Strengthening Local Political Accountability in Uganda
In Uganda, district governments are responsible for providing vital public services such as healthcare and education, but government accountability is relatively weak, leading to under-provision and low quality of services. Researchers worked with IPA to evaluate the impact of a program designed to improve local government accountability and the provision of services: disseminating information to citizens about district politicians’ performance via scorecards. Results suggested that disseminating scorecards to citizens was effective in improving district politicians’ job performance, but only when those politicians’ electoral districts were already competitive. The scorecard dissemination did not lead to corresponding improvements in public service delivery in those districts, however.
Government institutions are widely understood to be more effective purveyors of public services when they are accountable to citizens. Voters in many electoral democracies in low-income countries have limited information about their representatives’ political decisions and actions.1 One potential solution to this problem is a strategy of decentralization, in which decisions are increasingly made by local councils.2 However, as with national governments, there is limited accountability for local councils as well.3 Policymakers are interested in strategies for improving accountability on a local level. Recent research questions whether providing citizens with information about elected officials’ performance may be a potential mechanism for improving accountability and awareness of political issues.4This matter is politically sensitive, as politicians have the incentive to derail civil society transparency initiatives aiming to disseminate such information.5 Recent meta-analysis evidence finds that disseminating politician performance information to citizens directly prior to elections does not affect the reelection of poorly performing politicians.6 Unlike previous research, this study examines the effect of a civil society transparency initiative that disseminates politician performance information early in the electoral cycle. Further, this study examines the effect of such an initiative on politician performance in the years after the dissemination (and in a follow up study, the reelection of good and poor performing incumbents).
Context of the Evaluation
Uganda has decentralized its government by implementing a system of local councils. Unfortunately, there exist too few channels for ensuring that local governments are accountable to constituents. In practice, citizens’ engagement with local government officials is low. Village-level meetings rarely occur, few attend budget meetings, and the language and style of meetings effectively exclude many voters.7 As a result, many attempts to improve public services take place outside the existing government institutions. However, local governments largely control the allocation of resources and the standards for the provision of public services like healthcare and education, so outside interventions are greatly limited in their ability to have a large-scale or sustainable impact on public services.
Uganda-based Advocates Coalition for Development and Environment (ACODE) operates a program that explicitly aims to strengthen accountability and improve public services. The program, the Local Government Councilor Scorecard Initiative, collects information about politicians’ performance and disseminates it with the goal of improving citizen welfare. Scores measure politicians’ performance in four key areas: monitoring public services, interaction with lower local governments, execution of legislative duties and contact with their electorate.
Details of the Intervention
Researchers carried out a randomized evaluation to test whether local politicians improve their performance when citizens receive, and can share, information about their politicians’ performance. The study took place among 400 local government politicians in 20 (out of 112) districts. Within the districts, 208 sub-counties were randomly assigned to one of two groups:
- “Intense dissemination” of politician performance scorecards:
In these sub-counties, ACODE held yearly community meetings across two years where they provided information to citizens on how their politicians scored, as well as ranked in performance in comparison with other politicians in their district. Politicians were informed and invited to such meetings. Citizens also received regular updates on their mobile phones and through radio ads on their politicians’ performance. The scorecard was also provided to the politicians themselves in yearly plenary sessions.
- Comparison group:
These sub-counties did not receive the “intense dissemination” program. The scorecard was provided to the politicians themselves in yearly plenary sessions (the same sessions as the treatment group).
Researchers measured the relationships between the intense dissemination program and the level of politicians’ performance, including engagement in working with district technocrats to lobby for better services in their constituencies, contact with constituents, monitoring public services, and legislative activities in plenary sessions. The research team collaborated with the district education offices and a donor partner to examine effort in politician facilitation of school primary grant applications to a small grants program. Researchers also used random audits of schools and clinics in each sub-county in the study area to estimate whether greater effort by treatment politicians translated to better services at the local level.
Results and Policy Lessons
Impacts of intense dissemination:
Politician Performance: The more competitive a district was, the stronger the effect of intensely disseminating the scorecard to citizens above and beyond dissemination only to politicians themselves. Scorecards improved politicians’ performance the most in competitive constituencies, and effects were insignificant below a certain competitiveness threshold. The threshold was not particularly competitive—for the initiative to start to work, constituencies winners needed to have below a 0.22 margin of victory in vote share to the next runner up.
Public Service Delivery: Intense dissemination of politician performance scorecards did not lead to improvements in public service delivery above and beyond dissemination only to politicians themselves. This suggests that accountability for legislators alone may not be a sufficient strategy for improving service delivery. The authors note that many branches and levels of government are responsible for service delivery, and accountability mechanisms for many actors may need to be improved to see results in service delivery.
Elements of Politician Performance: Positive impacts were driven by increases in politicians’ monitoring of public service delivery and their involvement in local government. Additionally, the number of development projects increased in competitive constituencies, but budget size did not. There were no corresponding increases in those politicians’ contact with the electorate or fulfillment of legislative duties.
Timing of Politician Performance: Politicians who improved their performance did so in the years following the two intense disseminations to citizens, well in advance of elections. This finding suggests that transparency initiatives can help improve politician performance and accountability well in advance of elections, when it is often weakest.
1 Pande, Rohini. "Can informed voters enforce better governance? Experiments in low-income democracies." Annu. Rev. Econ. 3, no. 1 (2011): 215-237.
2 Pande, 2011.
3 International Development Department, School of Public Policy, University of Birmingham. “Local Government Decision-Making: Citizen Participation and Local Government Accountability.” IDD Research News, May 2002: 1-3.
4 Banerjee, Abhijit, Kumar, Selvan, Pande, Rohini, & Su, Felix. "Do informed voters make better choices? Experimental evidence from urban India." Working paper (2011).
Ferraz, Claudio, and Frederico Finan. "Exposing corrupt politicians: the effects of Brazil's publicly released audits on electoral outcomes." The Quarterly Journal of Economics 123.2 (2008): 703-745.
5 Humphreys, Macartan, and Jeremy M. Weinstein. "Policing Politicians: Citizen Empowerment and Political Accountability in Uganda Preliminary Analysis." IGC Working Paper (2012).
6 Dunning, Thad, Guy Grossman, Macartan Humphreys, Susan D. Hyde, and Craig McIntosh. Metaketa I: Information, Accountability, and Cumulative Learning.
7 Devas, Nick, and Ursula Grant. "Local government decision‐making—citizen participation and local accountability: Some evidence from Kenya and Uganda." Public Administration and development 23, no. 4 (2003): 307-316.