Real wages at the Cape of Good Hope: a long-term perspective, 1652-1912

Literature › Works

Pim de Zwart (2012). Real wages at the Cape of Good Hope: a long-term perspective, 1652-1912. International Institute of Social History (draft working paper).

Type	working paper
Scope	Economics and economic history of apartheid
Method	econometric
Period studied	1652–1912
Regions	Cape Colony, Cape Town
Themes	origins segregation, labour markets wages, inequality poverty unemployment, methods data sources, comparative theory, agriculture land
Citation key	zwart2012

Argument

De Zwart constructs the first long-run real wage series for the Cape (1652-1912) using Allen's subsistence-basket method, as a proxy for living standards where GDP data are lacking. He finds near-stagnant European real wages under the VOC, which does not support the view of a dynamic eighteenth-century economy. Welfare rose sharply in the first decades of British rule, which is consistent with Brunt's institutional account. Slope-dummy regressions show no marked acceleration after 1870, so the traditional mineral-driven 'take-off' is not supported by wage evidence. Cape workers' standard of living is placed against London, Amsterdam and Leipzig.

Evidence and method

Wage and price data from the VOC muster rolls and Cape cashbooks (Dutch National Archives, Western Cape Archives), Du Plessis and Du Plessis VOC wage data, Theal's Records of the Cape Colony, British Colonial Blue Books and Statistical Tables (1835-1913), and the South African Commercial Advertiser. Prices come from VOC expense bills, trade books, ship pay-ledgers, rendementen and Orphan Chamber auction lists. Comparative European series are from Allen et al. (2011).

Place in the debates

• Was the Cape economy stagnant before the 1870s mineral discoveries? (traditional/Feinstein vs Van Duin and Ross revisionists) (see Pre-Union colonial economy, the mineral revolution and the South African War). Takes a middle position: real wages were stable, not rising, in the VOC era, which does not support a dynamic eighteenth-century economy. Stable welfare at about twice subsistence is still seen as respectable given population growth. Rejects a post-1870 take-off.
• Institutions and property rights vs resource windfalls in colonial growth (Brunt; Acemoglu et al.) (see Pre-Union colonial economy, the mineral revolution and the South African War). Supports Brunt on the importance of early British reforms, since wages jumped after 1795. Finds the 1843 land-tenure change neither confirmed nor contradicted by wage data, and finds a limited effect of the resource windfalls on workers' real wages.
• VOC as stimulant or constraint on the Cape economy (see Pre-Union colonial economy, the mineral revolution and the South African War). Doubts Van Duin and Ross's claim that the VOC stimulated growth, noting that production growth could have occurred despite the Company and that population growth could offset rises in per-capita income.

Findings

• European wages doubled in the first decades of British rule after a long period of stable VOC wages, while wage rises for coloured labourers were far smaller. (p. 9)

  after the VOC era of stable wages, wages of the European workers doubled in the first decades of British rule, and continued to rise after that.

• Nearly all worker categories lived above subsistence, mostly at twice subsistence or higher; the exception was coloured rural workers in 1826-50, based on only three observations. (p. 12-13)

  In fact, most labourers had a standard of living twice subsistence level or higher.

• Unskilled real wages grew only marginally in the Dutch period and skilled trend was not significantly different from zero in some models, so the evidence does not indicate a dynamic eighteenth-century Cape. (p. 14)

  unskilled real wages grew marginally in the Dutch period (0.16 per cent per year or less)

• The 1795 slope dummy is significant in all models, with larger coefficients for shorter periods, suggesting an immediate effect of the British takeover and supporting Brunt. (p. 16)

  the 1795 dummy is significant for both skilled and unskilled real wages in all ‘models’ shown in table 3.

• Wages rose more than prices after 1795, producing real welfare gains during a trade boom after British liberalisation. (p. 16)

  Both wages (table 1) and prices (figure 2) increased tremendously during this period, but wages clearly increased the most which led to growing welfare.

• The pace of welfare growth did not change significantly after 1870 once dummies for other years are introduced. (p. 16)

  the pace of welfare growth did not change drastically after 1870.

• Cape unskilled living standards moved from lowest among the four cities to above Leipzig by mid-eighteenth century and on a par with the continent from c.1800. (p. 16-17)

  by the middle of the eighteenth century their standard of living had surpassed that in Leipzig and began to close the gap with London and Amsterdam.

