Missionaries and female empowerment in colonial Uganda: New evidence from Protestant marriage registers, 1880–1945

ABSTRACT

Protestant missionaries have recently been praised for their comparatively benign features concerning their support of women's education in Africa. Using a novel dataset of 5,202 Protestant brides born between 1880 and 1945 from urban and rural Uganda, this paper offers a first pass at analysing empirically the role of mission education on African women's socio-economic position within the household. The paper finds that although mission education raised the sampled brides' literacy skills way above female national levels, they were largely excluded from participating in the colonial wage labour market. In this context, the missionary society presented an almost exclusive source of female wage labour in areas of religious service, schooling and medical care. While literacy per se did not affect women's marriage behaviour, women who worked for the missionaries married significantly later in life and married men closer to their own age, signalling a shift in the power balance between parents and daughters and between husband and wife. On average, daughters of fathers deeply entrenched in the missionary movement had the highest chances to access wage employment, emphasizing the importance of paternal mission networks for Protestant women's work outside the household during colonial times.

Keywords:

• Missions
• Protestantism
• Uganda
• colonial economy
• African women
• education
• wage labour
• marriage patterns

JEL classification:

• J12
• J16
• N37
• N97
• Z12

ACKNOWLEDGEMENTS

This paper has benefitted from comments by Jan Luiten van Zanden, Tine de Moor, Ewout Frankema, Ellen Hillbom, Jacob Weisdorf, Edward Rugumayo, Angus Crichton, Oscar Gelderblom, Tracy Dennison, Miguel Laborda Pemán, Selin Dilli, Michiel de Haas, Sarah Carmichael, Andrew Roberts, David Martyn, two anonymous referees, and the seminar participants of the Utrecht Social and Economic History Graduate Seminar (March 2014), and the 10th European Social Science History Conference (Vienna, April 2014), the session, ‘Agency, Gender, Human Capital and World Economic Development’. I wish to thank Bernard Asiimwe and Christopher Byomukama from Mountains of the Moon University (Uganda), for excellent research assistance (funded by the University of Southern Denmark). I am particularly grateful to the Church of Uganda, in particular Bishop Reuben Kisembo (Rwenzori Diocese) and Bishop Wilberforce Kityo Luwalira (Namirembe Diocense) for opening their marriage books. The usual disclaimer applies.

Table A1: Fathers' occupational groups with most frequent occupations

Table A2: Correlation matrix

Table A3: Regression results (Kampala) - OLS (dependent variable: bride's age at first marriage)

Table A4: Regression results (Toro parishes) – OLS (dependent variable: bride's age at first marriage)

Table A5: Regression results (Kampala & Toro) – OLS (dependent variable: spousal age difference)

Figure A1: Share of Protestants among Christian followers in Uganda and Buganda, 1901–59

Source: Uganda Protectorate Blue Books 1901–45; Uganda Protectorate Census 1959

Figure A2: Women's age at first marriage frequency distribution (Kampala and Toro), birth cohorts 1880–1945

Notes

2 Hereafter the shorter term ‘Africa’ is used for ‘sub-Saharan Africa’.

3 There are no population numbers for the pre-colonial era. One of the earliest population estimates were provided by the Colonial Blue Book of 1910, revealing that about a quarter of the total population of Uganda resided in Buganda (667,387), while less than 5% of the Protectorate's population belonged to the Toro kingdom (93,946).

4 This contrasts with West Africa where women participated in local trade in colonial Nigeria (Ekechi 1995) and Ghana (Robertson 1984a).

5 Those traditions vary considerably in their manifestation between rural and urban areas today.

6 Admittedly, the relationship between Christian missionaries and British colonialism was much more complex than can be portrayed in this article. Generally, when imperial Britain and missionary interests aligned they worked together – when they conflicted, they competed (e.g., Hansen 1984; Stanley 1990; Porter 2004; Etherington 2005).

7 Data were taken from various issues (1901–45) of the Blue Book of the Uganda Protectorate.

8 Girls' primary school enrollment was calculated on the basis of annual primary mission school attendance from Ugandan Blue Books derived from Frankema (2012).

9 Data were taken from the Blue Book of the Uganda Protectorate (1922).

10 Some notable exceptions include: Nhonoli (1954); Thornton (1977); Katzenellenbogen et al. (1993); Notkola and Siiskonen (2000); Walters (2008); Doyle (2013); Meier zu Selhausen and Weisdorf (2014).

11 Over the course of the colonial era Kampala grew in size, from an urban population of 2,850 in 1912, to 24,000 in 1948 and 46,735 in 1959 (Omolo-Okalebo et al. 2010).

12 This cathedral burned down due to lighting and was replaced by another brick-walled and tile-roofed structure, completed in 1919 (Moon 1994).

13 Only recently the first marriage register (1891–95) was located in the Africana section at Makerere University library (Kampala), indicating that the first marriage solemnized at Namirembe took place on 17 February 1891.

14 According to Rachal (1987) and Schofield (1968, 317) when the motives for literacy were largely religious, there were greater benefits for being able to read (the Bible, tracts, newspapers) than knowing how to write.

15 Their index ranges between 0 and 100 (100 = no age-heaping, 0 = only ages ending on multiples of 5 are reported).

16 Note that this estimate is affected by how the WI is constructed, including only ages ranging 23–72, which as a result fails to capture 82.3% of female marriage ages from Kampala and 70.9% of those from rural parishes as mean marriage age was below 23 in both locations.

17 See Meier zu Selhausen and Weisdorf (2014) for comparative long-term Protestant male literacy statistics for Kampala.

18 Already, Robertson (1984b, 37) emphasized that in African capital cities women ‘do slightly better in the professional category alone, because of their roles as teachers and nurses’. Anecdotal evidence brought together by Little (1973, 31–3) holds that his has been the case in a number of other African countries.

19 For international comparison with the homeland of the Anglican missionaries, Humphries and Sarasúa (2012) offer British female adult labour market participation rates from 1920s for Leeds (37%), Bradford (54), and Middlesbrough (20%). It indicates that labour market participation was significantly higher in Britain than for the sampled female Protestants from Uganda born around 1900. Note that those values are for total females (i.e., married and unmarried).