Monetising human life: slave valuations on US and British West Indian plantations

Introduction

Since the appearance of Fogel and Engerman's seminal econometric study, Time on the Cross (1974), in which the authors attempted to demonstrate the economic viability of the plantation system and the productive efficiency of slavery, economic historians have examined and dissected the institution of slavery from virtually every conceivable perspective. However, the contribution of accounting historians to this project has been both sparse and non-critical. Pioneering studies by Flesher and Flesher (1981), Razek (1985), Heier (1988), Barney and Flesher (1994), and Vollmers (2003) have formed a valuable base upon which to build future research. However, the breadth of these works has been limited. In some cases, the accounting historians have dealt only with the records of a single plantation or enterprise; in all instances, the authors have not raised important issues as to the morality of slavery and whether or not accounting and/or accountants were implicated in supporting racist regimes. Critical scholars have not as yet adopted the accounting for slavery as suitable for their analysis, concerned perhaps with more contemporary issues such as gender, race as it relates to entrance into the accounting profession, environmentalism/sustainability, and corporate social responsibility. If, as Laughlin (1999) wrote, the critical agenda is to change the world, or at least to try, slavery in the traditional sense has largely passed from the scene. Notwithstanding, work on the Holocaust by Funnell (1998) and Walker (2000), as well as initial efforts on plantation labour (Fleischman & Tyson, 2000, 2004; Tyson et al., 2004a, 2004b), have value for bridging accounting's past and present and in helping to understand the potential of accounting to support racist or otherwise repressive regimes.

The owners and managers of slave plantations of the antebellum US South and the pre-emancipation British West Indies (BWI) deployed accounting in many ways. Most surviving records are unexceptional and typify a highly commercialised agriculture. Substantial attention was paid to crop yields, the productivity of individual fields, and financial arrangements with factors. Far more interesting to our research are those plantation records that relate to the slave labour force. These documents include lists of slaves (inventories) with a variety of data categories, daily logs of the whereabouts of individual slaves, accounts of money owed to and by the slaves, distributions of clothing and tools to slaves, and a variety of records that chronicled their demographics (yearly increase/decrease, conditions of health, ages). Fleischman and Tyson (2004), in an analysis of plantation accounting in the US, identified the valuation of slaves and the tracking of their productivity as the two most important components of that accounting.

This article focuses specifically on the valuation processes of slaves as assets. In the following section, we will consider the question of why slaves were valued at all since it is not immediately evident why the exercise was a cost-beneficial undertaking for the plantation. In succeeding sections, we will review the mechanisms by which the valuations were undertaken and the factors that went into the determination of the perceived worth of an individual slave. Finally, we will consider certain moral issues of slavery, with particular reference to how valuation constituted a dehumanising of the slaves and the role accounting and accountants played in the process. Throughout these sections, comparisons between the antebellum US South and the pre-emancipation BWI will be an ongoing focus.1

Slave valuations: the rationale

An annual revaluation of a plantation's human assets was not invariably undertaken either in the US or the BWI. Three of the most voluminous, surviving collections of British Caribbean archives, those of Worthy Park, the Codrington estates, and the Newton Plantation on Jamaica, Antigua-Barbuda, and Barbados respectively, maintained lists of slaves without ascribing values to individuals.2 Whilst this lack of detail was a distinct disappointment, a substantial number of other enterprises did provide more information. The Gale-Morant and Tharp Family archives contain individual slave valuations for substantial plantations; the Crawford papers (MSS 9769/23/14/8) include an annual valuation for all slaves collectively for three Jamaican plantations over a 17-year time-series. Table 4, which summarises our valuation data for Caribbean slaves by age, was derived from 17 plantations; comparable data for the US (Table 3) was drawn from 31 plantations. In the case of the US, there appears to have been a lower incidence of valuations in the period prior to 1850 and a greater attention to it thereafter. However, even in the later period, the practice appears relatively infrequent, notwithstanding the vagaries of record survival that prevent the archival researcher from knowing how much of the original record has been lost.

The fact that immaculately maintained slave lists exist for both Worthy Park and Newton Plantation, but without individual values, leads to the question of why some plantations undertook a more detailed evaluation while others did not. There appears to have been no compelling reason, such as governmental fiat for taxation purposes. In the US, there was a state property tax on slaves to be sure, but it was a per-capita (poll) levy rather than a percentage of value. The determinant of the taxes owed in several states was the age and gender of the slave so that the assessment was based on the number in various categories (for example, Prudhomme Papers, No.613, folder 33). The levy on slaves in the BWI was essentially the same, but here it was typically a straight per-capita tax without the age and gender classifications.

Government might also have induced valuation if linked to the slaveholders' expectations for compensation. South Carolina, for example, established a formal valuation procedure for slaves conscripted into Confederate service. Valuation was undertaken by the slave owner and was sworn to by two disinterested parties under oath (Perrin Papers, 2471, folder 5). Of course, the outcome of the Civil War mooted that valuation scheme. It also appears that some states mandated valuations when rendering judgements on the level of compensation for the owners of slaves who had committed capital crimes.3 In the BWI, meanwhile, the British government offered compensation for slaves emancipated in 1834. It turned out, however, that the system was not based on a valuation of individual slaves. Rather, slave owners classified their slaves into 20 categories, each of which was assigned a standard value.

