What we learn from Cape wheat farmers in 1825
Labour (management) may be more important than access to capital
It seems odd that a simple productivity calculation of wheat farmers in the Cape Colony can be relevant today. But I’ll argue below that a new research paper by Jan Greyling and myself offers more than just factoids about agricultural history.
The early 19th-century Cape Colony (see map) is often depicted as an ‘economic and social backwater’ characterised by poverty and slow development. Recent research challenges this view, revealing the Cape economy’s dynamic nature and the prosperity of its farmers. Jan and I investigated the productivity of Cape wheat farmers in 1825 using a newly transcribed series of household tax censuses for the entire Colony. The findings reveal considerable heterogeneity in wheat productivity at the farm level between and within districts. Larger districts were generally more productive, but even within these districts, some farmers were substantially more productive than others.
We then try to identify factors that explain these intra-regional productivity differences. Let's take a closer look at the figure below. We plot the relationship between farm size (proxied by farm output) and efficiency, our measure of productivity. The first thing to observe is that the marginal rate of return to size flattens at different output levels and even starts to decline for some districts, such as Albany. This means that increasing the farm size beyond a certain point does not significantly increase productivity; economies of scale are limited on early eighteenth-century wheat farms.
The second observation is that farms in districts with a lower median output level are, on average, more efficient than farms in districts with a larger median farm size, when compared at the same output level. To illustrate this point, let's consider an output level of 1 ton per farm. At this level, farms in the Cape district are almost 40% efficient, whereas farms of the same size in Albany are almost 70% efficient. This relationship even holds at an output of 100 tons per hectare, where farms in the Cape district are, on average, 80% efficient. In contrast, farms in Stellenbosch, the only comparative district at this output level, are almost 90% efficient.
Smaller farms may, therefore, have been better than larger farms in some places, like Albany, where there were diseconomies of scale. In contrast, the bigger districts, like Stellenbosch, see productivity increase as farms become larger, even if smaller farms in the big districts are less productive than those of a similar size in Albany.
In short, then, the best model of a productive farm, even for the same crop, may differ depending on the local context. There is no one-size-fits-all farm model. That remains true, I suspect, today.
A second lesson emerges from our analysis about the factors positively correlated to farm-level wheat productivity. We find that draught animals and Khoe labour are positively associated with wheat farm productivity, while, surprisingly, slave labour shows no significant correlation. I say ‘surprising’ because the finding contrasts the assumption, widely held in the literature, that slave labour was a crucial source of productivity in the Cape economy. Instead, enslaved labour, we argue, was primarily used as a form of capital and collateral in the dense but informal network of settler debit and credit transactions. Enslaved labour seems to have been a consequence, rather than a cause, of high productivity.
Today farmers continue to rely on capital and labour as major inputs into the production process. But not all, I surmise, are equally important in predicting productive versus unproductive farmers. The main challenge may still be management, specifically labour management on labour-intensive farms, rather than access to capital. Although capital inputs are associated with higher productivity, it is perhaps a consequence rather than a cause of high-productivity agriculture.