By Rebecca Stropoli
MIT’s Abhijit Banerjee and Harvard’s Nils Enevoldsen, Rohini Pande, and Michael Walton examined the effect that publicizing politicians’ records had on electoral results in the 2012 municipal elections in Delhi, India. They find that being issued public report cards caused politicians to shift their spending priorities.
With more than 18 million people, Delhi is the world’s second-largest city, behind Tokyo. Poor people living in slums form a significant share of the Delhi population. Slum dwellers, in fact, account for an electoral majority in many of the city’s 272 single-member wards, each of which elects a councillor to the municipal government every five years.
In addition to their powers as lawmakers, councillors have control over discretionary funds for projects in their wards. In the five years before the 2012 elections, these funds averaged $700,000 a year for each councillor, the researchers report. Surveys they conducted reveal little alignment between councillor spending―57 percent of which went for roads―and citizen concerns, which were clean water, sewage management, and garbage collection.
The researchers picked a random sample in 2010 of 110 councillors and told them they would receive report cards tracking their midterm performances. The findings would appear in the newspaper Dainik Hindustan. The councillors were told another report would be published one month before the 2012 election.
The report cards were generated by the NGO Satark Nagrik Sangathan, or Society for Citizens Vigilance Initiatives, and included the councillors’ total spending and their spending priorities, including public toilets, garbage removal, and education.
The “effective spending” on the needs of the poor by councillors in high-slum wards whose report cards were published increased more than 13 percent.
Another sample of 58 councillors were told they would receive only a 2012 preelection report. A control group of 72 were informed they would not receive a report card until 2014, two years after the election.
Yet another random sample of councillors received reports based on audits of toilet and garbage conditions in poor neighborhoods in their wards. These reports weren’t published, allowing the researchers to test whether “for your eyes only” information would cause politicians to reprioritize spending and policy.
The anticipation of media reports did influence the policies of politicians representing poorer areas, the findings suggest. Councillors in high-slum wards whose report cards were published shifted their spending priorities to better match the needs of their constituents. The “effective spending” on the needs of the poor by these councillors over two years increased by about $5,000 on average, or more than 13 percent, Enevoldsen says.
The politicians who received “for your eyes only” reports on toilet and garbage conditions didn’t exhibit a measurable shift in policy, and conditions were unchanged, the researchers find. “Simply improving information on policy issues for politicians per se seems not to matter,” they write.
Before the election, the government unexpectedly raised the number of council seats randomly reserved for women to 50 percent from 33 percent. As a result, some incumbents would not be eligible for reelection in their old wards. However, political parties had the option of reassigning such councillors to other wards. The researchers find that among councilors affected by the quota change, those whose report cards indicated high levels of “pro-poor” spending were 15 percentage points more likely to be assigned to the party’s ticket in a different ward.
A report card demonstrating high pro-poor spending was especially important for those councillors who were forced to run in new wards, as they saw their vote share increase 5 percentage points. These results suggest that positive media disclosures about candidates are effective in attracting new voters, the researchers report.