• After the mineral discoveries, ordinary Cape Town workers benefited only marginally and real wage growth lagged behind European cities, contradicting a clear take-off from the 1870s. (p. 17)

  Cape real wage growth after the mineral discoveries stayed behind compared with that in Amsterdam, Leipzig and London, which did not have the advantage of resource windfalls.

• Rising welfare in the nineteenth century came with greater inequality between European and coloured labourers and between skilled and unskilled workers. (p. 13)

  During the nineteenth century, this rise came at the cost of greater inequality, both between European and coloured labourers, as well as between skilled and unskilled workers.

• The paper concludes that institutional change under the British directly affected wage workers' welfare, while resource windfalls had limited impact on real wages. (p. 18)

  institutional changes implemented by the British after their takeover of the Cape had direct effects on the welfare of wage workers, while the impact of resource windfalls on real wages was limited.

Evidence ledger

Sources used

Source	Kind	Period	How used
VOC muster rolls (Cape establishment, 1657, 1658 and five rolls 1699-1790), Dutch National Archives	company records or reports	1657-1790	Gives fixed monthly wages of Company employees, the baseline for Dutch-period wages.
Report of the Council of Policy 1717 (WCA C 2749) and VOC cashbooks 1754 and 1786 (WCA C2734)	state archives	1717-1786	Documents additional emoluments (board, wheat) added to VOC monthly wages.
Wagenaar 1666 note on free servants' wages (Moodie, The Record) and Van Imhoff 1743 report (Merriman, Reports of Chavonnes and Van Imhoff)	commission or official report	1666-1743	Scattered evidence that free knechts and free craftsmen earned much more than VOC employees, used to correct VOC wages upward.
Van Duin and Ross (1987) data on shares of VOC employees, farmers and knechts among the free male population	secondary literature compilation	1701-1791	Supplies the occupational shares used to weight VOC and free-knecht wages and to argue wage labour was representative; also the eighteenth-century output and population series.
Western Cape Archives financial documents (Burgher Council, Burgher Senate, Civil Engineer, Colonial Office, First British Occupation, Master of the Supreme Court)	state archives	1795-1835	Source of wage observations (and prices) for 1795-1835, averaged within years.
G. M. Theal, Records of the Cape Colony, 36 vols.	commission or official report	1795-1835	Supplies wage observations, including Khoikhoi and coloured wages, for the early British period.
Colonial Blue Books of the Cape Colony, CO 53/72-CO 53/88 (1837-1854)	official statistical series	1837-1854	Average wage rates for European and coloured workers and average market prices of consumables.
Statistical tables relating to British self-governing dominions, colonies, possessions and protectorates, CO 442/27-CO 442/63 (1854-1913)	official statistical series	1854-1913	Wage rates by occupation and race, plus prices of consumables, produce, imports and exports, for 1854-1913.
VOC Received letters and documents from the Cape of Good Hope (VOC 3988-4355): expense bills and trade books	company records or reports	1653-1795	Seventeenth-century price series for the subsistence basket.
VOC Scheepssoldijboeken (ship's pay-ledgers), VOC 5269-5270	company records or reports	1633-1795	Eighteenth-century prices for basket goods from expenses of resupplying ships at the Cape.
VOC rendementen for the Cape, 1747-1775	company records or reports	1747-1775	Additional eighteenth-century price data from profits on sales of goods.
TEPC Transcription Project, Inventories of the Orphan Chamber (auction lists)	state archives	eighteenth century	Auction prices for goods from estates, used to supplement price series.
South African Commercial Advertiser, 1826-1853	newspapers periodicals	1826-1853	Fills gaps in nineteenth-century price series; prices supplied by Jorrit Bakker.
Allen et al. (2011) data and basket methodology for London, Amsterdam and Leipzig	secondary literature compilation	1653-1913	Barebones basket, 3.15 family multiplier, and European benchmarks for comparing Cape subsistence ratios.
Currency and exchange data: Van Zanden IISH prices and wages site; Clark England prices and wages; Jacobs (2000); Arndt (1928) and Neumark (1957)	secondary literature compilation	1652-1913	Converts guilders, rix-dollars and pounds sterling into grams of silver.
Conversion sources: Verhoeff (1983), GPIH conversion datasheets, Böeseken (1938), Hariharan (2002), Wisconsin energy tables	secondary literature compilation	various	Converts old Dutch and British units to metric and calculates cotton and fuel prices per basket unit.
Other published comparisons: Brunt (2007) output and real wage series; Boshoff and Fourie (2010); Fourie and Van Zanden GDP; Feinstein (2005) trade and immigration figures	secondary literature compilation	1701-1913	Provides hypotheses and breakpoints (1795, 1814, 1827, 1843, 1870) tested against the real wage series, plus population and trade figures.

