- What We Do
- Agriculture and Food Security
- Democracy, Human Rights and Governance
- Economic Growth and Trade
- Ending Extreme Poverty
- Environment and Global Climate Change
- Gender Equality and Women's Empowerment
- Global Health
- Water and Sanitation
- Working in Crises and Conflict
- U.S. Global Development Lab
- Cornerstone Partners
- Work with the Lab
- Development Innovation Ventures
- Data & Analytics for Development
- Digital Development
- Global Partnerships
- Grand Challenges for Development
- Higher Education Solutions Network (HESN)
- International Research & Science Programs
- Research and Innovation Fellowships
- Science at USAID
- Monitoring, Evaluation, Research and Learning Innovations Program (MERLIN)
THE OPPORTUNITY: To design and test the effectiveness of a voter education campaign as a way to improve government accountability.
THE PROJECT: Professors Abhijit Banerjee (MIT) and Rohini Pande (Harvard), with the Jameel Poverty Action Lab South Asia at the Institute for Financial Management and Research, are running a randomized control trial to rigorously evaluate whether pre-election voter-education campaigns that provide incumbent qualifications and performance report cards can improve the quality of elected politicians. This project will provide rigorous evidence on the impact of the report cards on electoral outcomes in the 2010 Bihar Assembly elections. In addition to the $100,000 DIV contribution, this project has leveraged $71,000 in other funding and in kind contributions from local NGO Satark Nagrik Sangathan (New Delhi, India) Hindustan, a privately owned newspaper based in New Delhi, India, and other partners.
In Bihar, the poorest state in India, the researchers conducted a voter information campaign that disseminated information on the performance and political backgrounds of 60 randomly selected incumbents through a local high-profile newspaper. The research team used a randomized control trial to evaluate the impact of this campaign on electoral outcomes. These outcomes included incumbent vote-share, voter turnout, campaign expenditures, vote buying and other illegal campaign practices.
THE RESULTS: Preliminary findings suggest that, on average, increased information about incumbent qualifications and performance decreased support for poor-performing candidates. Incumbents who had report cards published about them received 4.6% fewer votes than those who did not. The results will be the basis of a political science paper, currently being prepared, on the role of information in shifting electoral behavior and translated into a policy memo and disseminated to policymakers and civil society organizations, so that future report cards campaigns across India and globally can be designed more effectively.
Last updated: May 26, 2016