefficient, the authors found that slave agriculture was actually 35 percent more efficient than free agriculture. Moreover, the purchase of a slave was a highly rational investment, just the same as an investment in a manufacturing company. Further, the authors noted, slavery continued as an economically viable and even expanding system up to the Civil War. Finally and perhaps most problematically, Fogel and Engerman argued that the costs of owning a slave amounted to 90 percent of the profit derived from his or her labor. Many historians and economists have disputed Fogel and Engerman’s analysis and conclusions, but some have come to their defense, and Fogel and Engerman themselves have, in the intervening years, published several books of essays and evidence supporting their initial claims. The result has been a large and diverse literature. What this literature reveals broadly, without retracing the specific lines of argument, is that although there is a good deal of general evidence supporting most of Fogel and Engerman’s analysis, especially their more strictly economic conclusions, there are also enormous regional and historical disparities underlying the American experience of slavery, which are masked by an aggregate quantitative focus. Statistical analysis needs to be balanced with historical, regional, and microlevel examination if it is to be illuminating. Although microlevel analyses of individual plantations are somewhat beyond the parameters of this short interpretive essay, a brief historical overview sensitive to regional divergences and the evolution of slavery over time should help to contextualize Fogel and Engerman’s findings. Historical Overview American slavery was a system of labor based on ownership of persons rather than on consent and contract. It applied almost exclusively to blacks (mainly Africans or people of African extraction). It ended formally in 1865 with the Northern victory in the Civil War and the passage of the Thirteenth Amendment to the U.S. Constitution. The more difficult and controversial question is when slavery effectively began as a general system of labor in the British colonies that would eventually become the United States. Slavery was, of course, permitted virtually everywhere in the colonies during most of the seventeenth century, and it was formally legal throughout the colonies by 1750 (when Georgia overturned its short-lived antislavery statutes of 1733). But until the Anglo-Dutch War (1664–1667), the slave trade was dominated by the Portuguese and the Dutch, and there was little commerce with the American colonies. Records show that one Dutch captain sold 20 slaves in Virginia in 1619, but this transaction seems to have been rather exceptional. Although the historical records are unclear about the numbers of black slaves in the early colonies, the scholarly consensus is that although they were not unknown, they were a rarity. In general, the colonists showed a marked preference for importing indentured servants from Europe to fill their growing demands for labor. African slaves were considered too expensive, too difficult to acclimatize and train, and too time-consuming to supervise in comparison with European servants. The American colonists did experiment with using native Indians as slaves (for example, the government of South Carolina in 1708 estimated the colony had 1,400 Indian slaves in a total population of 12,580), but they were frustrated by cultural barriers (the men frequently refused to perform agricultural tasks they regarded as women’s work), the proximity of escape, and the vulnerability of natives to European diseases. The initial experiment with slavery was a failure. In the latter part of the seventeenth century, however, the economic incentives concerning the importation of labor began to change. This shift in incentives had three main causes. First, there was an enormous increase in the demand for labor. In the Virginia colony, for example, the highly labor-intensive tobacco agriculture took off economically while at the same time the overall population of the colony and of those attempting tobacco cultivation tripled in the years from 1750 to 1800. The demand for labor correspondingly increased. Second, political stabilization and economic growth in Europe and particularly the United Kingdom led to a sharp reduction in the availability of indentured servants as well as a sharp rise in the cost of importing them. Third, the successful English war against the Dutch and the consequent British takeover of the Dutch slave trade led to a sharp reduction in the price and an increase in the availability of African slaves to American colonists. Historian Russel Menard, for example, has calculated that the comparative price of African slaves to indentured servants fell between 1674 and 1791 from a ratio of 2.88 to 1 to 1.83 to 1. When the permanence of slave labor (and the slaves’ progeny) is factored into these comparative costs, it is easy to see why the colonists began to rationally opt for slave labor over the importation of indentured servants. The result of this combination of a new structure of incentives and the new accessibility of the African slaves was the development of a triangular system of trade. Ships would typically depart from ports such as Liverpool and Boston loaded with weapons, manufactured goods, and rum and sail for the coast of northwest Africa, where they would trade these goods with coastal forts, sometimes called factories, and local tribes in exchange for slaves. The ships would then sail for the New World laden with a human cargo that would later be sold in South America (most notoriously in Brazil), in the Caribbean Islands (for instance, Saint Domingue [later Haiti], Jamaica, Cuba, or Bermuda), or in the American