New data

A new long-run Cape real wage dataset, 1652-1913. It has nominal daily wage series for European and coloured workers (unskilled, skilled, rural) and a consumer price index for a barebones basket. Both are built from VOC, Western Cape and British archives, and combined into subsistence ratios. The author says the data will be made available on the IISH HPW and GPIH websites.

Inference

Quantification: Econometric (causal design).

Allen-style subsistence-basket real wages are constructed from archival wage and price data and compared with European cities. Breakpoint regressions with slope dummies for 1795, 1808, 1814, 1827, 1838, 1843 and 1870 are run on log subsistence ratios of European skilled and unskilled workers. The estimated trends are used to test competing historiographical claims (stagnation, VOC dynamism, British institutional take-off, resource-windfall take-off).

Key claims

Claim	Basis	Strength
Real wages were roughly stagnant under the VOC, so the data do not support a dynamic eighteenth-century economy. (p. 14)	Table 1 wages and Table 3 trend coefficients, built from VOC muster rolls and VOC price documents.	Partial data
Welfare rose markedly in the early decades of British rule, supporting Brunt's claim of output growth after 1795. (p. 16)	Significant 1795 slope dummies in all models, built on wage observations from the Western Cape Archives and Theal, and on price data for 1795-1835.	Partial data
There was no take-off in welfare after 1870; resource windfalls had limited impact on workers' real wages, which lagged behind Europe. (p. 17)	1870 slope dummy insignificant once other breakpoints are added, plus comparison with the Allen et al. European series (Figure 6).	Partial data
Most Cape workers lived at about twice subsistence or more, and Cape unskilled living standards caught up with continental Europe by the early nineteenth century. (p. 13)	Subsistence ratios (Fig. 4) and the comparison with London, Amsterdam and Leipzig (Fig. 6).	Partial data
Wage labourers, especially the European workers analysed, are representative of the lower classes at the Cape. (p. 8)	Shares of wage earners from Van Duin and Ross for the eighteenth century; for the nineteenth century, inference from population growth and labour legislation, with no direct data.	Illustrative or anecdotal
Inequality grew in the nineteenth century between European and coloured workers and between skilled and unskilled workers. (p. 13)	Table 1 and Figure 4 comparisons across worker categories; the coloured series rest on few observations.	Partial data
The post-1814 decline and later wage movements are tied to British indentured migrants, the 1828 legal equality and the abolition of slavery. (p. 16)	Slope dummy signs and significance in Table 3, interpreted using secondary literature (Newton-King); no direct test of mechanisms.	Illustrative or anecdotal

The author's caveats

Early European rural wages and all coloured series before 1854 rest on few observations (24 for early rural wages). Prices mix wholesale, institutional and consumer prices, and no retail mark-up is applied. Unskilled urban wages for the 1830s-40s are extrapolated from skilled wages, and gaps in the wheat and bean prices after 1870 are filled by regression. Unrealistically high knecht observations are dropped, and the analysis is limited to European workers because other series are inadequate. The author notes wages are only a proxy for living standards.

Gaps in the evidence

The analysis covers only European urban workers in Cape Town, so the experience of Khoisan, slave and coloured labourers, and of rural workers, is barely tested. The small and uneven wage observations in key periods (1795-1850) and the use of wholesale prices and interpolation make the breakpoint regressions sensitive to data construction. The slope-dummy approach has no controls and cannot separate institutional change from migration or price shocks. Independent output or income data (the Fourie-Van Zanden GDP estimates, tax rolls, census and opgaaf data) and more series for rural and coloured wages exist in the archives and could be used to cross-check.

Replicability: Data described but not public.