The probating of wills was an occasion when the valuation of slaves on an individual basis might be required for an equal distribution of a plantation owner's assets among the heirs. Although the physical plantation itself was rarely divided in this age of primogeniture and entail, settlement was frequently made for the siblings of the primary heir. Fogel and Engerman (1974) based much of their valuation research on probate data.4

A number of business reasons exist that could explain why some planters chose to value their slaves more regularly than others. Many plantation owners in both venues found themselves heavily in debt, either because of poor estate management and/or because of their "conspicuously consumptive" lifestyles, whether they were resident on their plantations or absentee in the British Isles. Consequently, slaves were used as collateral for a cycle of bank loans that plantation owners required. While we have not examined bank-loan indentures for US plantations, we have seen a substantial number for BWI estates. In most instances where slaves constituted a portion of the collateral, they were named, sometimes with accompanying individual valuations. However, one might have expected an even higher incidence of valuation for this purpose than what actually occurred.

There is some, albeit not overwhelming, evidence that some planters insured their slaves, particularly with respect for compensation in the event of runaway. The permanent desertion of slaves was a problem even on Caribbean islands such as Barbados and Antigua where there was little place to hide, but in the vastness of the US and in the wilds of Jamaica, losses due to runaways could be substantial. An Antiguan plantation manager complained that the value of an able, Negro male would decline from 90-100 to 40-45 if he displayed runaway tendencies (Tudway Papers, letter from M.S. Walroad to C. Tudway, 2 July 1769). On Jamaica, the quarterly vestry reports required by the government contained reports on runaways. Insuring against this loss would require the valuation of slaves on an individual basis.

More substantial evidence exists that the renting of slaves was a very common occurrence. Plantations were frequently involved on both sides of hire transactions, renting out slaves during slack periods and hiring them in at peaks. Most slave rentals did not require individual valuations because these tended to be large-lot hires at a stipulated rate per day per slave. These rates were possibly derived from knowledge of the current market prices of slaves. There were also instances when individual slaves were rented. These hires tended to be skilled slaves who commanded substantially higher rates reflective of their skills.

The hiring out of slaves was much more extensive in the Caribbean than in the US South. Here, the majority of plantation books recorded the periodic hiring in and renting out of slaves. This development might have been expected in the BWI with its higher rates of mortality, especially after the cessation of the slave trade. While the quoted rental rates vary considerably, several trends emerge. Fees tended to be higher in Antigua and lower in Barbados. Rates also corresponded to the fluctuating slave values as calculated in Table 5 below. Rentals were higher in the 1800-1820 period when the abolition of the slave trade drove prices higher and lower afterwards with emancipation in the offing. The skill premium was clearly evident as in the case of Colleton Plantation on Barbados in 1828 when the going rate for unskilled slave labour varied from 6_ to 10_d per day whilst carpenters were hired out for 2s and masons for 2/6s (Colleton Plantation Book 1818-1844, No.44).

On the Blubber Valley Plantation on Antigua in 1784, the going rental rate was 10 per cent of appraised value with any insurance premium or risk lying with the renter. However, if hired out at 12 per cent of appraised value, the insurance cost or risk was retained by the proprietor (89/4/36). This singular arrangement was the best piece of evidence that we have seen of how the valuation process was linked to slave rentals. Another interesting rental development was the payment of 75 by Grange Estate as compensation for a drowned Negro where the rental agreement had specified light and easy work while the task assigned was deemed to have been "laborious and very risky" (GD 1/8/36, p.2). In toto, the evidence supports Craton's (1994, p.31) assessment:

the very fact that slaves were commonly hired out, either as jobbing gangs or as individuals with special skills, brought home the measurability of their value, not just to the owners but to the slaves themselves. Estate record books that list the slaves in their gangs and include alongside each slave his or her assessed market value are therefore not merely an indication of an owner's assessment of his capital assets, but an indicator of the slave's actual value, potential bargaining power, and consequent status.

From our perspective, those factors which have the greatest explanatory power as to why valuations were undertaken were distinctly particular to each slavery environment. In the US, there were infrequent slave valuations until the appearance of Thomas Affleck's The Cotton Plantation Record and Account Book in 1851.5 This popular plantation journal was a remarkable marketing venture. Subscribers were expected to purchase a blank journal each year and to enter the various pieces of data required. The impression was created that a religious attention to recording that information would signal a well-run plantation. Several of the later pages were earmarked for entering the names, ages, and beginning/ending values of the slaves. While these pages were perhaps the most frequently ignored in the volume, they were conscientiously done on some occasions. It may have been that there were as many valuations before the appearance of the Affleck books, but that since this plantation record was a bound volume, the chances of survival were far greater. Other unbound records have survived, however, and it would appear that this widely distributed journal induced a substantial increase in the number of slave valuations undertaken in the US.6 Also, the Affleck books permit the construction of the occasional slave-value time series whereas valuations in the Caribbean tended to be too sporadic for that purpose.

