The following is an excerpt from Jubilee: The Emergence of African-American Culture by the Schomburg Center for Research in Black Culture of the New York Public Library (National Geographic Books,2003). Order it here. "Is life so dear or peace so sweet as to be purchased at the price of chains and slavery? Forbid it, Almighty God! I know not what course others may take, but as for me, give me liberty, or give me death!"—Patrick Henry, Speech in the Virginia Convention, March, 1775. African peoples were captured and transported to the Americas to work. Most European colonial economies in the Americas from the 16th through the 19th century were dependent on enslaved African labor for their survival. According to European colonial officials, the abundant land they had "discovered" in the Americas was useless without sufficient labor to exploit it. Slavery systems of labor exploitation were preferred, but neither European nor Native American sources proved adequate to the task. The trans-Saharan slave trade had long supplied enslaved African labor to work on sugar plantations in the Mediterranean alongside white slaves from Russia and the Balkans. This same trade also sent as many as 10,000 slaves a year to serve owners in North Africa, the Middle East, and the Iberian Peninsula. Having proved themselves competent workers in Europe and on nascent sugar plantations on the Madeira and Canary Islands off the coast of Africa, enslaved Africans became the labor force of choice in the Western Hemisphere—so much so that they became the overwhelming majority of the colonial populations of the Americas. Of the 6.5 million immigrants who survived the crossing of the Atlantic and settled in the Western Hemisphere between 1492 and 1776, only 1 million were Europeans. The remaining 5.5 million were African. An average of 80 percent of these enslaved Africans—men, women, and children—were employed, mostly as field-workers. Women as well as children worked in some capacity. Only very young children (under six), the elderly, the sick, and the infirm escaped the day-to-day work routine. More than half of the enslaved African captives in the Americas were employed on sugar plantations. Sugar developed into the leading slave-produced commodity in the Americas. During the 16th and 17th centuries, Brazil dominated the production of sugarcane. One of the earliest large-scale manufacturing industries was established to convert the juice from the sugarcane into sugar, molasses, and eventually rum, the alcoholic beverage of choice of the triangular trade. Ironically, the profits made from the sale of these goods in Europe, as well as the trade in these commodities in Africa, were used to purchase more slaves. During the 18th century, Saint Domingue (Haiti) surpassed Brazil as the leading sugar-producing colony. The number of slaves brought to the tiny island of Haiti equaled more than twice the number imported into the United States. The vast majority came during the 18th century to work in the expanding sugar plantation economy. The Haitian Revolution abolished slavery there and led to the establishment of the first black republic in the Americas. It also ended Haiti's dominance of world sugar production. Cuba assumed this position during the 19th century, and even after slavery was abolished there in 1886, sugar remained the foundation of its economy and its primary export commodity throughout the 20th century. Sugar was also produced by slave labor in the other Caribbean islands as well as in Louisiana in the United States. During the colonial period in the United States, tobacco was the dominant slave-produced commodity. Concentrated in Virginia and Maryland, tobacco plantations utilized the largest percentage of enslaved Africans imported into the United States prior to the American Revolution. Rice and indigo plantations in South Carolina also employed enslaved African labor. The American Revolution cost Virginia and Maryland their principal European tobacco markets, and for a brief period of time after the Revolution, the future of slavery in the United States was in jeopardy. Most of the northern states abolished it, and even Virginia debated abolition in the Virginia Assembly. The invention of the cotton gin in 1793 gave slavery a new life in the United States. Between 1800 and 1860, slave-produced cotton expanded from South Carolina and Georgia to newly colonized lands west of the Mississippi. This shift of the slave economy from the upper South (Virginia and Maryland) to the lower South was accompanied by a comparable shift of the enslaved African population to the lower South and West. After the abolition of the slave trade in 1808, the principal source of the expansion of slavery into the lower South was the domestic slave trade from the upper South. By 1850, 1.8 million of the 2.5 million enslaved Africans employed in agriculture in the United States were working on cotton plantations. The vast majority of enslaved Africans employed in plantation agriculture were field hands. Even on plantations, however, they worked in other capacities. Some were domestics and worked as butlers, waiters, maids, seamstresses, and launderers. Others were assigned as carriage drivers, hostlers, and stable boys. Artisans—carpenters, stonemasons, blacksmiths, millers, coopers, spinners, and weavers—were also employed as part of plantation labor forces. Enslaved Africans also worked in urban areas. Upward of ten percent of the enslaved African population in the United States lived in cities. Charleston, Richmond, Savannah, Mobile, New York, Philadelphia, and New Orleans all had sizable slave populations. In the southern cities they totaled approximately a third of the population. The range of slave occupations in cities was vast. Domestic servants dominated, but there were carpenters, fishermen, coopers, draymen, sailors, masons, bricklayers, blacksmiths, bakers, tailors, peddlers, painters, and porters. Although most worked directly for their owners, others were hired out to work as skilled laborers on plantations, on public works projects, and in industrial enterprises. A small percentage hired themselves out and paid their owners a percentage of their earnings. Each plantation economy was part of a larger national and international political economy. The cotton plantation economy, for instance, is generally seen as part of the regional economy of the American South. By the 1830s, "cotton was king" indeed in the South. It was also king in the United States, which was competing for economic leadership in the global political economy. Plantation-grown cotton was the foundation of the antebellum southern economy. But the American financial and shipping industries were also dependent on slave-produced cotton. So was the British textile industry. Cotton was not shipped directly to Europe from the South. Rather, it was shipped to New York and then transshipped to England and other centers of cotton manufacturing in the United States and Europe. As the cotton plantation economy expanded throughout the southern region, banks and financial houses in New York supplied the loan capital and/or investment capital to purchase land and slaves. Recruited as an inexpensive source of labor, enslaved Africans in the United States also became important economic and political capital in the American political economy. Enslaved Africans were legally a form of property—a commodity. Individually and collectively, they were frequently used as collateral in all kinds of business transactions. They were also traded for other kinds of goods and services. The value of the investments slaveholders held in their slaves was often used to secure loans to purchase additional land or slaves. Slaves were also used to pay off outstanding debts. When calculating the value of estates, the estimated value of each slave was included. This became the source of tax revenue for local and state governments. Taxes were also levied on slave transactions. Politically, the U.S. Constitution incorporated a feature that made enslaved Africans political capital—to the benefit of southern states. The so-called three-fifths compromise allowed the southern states to count their slaves as three-fifths of a person for purposes of calculating states' representation in the U.S. Congress. Thus the balance of power between slaveholding and non-slaveholding states turned, in part, on the three-fifths presence of enslaved Africans in the census. Slaveholders were taxed on the same three-fifths principle, and no taxes paid on slaves supported the national treasury. In sum, the slavery system in the United States was a national system that touched the very core of its economic and political life.
