Skip to main content

Advertisement

Log in

Slave prices and productivity at the Cape of Good Hope from 1700 to 1725: Did everyone win from the trade?

  • Original Paper
  • Published:
Cliometrica Aims and scope Submit manuscript

An Erratum to this article was published on 16 December 2014

Abstract

This paper analyses the economic viability of slavery in the Cape Colony in southern Africa. It has been extensively documented that the affluence of elites was built on the importation of slaves. However, the Dutch East India Company or Verengide Oost-indische Companje, which administrated the colony, expressed concerns that free settlers had invested too much capital in the trade, so that some indications exist that profitability was not certain for all farmers. In this paper, hedonic slave price indices and the value of their marginal productivity have been estimated, to construct annual returns, which are in turn compared with returns on other investments for the period 1700–1725. Hedonic price functions were estimated to remove the anticipated lifetime returns that slaves would yield and to isolate buyers’ perceived depreciation of the slave for 1 year. Cobb–Douglas production functions were estimated for average farmers, as well as at various quintiles along the distribution, to evaluate scale effects. Large farmers enjoyed high returns to slavery over most of the period, confirming the assertions that the elite used slaves profitably. Small farmers, however, did not recoup slave costs from agricultural production: this suggests either that they overinvested in slavery relative to other capital goods (e.g. ploughs or wagons), or that they used slaves profitably outside of agriculture.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Notes

  1. Giliomee (2004), Van Duin and Ross (1987), Guelke and Shell (1983), Fourie and von Fintel (2012), Fourie (2013), and Muller (1983). Worden (1985) estimated an annual return to slavery, though he used aggregate figures to achieve this instead of micro records (see the methodology section for more details on his estimation and results).

  2. Dahomey was situated in the present-day Benin.

  3. Nunn (2008) produces a table that documents the origin of slaves going to the New World (1400–1900).

  4. Slavery was illegal in the Netherlands itself, although it was allowed in its colonies (De Beer 1993).

  5. The data used in this study does not contain information on land size for individual farmers. Where we do analyse “scale”, we refer to the value of production volumes (the output), which is potentially highly correlated with plot sizes (the inputs). Additionally, we show in Tables 4, 5, 6 and 7 that other production inputs, such as vines planted, are highly correlated with this definition of scale. Hence, agricultural outputs are correlated with capital. We therefore assume that production volumes are good proxies for scale and capital availability. In productivity estimates we control for other proxy inputs to deal with the omission of land size from the records.

  6. Shell (1997) provides the account of a traveller, Admiral Stavorinus, on how Melck’s slaves were housed—with separate housing for married slaves.

  7. Most farmers in the wine and grain producing regions used more than 15 slaves, with some wine farms even using more than 80 slaves (Muller 1983).

  8. We are indebted to Robert Shell for sharing the databases in a user-friendly format with us.

  9. Though micro-level records over time are contained in this database, they do not constitute a panel dataset. We use pooled cross-section data with various time fixed effects to model the price index.

  10. The micro records reveal that many sales were conducted as “batch sales”. In these cases slaves were clearly not bought on the basis of their personal characteristics, but a discounted price was offered. This does not concur with the notion of value that accrued to buyers, nor with expected returns to characteristics. However, much information on these observations was omitted (such as productive characteristics), so that these peculiar cases were excluded from estimates by default. Furthermore, many characteristics were omitted from the records in later years, so that controlling for the hedonic features dramatically reduced the sample size in that period. Estimating model (2) with the smaller sample introduced sample selection bias for later years. Additionally, estimates of \(\grave{\lambda}_{t,}\) were sporadic in later years and imprecisely estimated. For this reason, the study period is truncated to 1725, where these biases are absent or insubstantial.

  11. By TR we mean that even quantities of products that were consumed by the household could potentially have been sold at market prices. The data record the total of home consumption and products that were sold, without distinguishing between them.

  12. Because many farmers have zero inputs on some items, the Cobb–Douglas production function would result in sample selection bias once variables are logged. As a result, we impute a “small” value of 0.020 to each zero input, in order to maintain the sample size.

  13. The Orphan Chamber was established at the Cape to administer the estates of individuals who died intestate and left heirs that were either too young or untraceable. While all wills and deaths therefore had to be registered with the Orphan Chamber, it only inventoried and acted as executor for the estates of free people (Fourie 2012). In this process of valuating estates, they collected detailed prices of multiple products.

  14. Quantile regressions were estimated in quantile intervals of 5, from 5 to 95. This yields a series of estimates of \(\hat{\beta}_{1,}\), weighted at the different values of the TR distribution. Farmers were classified into intervals of a five percentile range, centred at the respective quantile regression weights—these observations were associated with respective coefficients from the various regressions. Relevant coefficients were multiplied by each farmer’s average revenue with respect to slaves, so that each farmer obtained an individual VMP that hence also reflected their position in the scale distribution. Results for selected years are available in Tables 5, 6 and 7, in the “Appendix”. Additional results are available from the authors on request.

