Incentivizing Pro-social Behavior of Gig Workers - Evidence from an Experiment

How to cite

han, ruobing; Liu, Ying; Xue, Peng. 2024. "Incentivizing Pro-social Behavior of Gig Workers - Evidence from an Experiment." AEA RCT Registry. AEARCTR-0015064. https://doi.org/10.1257/rct.15064-1.1

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

ruobing han* , Ying Liu , Peng Xue

* Primary investigator

Registry ID
AEARCTR-0015064
DOI
https://doi.org/10.1257/rct.15064-1.1
First registered
Last updated
Status
Completed
Trial dates
2024-07-01 – 2024-09-30
Countries
China ()
Registry record
http://www.socialscienceregistry.org/trials/15064

Abstract

An on-demand courier company conducted city-level randomization from July to September 2024, dividing its more than 300 operating cities in mainland China into three groups: A, B, and C. In the Group A cities, the platform’s flexible workers (riders) are informed that they can earn platform tokens (which can be redeemed for priority order access) by performing good deeds. In the Group B cities, couriers are told that by doing good deeds, they have the chance to earn platform badges (both physical and displayed in the app). In the Group C cities, couriers are encouraged to do good deeds, but no reward is given. We studied how the mood of the riders affects their pro-social behavior and how their pro-social behavior interacts with their normal business, and how the company's pro-social campaign helps with its brand image, hence affecting its competition with other platforms.

Experimental design

An on-demand courier company conducted city-level randomization from July to September 2024, dividing its more than 300 operating cities in mainland China into three groups: A, B, and C. In the Group A cities, the platform’s flexible workers (riders) are informed that they can earn platform tokens (which can be redeemed for priority order access) by performing good deeds. In the Group B cities, couriers are told that by doing good deeds, they have the chance to earn platform badges (both physical and displayed in the app). In the Group C cities, couriers are encouraged to do good deeds, but no reward is given. We studied how the mood of the riders affects their pro-social behavior and how their pro-social behavior interacts with their normal business, and how the company's pro-social campaign helps with its brand image, hence affecting its competition with other platforms.

Design details

Randomization method
Determined by company's algorithms
Randomization unit
City
Clusters
3 groups of cities, 100+ in each group
Observations
about 10000+ riders
Arms
3 groups of cities, 100+ in each group, riders in each city have the same treatment.
Primary outcomes
The take-up of pro-social behaviors among three groups/arms.
Secondary outcomes
How the mood of riders affects their pro-social behaviors.

Cited by

No citation data available for this registration yet.

Keywords