In the BWI, meanwhile, the primary explanation for slave valuations lay in the ownership structure of the plantations there. With the Caribbean climate not salubrious for Europeans and with social life so firmly rooted in the UK, virtually all plantations were absentee owned. Consequently, a flow of data was required to keep the owners apprised of developments in the BWI. In a 1786 manual on how to manage a plantation, ten Barbadian planters listed "the increase or decrease of the flock, viz. what is born, bought, or received, also what is dead, worn out, or disposed of (Lascelles et al., 1786, pp.63-4) as one of the pieces of information absentee owners should have accounted to them. Slaves were a vital component of the flock. Most US plantation owners, by contrast, were resident on their estates for at least part of the year and, thus, required less formal information for control or decision-making purposes.

The prescriptions of Arthur Young show that it was best agricultural practice in Britain by the end of the eighteenth century to calculate income, taking into account the increase or decrease in the valuation of implements and livestock (Juchau, 2002). Loudon's (1831, p.793) Encyclopaedia of Agriculture confirmed that, "no form of books, or mode of procedure will enable a farmer to know whether he is losing or gaining but that of taking stock". Likewise, John Mair's Bookkeeping Methodized recognised the relationship between profits and goods in stock in recommending that profits should only be charged with the cost of goods consumed (Winjum, 1972, pp.72, 110-11). The work was highly popular in both Britain and America. George Washington himself kept a copy of the successor volume, Book-keeping Moderniz'd, at his Mount Vernon estate (Mepham, 1988; Forrester & Vangermeersch, 1996). Confirmation that performance measurement was a motive behind some of the valuations in the BWI is evident from the records. Thus, in the case of several prominent Barbadian plantations, including Newton, an adjusting entry was made each year where the net increase (decrease) in slaves was multiplied by a standard value (usually 70-75) and added (subtracted) to the "stock" [capital] accounts.7 Elsewhere, the annual increase (decrease) in the number of slaves was taken into the profit and loss statement as in the Nisbet papers for 1799 (MSS 4582). In view of the lack of joint-ownership, which made profit apportionment unnecessary, it is probable that these calculations were part of the information required by the owners at home to monitor the performance of their attorneys and managers overseas.

One of the best explanations of this process comes from the Earl of Balcarres, a plantation owner in Jamaica. He demanded annual returns from his agents which listed and valued all the property on his plantations, including the slaves. Writing of these valuations in 1811, he opined that they were unsuitable for evaluating rate of return from the plantations, which he maintained should be based on initial investment:

I have never paid nor am out of pocket anything like the sum on which these estimates are taken, and whether this estimated capital shall rise hereafter to a greater sum, or fall considerably short of it, is a matter of little moment.

Thus, for decision-making purposes,

... these estimates seem very specious and encouraging, but they are fallacious and delusive, and little dependence ought, in my opinion, to be placed upon them.

However, when it came to controlling the activities of his agents, it was a different matter:

I find them, however, extremely useful as they tend to show the comparative industry of the current year contrasted with that of the former year. I also find them a powerful check, not only on the overseers, but on my attorneys ... . I have every reason to consider this opinion as firm, because I have consulted the most scientific planters in Jamaica, who confirmed, and assured me, that I had formed a just conception of the nature of these estimates (Crawford, 23/14/8).

Apart from the land itself, the slaves were the most important productive asset on the plantations, and the absentee owners in Britain required a range of reports to ensure that its value was being safeguarded. In point of fact, the capital investment in slaves was actually greater than in land in Antigua whereas in Jamaica and Barbados, land values were usually only 10 to 20 per cent greater than the human assets. As well as the occasional valuations, landowners would typically demand annual returns of the increases or decreases in the work force, with explanations of any deaths that had occurred. By forcing their agents to compile these reports, the owners were able to hold them to account.

Another important factor contributing to the relative lack of valuations in the US prior to 1850 compared to the BWl was the difference between the two plantation environments in the personnel who maintained the records. In the American South, the planters themselves or their managers/overseers kept the books. They had no formal accounting training or knowledge as a general rule and might not have seen any purpose in undertaking an annual revaluation of the slaves. In the BWI, the non-resident owners depended on bookkeepers of varying degrees of training. There were so few whites on the Caribbean islands that the governments attempted to use the tax structure to force plantation owners to keep white staff on their estates, presumably to maintain racial control. Some governments established a white-to-black ratio for plantations under their auspices and proceeded to levy a 40 annual tax, called a "deficiency", for every white-male shortfall.8 Additional white women could reduce the deficiency, but only by half the amount. In Antigua, for example, the amount of the deficiency tax dwarfed the property levy on land and slaves as a state revenue source (Tudway Papers, DD/TD, box 10). The upshot was that a number of low-talent whites, whose annual salary was less than the tax penalty, staffed the plantations as secondary bookkeepers or "posters". Since the names of whites had to be reported to the government, the remarkably high turnover of these bookkeepers is evident. The larger plantations had as many as four or five bookkeepers, some of whom received such low pay that it is doubtful that they brought much accounting expertise to the job. A chief bookkeeper, by contrast, could expect a salary that parallelled that of a second-in-command overseer. In any event, with a substantial staff in residence and some professional expertise on board, the higher incidence of valuations and other agency data in the BWI prior to 1834 is not surprising, if for no other reason than to keep the administrative employees occupied.

