Peer Information and Incentives in the Gig Economy

How to cite

Lim, Zhi Hao. 2025. "Peer Information and Incentives in the Gig Economy." AEA RCT Registry. AEARCTR-0016393. https://doi.org/10.1257/rct.16393-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Zhi Hao Lim*

* Primary investigator

Registry ID
AEARCTR-0016393
DOI
https://doi.org/10.1257/rct.16393-2.0
First registered
Last updated
Status
In development
Trial dates
2025-08-13 – 2026-12-31
Countries
Private
JEL codes
C93, D91, J24
Registry record
http://www.socialscienceregistry.org/trials/16393

Abstract

Peer information is pervasive in the workplace, but recent work shows that workers differ in how they value and respond to it (Lim, 2025). We conduct a field experiment with rideshare drivers to study demand for peer earnings information and identify the mechanisms behind it. Drivers are randomly assigned to one of three information treatments: no peer information, peer information, or an endogenous choice condition. Over the intervention period, they are also assigned all three incentive conditions: no bonus, a target-based bonus, and a proportional bonus, with the order randomized. Using administrative records and survey responses, we examine heterogeneity in information demand and labor supply responses across treatment arms. We benchmark the effects of peer information against financial incentives, examine potential interaction effects, and additionally provide experimental estimates of labor supply elasticities leveraging on exogenous variation in wages.

Intervention

Refer to experimental design.

Experimental design

Drivers are randomly assigned to one of three information treatment groups:

1. No Info group: Drivers receive only their own earnings updates each week.
2. Info group: Drivers receive both their own earnings and peer group earnings updates each week.
3. Endogenous Info group: Drivers receive their own earnings updates each week; whether they also receive peer earnings updates depends on their stated preferences.

In addition to the information treatments (whose assignments are fixed throughout), each driver faces three different incentive conditions over the intervention period. These are:

A. No Incentive: No additional bonuses are provided.
B. Target Incentive: Drivers receive a bonus of $X for completing Y trips on a specified day.
C. Multiplier Incentive: Drivers earn an additional Z% bonus on their total earnings on a specified day.
(X, Y, and Z are placeholders that are calibrated based on the company's current incentive structure.)

Each incentive condition lasts for one week, and the order of incentives is randomized for each driver.

Design details

Randomization method
Randomization is done by a computer using Python.
Randomization unit
Randomization is done at the individual level, stratified by driver baseline activity level (measured by average weekly hours worked) and tenure with the company (above- or below-median).
Clusters
We plan to recruit approximately 2,000 rideshare drivers for the study. All eligible drivers who sign up for the program will be enrolled. As a result, the final sample size may change depending on actual enrollment at the time of the baseline survey launch.
Observations
Over 644,000 driver-day level observations (2,000 workers * 7 days/week * (6 weeks of intervention + 28 weeks pre-intervention + 12 weeks of post-intervention))
Arms
Our sample is evenly split between the three information treatment groups, with approximately 670 drivers in each group.
Primary outcomes
i. Labor supply decisions (i.e. total earnings, hourly wage, total trips completed, total distance travelled, total utilization hours, number of days worked per week, average hours worked per day, working hour patterns, hazard rate of stopping work, proportion of trips from each rideshare platform, indicator for whether driver works on a given day, indicator for whether driver has attrited from the platform) ii. Survey measures of driver well-being (i.e. stress, work meaning, satisfaction, and motivation)
Secondary outcomes
To complement the primary outcomes, we will request data on driver online status (i.e. timestamps for when drivers turn the app on or off each day, net active hours on app each day) to study drivers' search behavior at work and their willingness to accept rides. We also request data on driver locations while they are online to further examine their search behavior. Our ability to study these outcomes are subject to data availability by the company.

Cited by

No citation data available for this registration yet.

Keywords