Poor governance due to lack of political accountability is often cited as an explanation for the low level of economic development in sub-Saharan Africa. Lack of political accountability can emerge when voters do not choose their candidates according to their expected performance. In sub-Saharan Africa, voters often use the ethnic profile of a candidate as an informational shortcut for the candidate’s political agenda (Ichino and Nathan 2013). As a consequence, politicians rely on tribal allegiances that deliver the votes of co-ethnics irrespective of performance (Casey 2013). In the 2013 Kenyan elections for instance, co-ethnicity was a strong predictor of votes.
Tribal allegiances may come from undeveloped media markets that leave citizens without relevant information on which to base their vote. Casey (2013) gives an illustration with the case of Sierra Leone – using differential radio coverage, she shows that increased information about candidate competence presents a counterpoint to tribal loyalty and strengthens accountability.
Therefore, more developed media markets should be associated with better governance and higher economic development in sub-Saharan African countries. However, any empirical evaluation of this hypothesis faces the challenge that common features may drive simultaneously media and economic development.
In a recent paper (Cagé and Rueda 2014), we circumvent this empirical challenge by using a sharp empirical strategy relying on our knowledge of history. We analyse the relationship between media, democracy, and economic development. We show that the introduction of the printing press in the 19th century in sub-Saharan Africa durably affected the newspaper industry. More developed media markets enhanced, in turn, local political participation and economic development. Our results suggest that a better understanding of African media development will be key for the future of African democracy and economic change.
Protestant missionaries pioneered the development of a written tradition for sub-Saharan African languages. Wherever they went, Protestants quickly formalised indigenous languages and printed Bibles and educational material in these languages. They facilitated access to the printing press, acting as intermediaries for its diffusion. Therefore, most of the first indigenous newspapers were printed and sponsored at mission centres. The first newspaper intended for black readers, the Umshumayeli Wendaba (‘Publishers of the News’), written in Xhosa, was published as an irregular quarterly in 1837 and printed at the Wesleyan Missionary Society in Cape Colony. Isigidimi samaXhosa (‘The Xhosa Messenger’), the first African newspaper edited by Africans, was first released in 1876 and printed at the Lovedale Mission Press in South Africa. In 1884, the English/Xhosa weekly Imvo Zabantsundu (‘The African Opinion’), the first black-owned and controlled newspaper in South Africa, was published. On the contrary, in regions where Protestant missions were less active, the first newspapers appeared only at the beginning of the 20th century, and no indigenous newspapers were created before WWI. The first paper in Ivory Coast to be owned and edited by an African, the Éclaireur de la Cote d’Ivoire, only appeared in 1935. This lag of more than one century in the creation of the first indigenous newspapers might explain the persistent effect of the proximity to a printing press on newspaper readership today.
To identify this persistent effect, our econometric analysis attempts to move beyond two forms of selection. First, historical and geographical characteristics might have determined mission station location, preventing us from comparing regions close and far from these settlements. Protestant missionaries did choose to locate in geographically favoured areas. To address this issue, we restrict our sample of analysis to regions near missions. Because regions near Protestant missions shared similar geographic, institutional, and cultural environments, this restriction isolates the specific effect of the printing technology from other possible long-term determinants of newspaper readership embedded in specific mission locations.
Second, Protestant stations invested in different activities such as printing, health, and education. There may be endogenous selection of missions into printing. To address this concern, we instrument proximity to the closest printing press. In the 19th century, missionaries formed numerous societies that were not equally inclined to the same activities. For each mission, we estimate the probability that it was endowed with a printing press using the share of missions from the mission’s society equipped with a printing press in all the regions of the world outside sub-Saharan Africa.
New data on missionary activity
We built and geocoded an entirely new dataset of Protestant mission locations and of their investments in education, health, and printing, using historical archives. Producing these data is the first contribution of this research. We locate all the Protestant mission stations in 1903. For each mission, we provide information on the nature and size of the investments performed. Among others, we identify the mission’s size (number of students, foreign missionaries, and native workers) and its investment in schooling, health facilities, and printing presses. We also identify the society to which each station is affiliated. Our sample of sub-Saharan African missions includes a total of 723 Protestant missions, of which 27 were equipped with a printing press in 1903. Figure 1 shows the location of mission stations and their printing presses in 1903.
