The African Political Public Goods Cycle

• In Event: The Political Economy of the Developing World

Abstract

We investigate the existence of Political Business Cycles (PBCs) in Sub-Saharan Africa. There are good theoretical reasons to expect such effects in Sub-Saharan Africa, and evidence for them in other regions, but no recent studies of PBCs in Sub-Saharan Africa. With rare exceptions the valuable evidence that does exist focuses on policy instruments, especially fiscal expansions, but does not look at realised outcomes on which citizens are likely to condition their votes.

In order to examine the existence of PBCs in Sub-Saharan Africa we use geocoded Afrobarometer surveys and Demographic and Health Surveys going back to the late 1980s to create two country-district-round panel datasets. Unlike much of the older PBC literature which use crude measures such as an election-year dummy variable, we use proximity to elections as our main RHS measure inasmuch as we would expect the effects of public goods provision close to an election to linger afterwards. We focus our attention on objective measures of access to public goods like electricity and piped water and the existence of schools and health clinics, as well as subjective assessments of personal and national living conditions.

Using two-way fixed effects and country-specific time trends, while also controlling for urbanization and democracy and weighting our results by country-round, we find clear evidence of PBCs across both datasets and a number of different outcomes. These results are robust to disaggregating by parliamentary and presidential elections, adding further control variables and using country-regions rather than country-districts. In accordance with our expectations we also demonstrate evidence for a relationship between electoral proximity and subjective measures of performance, trust and levels of corruption of government officials. Moreover, we find that PBCs are more likely in places with higher levels of political competition, as measured either by the average margin of victory or the previous election’s margin of victory, which is consistent with the incentives driving PBCs. These findings have important implications for our understanding of the link between elections, political competition and development outcomes in Sub-Saharan Africa and beyond.

Authors

• Elliott D. Green, London School of Economics

• Robin Harding, University of Oxford