Evaluation process

The plantation records do not provide precise information as to how the periodic revaluation of slaves was done. In the US, there was a professional group called "appraisers" who were typically hired when probate services were required. There were no record keepers identified as either accountants or bookkeepers. It appears as if the appraisers were called in on special occasions only because the appropriate pages of the Affleck plantation journal given over to slave valuations were invariably signed by the plantation overseer. On BWI plantations, by contrast, the bookkeepers were in residence and obviously available to participate whenever a valuation was required. There are documents in the Caribbean archives reflective of more formal appraisals than what plantation managers and bookkeepers might do. On such occasions, the attorneys representing the absentee owner would authorise the valuations. Strangely, the individuals undertaking these appraisals were never identified as to any formal licensure or even title that would qualify them for this function.

Despite differences in the titles of available professionals and their physical locations on or off the plantations, we suggest that the valuation process was essentially the same in both venues. The slaves were paraded past the overseeing staff, perhaps in the presence of an appraiser in the US and certainly in the company of the bookkeeper(s) in the Caribbean. A certain modal value was known for prime field hands, men and women, which as the next section relates was adjusted to take account of a range of factors.

The modal value was probably linked to the market price from recent sales or purchases. We have not tracked in detail the movement of slaves on US and BWI plantations. Yet, certain indications do exist to gauge the degree of buying and selling. On the large York Plantation in Jamaica, 70 slaves were purchased between 1785 and 1793, 100 were born on the estate, and 160 died (Gale-Morant Papers, 3d). On the smaller Grove Plantation in Mississippi (100 slaves compared to 450 at York) between 1853 and 1861, 31 slaves were apparently purchased and 10 were either sold or died. In the BWI especially, the planters tended to rely on imports rather than natural reproduction to replenish the work force. Hence, they could either buy newly arrived slaves or purchase ones already trained from other plantations, depending on availability. The former tended to be sold in job lots at a standard price. If the slaves were already trained, the price paid reflected the occupational differentials that existed in the valuations, with skilled occupations attracting the highest price (for example, Nisbet, MS5478-5479). In the case of the US, the modal price was more likely to have been determined by the going rates at regional slave-market auctions.

A 1750 appraisal on Nevis (Stapleton Manuscripts) provides evidence of the physical inspection procedure described above. Here the appraisal was signed by four gentlemen whose names appear at the bottom without a designated title. They were retained for the assignment by lawyers on behalf of the plantation owner. The signatories averred that the items constituting the valuation had been "shewn us by the aforementioned attorneys". In a formal valuation of Derry Pen (Jamaica) in 1817, the signatories said they were not able to value four slaves who had failed to appear before them (Worthy Park Plantation Book 1811-1817, 4/23, No.4). Two slaves who had been absent from the valuation procession at Prospect Pen, Jamaica in 1784 and, thus, could not be physically examined were accorded the same value as estimated at the former appraisement (Tharp Papers, R.55.7.123.3).

The importance of the valuation process to certain absentee owners is underscored by the several incidences we have found when two different individuals appraised the slaves with the valuation for the total work force being the average between the two. The Fitzherbert collection includes a 1784 appraisal of the estate of George Cammac that featured evaluations by both Robert Grant and William Sutherland (D289/E18036). The two appraisers were close in most instances; yet, it is clear that there was no collusion as in others the disagreement was substantial. A more elaborate joint venture was the dual valuation of 272 Negroes at the Albany (64) and New Hope (208) estates in 1814 by Dr Samuels and Mr Forbes (DM 89/7/68 and 80). The two appraisers matched one another on 173 of the 272 valuations. At Albany, the two totals were 200 apart; at New Hope, the disparity was 930. In both instances, the valuators agreed to "let the difference be divided". The custom in certain Caribbean colonies to have multiple slave valuations was reflected post-emancipation in British Guiana in a process whereby the freedmen, now tied to the land as apprentices, continued to be bought and sold. Valuations were initially made by both the planter and a representative of the Crown. In the event that consensus was not forthcoming, a colonial, court-appointed umpire made a third valuation, thus resolving the conflict. The system clearly favoured the plantation owners as the umpires invariably sided with them (see British Parliamentary Papers, 1837-38 for the raw data and Tyson et al., 2004b for the analysis).

Factors in the determination of value

If, as the work of Arthur Young suggests, it was best agricultural practice by the end of the eighteenth century to take an annual valuation of implements and livestock, there was little guidance in the available literature on what basis to use. Young, himself, recommended using "the price they would sell for at the moment" (Juchau, 2002, p.377). John Mair (1763, p.331) was less specific. In listing the accounts that plantations should keep, he merely said that these should include "an account of the number of servants and negroes belonging to the plantation". Eighteenth century estate guides were equally imprecise, focusing on the recording of transactions and the rental value of tenants' farms (Oldroyd, 1999). Even in Roger North's The Gentleman Accomptant, which was perhaps the most comprehensive guide of the genre, the basis of his prices in the "Accompt of Flocks of Sheep" is unclear (Parker, 1997). One of the closest descriptions in an agricultural textbook to what seems to have taken place in both the BWI and the US in valuing slaves comes from the early ninetheenth century work, Encyclopaedia of Agriculture, by J.C. Loudon (1831, p.540):

In valuing live stock, a variety of circumstances require to be taken into consideration. The value of all young animals may be considered as prospective ... . Draught cattle may be valued on an abstract principle, derived from the probable value of their lives and labour, but in general nothing is to be depended on but a knowledge of the market price, and this ought to be familiar to every valuator.