ABSTRACT What was the rationale of exploiting slave laborers instead of employing non-slave laborers? The question has given rise to a long-standing debate among historians. One key controversy in this debate is whether it was economically cheaper to obtain labor from slaves than from non-slaves. This article contributes with the first quantitative estimates comparing the cost of slave and non-slave labor, empirically studying the critical case of the antebellum Southern United States. The findings suggest that the cost of obtaining slave labor was much lower than the cost of obtaining non-slave laborers in this case, and that the difference was large enough to have had important consequences for the production involved, primarily of cotton. Economists and historians have for decades discussed the rationale of the institution of slavery. The institution was multi-faceted and there were several factors – including social, political, cultural and religious – that shaped the institution over time. Most scholars probably agree that economic motives also was one fundamental factor behind the existence of slavery in most slave societies. In many historical societies, slavery co-existed with other labor relations. Under such circumstances, slavery would not have persisted for so long time, and despite resistance from the enslaved, had it not been profitable for the masters. At issue has however been exactly what the economic rationale was, i.e. why slavery was profitable for the masters. One long-standing debate has been particularly concerned with the productivity of slaves. Slavery made economic sense to the masters, an argument goes, as the slaves could be made to work longer or harder under coercion than free laborers or small-farmers would do. Whether productivity really was higher on slave-plantations than on non-slave-holding farms has since been the subject of a heated scholarly debate. Another line of thought is instead that slavery made economic sense to masters because slavery constituted a cheap labor force. This argument has received a lot of attention in recent years, not the least in the wake of new research in the field of the ‘New History of Capitalism’. While there are many scholars who have alluded to this rationale, there are also several who have heatedly contested this claim, too. The debate has to a large extent been marred by vagueness about what is meant by labor being cheap. This article contributes to this debate by comparing the wage costs for non-slave laborers to the cost of hiring slave laborers in the antebellum U.S. South. The article draws on evidence from four large samples of data on slave-hires in the antebellum South in order to triangulate between different possible biases in the samples. The empirical evidence presented suggests that slave labor was, indeed, acquired at a substantially lower cost than free labor in the antebellum United States, despite the purchase price of slaves rising dramatically during the period. The costs of slave and free labor do not seem to converge during the period under study, as theoretical models of price convergence on factor markets would assume. The substantial and lingering differences can potentially be explained by racial prejudices among the master class, putting a too low value on the work the slaves performed. The relatively low cost for enslaved labor could in turn have had great consequences for industries relying upon slave-produced commodities as inputs, most importantly the emerging cotton textile industries in Europe and the U.S. The role of slavery for the development of the modern world, and the development of capitalism in particular, has in recent years attracted a new wave of scholarship (Baptist, 2014; Beckert, 2014; Beckert & Rockman, 2016; Clegg, 2020; Harvey, 2019; Johnson, 2013; Schermerhorn, 2015; Stelzner, 2020). Much of the earlier economic-historical research into the economics of slavery was devoted to more specific issues, such as the profitability of slavery (Aitken, 1971; Conrad & Meyer, 1958; Evans, 1962; Foust & Swan, 1970). The scholarship later became more preoccupied with the issue of whether slaves were more (or less) productive than free laborerers were. Theoretically, that debate goes back at least to Adam Smith, who argued that slaves always were less productive than free laborers, because the slaves had few incentives to actually work hard (Engerman, 2007, p. 83). Today, it seems as if no scholar believes that Smith got this right. The cliometric revolution, including the highly debated Time on the Cross by Stanley Engerman and Robert Fogel, lead to a protracted debate on the issue, as the results from these studies of antebellum United States suggested that slave labor there and then actually was more productive than free labor (Fogel, 1989, Chapter 3, 2006; Fogel & Engerman, 1974b). Several scholars justifiably questioned many of the estimates and claims made by Fogel and Engerman (see for example David & Stampp, 1976; G. Wright, 1978). Nowadays most leading scholars of slavery do, however, seem to accept that slave labor often might have been more productive than free labor, for various reasons (Davis, 2006, p. 181; Eltis, 2000, p. 18; Smith, 1998, Chapter 5). There are, nonetheless, some scholars who still argue that the evidence on the productivity is inconclusive (G. Wright, 2006, pp. 23–26, 94–99, 2020, pp. 358–359). Compared to the debates on the productivity of slaves, less research has been devoted to the issue of the cost of acquiring slaves. There is certainly research into the prices of slaves in various historical contexts (Eltis et al., 2005; Kotlikoff, 1979; Sutch, 2006). But no scholar has attempted to systematically compare the cost of acquiring slave versus free labor. Many historians have, nonetheless, claimed that slaves were cheap laborers (Baptist, 2014, p. 318; Beckert, 2014, p. 108; Blackmon, 2009, p. 40; Genovese & Fox-Genovese, 1979, p. 12; L. C. L. C. Gray, 1941, p. 464; Huston, 2003, p. 10; Merritt, 2017, p. 99; Stampp, 1956, p. 400). These authors do not always make it explicit which laborers the slaves supposedly were cheaper than, but implicitly seem to be comparing slave labor to free labor. Virtually none of them provide any direct empirical evidence to support their claims that the slaves really constituted a cheaper labor force, but rather make the claim as something more or less self-evidently true. There are only a few exceptions of scholars who at least have provided some evidence on the topic (Bonacich, 1975, pp. 603–605; L. L. Gray, 1971, pp. 72–73; Starobin, 1970, pp. 141–143; Takagi, 1999, pp. 80–81). Yet, several scholars in the field of economic history have vigorously opposed the idea that slaves were cheap laborers. That slaves were paid little for their hard labor is beyond doubt. What positive rewards most slaves could hope for, beyond subsistence costs, was at best some small wage or token bonus (Rönnbäck, 2016, p. 2). This does, however, not necessarily mean that slave labor was cheap to acquire from the perspective of the master. The perhaps most important argument, made for a long time by several scholars (most recently for example by Gavin Wright), is that slave labor was not necessarily cheap, as the price to be paid for purchasing a slave in the first place might have been very high. Even though the slaves were paid no (or only very low) wages in addition to subsistence costs, it could then still be very expensive for a master to invest capital in purchasing a slave (Woodman, 1963, pp. 311–315; G. Wright, 2006, pp. 71–72). The cost of slave labor was therefore not necessarily low. None of these authors present any empirical evidence directly comparing what the actual labor costs of slave and free labor would have been to a master. A second line of argument against the idea that slave labor was cheap is that slave labor in one particular part of the world was not cheap labor as the cost of acquiring labor elsewhere in the world could be even lower. Alan Olmstead and Paul Rhode have thus shown some empirical evidence suggesting that slave labor was not cheap in antebellum United States, compared to the cost of recruiting (non-slave) agricultural