Evaluating the impact of information on electoral accountability

Feb. 21, 2012, 3:21 p.m.

At a 3ie Delhi seminar in December, Dr. Rohini Pande, Harvard University, presented preliminary findings from a 3ie supported evaluation of the impact of voter education awareness campaigns on voting behavior.

Nearly a quarter of the 543 elected members of the Indian Parliament have been charged with crimes, including rape and murder (Association for Democratic Reforms, 2009). Civil society movements have made corruption among politicians a topic of national importance using approaches such as voter education campaigns to empower citizens and demand more effective leadership. At the 3ie Delhi seminar Dr. Rohini Pande, Harvard University, presented preliminary findings from a 3ie supported evaluation of the impact of voter education awareness campaigns on voting behavior. “There is no reason to suppose that democracy delivers the best kind of governance. There is little information flowing through the electoral system though election campaigning is so wide spread,” said Dr . Pande. The Abdul Latif Jameel Poverty Action Lab (J-PAL) research team, and its NGO partners Satark Nagrik Sanghatan and Prayatn, used  ‘report cards’ and street theatre to generate awareness on the performance of elected representatives in the urban slums of Delhi and villages of Rajasthan. The report cards give an objective profile of the candidate's education, criminal record, assembly or parliament meeting attendance and their performance as measured by expenditure on public works. The findings from this randomised control trial are promising. “Our evaluation in Delhi shows that the turn out for the election was 3-5 % higher in treatment groups than the control groups. Better performing incumbents received more votes in the treatment group. Cash-based vote-buying also declined by 19 % in treatment polling stations,” said Dr Pande. In Rajasthan, the findings show that the campaign motivated more villagers to stand for elections. While it discouraged the incumbents to run, female incumbents were more encouraged to contest. The Rajasthan results suggest that informational incentives have an impact on not just voters but also candidates. During the discussion at the seminar it was asked whether the results of this evaluation were purely attributable to objective information from the campaign. “What we still need to understand is whether the impact on voting behavior was a result of the information from the campaign or a general priming effect which led to voter perceptions being formed,” clarified Dr. Pande.  

comments powered by Disqus