In short, the basis for valuing draught animals (that is, working beasts rather than beasts reared for food) was their value in use as impounded in the market price. While London's ideas regarding valuation applied to farm animals, we can see these principles being applied with equal force to the slave inventories as well. Valuators, particularly in the BWI where a greater expertise was brought to bear on the process, were keenly aware of market prices, individual slave skills, and general economic conditions.

For the BWI, for instance, although not for the American South, it is possible to gauge the valuation premium that corresponded to particular skill sets. It was often the case that slave lists in the Caribbean grouped individual slaves according to their occupational functions on the plantations. In the US, by contrast, slaves were most often listed alphabetically with a skill, noted in parentheses, specified only on rare occasions. Table 1 identifies the basic male skills which are compared to the mean value of field hands in order to appreciate the incremental values of particular skills. It should be noted that the sample excludes "boys", as distinct from "men", so that the values of apprentices are not included in the means.

Plantation owners in the American South were relatively unconcerned about the skill sets of male slaves compared to their BWI counterparts. Where individuals on slave lists are identified as blacksmiths, carpenters, and so on, their values are higher although not reflective of the substantial Caribbean premiums. The difference between the two venues might be attributable to the availability of cheaper free labour in the US because of the existence of a rural, white population apart from the plantation economy. On BWI plantations, by contrast, the slave labour force necessarily possessed these skills and, consequently, much higher premiums were impounded into slave values.

On those occasions when formal Caribbean appraisals were done, such as when a plantation was being sold or for probate purposes, the state of the economy was a consideration of some import. One such valuation of slaves for probate purposes took place at the Holland Plantation on Jamaica in 1808 (James Lyons, 1802-1816, first MSS bay, bound volumes, first shelf). Here the individuals preceded their signatures by averring:

In making this Estimate we are very strongly biased by the present circumstances of the Island in relation to our intercourse with the American States, which in a particular manner affects the Value of the property, also the depressed state of the Sugar Market and the little prospect of amendment.

In both the US and the BWI, further adjustments could be made in respect of atypical physical characteristics (beauty, strength, illness, skin color), collateral skill, and/or productivity potential. Fleischman and Tyson (2004) have compiled slave values for the Grove Plantation in Mississippi from a lengthy and continuous series of Affleck books spanning nearly a decade. They found narrow value differences amongst slaves of the same age and gender. Similarly, there are data for the Old Montpelier Plantation in Jamaica where a single plantation book has slave lists for 1825-1827 (MSS 217, Old Montpelier Estate Account Book) and the Gale-Morant Papers which include valuations at York Plantation for 1778 and 1782 (3c). As at Grove, there was a modal value that apparently approximated a market price from which large deductions were made if a slave was an invalid or incapacitated, with smaller adjustments up or down reflecting physical and/or productivity differentials. It might be expected that less would be known of an individual slave's productive potential since the gang structure of labour was used more prominently in the sugar cultivation of the BWI. Yet, the presence of the overseers and perhaps the drivers as well permitted micro-level assessments of value.

Slave lists in the BWI typically included a column for remarks or the articulation of slave characteristics. These qualitative observations did not appear in US valuations; there was no column for them in the Affleck books. One of the most complete BWI slave lists we have seen was done at the Tharps' Good Hope Plantation in Jamaica in 1804 (R.55.7.123.2). Table 2 demonstrates how favourable remarks in the "condition" column resulted in substantially higher individual slave values. Positive comments identified those slaves who were "able", "well-disposed", or "good", whilst negatives characterised those who were "weakly", "indifferent", or "lazy". Consequently, the 13 masons who impressed the evaluators had a distinctly higher mean value than the six who did not.

Slavery researchers, particularly Fogel and Engerman (1974), have attempted to link slave values to the life cycle. Fogel and Engerman utilised probate data to secure information about 76,888 slaves in eight states (Friedman, 1992, p.72). By contrast, our survey of 31 sets of plantation records in four states has yielded ages and values of some 2,500 slaves. Table 3 summarises these data for five-year periods from 1845 through the conclusion of the Civil War. Table 4 contains similar age/valuation comparisons for the BWI and is gleaned from returns on 3,000 slaves from 17 plantations.9

While it is not our primary intent to replicate the massive demographic project of Fogel and Engerman (1974), our calculations confirm a number of trends reflective of life-cycle stages. Values increased at a steady rate ($100 a year in the US) as slaves approached prime productivity. Likewise, values declined for both sexes, but much more sporadically, as they entered their later thirties. While this pattern might have been expected, what was unanticipated was the failure of female slave values to decline as the years of prime fertility passed. We hypothesise that the failure to factor in the reproductive potential of females into their values was the high cost associated with supporting slave children until such time as they could earn their keep in the fields. Alternatively, the explanatory factors may have been a perilous mortality rate for young children and/or a low fecundity rate for females in the slavery environment. But it is clear that female slave values did not escalate rapidly in the late teens and early twenties as the prime fertility years were reached. Furthermore, a precipitous decline did not occur corresponding to the passage out of prime fertility. Tables 3 and 4 demonstrate that the ebb and flow of values corresponding to the life cycle were parallel in both the US and the BWI.