laborers in India (Olmstead & Rhode, 2018, p. 5; but see also Rönnbäck & Theodoridis, in press, for a critique). Some scholars have argued that slavery arose in the Americas since wage labor was ‘impossible’ there (see for example G. Wright, 1978, p. 43). As is well-established from previous research, the early settlement of the American colonies was to a large extent undertaken by indentured servants who had been recruited in Europe. The servants were offered a free voyage across the Atlantic in exchange for then serving a number of years as a laborer for the holder of the contract. The contract that European indentured laborers agreed to when recruited to work in early modern America included a form of compensating wage differential, as the contract periods were shorter for those willing to serve in the more risky Caribbean, rather than on the American mainland (Galenson, 1978, pp. 62–63, 1981a, , 1981b, table 7.1). Despite such compensating differentials, there is evidence that there was adverse selection among indentured servants going to the regions associated with greater disamenities (Abramitzky & Braggion, 2006). The fact that it many times was hard for landholders to enforce indentured servants’ contracts, as many of the servants for example attempted to run away, made the institution costly (Lebergott, 1964, p. 23). Download CSVDisplay Table The switch from indentured servants to slaves was in all probability not because it became completely impossible to recruit indentured servants. It was, after all, possible to recruit Europeans to work under far more dangerous conditions elsewhere in the world. One example is the recruitment of laborers to work for the European slave-trading companies in the so-called ‘White Man’s Grave’ (in West Africa) during this period. The risk of dying in this context was extremely high. Nonetheless, the European companies managed to recruit laborers of various ranks. In order to do so, they were, however, forced not only to pay the laborers a wage in addition to board and lodging, but also to offer a monetary wage premium (i.e. a compensating wage differential) above what the recruited laborers could have earned in Britain (Rönnbäck et al., 2019). It would thus not have been impossible to continue recruiting indentured servants for the American colonies, but slaves were in comparison considered cheaper (Eltis, 2000, p. 23). They were initially comparatively cheap to obtain in many places in Africa (Eltis, 2000, p. 14; Richardson, 1991). They were also in all probability less costly to transport across the Atlantic than it would have been to transport voluntary recruits, as it is well-established that the transatlantic slave trade suffered from extreme crowding on board many of the slave ships (Duquette, 2014; Solar & Duquette, 2017). As David Eltis has argued, indentured servants would hardly have accepted such crowding. It would thus have been considerably more costly to transport as large a labor force across the Atlantic on terms that the laborers would have agreed to voluntarily (Eltis, 2000, pp. 14–15). That slaves were a cheaper supply of labor than non-slaves was something many early colonists believed. One early colonist did, for example, in 1645 argue that slaves were necessary not because it was impossible to recruit free laborers, but because these servants ‘will still desire freedome to plant for them-selves, and not stay but for verie great wages’ (quoted in Walsh, 2011, p. 408). Other early colonists maintained the exact same idea (Jordan, 1968, pp. 69, 130). The Dutch West India Company, likewise, considered slaves to be the cheapest source of labor for its American possessions (Davis, 2006, p. 127). The European colonizers were thus well aware of the fact that European servants certainly could be enticed to work for a wage – but that the wages they would require then also would be very high. Indeed, recent comparative research has shown that wage levels paid to free laborers in labor-scarce colonial America – and particularly in North America – possibly were the highest in the world at that time (Allen et al., 2012). Contrast this to the case of early modern Asia, for example, where comparative research has suggested that the abundance of labor in the most densely populated countries had pushed down wage levels for unskilled laborers almost to subsistence levels (Allen et al., 2011). Whereas it certainly is clear that free laborers were scarce in the antebellum southern United States (Bailey, 1994, p. 38), it is an exaggeration to claim – as some previous scholars have done (e.g. G. Wright, 1976, p. 321) – that there were no wage laborers at all that could be hired. Poor whites might have comprised around one-third of the white population in antebellum South (Merritt, 2017, p. 3), and many of them would most probably have been willing to work for a wage, had the wage offered met their reservation wage. An additional crucial feature of slavery is that the slave is fixed capital to the owner. A slave-owner would therefore face the challenge of keeping all hands busy all the time in order to maximize profits (Anderson & Gallman, 1977). But the eighteenth and nineteenth century saw the growth of slave hires, whereby an owner of a slave could hire out the slave’s labor to someone looking for labor to hire. As Martin has put it, ‘Owners earned money on slaves they could not use, and hirers acquired needed labor that they could not afford to purchase in the slave pens’ (Martin, 2004, p. 3; see also Pargas, 2015, p. 37). A typical contract, notes Martin, was year-long, with particular hiring days regularly every year, but slaves could also be hired by the day, week or month. There were also rules and regulations as to how the hires slaves were to be treated, and further aspects of the treatment of the slaves were often specified in the hiring contracts (see also Evans, 1962, pp. 192–195; Martin, 2004, p. 6). For those hiring the slaves, this was a system that could allow for a much greater degree of flexibility in comparison with having to purchase a slave (Martin, 2004, p. 107). This was particularly important in the case of short-term hires of slave, for example during the season of peak labor demand in agriculture (Eaton, 1960, p. 676; Martin, 2004, p. 30). I will for simplicity henceforth use the term ‘master’ both for slaveholders and for employers of free laborers. The term ‘free labor’ will be used as a convenient shorthand for legally free (i.e. non-slave) labor. The term ‘cheap’ is here used in the meaning of ‘bearing a relatively low price’.1 Key to this definition is the word ‘relatively’. Something cannot be cheap on its own; it is cheap relative to some (relevant) alternative. The basic comparison used in much previous research has been between slaves and independent farmers. That comparison can be misleading. Working on a family farm does not entail the same incentive problems as working for someone else. If we are to understand the economic rationale of slavery, the most relevant comparison is therefore not between slaves and independent farmers, but between slaves and hired laborers. But we can, furthermore, not just compare slave and free labor in the abstract. Laborers have historically been paid very different wages at different geographical locations, for different tasks and at different points in time. The fundamental proposition of this article is therefore that slaves could constitute a cheap source of labor relative to what it would have cost to attract free laborers to undertake the very same task. The proposition here is to analyze this through what in labor economics is called the reservation wage of labor (compare with van der Linden, 2008, pp. 65–72). The reservation wage is the minimum wage required by an individual (who is able to choose) to voluntarily accept a job (e.g. Borjas, 2016, pp. 39–42; Mortensen, 2018; Steedman, 2018). The choice stands between the alternative economic opportunities available. A number of different factors determine the opportunities available, and thereby the reservation wage. One of the important factors of those outside opportunities in pre-industrial societies involve the availability of land in a