To What Extent Can Indirect Elections Reduce the Frequency of Contested Presidential Election Results in Africa?

Abstract

A contested election result can undermine the progress of a state, including fueling lengthy and costly litigations and electoral violence. Given the rise in cases of contested election results despite some improvements in the quality of electoral processes in Africa over the years, this study investigates the extent to which the adoption of indirect elections by African states could reduce the frequency of contested presidential election results in these countries. Evidence was gathered by using non-probability stratified sampling to select 25 African countries and 6 non-African countries and collecting secondary data on the frequency of contested elections, choice of electoral system, median Gross Domestic Product per capita, quality of election processes and the extent of power sharing from 1989 to 2020. The dissertation regresses the frequency of contested elections (dependent variable) on the other variables. The choice of electoral system and the interaction between the electoral system and the quality of election processes are significant at p-levels of .003 and .029 respectively. This means that the correlation between electoral systems and the frequency of contested elections had only a 0.3 percent chance of being observed if the occurrence was due to mere chance. The probability was 2.9 percent for the correlation between the interaction term and the frequency of contested elections. For Senegal and Ghana, two African countries that use direct elections and possess the highest quality of electoral processes in the sample, the regression model predicts that the historical frequencies of contested elections in these countries are more than 5 times the mean frequency of contested elections of indirect election countries. As the quality of election processes decline among direct election countries within the sample, the gap between the predicted frequencies for direct election countries and the predicted frequencies for indirect countries widens, with the latter always being lower. This signals greater benefits for African countries with lower quality of election processes than those of Ghana and Senegal, if they do adopt indirect elections. It follows that African countries using direct elections are expected to report lower frequencies of contested election results if they adopt indirect elections. The policy implication is that African states relying on direct elections need to explore how they might transition to indirect elections in order to limit the frequency of contested presidential election results.