Session 2: State Effectiveness – Strengthening Governance

The session was opened by Professor. Burgess (LSE) who presented a paper on Kenya, trying to identify the link between the the President’s ethnicity and the amount of roads building in his coethnic district, i.e. the district where the president’s ethnic group is majority. He showed that over the 1964-2002 time span, when democracy was in place, the districts that are coethnic with the president receive on average twice as much expenditure on roads than others, although they receive three times more during autocractic periods. As a consequence, he concludes that even an imperfect democratic process can improve the quality of decision making, ultimately forcing politicians to share more across ethnic groups. Finally he stresses that the role of civil society is crucial in this whole democratization process.

Mr. James then followed with a presentation on a set of randomized control trials that he and this team conducted in Sierra Leone, trying to understand how information availability can affect the voting behavior of citizens. Exploratory research in Sierra Leone showed that 70% of voters are illiterate, and that the majority of them rarely know the name or the programme of the different candidates. As a result of the experiments, Mr.James shows that encouraging public debates between the candidates does translate into a voting behavior which is more in line with the voter’s preferences, ultimately making the political contest more competitive and fair.

The last presentation was given by Professor. Suri (MIT) who started acknowledging that democratic participation is a major challenge everywhere, but more so in low income countries where low alphabetization rates and the novelty of democratic institutions itself results in extremely low rates of electoral participation. As a consequence, Prof.Suri implemented a series of randomized control trials in Kenya across the 2013 election period, using mobile technology to send different types of text messages to incentivate voters’ participation to the local elections. It was a fortunate coincidence that her experiments coincided with the establishment of a new biometric voters’ register and of an independendent authority to monitor the electoral process.The results were positive overall, in so far as participation substantially increased as a result of the different treatments.

Finally, the session was closed by a stimulating discussion between policymakers, panelists and attendees. In particular, Mrs. Kafarama (Advisor for the Office of the Presidency in Rwanda) underlined how using home-grown solutions to increase participation is often the best strategy. She brought the example in Rwanda of regular community meetings which seem to have improved on the voters’ perception of the candidates’ accountability. In this way, the citizens’ level of trust towards institutions was substantially increased, ultimately resulting in higher participation rates.

By Novella Maugeri, Country Economist, IGC Mozambique