The sample for the study was composed of 250 villages–each with an average population of roughly 1,000 people–selected from ten districts spanning northern, northeastern, eastern, central, and western Afghanistan (southern areas were excluded due to security concerns). Half of the villages was randomly assigned to hold district elections and the other half to hold at-large elections. Under district elections, the village was split into geographically-defined districts and each villager could only vote for a single candidate residing in the same district.
The authors estimate the effects of an anti-vote buying campaign carried out in the five weeks prior to the election by the Alliance for Election Campaign Finance Monitoring (ACFIM)—a coalition of Ugandan civil society organizations. The anti-vote buying intervention included five elements: (1) a community visit from an NGO, (2) a leaflet drop, (3) a community meeting, (4) targeted robocalls, and (5) publicly displayed posters. The experiment took place in 918 parishes, which included 2,796 surveyed villages.
The authors conducted their study in three constituencies selected to include communities that were electorally competitive or strongholds for either of the major parties. They videotaped and edited the debates to show 1,991 respondents different portions of the debates on smartphones. The treatment entailed showing respondents different combinations of segments. In one treatment condition (N=391), participants watched only the personal background segment. In another treatment condition (N=409), participants only viewed the policy segment.
This study was implemented during two local elections for county legislature during the 2010 general election in the U.S. Registered voters who had participated in the last 3 party primaries were randomly assigned into three groups—one to receive a positive message about the candidate, one to receive a negative message about the opponent, and one as control (which received no messaging). The messages were delivered by mail, and also provided the individual with the option of donating to the campaign.
We conducted six Randomized Controlled Trials (RCTs) across five countries to answer this question. The types of information on incumbent behavior provided to voters include legislative performance (Benin), municipal spending irregularities (Brazil), quality of public services (Burkina Faso), municipal government malfeasance (Mexico), candidate quality via debates (Uganda 1), and budget irregularities (Uganda 2). A planned seventh study on incumbent criminality in India did not take place due to implementation challenges.
In the initial stage of the nationwide field experiment, the researchers held a series of public lotteries to select the Members of Parliament (MPs) who would be treated and receive access to the uSpeak system - a platform that allowed MPs to log onto a dashboard where they could read tagged SMS messages from constituents, reply, and see simple descriptive statistics about the messages they received, such as what the priority issues in their constituency were within a selected time-frame.
To test the ways in which politicians influence public opinion, the authors conduct two studies. The structures of the two studies are nearly identical. The authors partner with state legislators in a Midwestern state and identify contentious policy positions that they hold (support for tax increases or support for undocumented immigrants, for example). Next, they conduct interviews with a randomly selected sample of constituents to identify their opinions on these issues.
In the lead-up to the 2014 elections, Green and Vasudevan hired an Indian agency to script and record advertisements to discourage voting for vote-buying parties. The advertisements were skits in which actors discussed why vote-buying politicians were untrustworthy and unlikely to make good on their promises. The 60-second spots were recorded in Hindi and four regional languages (Kannada, Marathi, Oriya, and Telugu). The authors chose 60 AIR stations covering 665 ACs in ten states.