Policy evidence and political accountability in Liberian education reform

Last registered on October 10, 2017

Pre-Trial

Trial Information

General Information

Title
Policy evidence and political accountability in Liberian education reform
RCT ID
AEARCTR-0002506
Initial registration date
October 07, 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
October 10, 2017, 12:12 AM EDT

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

Locations

Region

Primary Investigator

Affiliation
ITAM

Other Primary Investigator(s)

PI Affiliation
UC San Diego
PI Affiliation
Center for Global Development

Additional Trial Information

Status
Completed
Start date
2017-10-02
End date
2017-10-08
Secondary IDs
Abstract
We study the relationship between policy evidence and political accountability in Liberia. Specifically, we study 1) candidates for the Liberian House of Representatives in the October 2017 election and 2) households whose children were a part of the authors' previous evaluation of the Partnership Schools for Liberia (PSL) program. The study varies the provision of RCT results to candidates for the Liberian House of Representatives and bureaucrats, and varies the provision of information about candidates' public positions on PSL to households. The primary outcomes of interest for candidates are stated public positions on PSL, codified in two ways: mentions of PSL in representative debates for a subset of the sample whose debates happened after the PSL report became public, and responses to our survey question asking which statement they would like us to publicly tell their voters they support). The primary outcomes of interest for households are self-reported voting intentions and candidate approval rates. We plan to examine heterogeneity in treatment effects along several dimensions, but the most important among candidates is the density of PSL schools in their district; among households, it is PSL treatment assignment status.
External Link(s)

Registration Citation

Citation
Romero, Mauricio, Justin Sandefur and Wayne Sandholtz. 2017. "Policy evidence and political accountability in Liberian education reform." AEA RCT Registry. October 10. https://doi.org/10.1257/rct.2506-1.0
Former Citation
Romero, Mauricio, Justin Sandefur and Wayne Sandholtz. 2017. "Policy evidence and political accountability in Liberian education reform." AEA RCT Registry. October 10. https://www.socialscienceregistry.org/trials/2506/history/22147
Experimental Details

Interventions

Intervention(s)
For both candidates and households, we randomize treatment status conditional on the respondent agreeing to take part in the survey. For candidates, we vary the presentation of information from the independent evaluation's report. As part of the survey, all candidates receive a ''placebo'' condition describing the PSL program and noting some of the main arguments advanced by its supporters and detractors. Some candidates then receive more information. The ''Results'' condition provides a brief summary of what the independent PSL evaluation discovered: learning gains of 60%, increased teacher attendance and time on task, and the fact that in some schools teachers and students were forced out. Other candidates receive the ``Popularity'' condition, which presents our findings that PSL schools were broadly popular among parents and teachers. Some candidates get both the ``Results'' and the ``Popularity'' conditions.


For households, we vary the presentation of information about candidates' stated public positions on PSL. Each household randomly selected into treatment receives information on the positions of candidates for President of Liberia and Representative of their own self-reported district (there are 73 districts throughout the country). Public positions for presidential candidates were taken from a presidential debate held 26 September, in which moderators asked the three participating candidates for their position on the PSL program. Stated public positions for representatives came from the candidate survey. A question in that survey asks the following: ``Let me remind you that all the answers you have given will be confidential. However, I wanted to let you know that we are planning to conduct an information campaign to sensitize voters about Partnership Schools. If you like, we are willing to communicate the campaign message you choose to voters in your district on your behalf. I will read three statements, and I want you to tell me which one, if any, is your position which you want us to sensitize to the people in your district on your behalf. 1. The Partnership Schools program should be expanded and funded through the national budget. 2. The Partnership Schools program should continue to be tested before being scaled up. 3. The Partnership Schools program should be discontinued, and traditional public schools should get those resources.''



Because the candidate survey began a few weeks before the household survey, at least one candidate from each of the country's 73 districts had provided an answer to this question by the time we implemented the household survey. Treated households received a statement saying ``Some of the candidates for Representative in YOUR district also have made statements about PSL, which they wanted us to share with you. Please listen carefully:

These candidates say PSL should be taken into more schools, and supported by the national budget:
[list candidates]
These candidates say PSL needs to be tested more before making a decision:
[list candidates]
These candidates say PSL should be stopped immediately, and normal government schools should get that support:
[list candidates].''
Intervention Start Date
2017-10-02
Intervention End Date
2017-10-08

Primary Outcomes

Primary Outcomes (end points)
We will test the effect of information candidates' self-reported beliefs about what PSL is and what it should be, plans to incorporate PSL into the campaign message, and a statement for us to share with their potential voters. The main outcomes are voters' approval of 6 of the leading presidential candidates, and self-reported likelihood of voting for them and candidates for representative. We aim to test the hypothesis that giving voters information about candidates' positions causes those voters' attitudes and voting intentions to shift to a candidate morealigned with their own views. We also examine the effect of the info on a more generic question: is the country moving in the right direction?
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Our intervention consists in providing information to candidates and households in the context of a phone survey. For both candidates and households, we randomize treatment status conditional on the respondent agreeing to take part in the survey.

For candidates, we vary the presentation of information from the PSL evaluation's report. As part of the survey, all candidates receive a ``placebo'' condition describing the PSL program and noting some of the main arguments advanced by its supporters and detractors. Some candidates then receive more information. The ``Results'' condition provides a brief summary of what the independent PSL evaluation discovered: learning gains of 60\%, increased teacher attendance and time on task, and the fact that in some schools teachers and students were forced out. Other candidates receive the ``Popularity'' condition, which presents our findings that PSL schools were broadly popular among parents and teachers. Some candidates get both the ``Results'' and the ``Popularity'' conditions. The probability of being assigned to each of these four conditions (placebo, results, popularity, both) is 0.25.

For households, we vary the presentation of information about candidates' stated public positions on PSL. All households receive a brief summary of the PSL evaluation's results, similar to that in the ``results'' condition described above. With probability 0.50, households are randomly selected to also receive information on the positions of candidates for President of Liberia and Representative of their own self-reported district (there are 73 districts throughout the country, 40 of which contain at least one PSL school). Households receive either no candidate information, OR information about both presidential and representative candidates. We culled presidential candidates' public positions from a presidential debate held 26 September 2017, in which moderators asked the three participating candidates for their position on the PSL program.

Public positions for representatives came from a question in our candidate survey which reads as follows: ``Let me remind you that all the answers you have given will be confidential. However, I wanted to let you know that we are planning to conduct an information campaign to sensitize voters about Partnership Schools. If you like, we are willing to communicate the campaign message you choose to voters in your district on your behalf. I will read three statements, and I want you to tell me which one, if any, is your position which you want us to sensitize to the people in your district on your behalf.
The Partnership Schools program should be expanded and funded through the national budget
The Partnership Schools program should continue to be tested before being scaled up.
The Partnership Schools program should be discontinued, and traditional public schools should get those resources.''

Because the candidate survey began the week before the household survey, at least one candidate from each of the country's 73 districts had provided an answer to this question by the time we began the household survey. Treated households received a statement of which candidates in their district had selected which statements.
Experimental Design Details
Randomization Method
Randomization done by computer
Randomization Unit
Individual household/candidate
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
No clusters
Sample size: planned number of observations
~1000 Candidates ~1000 Households
Sample size (or number of clusters) by treatment arms
Candidates: 250 per arm, 4 arms
Households: 500 per arm, 2 arms
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
UC San Diego
IRB Approval Date
2017-09-15
IRB Approval Number
#171544SX
IRB Name
UC San Diego
IRB Approval Date
2017-09-15
IRB Approval Number
#171543XX
Analysis Plan

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

Request Information

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