colonies. Historian Philip Curtin estimated in 1969 that, despite the expiration of between 5 and 20 percent of the human cargo (mainly in the infamous “middle passage” portion of the journey), around 9.5 million Africans were transported as slaves to the New World. Current estimates range as high as 11 million. Around 85 percent of the slaves transported to the New World were sold in Brazil and many more in the Caribbean Islands (Jamaica, for example, is estimated to have imported 750,000 slaves to work on its sugar plantations). Before the banning of the slave trade by Congress in 1808, the American colonies (and later the United States) probably imported somewhere between 600,000 and 650,000 slaves, about 6 percent of the New World total. It is estimated that by 1680, the American colonies, though still overwhelmingly white, contained around 7,000 African slaves. By 1790, however, the population of African slaves in the colonies had increased almost a hundred times, to close to 700,000. By 1810 the number had risen to 1.1 million, and by 1860, on the eve of the Civil War, the population stood at almost 4 million. The rapidly rising number of slaves in the United States begs an important question, which, in turn, illuminates one of the highly distinctive characteristics of American slavery. If only around 650,000 slaves were imported into the States and if the trade was ended in 1808, what accounts for the fast and continuing growth of the slave population? The answer is simply that, unlike virtually every other slave society in the New World (except Bermuda), the American slave population grew naturally through reproduction, and it grew very rapidly—by four times between 1810 and 1860 alone. The remarkable character of this feature of American slavery can be illustrated by briefly comparing the demographics of slavery in America and in Jamaica. Of the 750,000 slaves imported into Jamaica, only 311,000 remained at the time of emancipation in 1834, whereas the smaller population imported into America had already grown well into the millions. A number of factors have been cited to explain the remarkable fertility of the American slave population, four of which have received the most attention. First, the food selfsufficiency of the American mainland is thought to have allowed slaveowners to provide their slaves with a larger, healthier, and more consistent diet than was practicable for most other New World slave populations. After all, the owners had an important vested interest in the health and strength of their slaves. Second, the absence of tropical diseases has been frequently identified as an important contributor to the high growth rate of the slave population. Third, the fact that slaves in America were largely involved in the cultivation of tobacco, rice, and later cotton rather than sugar (with the exception of a few large plantations in Louisiana) is thought to help explain a comparatively lower mortality rate, which contributed to the rate of overall population growth. Sugar cultivation typically exposed workers to grim and harsh conditions and an exhausting pace of labor, which raised mortality rates and permitted little time for raising families. Finally, it is often pointed out that there was a self-reenforcing quality to the natural growth of a slave population. In short, although the initially imported populations tended to be, for obvious reasons, disproportionately male, reproduction over generations tended to rapidly balance out the gender gap, encouraging further population growth. Of course, slaveowners also had a vested interest in the numerical increase of their slaves for the simple reason that it augmented their property and personal worth and the amount of labor under their control. It has correspondingly sometimes been argued that owners deliberately bred their slaves (or bred with their slaves) as a sort of investment. Although there can be no doubt that many slaveowners often took advantage of their female property and that at least some of them encouraged shorter lactation periods (often only a year), which encouraged more rapidly renewed fertility, there is little evidence that these behaviors were carried out systematically in a manner that would explain the pervasive phenomenon of natural population growth. Moreover, many of the same behaviors were recorded in other slave societies, in which the population shrank precipitously. At any rate, regardless of the precise explanation (and it is likely, in fact, some combination of all the factors mentioned here), the phenomenon of rapid natural population growth is a distinctive and unambiguously established feature of American slavery, which tends broadly to support Fogel and Engerman’s thesis that American slaves enjoyed a significantly better material condition than slaves elsewhere in the New World. Two further important and distinctive features of American slavery may be noted at this point. First, most owners tended to run or at least to personally oversee their own business affairs (as compared to the phenomenon of absentee ownership that characterized the bulk of New World slavery). Second, African slaves were always dispersed in America among a large white population. Even in the South, slaves never accounted for much more than a third of the total population, whereas in much of the Caribbean, they ended up outnumbering whites by ratios as high as 10 to 1. Still, although these statistical generalities are useful in establishing a framework for exploring American slavery, they also conceal a great deal of the very real diversity that developed on the ground. To understand that diversity, it is essential to distinguish the growth of different regional concentrations of slave labor organized around the cultivation of different crops. As Table 