The Impact of Voter Knowledge Initiatives in Sierra Leone

Last registered on March 06, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

The Impact of Voter Knowledge Initiatives in Sierra Leone

RCT ID

AEARCTR-0000026

Initial registration date

July 23, 2013

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

July 23, 2013, 9:57 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

March 06, 2017, 10:42 AM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Sierra Leone

Region

Africa

Primary Investigator

Name

Katherine Casey

Affiliation

Stanford GSB

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Rachel Glennerster

PI Affiliation

Abdul Latif Jameel Poverty Action Lab

Contact Investigator

PI Name

Kelly Bidwell

PI Affiliation

Post-Conflict Recovery and Fragile States Initiative, Innovations for Poverty Action

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2011-09-07

End date

2014-06-30

Keywords

Electoral, Governance

Additional Keywords

transparency, accountability

JEL code(s)

Secondary IDs

Abstract

This research program explores two main questions: (i) Does providing information on incumbent performance to centralized political parties influence candidate selection? and (ii) Do structured debates provide credible information to voters about candidates’ competencies, and if so, does this improve voting returns for better qualified candidates in the short run and public sector performance in the long run? In order to answer these questions, we are working with local partners to execute a series of initiatives during the 2012 Parliamentary and local government elections in Sierra Leone: i) a nonpartisan incumbent performance evaluation campaign; ii) structured, inter-party debates between candidates; and iii) several individual-level treatment arms that unpack the different types of information delivered by the debates. We will use exit polls and official electoral data to assess the impact of these programs on candidate selection, and voter knowledge, engagement and behavior.

External Link(s)

J-PAL evaluation summary

Registration Citation

Citation

Bidwell, Kelly, Katherine Casey and Rachel Glennerster. 2017. "The Impact of Voter Knowledge Initiatives in Sierra Leone." AEA RCT Registry. March 06. https://doi.org/10.1257/rct.26-7.0

Former Citation

Bidwell, Kelly, Katherine Casey and Rachel Glennerster. 2017. "The Impact of Voter Knowledge Initiatives in Sierra Leone." AEA RCT Registry. March 06. https://www.socialscienceregistry.org/trials/26/history/14733

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

Part I. Incumbent Scorecards
In collaboration with Campaign for Good Governance, we compiled scorecards on incumbent Members of Parliament (MP) and Local Councillors that combine objective performance measures with public opinion data. We distributed scorecards for a randomly selected subset of politicians to political party officials in advance of candidate nominations.

Part II. Debate Screenings
In partnership with Search for Common Ground, we hosted inter-party debates between Parliamentary candidates in 14 randomly selected constituencies. Debates covered a series of national and local policy issues. We took videotapes of the debates on a community "road show," where we screened the debates in a randomly selected sample of polling centers within the targeted constituencies. We estimate that 19,000 people saw the debates.

Part III. Individual Screenings
To unpack the effects of different types of information delivered by the debates, we implemented the following treatment arms at the individual level in a separate set of polling centers.
1. Debate: Individuals were shown the exact same debate screened in polling centers on a personal handheld device.
2. Getting to Know You: Individuals were shown a “getting to know you” video of the same two candidates speaking informally about their hobbies and interests.
3. Radio Report: Individuals listened to a recording of an independent moderator or journalist summarizing the main policy positions articulated by the two candidates during the debates.
4. Thin Slice Evaluations: Individuals participated in a “lab” experiment where they were exposed to pairs of isolated images, voice recordings, video clips, and names of candidates from other constituencies across the country and asked to rate them along a variety of metrics, such as who they thought would be a better leader.
5. Control group: No media. Some respondents were surveyed before the election, others only after the election.

Intervention Start Date

2012-05-01

Intervention End Date

2013-12-31

Primary Outcomes

Primary Outcomes (end points)

Part I: Candidate retention on the party symbol; Parts II and III: vote choice (including across ethnic lines), voter knowledge and behavior

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Part I: Incumbent Scorecards
Randomization into treatment (scorecard shared with party officials) and control groups (no scorecard) was stratified by incumbent political party, competitiveness of the constituency and overall performance rank, conducted separately for MPs and Local Councillors.

Part II: Debate Screenings
We randomly selected 14 of what we estimated to be the 28 most competitive MP constituencies for participation in the debates. Within these constituencies, 112 polling centers were randomly assigned to the debate screening treatment, 40 to the individual-level treatments, and 112 to the control group. Polling center randomization was stratified by number of registered voters and distance to the nearest neighboring polling center.

Part III: Individual Screenings
Within each polling center assigned to individual-level treatments, we randomly assigned households into the treatment arms above and randomly selected individual participants (based on an earlier household listing) stratified by age and gender.

Experimental Design Details

Randomization Method

All randomizations were conducted in the office using Stata.

Randomization Unit

Part I Scorecards: individual politicians; Part II Debate Screenings: first at the level of constituency, second at the level of polling centers; Part III Individual Screenings: first at the level of polling center, second at the level of individual

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

Part I: N/A; Part II: 224 polling centers in 14 constituencies; Part III: 40 polling centers in 8 constituencies

Sample size: planned number of observations

Part I: 108 MPs and 210 Local Councillors; Part II: 5,600 voters; Part III: 2,600 voters

Sample size (or number of clusters) by treatment arms

Part I: 159 T, 159 C politicians; Part II: 112 treatment polling centers (2,800 voters) and 112 control polling centers (2,800 voters); Part III: 400 voters per treatment arm plus 600 controls with pre-survey and 400 pure controls with no pre-survey

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

For Part II: MDE is 7.7 ppt increase in voting across party lines with alpha 0.05 and power 0.80; Part III: same MDE as in II for T versus C as well as continuous outcome MDE of 0.5 point shift on 10 point likeability scale across T arms

Supporting Documents and Materials

IRB