specific geographical location. One aspect of this is the land/labor-ratio, but the social relations in place also matter: if all land is formally appropriated by a small elite, the land available to a particular landless individual might still be low even in a sparsely populated region. A high effective land/labor-ratio would, all else equal, lead to a high reservation wage when it comes to accepting to work for any particular master. Many scholars will undoubtedly recognize the logic of this aspect as the Nieboer-Domar hypothesis (Domar, 1970; Nieboer, 1910). In some contexts, Malthusian pressures might already have reduced the reservation wages of large groups of the population to the bare, physical minimum – the cost of the subsistence needs – as Evsey Domar himself pointed out (Domar, 1970, p. 23), making coercive institutions such as slavery rather superfluous. A second factor of the model is that reservation wages also differ between different kinds of work. Another key theory in labor economics is the theory of compensating wage differentials for job disamenities (Borjas, 2016, pp. 196–203, 213–214; Rosen, 1986). Some work is not very pleasant to undertake: it can be dangerous, dirty or degrading. The theory of compensating differentials suggests that in order to attract individuals to voluntarily undertake such work, the master would have to offer a wage premium above what the individual can earn from alternative opportunities. The reservation wage for accepting work characterized by such disamenities is, thus, higher than for work without such disamenities. Slaves could, in contrast, be coerced to undertake whatever tasks were required, regardless of such disamenities (G. Wright, 2006, pp. 14–15, 20–21; R. E. R. E. Wright, 2017, pp. 124–125). Many laborers can also be reluctant to move since they might have social ties (family, friends) that bind them to particular locations. One key aspect of slavery suggested by Orlando Patterson is the ‘social death’ of slaves, whereby the institution of slavery effectively eliminated many of the social ties of the individual slaves (Patterson, 1982). Slaves could thereby be coerced to move geographically to wherever labor was demanded without taking such social relations into any deeper consideration (Engerman, 2007, p. 19; Pargas, 2015; G. Wright, 1986, p. 17, 2006, pp. 67–68). This framework thus suggests that the higher the reservation wages paid to non-slaves laborers for a particular job are above subsistence costs – for example because of plentiful outside options (such as a high effective land/labor-ratio) or because of a demand for compensating wage differentials for job disamenities – the higher can the price of slaves get with slaves still being the most economical option to the master acquiring labor. The agency of both the slaves and non-slave laborers can also play a role in this framework, as any resistance against slavery or workplace struggles to improve working conditions for non-slaves could be interpreted as influencing the costs of these respective types of labor. The suggested framework is also fully compatible with alternative theories of slavery, such as the suggestion that slavery might have led to lower turnover costs (Hanes, 1996), as well as with models trying to explain the choice between using the whip (the physical torture of slaves, but in many cases also of non-slaves) or the carrots that could be offered, to get people to work (Acemoglu & Wolitzky, 2011; Chwe, 1990; Fenoaltea, 1984). The reasons why slavery arose in a particular place might, as Stanley Engerman has argued (Engerman, 1973), however, not be the same reason why slavery persisted there over time. If slaves indeed were cheap to acquire at some point in time, that does not necessarily mean that slaves remained cheap labor over time. A basic expectation from economic theory is that prices for identical commodities or for perfect substitutes converge on markets, including on factor markets. This is the theory of the Law of One Price. If slave and free labor were (believed to be) substitutes by masters wanting to acquire labor, and markets were functioning well, the cost of slave and free labor ought to converge over the long run, as masters would seek the cheapest source of labor. They would thus not continue to pay more for a free laborer if a slave could do the same work at a lower cost, and vice versa. If slave labor then could be acquired at a lower cost than free labor in a specific historical setting could thus indicate several different possibilities: that slave and free labor in reality were not good substitutes for each other, that the masters did not believe that they were good substitutes (even if they in reality were), or that the process of convergence took time and had not come to its conclusion yet. There is very little empirical evidence that would allow for a comparative quantitative estimate of the cost of non-slave and slave labor in the Americas during the early modern era. Such data is, however, available from antebellum United States. Much research has been undertaken into the wages paid historically to free laborers in various parts of the United States, primarily from the nineteenth century onwards (Coelho & Shepherd, 1976, 1979; Margo, 2000, 2006; Margo & Villaflor, 1987). Robert Margo’s research has led to the, to date, largest dataset of wage observations from the antebellum South, which is the region of interest here. The data was assembled from primary sources on wages paid to civilian employees at U.S. forts around the country (Margo, 2000, pp. 25–29). The dataset contains more than 15,000 observations of wages paid to unskilled laborers in United States. The sample pertaining to states in the U.S. South is to a large extent dominated by observations from Florida, Virginia, Louisiana and Arkansas (Margo, 2000, table 2.2). The data refers to common laborers, rather than farm laborers. For this article, the contracts for daily and monthly wages, respectively, paid to male, non-slave, unskilled common laborers (excluding observations where the laborer was provided with rations in addition to the wages paid) have been used for the comparisons. The duration of the contracts were not necessarily only a single month, but could in some cases be longer. In such cases, each month of a longer contract has been treated as a single person-month observation (Margo, 1992, p. 181, 2000, p. 38). As for the cost of acquiring slave labor, there is much quantitative data on the purchasing prices of slaves, and how they developed over time in antebellum United States (assembled in Sutch, 2006). Most important for this study, however, are historical data on slave hire rates, as it is possible to compare these directly with the cost of hiring free laborers.2 One key strength of comparisons based on these direct observations of slave hire costs is that they would not have to rely upon estimates or assumptions, for example about the discount rate for the masters’ investments in slaves or on the fertility and mortality rates among the slaves. The masters would presumably have incorporated such aspects into the asking price when hiring out slaves. Recent scholars have argued that slave hires were very common throughout the Southern economy, at least during the nineteenth century (e.g. Martin, 2004). Quantitative data on slave hire rates have been collected independently in four different databases by Robert Evans, by Robert Fogel and Stanley Engerman, by Claudia Goldin, and by Robert Margo, respectively. The data on slave-hires in all these datasets have only rarely been used in previous research, and the data has in particular never been used for a comparison with the cost of acquiring free labor in the United States. Robert Evans’ sample of slave hire contracts is made up of 5,768 observations of year-long contracts, the overwhelming majority (94 per cent) from the Upper South, and only a smaller sample from the Lower South (Evans, 1962, tables 29–42). Evans also report data on smaller samples of monthly slave hires (in total 883 observations, 61 per cent of which are from the Upper South), as well as data on daily slave hires (the sizes of these samples are unfortunately not reported). Evans only included male slaves in the dataset. The majority of the observations come from institutions such as railroad companies, hiring slave laborers from the owners. Evans’ dataset does not include information on the occupation of the slaves hired. The vast majority of these slaves, Evans argued, were most probably unskilled laborers, but he was not always able to determine this for certain. In his dataset, Evans eventually included data on slave hires when ‘the source indicated that it probably represented a healthy adult male performing relatively unskilled labor’ (Evans, 1962, p. 196). Robert Fogel and Stanley Engerman’s dataset is made up of 20,253 observations of slave hires, primarily from rural areas of the South; 61 per cent from the Upper South and 39 per cent from the Lower South (Fogel & Engerman, 1974a, 1976). 