1 illustrates, the slave population was by no means evenly spread through the colonies, and indeed, following the War of Independence, African American slavery quickly became a wholly Southern phenomenon. In 1790 the Northern states contained just over 40,000 slaves (mainly concentrated in New York State and Rhode Island) out of a broader population of 697,897, accounting in total for just under 6 percent of the American slave population as a whole. By 1860, however, slavery had been effectively eliminated in the North, whereas the total slave population, now entirely in the South, continued to rise to close to 4 million. Even within the South, however, slaves were not evenly distributed. In 1790 a little under half of the slave population (293,427) was concentrated in Virginia (accounting for close to 40 percent of the state’s total population), with Maryland, North Carolina, and South Carolina accounting for most of the other half. By 1860 slavery had expanded geographically along with the South. Although Virginia and North Carolina continued to lead the states of the Upper South and South Carolina and Georgia continued to be among the leading states of the Deep South (with well over half of South Carolina’s population being made up by slaves), Alabama, Mississippi, and Louisiana had also emerged as major slave states, and, equally important, the institution had infiltrated every Southern state. The distribution of slaves throughout the United States and its change over time reflected basic economic realities on the ground. In the preindependence period, the colonies could be usefully divided into three basic groups: the North, the Upper South, and the Deep South. The primary early demand for slaves was concentrated in Virginia, Maryland, and the upper part of North Carolina and was mostly driven by the development of commercial tobacco farming, which grew rapidly through the latter part of the seventeenth century (increasing from exports of 20,000 pounds in 1619 to 38 million pounds in 1700). Toward the end of the century, South Carolina and later Georgia emerged as a second major source of demand for slaves as the commercial farming of rice developed in the low country (growing from exports of 12,000 pounds in 1698 to 18 million pounds in 1730 and 83 million pounds in 1770). Finally, at the turn of the century and into the antebellum period, technological advances (such as the harnessing of steam power and the invention of the cotton gin in 1793) resulted in a sharply rising demand, particularly in England, for cotton, a crop for which the conditions of the Deep South were particularly well suited. Although America exported only 3,000 bales of cotton in 1790, total exports rose to 178,000 bales by 1810 and surpassed 4 million bales by 1860. The cultivation of cotton initially was restricted to South Carolina and Georgia but quickly expanded into newly settled states, such as Arkansas, Florida, Texas, and, most prominently, Alabama, Mississippi, and Louisiana (which together grew more than half of the nation’s cotton by 1834). The highly labor-intensive character of cotton cultivation generated an insatiable demand for labor. The steep growth of the cotton industry and its rapid expansion westward through the Deep South correspondingly help to account for both for the spread of slavery throughout the states of the South, especially the Deep South, and the remarkable increase in the numbers of slaves working in these states. The distinctive character of the intensive slavery that emerged with the opening of the Deep Southwest suggests a final subdistinction that is useful to keep in mind. The Deep Southeast continued to mix limited cotton cultivation with traditional rice (and indigo) cultivation, whereas the Deep Southwest concentrated on intensive cotton cultivation. The resulting demand for labor in the Deep Southwest generated high prices for slaves and resulted in enormous sales of slaves “down the river”—farther into the West and deeper into the South. Although statistics are not precise, scholars estimate that over a million slaves were sent westward between 1790 and 1860—perhaps up to twice as many as made the transatlantic passage. The danger of being sold down the river into the “new” South represented a genuine horror for slaves, not only because of the trauma of adjusting to unknown owners and the separation from family that was often implied (usually permanently) but also because of the rumors they heard of a harsher and more brutal slavery awaiting them in the West. The rumors were not wrong. It was generally better to be a slave in the Southeast than in the Southwest (although rice cultivation in the Southeast was probably worse than tobacco cultivation in the Upper South or the kind of slavery that had developed in the North). This state of affairs evolved for Table 1 Slave population and distribution, 1790 and 1860 1790 1860 United States 697,897 (17.8%)* 3,953,760 (12.6%) North 40,370 (2.1%) 64† (0.0%) Regional share 5.8% 0.0% South 657,527 (33.5%) 3,953,696 (32.1%) Regional share 94.2% 100.0% Upper South 521,169 (32.0%) 1,530,229 (22.1%) Regional share 74.7% 38.7% Deep South 136,358 (41.1%) 2,423,467 (44.8%) Regional share 19.5% 61.3% Upper South by state Delaware 8,887 (15.0%) 1,798 (1.6%) Maryland 103,036 (32.2%) 87,189 (12.7%) District of Columbia 3,185 (4.2%) Virginia 293,427 (39.2%) 490,865 (30.7%) North Carolina 100,572 (25.5%) 331,059 (33.4%) Kentucky 11,830 (16.2%) 225,483 (19.5%) Missouri 114,931 (9.7%) Tennessee 3,417 (9.5%) 275,719 (24.8%) Deep South by state South Carolina 107,094 (43.0%) 402,406 (57.2%) Georgia 462,198 (43.7%) Florida 61,745 (44.0%) Arkansas 111,115 (25.5%) Alabama 435,080 (45.1%) Louisiana 16,544 (51.6%)‡ 331,726 (46.9%) Mississippi 436,631 (55.2%) Texas 182,566 (30.2%)