The escalation of slave values chronologically in the US in each five-year period was reflective of raw cotton prices and the wholesale price index as generated by the Department of Commerce (Fleischman & Tyson, 2004). The Caribbean values, on the other hand, corresponded to the preeminent events in nineteenth century, British slavery history - the abolition of the slave trade in 1807 and emancipation itself in 1834. Prices rose to unprecedented heights when the supply of slaves from Africa was cut off and fell precipitously as freedom neared. It is interesting that there was no corresponding value decline during the US Civil War, suggesting either that the market did not expect the Confederacy to lose or that the abolition of slavery was not going to be imposed upon a vanquished South.

That which is of greater interest from our perspective are the substantially greater values for male slaves compared to females of the same age in the US, a premium that does not appear to have existed in the BWI. There was no evidence that there were marked productive differences between males and females picking in the cotton fields although the largest single-day producers were invariably men. For example, at Ballard's Magnolia Plantation in Louisiana for a week in September 1848, 30 men averaged 158.66 pounds daily, while 32 women averaged 152.06 pounds (Ballard Papers No.442). Yet, there was a marked differential in valuation along gender lines during the prime productive ages (15-34 years) for all five-year periods examined (see Table 5). Male values were consistently 25 per cent higher, corresponding to the differential between male and female earnings elsewhere that Fogel and Engerman (1974, I, p.77) found for most years beyond age 18.

Part of the explanation for this valuation premium on males may lie in the fact that they held the vast majority of the skilled positions, and therefore commanded additional value. As we have already noted, it is not possible to know whether these skills were factored into the US valuations as slaves there were not generally identified by their occupations. In the BWI, however, slaves were most often grouped by occupation on the slave lists. Table 6 is a gender values comparison for a large Jamaican sugar plantation, Old Montpelier. Here the slaves that formed the comparison groups are those identified as "first class" field hands for 1825. It will be noted the degree to which the premium has narrowed. The male sample is quite small, however, because the majority of males on this plantation had an identified skill rather than "work in the field".

The findings at Old Montpelier that BWI premiums for male field hands were narrow are verified by 1788 data from the Wallens estate (GD1/850/75/3). Like many Caribbean slave lists, values were given but not ages. The mean value of 30 male field hands was 87.0 while for the 37 females the average was 81.2, a premium of 7.1 per cent. In the case of the estate of Peter L. Brooke on Antigua, 24 field males averaged a valuation of 172.1 and 42 females similarly identified had a mean of 175.0. In both instances, the modal value was 200 (Brooke of Mare, Box III, No.5, No.855). The field-hand valuations on Wakefield Plantation in 1787 reflected virtually no disparity whatsoever - 79.58 for 12 males; 79.72 for 18 females (University of the West Indies, MSS Cabinet C, Drawer 3). Tables 3 and 4 show clearly the presence of the premium for maleness in the US and its absence in the BWI. The tables are a compendium of most slaves for whom we have both age and valuation data. No adjustment has been made to eliminate the higher values of male slaves with identifiable skills. Yet, the mean values are virtually identical for males and females in the BWI.

Accounting and the slave regime

As Baxter (1999, p.256) has reminded us, "accounting started as a means of keeping track of debtors, creditors and cash; and that this is still its main task, without which business would collapse". The same, of course, is true of inventories, and our findings indicate that the slave plantations could not have functioned without valuing the slaves from time to time. As we have seen, some valuations took place annually, some only on special occasions, such as for probate purposes, loan collateralisation, or when an estate was up for sale. In the case of the BWI especially, valuations formed part of the agency information absentee owners required, and these owners could not have invested in plantations without it. In this respect, the valuations supported the system of slavery on which sugar production in the Caribbean was based. However, accounting was performing a value-neutral function here in the sense that it would still have provided whatever information was necessary to trade irrespective of the system of production in force. One can see this in other British imperial locales, such as in the accounting information provided by coal-mining agents in Nova Scotia to the proprietors in London (Fleischman & Oldroyd, 2001).

Thus, one might excuse the agents who prepared the valuations from moral censure for their role in facilitating slavery on the grounds that they were simply acting as economically rational, racially neutral beings. By modern standards, the most overt manifestation of racism was the typical pattern on the BWI plantations of accounting for slaves precisely as the livestock was accounted for. It was a sad commentary on the times that multiple plantation books would follow a listing of slaves with identical information for horses, mules, and other farm animals. The data categories for both could include ages, conditions, and values. Particularly impacting was the appearance of these slave and livestock listings on the same journal page. An early letter from James Butler to Lady Stapleton, absentee owner of a Nevis plantation, read, "ther's 59 Negro's, 20 horses ..., I found the Negro's in but very ordinary plight, the horses extraordinarily well ..." (Stapleton MSS, 5/1/2, 19 May 1711). A 1732 list of Negroes continued with mules, mares, horses, and horned cattle which were tellingly referred to as "other stock" (Stapleton MSS, 5/1/2). Similarly, for those plantations on which an annual valuation ceremony took place with the slaves passing in front of an appraisal corps, the event must rank high as one of a slave's most dehumanising moments, and it appears particularly distasteful to us.10 Here an individual whose life had been stripped of value by the slave plantation system was being appraised for the benefit of persons totally foreign and frequently unknown to him. Decisions as to this value were being rendered on the basis of information that was typically incidental to a person's true intrinsic worth in any human terms. For the "superannuated" slave, the reward for a lifetime of toil was a valuation of zero and a descriptive of "useless". But is it fair to condemn accounting practitioners for engaging with an inherently immoral system whose values they did not invent, especially in the absence of a professional superstructure to set ethical standards?

