![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
For economic transactions, including debt transactions, to occur in a market system, property rights are essential. The literature has focussed on finding empirical proof of the effect of property right regimes, noting differences between de jure and de facto property rights. We use a novel combination of data on wealth and demographics to investigate the effects of property right regimes on economic outcomes at the individual level. At the Cape, de jure property rights between freehold and loan farms differed. Historians, however, suggest that de facto property rights between these two property types were the same. We exploit the random variation of the birth order, specifically being the eldest son, to estimate whether the type of farm and therefore the type of property rights, mattered for economic activity, in our case, debt transactions. Our results suggest that historians were correct: loan farms were as secure in their de facto property rights, despite differences in de jure property rights. Our results confirm that the local context in which property right regimes are embedded is at least as important as the property right regime itself.
Notes
1 The authors would like to thank Jeanne Cilliers, Claude Diebolt, Kara Dimitruk, Joost Jonker, Sumner La Croix, Richard Hornbeck, Dieter von Fintel and seminar participants at the universities of Arizona, Stellenbosch and Utrecht for comments on an earlier version of this paper. We also thank Economic Research Southern Africa and the National Research Foundation for financial support.
2 Although farm size, soil quality and distance from Cape Town certainly matter, these are difficult to control for various reasons. Farm size is not recorded specifically in the probates or if there is mention of the size, it is the Company’s prescribed 60 morgen. Land demarcation and the mapping of farms only happened in the midnineteenth century. Soil quality is also difficult to control for since the exact geographic location of the farms are unknown, which is also the reasons it is not possible to control for the distance from Cape Town. The descriptions are often vague, for example, ‘next to a river’, or ‘in the district of’, where districts often covered large areas.
3 It should be noted here that the loan farms system at the Cape was similar to the Dutch system of the sixteenth century. CitationDe Vries & Van Der Woude (1997:161-162) found the tenants in the Netherlands had strong legal support and it was often difficult for owners to replace tenants. They state that ‘tenants acquired de facto permanent possession while the owners held nothing more than an old right to collect a fixed money rental.’ CitationMitchell (2008 Chapter 3, p.4) calls the ‘loan farm system a remnant of Dutch feudal land tenure practice.’
4 See footnote 1 on controlling for distance from Cape Town.
5 Refer to footnote 1 on problems with the measurement of land for economic value, like water rights and soil quality. Without precise geographic data on these farms, we cannot control for these variables. Some information exists on the area of some of these farms, but not to the extent found in the probate inventories and less on the loan farms than the freehold farms. Surveys and information on these farms were more accurately captured toward the start of the nineteenth century and the British colonial period at the Cape, but not for this early period studied in the paper.
6 For detailed information on how the genealogies were compiled and can be used in economic and demographic studies, see CitationCilliers & Fourie (2014).
7 Another form quitrent (erfpagt) was observed, but only 60 are found in the inventories. These were mainly loan farms which were converted to freehold farms. Their tenure was closer to that of the freehold farms and we, therefore, include them as freehold farms. The results remain whether these are included or excluded.
8 If individuals owned both freehold and loan farms, we add them to individuals with freehold farms, because freehold farms were more scarce and valuable. Our results are robust whether we include these observations or not.
9 The results presented here are not sensitive to this grouping.
10 Because debt in the probate inventories does not distinguish between the debt of the wife or husband, it is safe to assume that both have the responsibility to repay these debts.
11 Ideally, we would like to do robustness checks for smaller periods as well, but the current sample size does not allow for this.
12 For more information on these records, see CitationCilliers (2015). She further provides information on which occupations are divided into which skill category.
13 How the settlers bypassed the system and how firstborn sons managed to obtain the freehold farms remain unknown.
14 We can investigate the location of the death of the eldest brother versus brothers born later in either data sources. However, the death locations are not also recorded.
15 The debt on a farm was not inherited with the fixed property, but all debts of the estate was repaid before any inheritance by children were received.
16 For more information on the significance of slaves and the credit market at the Cape, see CitationSwanepoel & Fourie, (2018).
17 Because we only have one endogenous regressor and one instrument, the specification is just identified. Other tests like AR and Kleinbergen Paap do not allow for just identified regressions and cannot be performed here.
18 To further investigate the positive coefficient, would require further information unavailable at the time of writing. This could include the type of farming activities between loan farms and freehold farms, or the profitability of these farms.