  15. Probate inventories were collected by the Master of the Orphan Chamber (see footnote 13). It was obligatory for every executor of a will to provide the registry with an inventory of the deceased’s goods, together with their value.

  16. While this does not directly measure a characteristic of slaves as is commonly understood in the hedonic framework, it does proxy for one of these, namely their susceptibility to disease.

  17. Liesbeeck valley, which is near to the coast, is the one region that does not fit this pattern. This was one of the first areas that Van Riebeeck allocated to free farmers to cultivate. However, recorded sales from this region appeared only in 1708, so that this may be an outlying result rather than part of a systematic pattern.

  18. “Small” is defined by the bottom quintiles of TR, rather than land size, due to the absence of indicators for the latter.

  19. In 1700 45 % of sellers were situated abroad, while by 1725 this figure had declined to 29.3 %.

  20. Unobservables have likely been omitted from the model and are still captured by this series. However, if one assumes that the most important characteristics which buyers based their decisions on would not have been omitted from the sale deeds, then most of the price formation process would have been captured by the included variables, apart from a premium for current productivity (which is what remained after exhaustive controls were introduced). Similarly, unobservables in the calculation of VMP remain, notably land size. However, as noted before, proxies have been introduced to account for this omission.

  21. The coefficients of the inputs in the OLS Cobb–Douglas production function in Table 4 in the appendix, however, do not add up to a value of larger than 1, suggesting that increasing returns to scale did not exist across all inputs. However, the results of quantile regressions in Tables 5, 6 and 7 in the appendix show that farmers with a larger output reaped higher benefits from slaves. Furthermore, these functions do not control for land size (which are not available in the current data), so that this omitted factor would potentially allow us to conclude overall increasing returns to scale.

  22. Probate inventories are lists of assets owned by deceased individuals or households. They included all deceased settlers, who were not necessarily farmers. However, since the type of capital goods we analysed would most probably have been used by farmers, we assumed this would not affect our analysis.

  23. Slave categories, in line with figures for each TR quintile from Table 2, were created in the Probate Inventories to proxy for farm size. There is no other way to match wealth between these records and the Opgaafrollen.

References

  • Armstrong J, Worden N (1988) The slaves 1652–1834. In: Elphick R, Giliomee H (eds) The shaping of South African Society. Wesleyan University Press, Connecticut, pp 1652–1840

    Google Scholar 

  • Boshoff W, Fourie J (2010) The significance of the Cape trade route to economic activity in the Cape Colony: a Medium-term business cycle analysis. Eur Rev Econ Hist 14(3):469–503

    Article  Google Scholar 

  • Chenny S, St-Amour P, Vencatachellum D (2003) Slave prices from succession and bankruptcy sales in Mauritius, 1825–1827. Explor Econ Hist 40:419–442

    Article  Google Scholar 

  • Conrad A, Meyer J (1958) The Economics of Slavery in the Ante Bellum South. J Polit Econ 66(2):95–130

    Article  Google Scholar 

  • De Beer CR (1993) The law of slavery: the predicament of the slave community at the Cape. Koers 58(4):417–430

    Google Scholar 

  • Dooling W (1999) The decline of the Cape gentry, 1838 ± c. 1900. J African Hist 40(2):215–242

  • Dooling W (2007) Slavery, emancipation and colonial rule in South Africa. Ohio University Press, Athens

    Google Scholar 

  • Du Plessis SA, Du Plessis SWF (2009) Happy in the service of the Company: the purchasing power of VOC salaries at the Cape in the 18th century, Stellenbosch Economic Working Paper 01/2012

  • Du Plessis SWF, Du Plessis S (2012) Happy in the service of the company: the purchasing power of VOC salaries in the Cape in the 18th century. Econ Hist Dev Reg 27(1):125–149

  • Fogel R, Engerman S (1974) Time on the cross: the economics of American Negro slavery. Little Brown, Boston

    Google Scholar 

  • Fourie J (2012) An inquiry into the nature, causes and distribution of wealth in the Cape Colony, 1652–1795. Unpublished PhD Dissertation, Utrecht University

  • Fourie J (2013) Slaves as capital investment in the Dutch Cape Colony. In: Hillbom E, Svensson P (eds) Agricultural transformation in a global history perspective. Routledge, Oxford, pp 1652–1795

    Google Scholar 

  • Fourie J, von Fintel DP (2010) The dynamics of inequality in a newly settled, pre-industrial society: evidence from Cape Colony tax records. Cliometrica 4(3):229

    Article  Google Scholar 

  • Fourie J, von Fintel DP (2011) A history with evidence: income inequality in the dutch cape colony. Econ Hist Dev Reg 26(2):16–48

