The Impact of Primary Conventions

Last registered on May 22, 2020

Pre-Trial

Trial Information

General Information

Title
The Impact of Primary Conventions
RCT ID
AEARCTR-0002485
Initial registration date
November 24, 2017

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
November 26, 2017, 3:07 PM EST

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

Last updated
May 22, 2020, 12:45 PM EDT

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

Locations

Region

Primary Investigator

Affiliation
University of Chicago

Other Primary Investigator(s)

PI Affiliation
Stanford University

Additional Trial Information

Status
Completed
Start date
2017-11-25
End date
2019-12-31
Secondary IDs
Abstract
Across the world, political parties have regional strongholds where competition in the general election is weak and the competition of import occurs at the primary stage. Internal party selection processes thus directly determine the characteristics and performance of elected representatives in many areas. In much of the developing world, these processes are opaque, controlled by elites, and likely hampered by information constraints. Working with the Political Parties Registration Commission in Sierra Leone, we propose to evaluate whether increasing citizen voice in primaries and hosting party primary conventions that feature informative debates between aspirants facilitate the election of more competent and accountable leaders.
External Link(s)

Registration Citation

Citation
Casey, Katherine and Rachel Glennerster. 2020. "The Impact of Primary Conventions." AEA RCT Registry. May 22. https://doi.org/10.1257/rct.2485-1.1
Former Citation
Casey, Katherine and Rachel Glennerster. 2020. "The Impact of Primary Conventions." AEA RCT Registry. May 22. https://www.socialscienceregistry.org/trials/2485/history/69021
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)
The Political Parties Registration Commission (PPRC) and Search for Common Ground (SFCG) will support political party primary conventions during which aspirants for the party symbol for a given Parliamentary constituency present their qualifications and platform. Conventions will be broadcast on local radio stations. Voters in the constituency will be polled and their preferences over aspirants aggregated and shared with political parties.
Intervention (Hidden)
Intervention Start Date
2017-11-25
Intervention End Date
2018-03-31

Primary Outcomes

Primary Outcomes (end points)
Representation: whether more locally popular aspirants are awarded the party symbol.
Candidate characteristics: characteristics of aspirants selected for the party symbol.
Political knowledge: voter and party leader knowledge of aspirant characteristics and policy positions and their rank order evaluation of aspirants.
Long run outcome: performance in office, including MP allocation of the constituency facilitation fund (CFF). (Whether this is worth collecting will depend in part on first round effects).

Update: long run primary outcomes are: total verified development expenditure (winsorized at 95th percentile), verified existence of constituency office; total public meetings (average of key informant reports)
Primary Outcomes (explanation)
Representation: There are two ways to think about whether choice of aspirant is popular--voter polling and final vote share.

Secondary Outcomes

Secondary Outcomes (end points)
Campaign expenditure.
Candidate entry: Change in number and characteristics of aspirants running (not relevant for all areas depending on timing of announcement of primaries by parties).

Update: secondary long run outcome: key informant performance evaluation (average of key informant reports)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Primary conventions and voter polling are randomized across Parliamentary constituencies, separately for each participating political party.

The timeline for the experiment is as follows. Parties submit a long list of constituencies in which they are interested in holding primaries. A lottery selects in treatment and control constituencies from this list. Treatment constituencies are announced publicly via radio and national party conventions. In treatment constituencies, voters are surveyed about their knowledge and views of aspirants for the party symbol and encouraged to listen to the upcoming convention via radio. Primary conventions featuring informative debates between aspirants are hosted by the Political Parties Registration Commission and moderated by Search for Common Ground. The conventions are broadcast over radio. A larger number of voters are polled after the convention about their views of aspirants and the aggregated results fed back to party leaders.

Voters in control constituencies are also polled at two points in time about their knowledge and views of aspirants. Aspirants in treatment and control constituencies are surveyed about their professional qualifications and other characteristics. Party leaders are surveyed in both treatment and control constituencies about their knowledge and views of aspirants.

Parties select their MP candidates for the general election. Voter chose among MP candidates in the general election.

Experimental Design Details
Data collection on candidate characteristics is designed to measure professional qualifications, cognitive ability, party loyalty and connections, pro-sociality and attention to detail, public service motivation, local knowledge and connectivity, development effort, personality, demographics and wealth
Randomization Method
All randomizations are done in the office on a computer.
Randomization Unit
The unit of randomization is the Parliamentary constituency, randomized separately for each participating party.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
The total number of constituencies depends on the participation of political parties. At present, there are 46 treated and 46 control constituencies, with some overlap across parties

Sample size: planned number of observations
100 voters will be surveyed per constituency. The number of aspirants is not yet determined for all parties in all constituencies, but all aspirants for a participating political party will be surveyed. Five political party leaders will be surveyed per constituency.
Sample size (or number of clusters) by treatment arms
At present there are 46 treated and 46 control MP constituencies, with some overlap across parties.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Our power calculations use 2012 data from 53 major party candidates in 28 constituencies and are based on 39 treated and 39 control constituencies. We will be able to detect an increase in the number of aspirants for a given party’s symbol from 3.7 to 5.0 with alpha=0.10, power=0.80 (SD=2.3). We do not have baseline data on most of the candidate characteristics of interest, so use proxies from our 2012 data. With alpha=0.10, power=0.80, we can detect an increase in knowledge of government counterparts from 0.94 to 1.49 (out of 3, SD=0.97) or in years of schooling from 15 to 16.2 (SD=2.1). For elected officials, we have power to detect an increase in the percentage of CFF verified as spent on development from 35.6% to 62.0% with alpha=0.10 and power = 0.80 (SD=46.8).
Supporting Documents and Materials

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

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
COUHES MIT
IRB Approval Date
2017-08-24
IRB Approval Number
1708065627
IRB Name
Human Subjects Research Stanford University
IRB Approval Date
2017-10-27
IRB Approval Number
42946
IRB Name
Sierra Leone Ethics and Scientific Review Committee
IRB Approval Date
2017-10-18
IRB Approval Number
N/A

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials