Merit and Job Allocation

How to cite

Bhattacharya, Puja; Rampal, Jeevant. 2025. "Merit and Job Allocation." AEA RCT Registry. AEARCTR-0016034. https://doi.org/10.1257/rct.16034-1.0

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

Puja Bhattacharya* , Jeevant Rampal

* Primary investigator

Registry ID
AEARCTR-0016034
DOI
https://doi.org/10.1257/rct.16034-1.0
First registered
Last updated
Status
In development
Trial dates
2025-05-23 – 2028-06-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/16034

Abstract

Different jobs usually require different skill sets. Hence, certain jobs are better performed by some than by others. A popular notion considers selecting workers for jobs based on individual merit – the ‘meritocratic’ principle – to be fair and efficient. However, in situations where multiple people need to be assigned to different jobs, assigning an individual who is the best available for a job may not always maximize productivity. We focus on a setting where two workers need to be assigned across two jobs. We show that when one of the workers is better at both the jobs, there arises a tension between a desert claim and an efficiency claim for justifying why an individual may merit a certain job. We conduct online experiments that test for and quantify whether individuals have such desert-based distributive preferences.

Intervention

We implement a primary elicitation using a 'spectator' design where "Spectators" are asked to assign two jobs between two anonymous "Workers". Spectators and Workers are participants recruited using an online survey platform (Prolific or similar). The jobs are selected to resemble familiar activities for workers on online platforms. Spectators are presented with a measure of 'skills' of the the two Workers on the two jobs. They observe the skills and choose a job-assignment for the pair of workers. A job assignment is a one-to-one matching of a worker to a job.

Experimental design

Spectators are matched to a pair of workers and are informed about the two jobs the workers will complete, for how long, and the piece-rate for completing individual tasks on the jobs. In addition, the spectators are informed about the individual performance of the two workers from a preliminary survey where the workers worked on the two jobs. The spectator observes a performance profile and decides which worker will be assigned to which job in the future. The spectator is told that within a month of them completing the survey, the workers will be invited back to work on the assigned job.

The experiment presents the spectators with different performance profiles to tease apart preference for 'desert-based' and 'efficiency-based' considerations behind job-assignment decisions.

Design details

Randomization method
Randomization done in Qualtrics
Randomization unit
Individual
Clusters
600 individuals
Observations
600 individuals
Arms
Treatments involve different profiles of workers' skills the Spectator is presented with. Barring the real skill profile of the workers the spectators are matched with, the spectators will be randomized into 5 hypothetical skill profiles, i.e., approximately 120 individuals in each profile or treatment.
Primary outcomes
Job-assignment decision of the Spectators

Cited by

No citation data available for this registration yet.

Keywords