62 per cent of the slaves hired out were male, and 36 per cent female (the remainder of unknown gender). The data was collected from probate records. The vast majority (89 per cent) of the observations are for yearly hires, the remainder generally being for shorter terms (from one to eleven months) and a tiny fraction for contracts longer than one year. The cost of hiring the slave was not reported in all cases. The dataset includes variables for the age and occupation of the slaves hired, but data is missing on these variables for the vast majority of the observations (97.2 and 99.8 per cent, respectively). The dataset also includes a variable with information on whether the slaves had a ‘defect’ or were handicapped in some way (e.g. aged, crippled or sick). The vast majority (87 per cent) were not reported to have any such ‘defects’, and were thus presumably able-bodied adults. Claudia Goldin’s dataset of slave hires is based on hiring data from cities in Virginia for five benchmark periods from 1820 to 1860. The dataset is comparatively smaller than the three other datasets (381 observations, 65 per cent male and 35 per cent female). Robert Margo’s dataset on laborers at U.S. forts, finally, contain a number of observations based on contracts for hiring slaves to work at the same forts. All slaves hired by these forts were male. In total, the sample includes 5,006 observations of slave-hires, out of which 66 per cent were for monthly slave hires and 34 per cent for daily slave hires. The vast majority of the observations (86 per cent of the sample) come from the Lower South. This is particularly the case for monthly slave hires, where 98 per cent of the observations are from the Lower South. Margo’s data also contain information on the occupational categories of all the slaves hired. The largest group (48 per cent of the sample) were simply categorized as ‘laborer’. Other important occupations among the slaves hired by the forts included carpenters and teamsters. The terms of slave hires were often quite standardized. The person hiring a slave for a full year would as a rule also have had to pay for the subsistence cost of the slave, including food, housing and medical bills. For shorter hires – daily or weekly – the owner of the slave as a rule seem to have paid for the subsistence costs (Evans, 1962, pp. 194–196). In some of the observations in Margo’s dataset (30 per cent of the daily hire contracts, and 11 per cent of the monthly hire contracts), the sources report that the slaves were provided with some rations, in addition to the hire costs. The samples either only contain observations for male slaves or allow for selecting slaves of a particular gender, so that the cost of hiring a male free laborer can be compared to the cost of hiring a male slave. All time-series of data on slave hires are unbalanced. Evans’ observations are highly concentrated to the years 1852–1860: 97 per cent of all observations in this dataset date from any of these years (Butlin, 1971, pp. 90–93). Margo’s data is in contrast highly concentrated to the period 1835–1842, with 92 per cent of all observations dating from this period. Fogel and Engerman’s dataset is not only the largest in absolute terms, but also the least unbalanced, with comparatively fair annual samples (of around 20–30 observations per year) regularly from the 1810s, and large samples (of more than 100 observations per year) for virtually every year from the late 1830s onwards. It is in addition important whether there might be any biases in the samples of slave hire rates. This comes down to two different issues: whether the hire rates in the dataset are representative of average slave hire rates, and whether the slaves hired out are representative of the general slave population. The hire rates might in particular suffer from two issues: skill premia and risk premia.
In essence, we should expect that the cost of hiring an able-bodied, adult male slave of average productivity to work in the rural South might have been lower than what is suggested by the datasets by Evans, Margo and Goldin, but possibly higher than what is suggested by the dataset by Fogel and Engerman. As we expect different biases depending on the sources employed by the different scholars, the different samples of data on slave hires will – for reasons of transparency – not be merged into a single dataset in the following analysis, but will be reported as separate time-series. The data on the hiring of free laborers is, as noted above, based on evidence on wages for common laborers working at the U.S. forts, but Margo argues that the wage gap between unskilled farm and non-farm laborers were small at this time (Margo, 2000, p. 92). If so, these wages can be used as a reasonable proxy for wages paid to free farm laborers in the antebellum South. The wages paid to the workers and the hire rates paid to the owners of a slave would not be all costs associated with acquiring free or enslaved labor, respectively. There were for example recruitment costs for acquiring free laborers (see for example Fleisig, 1976). There could, at the same time, be costs associated with the process of hiring slave labor, such as brokers’ fees (Butlin, 1971, p. 102). It is here assumed that these costs would not greatly bias the comparison between the costs free and enslaved laborers to be undertaken here. Finally, a note on the geographical aggregation of the data. Evans’ data is reported only for two broader regions: the Upper and Lower South, respectively. The data in the other datasets are reported by state. Wages and slave hire costs seem to differ somewhat between the states within these regions, but the very patchy data from most states prohibits drawing very strong conclusions. For the comparison, the data has here been aggregated by broad region. The Lower South is then taken to include the states of Alabama, Florida, Georgia, Louisiana, Mississippi, North Carolina and South Carolina. The Upper South is here taken to include Arkansas, Kentucky, Maryland, Oklahoma, Tennessee and Virginia. The price-trend for hiring a slave follows the price-trend for purchasing slaves quite well, as is shown in Figure 1. The data shows that the cost of purchasing a slave (measured on the right-hand axis) certainly increased substantially in nominal terms over the period from 1830 to 1860 – from around $600 to around $1600. This would undoubtedly have been a substantial investment that comparatively few people could have afforded at the time. The cost of hiring slaves did also increase in a similar manner, and the evidence suggests that the prices on the hire market for slaves moved in very similar patterns, with prices for example falling during the economic depression following the panic of 1837; similar to the sales market for slave (but potentially with a slight lag). The fact that both slaves and free laborers could be hired in the South enables us to compare the direct cost of slave and free labor. This is shown in Figure 2, based on the data on the cost of hiring unskilled slave laborers by the month (in the Lower South only), and in Figure 3, based on the data for yearly slave-hires in both the Upper and Lower South. Figure 2 reports data on monthly hires of free and enslaved laborers in the Lower South (see appendix for the equivalent from the Upper South, and