Therefore, in order to hold accountants responsible for supporting slavery, one would need to attribute motives other than economic rationalism. This is not easy in an historical setting given the loss of information or the fact that the motives may never have been recorded in the first place. In a parallel study of labour on Hawaiian sugar plantations in the late nineteenth and early twentieth century, for example, Fleischman and Tyson (2000, 2002) expressed the opinion that the failure of the managers to track the individual efficiency of workers was indicative of the fact that the system there was one of racial control. However, Burrows (2002) subsequently commented that the gang system of labour did not lend itself readily to the maintenance of statistics on an individual worker's productivity and that it would not have been cost-beneficial to do so. Hence, the Hawaiian plantation owners were simply acting in an economically rational way.

Although there were few accounting practitioners in the US South, there are indications in the premium for maleness that the valuations now discussed were not purely economically rational. Additional research is required to determine why the male premium found in the US data and verified in the larger Fogel and Engerman study was not evident in our BWI statistics. But, we can be confident that the lack of a premium in the BWI is not simply a random error due to a small sample size as our data cover a wide chronological and geographical scope. There was certainly a premium in the BWI linked to the relative importance of the slave's occupation, which in some instances valued the men higher than the women, but when they were performing the same tasks, such as in the field, they tended to be valued the same. As we have seen, the women were equally as productive as the men, and one must therefore conclude that in the BWI slaves were valued simply as economic resources, whereas in the US there were additional social forces at work.

Regarding the slaves purely as economic resources in the BWI is fully consistent with their being valued alongside the livestock, as well as their being listed and valued by occupation, with adjustments for ability. It also fits in with our primary explanation for slave valuations in the BWI (that is, to provide information for absentee owners in the UK) and the concern that the owners often expressed in their correspondence that their stock be properly maintained, again, in some instances, talking about the mules and slaves in the same breath (for example, James Stothert's letter-book, 1792-1793, Stothert of Cargen, GD 241/189/1).

The main puzzle, then, is why was gender significant in the US valuations? The answer lies plausibly in the fact that, unlike in the BWI, the owners lived among the slaves. In the Caribbean, the plantations were merely investment opportunities, and certainly not home for the owners. In fact, the number of European residents was very small, with the slaves outnumbering the white population by about ten to one. In the US, however, the slaves constituted part of the owners' society. As we have seen, plantations tended to be smaller than in the BWI with the owners directly involved in the management. As Holt (1992, p. 128) observed, planters in the BWI looked primarily to profit, whereas for planters in the US, slavery was a vital part of the social system they inhabited. There the slaves were closer to the owners than mere names on an inventory, and it is credible that the premium for male slaves is evidence of the planters translating their own social values to the slave population.11 The emphasis in the US valuations on gender rather than occupation suggests that the owners regarded their slaves more like humans, albeit inferior ones, than economic commodities as was the case in the BWI. Hence, males were more highly valued than females even if that premium was not justified in terms of economic rationality.

If the foregoing explanation is correct, the valuations reflected an important social difference between the two regions. Whether accounting also helped constitute the values of the proprietors is harder to say. As far as the BWI is concerned, the plantations operated against a tide of "massive popular mobilizations against colonial slavery" (Blackburn, 1988, p.465). In what Trevelyan (1945, p.599) described as "the first successful propagandist agitation of the modern type", the abolitionists were able to capture the moral high ground, as a letter from James Stothert to his agent in the Caribbean in January 1792 attested:

Mr W[ilberforce] (it is expected) will again bring forward his Bill of Abolition ... . Where they [the abolitionists] are in general so totally ignorant in the business, the frenzy spreads amazingly for which great pains are taken by some individuals, and religious motives are, in course, introduced. I am now told that Mr W. is sending copies of the evidence taken before the committee of the House of Commons to the Minister of every parish, no doubt to obtain charitable contributions, to carry out the Bill now depending. I am sorry to add that no great such appears on the planters' behalf to counteract these dangerous machinations against the Colonies (Stothert of Cargen, GD 241/189/1).

This letter shows a keen awareness on the part of the owners of the threat to the plantations posed by the campaign for abolition. One of the standard defenses in the propaganda war was that slavery was necessary for business, and that without it, sugar production would collapse. This was a real concern for government as can be seen from the letters coming to the Colonial Office during the transitional period in the 1830s leading to full emancipation (for example, West Indies Original Correspondence, CO 318). The economic-rationalist tenor of the slave valuations in the BWI re-enforced the view that slavery was first and foremost about business. Furthermore, one can see a parallel with Funnell's (1998) findings on the Holocaust. He argued that the accounting records made it easier for the Nazis to engage in mass genocide by reducing the victims to numbers, thereby disguising their humanity. Conceivably, the BWI valuations fulfilled a similar function in dehumanising the slaves by reducing them to the status of livestock and economic commodities. It is true that slaves were listed by name rather than serial number, but this was the most practical method of identification given the need for the agents to discuss particular slaves in the correspondence with the proprietor.

