x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
Information Frictions in the Labor Market: Evidence from a Field Experiment in Uganda
Last registered on March 14, 2019

Pre-Trial

Trial Information
General Information
Title
Information Frictions in the Labor Market: Evidence from a Field Experiment in Uganda
RCT ID
AEARCTR-0001005
Initial registration date
April 26, 2016
Last updated
March 14, 2019 8:20 PM EDT
Location(s)
Region
Primary Investigator
Affiliation
University College London & Institute for Fiscal Studies
Other Primary Investigator(s)
PI Affiliation
University College London & Institute for Fiscal Studies
PI Affiliation
BRAC
Additional Trial Information
Status
Completed
Start date
2014-07-01
End date
2018-01-01
Secondary IDs
Abstract
This project studies the labor market matching process between young workers and firms in Uganda. We focus on three core aspects of this process. First, we study the relative importance of cognitive and non-cognitive skills in determining the labor market outcomes of young workers; second, we examine whether asymmetric information on the skills of youth that first enter the labor market is a significant source of inefficiency in terms of reduced employment opportunities and mismatch; third, we study whether an intervention improving the signal to potential employers on the skills of job market entrants reduces the asymmetric information problem and limits the related inefficiencies. We carry out this analysis through a randomized control trial in the Ugandan labor market. Our intervention exogenously improves the signals employers receive about a given young worker’s non-cognitive skills, and hence is intended to reduce information asymmetries during the recruitment process. We then trace the impacts this has on worker and firm outcomes.
External Link(s)
Registration Citation
Citation
Bassi, Vittorio, Aisha Nansamba and Imran Rasul. 2019. "Information Frictions in the Labor Market: Evidence from a Field Experiment in Uganda." AEA RCT Registry. March 14. https://doi.org/10.1257/rct.1005-3.0.
Former Citation
Bassi, Vittorio, Aisha Nansamba and Imran Rasul. 2019. "Information Frictions in the Labor Market: Evidence from a Field Experiment in Uganda." AEA RCT Registry. March 14. https://www.socialscienceregistry.org/trials/1005/history/43362.
Sponsors & Partners

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

Request Information
Experimental Details
Interventions
Intervention(s)
The intervention has three main components: (i) Pre-screening assessments were conducted on both the cognitive and non-cognitive skills of workers about to complete a vocational training program, and interested in participating in the intervention. (ii) Upon completion of the training program, all eligible trainees were then matched to potential employers, who had expressed an interest in meeting potential job candidates. The matching process involved facilitating a job interview between the trainee and the firm. (iii) For a random subset of matches, information on the scores of trainees on their non-cognitive skills assessments was also disclosed to both the trainee and the firm during the matching process, by means of certificates.
Intervention Start Date
2015-04-01
Intervention End Date
2015-07-31
Primary Outcomes
Primary Outcomes (end points)
A first set of outcomes is at the worker-firm match level, and includes: probability of the match resulting in a job offer to the matched trainee; job characteristics, such as wages, mode of payment, tasks performed, productivity and duration for all matches resulting in the trainee being hired. A second set of outcomes relates to how the intervention may have affected the behavior of firms and workers outside the matching intervention. This second set of outcomes is at the individual worker and firm level. On the worker side, outcomes include: search effort, employment outcomes, labor market expectations and beliefs about own skills. On the firm side, they include: hiring strategies, firm size, expectations about the distribution of workers in the economy.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The evaluation uses a Randomized Control Trial design. On the worker side, we identified trainees that were about to complete training at 15 Vocational Training Institutes (VTIs) in Uganda. Trainees were told that as part of the programme they would be matched to available employers after the end of training, and that assessments would be conducted on their cognitive and non-cognitive skills. Trainees were explicitly told that there was the possibility of their test results being disclosed to employers during the matching. Trainees that voluntarily accepted to participate were included in the research sample. Skills assessments were then collected for all interested trainees. On the firm side, interested employers were identified by conducting a census of eligible firms in 15 urban areas of Uganda.

Upon completion of the training programme, a matching intervention was carried out, whereby our trainees were introduced to our sample of potential employers. Workers and firms were randomly assigned to two experimental groups before the matching took place: for trainees in the treatment group, additional information on their non-cognitive skills was revealed to both potential employers and the trainees themselves during the matching. This was done by means of certificates reporting the score of the trainee on the non-cognitive skills assessments conducted at the VTIs. For trainees in the control group, no additional information was revealed during the matching.

Experimental Design Details
Randomization Method
Randomization done in office by a computer (using STATA).
Randomization Unit
Treatment assignment: Individual level randomization into treatment groups was conducted on both the worker side and the firm side. So the unit of randomization is the individual worker and the individual firm.

Matching assignment: Trainees in the treatment group were randomly assigned to meet firms in the treatment group. Trainees in the control group were randomly assigned to meet firms in the control group. Randomization of the matching assignments was done at the worker-firm match level.
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
Worker side: 787 workers
Firm side: 422 firms
Match level: 1230 worker-firm matches
Sample size: planned number of observations
Worker side: 787 workers. Firm side: 422 firms. Match level: 1230 worker-firm matches
Sample size (or number of clusters) by treatment arms
Worker side:
Control group: 397 workers
Treatment group: 390 workers

Firm side:
Control group: 211 firms
Treatment group: 211 firms

Match level:
Control group: 614 worker-firm matches
Treatment group: 616 worker-firm matches
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We report power calculations for one of the main outcomes: the probability that a worker-firm match generates a job offer to the worker. We can divide the 1230 worker-firm matches into two groups: a first group of matches where the matched trainee has non-cognitive skills equal or greater than the median trainee, and a second group where the matched trainee has lower than median non-cognitive skills. Based on the actual data from the baseline skills assessments, the sample sizes in each group are the following: (i) 268 Control matches where the matched trainee has lower than median skills (ii) 346 Control matches where the matched trainee has skills that are equal or greater than the median (iii) 264 Treatment matches where the matched trainee has lower than median skills (iv) 352 Treatment matches where the matched trainee has skills that are equal or greater than the median We expect treatment effects to be heterogeneous, given the nature of the treatment. In particular, we expect the treatment to have a positive impact of the job interview outcomes of trainees that have skills that are equal or higher than the median, and a negative effect on trainees with lower than median skills. Therefore, the relevant comparisons in the analysis are between groups (i) and (iii) on the one hand, and between groups (ii) and (iv) on the other. For the purpose of power calculations, we consider the minimum detectable impact of the treatment on the probability that a trainee with skills equal or greater than the median receives a job offer. So the comparison is between matches in group (ii) and group (iv) above. Table 1 in the Supporting Documents and Materials section reports the power calculations, and shows how the minimum detectable effect size varies as a function of the proportion of Control matches ending up in a job offer.
Supporting Documents and Materials

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

Request Information
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
University of Southern California Institutional Review Board
IRB Approval Date
2018-06-04
IRB Approval Number
UP-18-00322 (Note: approval relates to secondary data analysis)
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers