The data revolution in African Economic History

A few years ago, I took my graduate Economics students to the Cape Town Archives. Most of them had never been there, and it was fun to show them the original manuscripts that contained the information some of them had used in essays and dissertations. While my students were exploring the archive, I found myself with about an hour to burn – so I did what many of us would do: I ordered a few manuscripts that had interesting descriptions but which I knew nothing about.

That is how I stumbled on the Cape Colony Voters’ Rolls, lists of all residents between 1872 and 1909 who could vote in the Cape Colony. They are extensive – we have only a very rough idea of how many individuals are listed – but there are almost certainly several hundred thousand. The Voters’ Rolls include individuals that are often excluded from other statistical sources; the names of black and colored residents who could vote are listed here, for example, with their addresses and occupations. With the help of some hard-working Ugandan colleagues, these lists are now being transcribed into digital format. One of my PhD students, Farai Nyika, will use them to analyze representation and malapportionment in late 19th century South Africa.

A revolution from below

Increased access to such data at the individual-level is driving what has been labelled the “renaissance in African economic history.” It is what I like to call a “revolution from below.” In a recent paper published in the Journal of Interdisciplinary History, I outline how the digitization and transcription of large historical datasets are reshaping the writing of African economic history. Of course, nothing is ever entirely new: the “revolution from below” stands on the shoulders of older generations who often also used large datasets, like Philip Curtin’s work on African slavery in the 1960s. Yet the introduction of the digital camera and advances in cheap computational power combined with new statistical techniques and packages have pushed the research supply-curve out considerably.

One aim of the data revolution is to use more representative, unbiased sources than the aggregate statistics published in colonial records, which was the staple of earlier research. Researchers often use such sources of information at the individual-level for purposes orthogonal to those for which they were originally collected. Consider, for example, Anglican marriage records that are now used to investigate gender equality and intergenerational social mobility (Meier zu Selhausen 2014, Meier zu Selhausen & Weisdorf 2016), or military attestation forms that provide evidence on individual heights, an anthropometric measure used to proxy for living standards. These sources themselves cannot escape issues of selection bias entirely – see, for example, the lively debates on the interpretation of colonial marriage markets (De Haas & Frankema 2016) or the use of attestation forms to assess living standards over time (Bodenhorn et al. 2015). But the advantage is that the direction and size of selection bias is now more explicit and can be tested with alternative data sources (like historical cohorts in contemporary census data) or accommodated with more advanced techniques (like truncated regressions in anthropometric research).

Individual-level records also allow for greater spatial variation than what could be obtained from colonial offices that, at best, provided statistics at the district level. Combining such spatial variation during the colonial or even pre-colonial era with modern-day satellite imagery, surveys, or censuses can yield surprising insights into the persistence (or not) of historical events or episodes (Michalopoulos & Papaioannou 2013; Moradi & Jedwab 2016). Many of the papers presented at the VI Annual Meeting of the African Economic History Network, held at the University of Sussex in late October 2016, showcased the popularity of such tools.

How inclusive is the data revolution?

The concern, though, is that African scholars are excluded from much of this research activity. Of the 50 authors I cite that have contributed to this data revolution in my paper, fewer than 10 are from Africa. The proportion at the recent conference was the same. The statistical methods are often not taught in African economic history departments, and large financial resources required to undertake digitization and transcription projects are not available in Africa. Scholars and students from African universities also find it difficult to attend the international conferences that will allow them to stay abreast of the latest methodological and scholarly developments.

There is no easy answer to these issues. Collaborative projects between European and African universities could help, but these are often limited to short-term interactions with little knowledge and skills transfer. A more sustainable solution, I argue, is to sponsor larger numbers of African students for graduate and PhD studies at the best universities in Europe and the US. Equipping African scholars with the tools to take part in the data revolution (and building the networks to partake in academic debate) is not only necessary to redress the inequalities of the past, but also vital to build a thriving interdisciplinary academic discourse, a true revolution from below.

References

Bodenhorn, Howard, Timothy W. Guinnane and Thomas A. Mroz (2015), “Sample-selection biases and the “industrialization puzzle,” National Bureau of Economic Research No. 21249.

De Haas, Michiel and Ewout Frankema (2016), “Tracing the uneven diffusion of missionary education in colonial Uganda: European influences, African realities and the pitfalls of church record data,” AEHN Working Paper Series No. 25.

Fourie, Johan (2016), “The data revolution in African economic history,” Journal of Interdisciplinary History 47(2), pp. 193–212.

Jedwab, Remi and Alexander Moradi (2016), “The permanent effects of transportation revolutions in poor countries: evidence from Africa,” Review of Economics and Statistics 98(2), pp. 268-284.

Meier zu Selhausen, Felix (2014), “Missionaries and female empowerment in colonial Uganda: New evidence from Protestant marriage registers, 1880–1945,” Economic History of Developing Regions 29(1), pp. 74-112.

Meier zu Selhausen, Felix and Jacob Weisdorf (2016), “A colonial legacy of African gender inequality? Evidence from Christian Kampala, 1895–2011,” The Economic History Review 69(1), pp. 229-257.

Michalopoulos, Stelios and Elias Papaioannou (2013), “Pre‐Colonial Ethnic Institutions and Contemporary African Development,” Econometrica 81(1), pp. 113-152.