Monitoring Employees: A Principal-Agent Approach to Contracts, Safety and Productivity in Kenya’s Informal Public Transport Industry

Last registered on April 16, 2017

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0001482

Initial registration date

September 14, 2016

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

September 14, 2016, 9:43 PM EDT

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

Last updated

April 16, 2017, 12:31 AM EDT

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

Locations

Country

Kenya

Region

Nairobi

Primary Investigator

Name

David Schoenholzer

Affiliation

UC Berkeley Economics

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Gregory Lane

PI Affiliation

UC Berkeley Agricultural and Resource Economics

Contact Investigator

PI Name

Erin Kelley

PI Affiliation

UC Berkeley Agricultural and Resource Economics

Contact Investigator

Additional Trial Information

Status

On going

Start date

2016-01-01

End date

2017-08-01

Keywords

Health, Labor

Additional Keywords

public transport, informal transit

JEL code(s)

Secondary IDs

Abstract

Asymmetric information between employers and employees is a key determinant of contractual arrangements and firm outcomes. We report on a randomized controlled trial estimating the effect of reducing information asymmetry on contracts, productivity and driver safety via a new monitoring technology in Kenya’s informal public transport industry. This industry is dominated by small-scale entrepreneurs hiring minibus (matatu) drivers on a daily basis using residual-claimant schemes to ensure effort. We fit minibuses in both the treatment and control group with monitoring devices equipped with GPS and accelerometers recording high-frequency data on productivity and safety performance of drivers. Minibus owners in the treatment group receive information on driver effort and output via smartphone app. Owners and drivers in both treatment group and control group submit daily data on business outcomes to document how contracts change and how this subsequently affects productivity and safety.

External Link(s)

Registration Citation

Citation

Kelley, Erin, Gregory Lane and David Schoenholzer. 2017. "Monitoring Employees: A Principal-Agent Approach to Contracts, Safety and Productivity in Kenya’s Informal Public Transport Industry." AEA RCT Registry. April 16. https://doi.org/10.1257/rct.1482-2.0

Former Citation

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)

Primary intervention: access to safety and productivity information for minibus (matatu) owners via smartphone app. Secondary intervention: cash incentives for drivers based on safe driving performance.

Intervention Start Date

2016-07-01

Intervention End Date

2017-04-01

Primary Outcomes

Primary Outcomes (end points)

safety: speed, acceleration, alerts, safety index; productivity: revenue, work hours, mileage, earnings, expenses, profit; contract: target, target gap, linear rate, separations, firings.

Primary Outcomes (explanation)

safety index is based on mapping speed and acceleration to human-measured driving behavior; profit equals revenue minus expenses.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Stratified randomization across routes.

Experimental Design Details

Randomization Method

computer randomization.

Randomization Unit

The minibus (matatu) is the cluster of randomization; we stratify on the route. We cross-randomize the secondary treatment across the primary treatment such that with equal probability primary-treatment and primary-controls are in the secondary treatment group.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

253 minibuses (matatus).

Sample size: planned number of observations

For driving behavior analysis, we generate a 30-second panel (when ignition on) over 6 months and 253 bus, resulting in about 76 million observations. For productivity and safety analysis, daily measurements from reports via smartphone and SMS, resulting in about 45,500 observations.

Sample size (or number of clusters) by treatment arms

120 treatment, 120 control and 13 pure control; secondary treatment is crossed with the first one after three months (i.e. within primary treatment and control there will be equal subgroups of secondary treatment and control).

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

Supporting Documents and Materials

IRB