Obscuring the humanity of the slaves through the accounts was not an option in the American South given their coexistence with the owners and white population. However, the popularity of Affleck's accounting instructions amongst Southern planters again suggests that accounting was being used to legitimise an institution under threat. By 1851, when the text was first published, the South was isolated at home and abroad in its use of slaves. The last vestiges of slavery had disappeared from the British Empire by 1838. In preaching the best methods of plantation accounting, one could argue the book helped legitimise slavery by promoting it as a normal and scientific business activity. For instance, the book recommended that profits should be calculated taking the increase or decrease in the annual valuation of slaves into account (Fleischman & Tyson, 2004). The lack of joint ownership and profit apportionments obviated the need for this procedure. Yet, it must have been reassuring to the owners to realise that recommended practice of the day also applied to their enterprises.

Conclusion

Surviving plantation records for the BWI contain many more slave valuations than those for the antebellum US. We have hypothesised in this paper that the explanation lays in the higher frequency of slave rentals in the Caribbean due to the dearth of available free labour and the even higher incidence of absentee ownership. Meanwhile, the Affleck journals, the factor we feel contributed most to the volume of US appraisals that did exist, were in print for little more than a decade before the South's "peculiar institution" was abolished. Not only were there more BWI valuations, but the information conveyed to the absentee owners was more exhaustive. BWI slave lists frequently included qualitative information about the condition and disposition of slaves, as well as their occupations as we have seen. More specific information about race was often provided - whether slaves were African or native-born (Creoles). Furthermore, additional bloodline data related to skin colour were occasionally chronicled - identification of slaves as sambos, mulattos, mustees, and quadroons (for example, Gale-Morant Papers, 3d). It is left for further research to see if valuations were linked to racial characteristics. In the US, the only distinction that one ever sees in the plantation records is the infrequent identification of certain female slaves as mulattos. Here, there was an associated premium as the yellowish skin cast was considered sexually desirable.

Further research is also required to explore the impact of demographic factors on the insistence of BWI plantation owners for more information about their slaves. Most current literature suggests that slave mortality and fertility rates were unfavourable in the Caribbean as compared to the US. With the cessation of the slave trade, the slave labour force could not be replenished except by natural increase. The Newton plantation books and the Tudway Papers provide indications of certain steps taken. Cash was paid to fertile women both on the occasion of their giving birth and again when their children first reported to work. Sick slaves were particularly well cared for. In the US, geographic, chronological, and demographic factors may have combined to produce an entirely different set of priorities.

Finally, we return to the larger question of the role of accountants vis--vis the moral aspects of slavery. There are two difficult ethical questions to consider before we can even begin to condemn accounting practitioners for active participation in an inherently immoral system. First, to what degree can personalities of the past be held responsible for actions that were legally and socially sanctioned and were not widely considered immoral, especially in the environment in which they operated? Central to our thinking about this question are the linkages we feel exist between the past and the present. We have argued elsewhere (Fleischman & Tyson, 1997) in response to Carnegie and Napier (1996) our conviction that historians cannot achieve objectivity since efforts to divorce their contemporary biases and prejudices from their narratives are ultimately futile. Consequently, the best that we can do is to disclose openly and forthrightly to our readers the nature and depth of our moral postures, attitudes about issues, and paradigmatic groundings so that the reader can judge the degree to which the past rather than the historian is speaking at key junctures. In the slavery context, moreover, it must be recalled that whereas the institution was considered the natural order of things by the plantocracy of the US South and the Caribbean islands, there was substantial anti-slavery sentiment in the American North and in Great Britain. So it might be argued that even in the heyday of slave plantations, the high moral ground among men of good will was abolitionism. Funnell (1998) and Fleischman and Tyson (2000) felt accountants had much to answer for with respect to the Holocaust's genocide and Hawaii's racial control respectively and, in this regard, they moved in distinctly more critical directions than was evident in their earlier work. It is a like sentiment that has now come to inform our reactions to the roles accounting practitioners played in sustaining US and BWI slave regimes. But even if it is only hindsight that clearly reveals the immorality of slavery, whereas those active in it were not so advantaged, arguably a higher standard of justice may be invoked for those persons who treat and account for slaves as if they were farm animals or real property.

Second, have accountants had the power historically to influence managers toward more moral action and decision making? Probably not! But, then again, there are no writings that we have seen that document any attempts at protest or reform by the keepers of the historical record. Furthermore, our findings suggest that far from condemning slavery, the accounts helped justify it in the face of mounting pressure for its abolition. The dispassionate valuation of slaves in the BWI as economic commodities reinforced the view that slavery was merely business, as did Affleck's journals in the American South. Clearly, these difficult ethical questions are upsetting to a discipline that has prided itself on objectivity and neutrality. Nevertheless, the further we delve into the archives and learn about particular accounting techniques like valuation, the more certain we become in asserting that accounting was instrumental in sustaining slavery's social institutions and basic practices.