for daily hiring rates). The samples are many years comparatively small, and there are large fluctuations between the years. It is noteworthy that the slave hire rates in the different samples differ as expected given the possible biases discussed above. The rural data from probate records assembled by Fogel and Engerman thus suggests very much lower hiring costs than the two other datasets on slave hires. This data suggests that the cost of hiring a slave quite consistently was lower than the cost of hiring a free laborer in the Lower South. Even if we compare the costs of hiring slaves estimated in the two other datasets – by Margo and Evans, respectively – the cost of hiring a slave was are in many years lower than the cost of hiring free laborers.3 All these samples of monthly data on slave-hires must, however, be interpreted with great caution, as the evidence is patchy, with small samples overall and a highly unbalanced series of data. The scattered evidence on monthly slave hires (including self-hires) that can be found in some previous research are normally in the range of $10-20 per month for male slaves, with only the exceptional contracts (e.g. hiring a slave to work on a steamboat, where the slave more easily could run away) exhibiting hire rates higher than $20 per month.4 The range of rates from this anecdotal evidence in previous research does thereby fall roughly in between the data-series on slave hires reported in Figure 2. The samples of data on annual slave-hires are substantially larger than the samples of monthly data, and are for that reason potentially more reliable. This data is presented in Figures 3(a-b), comparing data on the annual hiring cost for a slave (including subsistence and surveillance costs of the slave that the person hiring the slave had to pay during long-term hires) to Margo’s data on the cost of hiring free laborers for 12 months in the Southern United States. Again, the hire rates in the different samples differ as we would expect given the biases the respective datasets might suffer from. The anecdotal data on slave hire rates that can be found in previous research vary much, but are generally in the range of $75–120 per year during the early part of the period under study, rising to rates in the range of $100–200 per year during the later decades, and only in exceptional cases seem to have been above $200 per year.5 The anecdotal data from this previous research does therefore fit closest to Fogel and Engerman’s data-series from probate records at least for the early period, and somewhere in between the two data-series during the later part of the period. Comparing these slave hire rates to the cost of hiring free laborers shows even more clear-cut differences in costs than the monthly data shown in figure 2.6 In the Lower South, Evans’ dataset on slave hires by railroad companies is very small, but there is a comparably large sample of evidence from the probate records assembled by Fogel and Engerman. From the Upper South, there are comparatively large samples of data on slave hires both from railroad companies and from probate records (but unfortunately quite small samples of data on wages to free laborers). For both the Upper and Lower South, the cost of hiring a slave was as a rule substantially lower than the cost of hiring a free laborer for a full year, regardless of which dataset on slave hires we employ. Only in a few exceptional years was the cost that railroad companies paid for hiring a slave in the Lower South more or less on par with the estimated cost of employing a free laborer there, but this might be due to the very limited sample size from this region in Evans’ dataset giving rise to quite unrepresentative samples these particular years. Using the data reported in Figure 2 and Figure 3, we can make some estimates of how much higher the cost of employing a non-slave actually was on average. This is attempted in in the form of an average cost-premium for hiring non-slaves relative to the costs of hiring slaves in the Upper and Lower South, respectively, for all the years for which there is comparable data. The data in shows the labor cost premia for hiring free laborers compared to hiring slaves. It is important to note that the estimated mean premium for hiring free laborers instead of slaves is positive in all cases. Most of the estimates fall in the range of 35–75 per cent higher cost for hiring free laborers. Only one single data-series (using Margo’s data on monthly slave hires from the Lower South) exhibit a lower mean cost premium than this (+16 per cent on average). The estimates based on Fogel & Engerman’s dataset of slave hires on the other hand all exhibit considerably higher labor cost premia for free laborers (all exceeding 100 per cent), as the slave hire rates in this dataset were consistently much lower than those found in all the other datasets. Given what was discussed above about the possible biases in the respective datasets on slave hires, it seems reasonable to conclude that the cost-premium for free laborers relative to average slaves might have fallen somewhere between the estimates using Fogel and Engerman’s data on the one hand, and Evans’ data on the other. It would therefore not seem unreasonable to assume that masters on average might have had to pay around twice as much for free labor as they had to pay for enslaved laborers in the antebellum South. It would thus seem as if slave labor indeed was cheaper than free labor in the antebellum South. Two other factors might, however, theoretically contribute these results: if the slaves were less productive than the free laborers, or if the wages paid to free laborers at the U.S. forts were unrepresentative of rural wages in the Southern states. Both these explanations seem implausible as the sole explanations for the major differences shown in . The relative productivity of slaves would need to have been considerably lower than what has been suggested in previous literature, in order to overturn the empirical findings of lower labor costs for slave laborers. There is, to the knowledge of this author, no modern scholar studying slavery in the antebellum U.S. that has claimed that slaves actually were less productive than free laborers. As is well-known by now, pace Adam Smith, ideas that slaves had no incentive to work hard are simply untrue. Slaves were regularly driven to work hard by the whip, but sometimes also by various carrots. The scholarly debate following the publication of the Time on the Cross has – as was discussed at the outset of this article – therefore been over whether slaves were more productive than, or only as productive as, free labor. That the lower cost of slave labor shown above only would reflect a lower productivity of the slaves would therefore not seem a plausible explanation to the results arrived at in this study. Turning to the wages paid to free laborers, it is certainly possible that – pace what Robert Margo argues – the wages paid to common laborers working for the U.S. forts are a poor proxy for labor costs for free rural laborers, at least in the U.S. South. While previous research certainly has suggested that there were frictions on antebellum free labor markets (Rosenbloom, 2018), the markets for free laborers would, however, have had to be extremely poorly integrated in order for wage gaps of a magnitude such as shown in to occur within the group of unskilled free laborers living in the same region. The empirical evidence presented here thus shows that for a master hiring a slave in the antebellum South, the cost of the hire (including the subsistence and additional surveillance costs) was as a rule lower than the cost of employing a free laborer. Slaves were, it therefore seems, indeed a cheap source of labor in the antebellum U.S. South. These quantitative estimates thus in the main lend support to the anecdotal evidence cited by some scholars in support of the proposition that slaves indeed were cheap laborers, even in the United States by the mid-nineteenth century (Bonacich, 1975, pp. 603–605; L. L. Gray, 1971, pp. 72–73; Starobin, 1970, pp. 141–143). The evidence presented in Figure 2 and Figure 3 furthermore suggest that