Overconfidence and Biased Beliefs in Gig Economy Jobs

How to cite

Pires, Pedro. 2022. "Overconfidence and Biased Beliefs in Gig Economy Jobs." AEA RCT Registry. AEARCTR-0010144. https://doi.org/10.1257/rct.10144-1.3

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

Pedro Pires*

* Primary investigator

Registry ID
AEARCTR-0010144
DOI
https://doi.org/10.1257/rct.10144-1.3
First registered
Last updated
Status
Completed
Trial dates
2022-04-01 – 2024-09-30
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/10144

Abstract

Pay and work hours fluctuate from month to month in many jobs. A particular case -- growing in popularity over time -- is gig jobs on online platforms. Variation in job outcomes is not a problem as long as workers can, on average, understand and predict them accurately. As a result of limited attention or cognitive capacity, or a desire for self-deception, workers may persistently hold inaccurate beliefs that lead to inefficient labor supply decisions. We first test, over three surveys, whether gig workers understand their job outcomes by collecting data on their beliefs and actual job outcomes, such as pay, hours and expenses. We then conduct an experimental randomized information treatment, in which half of workers in our sample receive information about their mistakes in understanding their take home pay. We measure effects over time on both beliefs and labor market decisions.

Intervention

In our randomized information treatment, workers receives information on whether they are correctly assessing their take home hourly pay, are given an individualized example on how to calculate expenses and are told that overconfidence is common in many settings. Workers in the treatment group receive this information on the Baseline and Midline surveys. Workers in the control group receives the same information at the end of the Endline survey.

Experimental design

See pre-analysis plan for more details

Design details

Randomization method
Simple randomization
Randomization unit
Randomization at the worker level
Clusters
N/A
Observations
Between 400 and 600 workers
Arms
Between 200 and 300 for each of the treatment and control groups
Primary outcomes
Job outcomes: hours worked, hourly and weekly pay, expenses Other labor market variables: job search and information on other jobs Beliefs of job outcomes for oneself and for other workers: hours worked, hourly and weekly pay, expenses Misperception of job outcomes: difference between beliefs and actual outcomes

Cited by

No citation data available for this registration yet.

Keywords