    Google Scholar 

  • Fourie J, von Fintel DP (2012) The fruit of the vine? An augmented endowments-inequality hypothesis and the rise of an elite in the Cape Colony. In Amsden, Di Caprio, Robinson (eds.) The role of elites in economic development. Oxford University Press, Oxford

  • Fraginals M, Klein S, Engerman S (1983) The level and structure of slave prices on cuban plantations in the mid-nineteenth century: some comparative perspectives. Am Hist Rev 88(5):1201–1218

    Article  Google Scholar 

  • Giliomee H (2004) Die Afrikaners—‘n Bibliografie. Tafelberg Uitgewers, Kaapstad

    Google Scholar 

  • Green E (2013) The economics of slavery in the eighteenth-century Cape Colony: Revising the Nieboer-Domar Hypothesis. Int Rev Soc Hist. 1–32

  • Guelke LT (1974) The early European settlement of South Africa. Thesis (PhD), University of Toronto

  • Guelke L, Shell R (1983) An early colonial landed gentry: land and wealth in the Cape Colony 1682–1731. J Hist Geogr 9(3):265–286

    Article  Google Scholar 

  • Gutman H (1975) Slavery and the numbers game: a critique of time on the cross. University of Illinois Press, Urbana

    Google Scholar 

  • Holland J et al (eds) (1960). The Cambridge History of the British Empire, Vol. 1: the old empire from the beginnings to 1783. Cambridge University Press

  • Kotlikoff L (1979) The structure of slave price in New Orleans, 1804 to 1862. Econ Enq XVII:496–518

  • Levendis J (2007) Qualities and effective-quantities of slaves in New Orleans. Southwestern Econ Rev 34(1):161–178

    Google Scholar 

  • Muller A (1983) The impact of slavery on the economic development of South Africa. Kronos 5:1–24

    Google Scholar 

  • Nunn N (2008) The long-term effects of Africa’s slave trades. Q J Econ

  • Shell R (1997) Children of bondage: a social history of slave society at The Cape Of Good Hope, 1652–1834. University Press of New England/Johannesburg, Witwatersrand University Press, Hanover

    Google Scholar 

  • Shell R (2007) Changing hands—an omnibus database for the Cape, 1588 to 1888. CD-ROM publication ISBN 1-86918-062-3. Cape Town: Ancestry24

  • Shell R, Rama P (2007) Breeders or workers? The Structure of slave prices in the Cape Colony, 1823–1830 Safundi. 8(4):413–433

  • TEPC (2008) Verbatim copies in the Cape Town archives repository and digital images of Cape documents in the Nationaal Archief, The Hague, regarding VOC employees (1656–1790). TEPC Project and Sentrum vir Besigheids- en Taaldiens, Cape Town

    Google Scholar 

  • Van Duin P, Ross R (1987) The economy of the Cape Colony in the eighteenth century. Leidein, Intercontinenta no 7

  • Vedder R, Stockdale D (1975) The profitability of slavery revisited: a different approach. Agric Hist 49(2):392–404

    Google Scholar 

  • Ward JR (1978) The profitability of sugar planting in the British West Indies, 1650–1834. Econ Hist Rev 31(2):197–213

    Article  Google Scholar 

  • Worden, N. (1985). Slavery in Dutch South Africa. African Studies series, no. 44. Cambridge, New York. Cambridge University Press

  • Worden N, Crais C (1994) Breaking the chains: slavery and its legacy in the nineteenth century Cape Colony. Witwatersrand University Press, Johannesburg

    Google Scholar 

Download references

Acknowledgments

We thank participants of the ERSA/FRESH conference, the Economic History Society of Southern Africa session at the ESSA conference and the SANORD research workshop for their useful comments. We acknowledge Economic Research Southern Africa (ERSA), who published an earlier draft as Working Paper 383, for their financial support of this study. Opinions expressed here do not reflect those of ERSA or conference participants. The authors would like to thank Prof Robert Shell for his constant support of this project, for providing data and for clarifying some of the contextual issues. The authors would also like to thank three anonymous reviewers for their helpful comments and suggestions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Sophia Du Plessis.

Appendix

Appendix

See Tables 4, 5, 6 and 7.

Table 4 OLS Cobb–Douglas regressions of total value
Table 5 Quantile regressions of log (total value)—1700
Table 6 Quantile regressions of log (total value)—1709
Table 7 Quantile regressions of log (total value)—1723

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Du Plessis, S., Jansen, A. & von Fintel, D. Slave prices and productivity at the Cape of Good Hope from 1700 to 1725: Did everyone win from the trade?. Cliometrica 9, 289–330 (2015). https://doi.org/10.1007/s11698-014-0116-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11698-014-0116-8

Keywords

JEL Classification

Navigation