the costs were not converging much over time between slave and free labor. By 1860, the costs were still far from on parity. An important issue is therefore why the costs of slave and free labor did not converge more during this period. The period under study is admittedly comparatively short compared to how long slavery had dominated labor markets in the Americas, so it seems plausible that the difference in cost between slave and free labor could have been higher still at earlier points in time, and that a slow process of convergence indeed had taken place at an earlier date. General frictions on the labor markets – not the least on the free labor market – might, however, have retarded the process of convergence considerably (Rosenbloom, 2018). But the potentially most important factor as to why slaves remained cheaper in the nineteenth century might be racial prejudices among the masters. Slaves were, most certainly, on average not less productive than free laborers in the antebellum U.S., but many masters might – along with Adam Smith – have held incorrect ideas on this topic. Anecdotal evidence certainly suggests that some Southern entrepreneurs believed that slave laborers were not necessarily of any ‘lower quality’ or less productive than the free laborers (Bateman & Weiss, 1981, p. 87; Blackmon, 2009, p. 52; Merritt, 2017, p. 67; Miller, 1981; Starobin, 1970, p. 140; G. Wright, 1978, p. 46). These do, however, seem to be exceptional ideas held by a few individuals of the Southern master class. A most ubiquitous idea among the vast majority of the masters was, instead, a highly prejudiced conception of Africans and African-Americans as ‘inferior’ to people of European descent in several respects, including being indolent and ignorant (Fogel, 1989, pp. 155–162; Fredrickson, 1971, Chapter 2; e.g. Jordan, 1968, pp. 305–308; Smith, 1997, pp. 155–163). That the common master thus were so prejudiced would have led many of them to believe that the labor performed by the slaves was of low value (even when it was not), compared to how they valued non-slaves’ labor. This would have been reflected in the price they would have been willing to pay for the slaves (either when purchasing or hiring them). So what difference did it make if slavery was cheaper than free labor? Wages paid to free laborers in many parts of the South were generally low compared to the wages paid to free laborers in other parts of the United States. One of the factors contributing to this wage gap was the very institution of slavery itself. As the former slave and abolitionist Frederick Douglass once put it: ‘the white laboring man was robbed by the slave system of the just results of his labor, because he was flung into competition with a class of laborers who worked without wages’ (Douglass quoted in Huston, 2003, p. 88).7 Many modern scholars have in a similar manner argued that slavery depressed the demand for free labor, and hence potentially also lowered the equilibrium wages paid to free laborers (Baptist, 2014, p. 386; Cohn, 2008, p. 158; Glickstein, 2002, p. 154; Huston, 2003, Chapter 3; Stampp, 1956, p. 426). Keri Leigh Merritt has recently shown in great detail how this effect also created great tensions among poor whites in antebellum South (Merritt, 2017). It might also have influenced migration patterns. The number of slaves arriving in what became the United States outnumbered European immigrants by far until the nineteenth century (Eltis, 2000, table 1.1). European immigration to the United States started to rise only in the 1820s (Cohn, 2008, table 2.1). Immigrants did, however, avoid the U.S. South if they could: scholars have shown that only a tiny fraction of all immigrants arriving in antebellum times headed for any of the Southern states (Cohn, 2008, table 7.4; Ferrie, 1999, table 4.4). Much previous research has shown that nineteenth century immigration patterns to the United States to a very large extent were determined by the wages that the migrants expected to receive, and this has been shown to be a major explanation to the small number of immigrants heading for the Southern states (Dunlevy, 1980; Dunlevy & Gemery, 1978; Dunlevy & Saba, 1992; Gallaway et al., 1974; Lebergott, 1964, pp. 74–100). Yet another potentially important factor determining the patterns of immigration to the United States was the comparatively unhealthy climate in many parts of the South. Mortality rates among the white population was thus considerably higher in the South than in the North (Warren, 1997, Figure 1). Working with certain production, not the least the production of sugar, was more hazardous still because of the risk of contracting epidemic diseases prevalent in and around such plantations (Engerman, 2017, p. 156; McNeill, 2010, pp. 52–57). Many antebellum immigrants thus seem to have avoided the South, even when socio-economic factors are controlled for (Dunlevy, 1983, table 2). Enough free laborers to substitute for slaves could thus most probably only have been attracted to work in the antebellum South, had the wages to free laborers been considerably higher than what they actually were. For the millions of slaves, slavery of course meant a life of exploitation and suffering, as is well-known from much previous research. Slavery might also have retarded a more dynamic long-term development of the Southern economy, for example as the depressed labor costs might have created disincentives to increase mechanization (Hornbeck & Naidu, 2014) and limited the educational level of the slaves and their descendants (Bertocchi & Dimico, 2014). Slavery was, on the other hand, clearly fully compatible with innovation, and corresponding great increases in labor productivity, as for example the research by Alan Olmstead and Paul Rhode on biological innovation in the antebellum U.S. South has shown (Olmstead & Rhode, 2008; Rhode, 2021). The exploitation of slaves also enabled the enrichment of the Southern master elite, thus contributing to the highly unequal distribution of resources that came to persist long after slavery itself was abandoned. But the output that the slaves produced in the antebellum South – most importantly cotton – was cheaper thanks to slavery than it otherwise would have been (similar to Coclanis’ opinion in Coclanis & Engerman, 2013, p. 73). The low price of slave-produced cotton would have played a crucial role for the competitiveness of the industries dependent on these raw materials as inputs. Some scholars have argued that this possibly would have made little difference to the process of industrialization. John Clegg, for example, has argued that raw cotton only made up what he believes was a ‘small component’ of the price of the final textiles, and a ‘modest’ increase in price of the raw cotton would have had little effect (Clegg, 2015, p. 297). Clegg does not show any empirical data neither regarding how large a component raw cotton was in the final price of the textiles, nor any estimates of how much the price of raw cotton potentially would have increased, had there been no slave labor to exploit at a cheap cost. The estimates presented above do, however, suggest that the cost of labor might have been substantially higher – as was suggested in – in the absence of Southern slaves. Research by C. Knick Harley on the British textile industry – the main consumer of the Southern cotton – also suggests that the cost of raw cotton was far from a small component of the final price of the textiles. Harley’s data thus shows that the cost of raw cotton in Britain made up around 30 per cent (and in some years almost 50 per cent) of the cost of the final textiles by the late eighteenth century and early nineteenth centuries (Harley, 1998, table 5). A higher price of the raw cotton, possibly in the ranges shown in , would thus undoubtedly have translated into a substantially higher price of the final cotton textiles. A domestic industry could potentially still have developed in Britain, if nothing else supported by heavy protectionist policies. But would the British textile industry have been able to compete on international markets under such circumstances? A main competitor on the global market for textiles until well