Author: Igor Martins (Lund University)
In 1807, Britain outlawed the slave trade across its empire. Yet in the Cape Colony, households continued accumulating slaves for decades afterwards. New evidence from digitised tax censuses suggests the answer lies not in labour demand, but in the financial value of enslaved people as capital assets.
A supply shock and a puzzle
The 1807 Slave Trade Act was a landmark in the history of abolition. By banning the importation and transhipment of enslaved people within the British Empire, it imposed an immediate supply constraint on slave-holding colonies. In the Caribbean, slave populations began to decline. At the Cape Colony, the legal end of imports drove prices upward. The standard expectation would be that slaveholding contracted, or at least shifted toward those who could extract the greatest productive value from increasingly expensive labour.
That is not what happened. Using newly digitised household-level tax records (the opgaafrollen) from two contrasting Cape districts, I find that slave accumulation continued in both. Stellenbosch was a settled, densely populated agricultural region dominated by viticulture and crop farming. Graaff-Reinet was a sparsely settled pastoral frontier, characterised by livestock herding and greater access to alternative (Khoe) labour. The two districts differ in ecology, production structure, population density, and institutional setting. Standard theories of coerced labour predict divergent responses to a supply shock across such different environments.
Yet the data show convergence, not divergence. Both districts exhibit continued accumulation after 1807, at a slower pace than before, and with no statistically distinguishable difference in the post-abolition trajectory. This symmetry poses a problem: what could generate a common response in structurally dissimilar settings?
Five theories, one survivor
The paper tests five theoretical frameworks against the household-level evidence. The Nieboer-Domar hypothesis links slavery to land abundance and labour scarcity, predicting stronger reliance on coerced labour in open-frontier settings like Graaff-Reinet. Principal-agent models emphasise supervision costs and task observability, predicting sharper declines in districts with harder-to-monitor activities. Frontier-closure theories associate slavery with unsettled land and weak state consolidation. Outside-option models predict that coercion weakens where slaveholders have access to substitute labour or where the enslaved face fewer barriers to exit.
None of these production-based accounts survive the empirical test cleanly. The absence of district divergence undermines models that hinge on local ecological or institutional conditions generating differential responses to a common shock. The framework that best fits the evidence treats enslaved people not only as labourers but as capital assets: stores of value in a capital-scarce colonial economy with weak financial institutions. Wealthier households continued accumulating slaves after 1807, but persistence extended well into the middle of the wealth distribution, where production-based profitability alone is an implausible explanation. Retention under rising prices is also inconsistent with a pure status-consumption logic, which would predict divestment of luxury goods as their cost increases.
The capital interpretation is reinforced by the mechanisms sustaining slave stocks after the trade ban. Demographic reproduction increased the share of children in household slaveholdings, gradually replenishing adult stocks. At the same time, redistribution from exiting to continuing households sustained adult holdings, as those leaving the census divested while those remaining absorbed their slaves. Both channels are consistent with a logic in which slaveholding preserved a household’s asset base, not merely its labour supply.
Implications
These findings contribute to a growing literature that reframes slavery as a mechanism of capital accumulation alongside labour exploitation. Across African, Caribbean, and American contexts, scholars have documented how enslaved individuals were pledged as collateral, treated as appreciating assets, and used to access credit. This paper adds household-level behavioural evidence to that picture, showing that post-abolition persistence patterns align more plausibly with financial motivations than with production-based theories alone.
The results also shift the profitability debate. For large estates, productive returns from slavery may well have been substantial. But for small and middle households, persistence under a supply-driven price shock points toward wealth storage and collateral value rather than scale-based agricultural profits. The dual function of slavery as both labour and capital may help explain why coercive institutions proved so resilient, even when the economic logic of labour extraction alone would have predicted contraction.
Igor Martins is a researcher at the Department of Economic History, Lund University.
Contact: igor.martins@ekh.lu.se
Website: www.ibmartins.com
ORCID: https://orcid.org/0000-0002-6141-9273