Abstract
Research on decentralization in Africa and beyond has made clear that the quality of decentralized governance is highly variable across localities within countries. In light of that variation, this article has three goals: first, we critique existing academic research on the quality of governance in light of work on decentralized governance in Africa; second, we provide a conceptual map of how to theorize subnational variation in the quality of governance in settings characterized by considerable dependence on higher authorities for revenues; and third, we outline a series of data initiatives that offer the opportunity to study local and regional politics in new and exciting ways across the region. We conclude with great optimism about the prospects for innovative work on decentralized governance within countries across the region.
Key Words:
Acknowledgements
We would like to thank seminar participants at George Washington University, Texas A&M, and Brown University for feedback on elements of this article. Our special thanks go to Jan Erk for his patience with our slow-developing drafts and his detailed, insightful comments on an earlier draft of this manuscript.
Notes
1 See Kaufmann and Kray (Citation2008) for a discussion of some of these issues.
2 Where national indicators come with measures of uncertainty, as both the World Bank and Transparency International do in some cases, they are typically conceptualized as providing information on measurement error rather than as theoretically interesting variation.
3 These responses come from the World Bank's Enterprise Surveys, which have been conducted since 2002. We have pooled the survey waves as described in Barber and Wibbels (Citation2012).
4 We toggle the TI scores slightly to ease graphic presentation.
5 TI does report a standard deviation, but this is a standard deviation of the experts' opinion. Likewise, the World Bank reports the standard errors on its governance indicators, but they are of a complicated procedure for modeling governance as a latent variable.
6 Data from the OECD's International Development Statistics database: http://www.oecd.org/dac/stats/idsonline.htm
9 See: http://www.electiondataarchive.org/. A separate project, the Global Elections Database, includes data on subnational elections, but its African coverage is limited.
10 See, for instance, Wucherpfennig et al.'s (Citation2011) data on the settlement patterns of ethnic groups.
11 The discrepancy results from UCDP using a more restrictive definition of violent events.
14 One concern with satellite data on electrification: foreign direct investors in resource-intensive industries build electrical infrastructure of their own. In Liberia, for instance, two cities that experience a substantial increase in post-war electrification are Buchanan and Harbel. The former is a railroad terminus for delivering iron ore from the east, where Mittal Steel has invested; the latter is a Bridgestone-owned rubber plantation.
15 The shading intensity reflects the number of violent events that took place in that particular location.