into the nineteenth century was India. Harley has shown that Indian muslins were competing with English calicos on the British market, in many years only at a price some 40–50 per cent higher than the price of British calicos (Harley, 1998, appendix table A2.1). Presumably, the Indian muslins were competitive at such a high price due to their higher quality. Would consumers still have purchased British calicos of lower quality, had they been forced to pay almost as much for the British calicos as they had to pay for the finer Indian muslins? By the nineteenth century, many abolitionists hoped or believed that slavery had become an obsolete means of acquiring labor, or at least claimed this rhetorically in order to win support for the abolitionist cause. If this would have been the case, the institution of slavery would undoubtedly have gradually diminished in importance, and eventually disappeared completely. Alas, the historical evidence does not suggest that slavery actually was obsolete. Slavery did as a rule not disappear without great political struggles. At issue, however, is exactly why slavery turned out to be economically profitable for the slave masters for so long. Much of the scholarly research into the economics of slavery has been dedicated to the issue of whether slavery was profitable since slaves were more productive than free laborers. Much less attention has been devoted to the issue of how costly it was to acquire slave labor, in comparison with free labor. While a number of scholars have more or less taken it as self-evident that slaves were cheap labor, others have taken it as equally self-evident that slaves were not cheap, since the capital cost of purchasing a slave could be high. The aim of this article has been to contribute to this debate with a systematic comparison based on quantitative empirical evidence. In the article, a theoretical proposition is made explaining why slavery indeed could be a cheap source of labor; as it could reduce labor costs for a master under conditions of high reservation wages for free laborers. High reservation wages could be caused for example by a high effective land/labor-ratio in a specific location, as suggested by the Nieboer-Domar-hypothesis. It could also be caused by compensating wage differentials for disamenities associated with specific tasks or locations of work. The higher the reservation wage was among common laborers for a particular work, the higher the price of slaves could be with slaves still being the cheapest option available for a master who needed to acquire labor. The article explored this proposition empirically using data from antebellum Southern United States. The case was chosen as it was deemed critical to the theory proposed, but also since it was this particular slave society, during this period of time, which has been the focus of most of the controversy on the topic. Many scholars have then pointed to the high and rising prices of slaves as evidence suggesting that slaves were in reality not very cheap labor there and then. No scholar has, however, undertaken a systematic, quantitative comparison of the cost of acquiring slave labor relative to the cost of acquiring free labor. This was done in this article, using large samples of quantitative data on the costs of hiring both slave and free labor in the antebellum South. Purchasing a slave certainly required a substantial capital investment in the antebellum U.S. South. But despite rising purchase prices for slaves during the nineteenth century, the cost of hiring slave labor from an owner was consistently lower than the cost of hiring free labor throughout the period. A master in the South on average might have had to pay twice as much for hiring a free laborer than what they had to pay in total for hiring a slave laborer, including both the subsistence and extra surveillance costs for the slaves. If the slaves furthermore could be coerced to work harder or longer hours than free laborers, as much previous research has suggested, this would be a further bonus adding to how relatively cheap slave labor could be. It is therefore not surprising that there were few free wage laborers in the antebellum South’s cotton fields or sugar plantations, and that so many immigrants avoided the Southern states: having to compete with such cheap laborers put a downward pressure on the wage levels for anyone voluntarily seeking work in the antebellum South. Low labor costs would have many negative socio-economic consequences – most obviously, of course, for the slaves who were robbed of the fruits of their labor, but potentially also for long-term socio-economic development of the slave economies. For industries using the products from the plantation complex as inputs, in the United States as well as in Europe, the cheap cost of the products of slave labor did on the other hand provide a very important benefit, so large that it could have been critical for making these industries internationally competitive at the time. Table 1. Cost-premium for hiring free laborers relative to slave laborers (average 1824–1860), by length of contract and region
The author would like to thank professor Robert Margo for very generously sharing his dataset on U.S. wages, and several colleagues at the unit for economic history for constructive comments on a draft version of this article. No potential conflict of interest was reported by the author(s). Notes1. Oxford English Dictionary, ”cheap”, adj a.1 [accessed 2020–05-28]. 2. For a theoretical discussion on what slave hire rates can reveal about labor markets, see (Fogel & Engerman, 1974a, pp. 54–58) 3. Margo’s own regression estimates of the determinants of wage costs in his sample of data also confirm a negative coefficient for the slave-variable, see (Margo, 2000, tables 3A.3–3A.4). 4. See (Eaton, 1960, p. 664; L. C. L. C. Gray, 1941, p. 469; Martin, 2004, pp. 35, 39, 40, 68, 79, 80, 81, 108, 124, 168, 180; Zaborney, 2012, p. 101) 5. See (Conrad & Meyer, 1958, table 15; Eaton, 1960, p. 667; L. C. L. C. Gray, 1941, p. 469; Martin, 2004, pp. 39, 40, 42, 81, 85, 102, 118; Murphy, in press, p. 22; Takagi, 1999, p. 23; Zaborney, 2012, pp. 73, 125, 142, 154, 156) 6. Saraydar has suggested a surveillance cost of around $5-10 per slave per year on plantations (Saraydar, 1964, p. 328), the higher figure has here been used. Robert Starobin has argued that the supervision cost of slaves employed in industry might have been higher than for those employed in agriculture, suggesting an average there of $27 per year (Starobin, 1970, p. 142). Masters certainly also had costs surveilling free laborers, but it has here been assumed that this was comparatively low, so that the costs estimated by Saraydar could be used as a proxy for the difference in surveillance costs between slave and free labor. 7. Karl Marx took the argument even further, when writing in Capital (volume I) that ‘In the United States of North America, every independent movement of the workers were paralysed so long as slavery disfigured a part of the Republic. Labour cannot emancipate itself in the white skin where in the black it is branded.’ (Marx, 1954, p. 284) Figure 2 in the main text showed monthly labor costs for free and enslaved labor from the Lower South. The samples of data from the Upper South are considerably smaller, but the data available in these samples are shown in Figure A1. Figure A2 turns to the daily hiring costs of slaves and free laborers, from both the Upper and the Lower South. As a rule, the cost of employing free laborers seem to be higher than the cost of hiring slave laborers, with a few exceptions some particular years, both for monthly and daily hiring rates. These data then confirm the general picture showed in the main article.
Permission is granted subject to the terms of the License under which the work was published. Permission will be required if your reuse is not covered by the terms of the License. To request a reprint or commercial or derivative permissions for this article, please click on the relevant link below. |
After slavery was abolished in the South, what type of cheap labor did plantations rely on
Related Posts
Copyright © 2024 ShotOnMac Inc.