Figure 1. Mission stations with and without a printing press in 1903
The printing press and newspaper readership
Using contemporary individual-level data from the Afrobarometer, we find that proximity to the closest location of a mission with a printing press has a positive and statistically significant impact on the probability of reading the news. A one-standard deviation increase in the proximity to a mission with a printing press increases the probability of reading the news on a monthly basis from 3% to 14% of a standard deviation, depending on the specifications (in contrast, proximity to a mission without a printing press has no significant impact on newspaper readership).
This coefficient is statistically significant and economically meaningful. Proximity to the printing press and the other covariates together explain 21.3% of the total variation in newspaper readership. Of these 21.3%, the distance to the printing press explains 0.4% to 2.9%. Moreover, the impact of the distance to the printing press is of the same order of magnitude as the intensity of cash constraints (a proxy for living standards) and the distance to the capital, and only slightly smaller than the urban area dummy. The effect of the respondent’s level of education is around ten times higher. This difference is not surprising, since distance to the printing press captures a historical effect that might have been attenuated over time.
The printing press and contemporary economic development
Finally, we document a strong association between proximity to a printing press and contemporary economic development as proxied by satellite images of light density at night and population density.
Figure 2 shows the evolution of population density around missions depending on whether they had a printing press. Missions start to diverge in the late 19th century. Population density is estimated to be systematically higher around missions with a printing press after missions started to settle (the average arrival date of missions in Africa is 1850).
Figure 2. Population density around missions
Figure 3 plots our measure of average light density for Afrobarometer towns as a function of the distance to the printing press. The plot presents suggestive evidence of a negative relationship between the two variables.
Figure 3. Proximity to a printing press and light density
Turning to the econometric analysis, we find that a one-standard deviation increase in the distance to the printing press significantly decreases light density today by 10.7% of a standard deviation. How to explain such an effect? Weber’s hypothesis of the ‘Protestant Ethic’ (Weber 1930) suggests that the Protestant religion per se might drive economic development. However, recent studies have challenged this interpretation and have emphasised the importance of human capital accumulation among Protestants (Becker and Woessmann 2009, Cantoni 2012). Building on Habermas (1989), Dittmar (2011) argues that proximity to the printing press would have enabled the development of a culture of information exchange and the development of an urban, bourgeois public sphere. Newspaper readers would be more informed citizens more likely to cross party lines when voting (Casey 2013). Political accountability is increased, and so the allocation of public spending is more equitable, leading to economic development. We examine these potential channels. We show that the only variables capturing the effect of the distance to the printing press on economic development are newspaper readership and Protestantism. Protestantism negatively affects economic development nowadays, contradicting Weber’s assumption. The development of the culture of information and the bourgeois sphere is the main channel through which the printing press had a long-term impact on economic development.
This research studies the impact of the early introduction of the printing press by Protestant missionaries in the 19th century. Our conjecture has been that the early introduction of the printing press had long-term effects on newspaper readership. The evidence we obtain is consistent with this hypothesis. Moreover, we document a long-term impact on contemporary economic development, and show that the early development of the newspaper industry and reading habits can explain part of this effect. This suggests that the promotion of well-functioning media institutions will be key for the future of African development.
Becker, S O and L Woessmann (2009), “Was Weber Wrong? A Human Capital Theory of Protestant Economic History”, Quarterly Journal of Economics, 124(2): 531–596.
Cagé, J and V Rueda (2014), “The long-Term Effects of the Printing Press in Sub- Saharan Africa”, PSE Working Paper.
Cantoni, D (2012), “Adopting a New Religion: The Case of Protestantism in 16th Century Germany”, Economic Journal, 122(560): 502–531.
Casey, K (2013), “Crossing Party Lines: The Effects of Information on Redistributive Politics”, Working Paper.
Dittmar, J E (2011), “Information Technology and Economic Change: The Impact of The Printing Press”, Quarterly Journal of Economics, 126(3): 1133–1172.
Habermas, J (1989), The Structural Transformation of the Public Sphere: An Inquiry into a Bourgeois Society, MIT Press.
Ichino, N and N L Nathan (2013), “Crossing the Line: Local Ethnic Geography and Voting in Ghana”, American Political Science Review, forthcoming.
Weber, M (1930), The Protestant Ethic and the Spirit of Capitalism, Unwin Hyman, London and Boston.