The Impacts of Worker Monitoring

How to cite

Kala, Namrata; Lyons, Elizabeth. 2023. "The Impacts of Worker Monitoring." AEA RCT Registry. AEARCTR-0011365. https://doi.org/10.1257/rct.11365-1.0

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

Namrata Kala* , Elizabeth Lyons

* Primary investigator

Registry ID
AEARCTR-0011365
DOI
https://doi.org/10.1257/rct.11365-1.0
First registered
Last updated
Status
In development
Trial dates
2023-05-10 – 2024-06-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/11365

Abstract

Remote hiring and work have the potential to expand the set of labor market opportunities across geographies. We will use an online platform to study how worker input monitoring affects worker productivity and morale, and how this relationship varies by the reasons cited for monitoring.

Intervention

Remote hiring and work have the potential to expand the set of labor market opportunities across geographies, and disproportionately benefit workers traditionally underemployed in their local labor markets (such as those in remote areas, or those preferring flexible work arrangements). However, which practices optimize worker productivity and well-being in remote work is less studied. In this paper, we focus on one management practice frequently and increasingly employed by remote work employers; digital monitoring of workers. Software that monitors digital activities can help managers better identify shirking and thus may make riskier hiring more feasible, but may also reduce worker motivation and well-being. Our study will test the impacts of digital monitoring on worker performance and job satisfaction among remote workers on an online platform.

Experimental design

We will hire workers in an online labor market for a data entry job. All workers will work for 2 hours in the same conditions, which will be in the "Team Room", which is an online room where their presence is monitored, and screenshot and activity summary is sent every 10 minutes that a freelancer tracks time. This is what we call "surveilled conditions" or “monitored conditions”, which in particular focus on input monitoring such as screenshots.

After two hours, we will randomly assign workers to one the following arms:
a) A random subset of high (above) median productivity workers (N=130), as measured by the number of accurate cells of data entered, will be told that they can work the other 6 hours out of the monitored conditions because they have performed well so far. We call this the "justified lack of monitoring" arm, since workers get the ability to work unmonitored as a reward for higher productivity.
b) A random subset of high (above) median productivity workers (N=130) will be informed about their productivity, and have the requirement to work in the monitored conditions digitally removed,. We call this the "lack of monitoring" arm, since workers get the ability to work unmonitored but without directly linking it to their productivity.
c) A random subset of high (above) median productivity workers (N=130) will be informed about their productivity, but there will be no change in their working condition.
d) A random subset of low (below median) productivity workers (N=130) will be told that they will work the other 6 hours under monitored conditions because they have made a high number of errors in the data entry. We call this the "justified monitoring" arm.
e) A random subset of low (below median) productivity workers (N=130) will be informed about their productivity, but there will be no change in their working condition. We call this the "monitoring" arm.
f) A random subset of low (below median) productivity workers (N=130) will be informed about their productivity, and have the requirement to work in the monitored conditions digitally removed. We call this the "lack of monitoring" arm.

Design details

Randomization method
Randomization will be done in the office with a computer
Randomization unit
Worker-level
Clusters
No clusters
Observations
800 workers
Arms
800 workers split equally into the above treatment arms
Primary outcomes
a) Productivity, which we will measure as the number of accurate cells of data entered during the hours that they are under the different treatment conditions. b) Subjective assessment of working conditions, and job satisfaction, as measured by a short survey administered at the end of the contract. We will form an index from the following survey questions, each of which will be standardized: 1. On a scale of 1-7, how much did you enjoy working on this text as compared to similar tasks you have worked on? 2. I feel like I was treated with respect on this job (1=not at all, 7=completely) 3. On a scale of 1-7, how likely would you be to accept another data entry task from us? We will also test whether workers reported feeling whether they were paid fairly for this job, which should not vary across treatments. We will test for heterogeneity by the following variables: a) Baseline productivity (whether above or median productivity) b) Internal and external motivation (by high or low of each of these measures, split at the median). These will be measured by the following survey questions at the end of the task: • Internal Motivation: How much do you agree with the following statement in response to the question of why do you work? “For the satisfaction I experience when I am successful at doing difficult tasks” • External Motivation: How much do you agree with the following statement in response to the question of why do you work? “For the income it provides me” c) Hofstede’s Power Distance Index, where each worker is assigned the power distance measure of their country d) Gender e) Worker wages (above vs. below median)
Secondary outcomes
Number of Mistakes Made in Data Entry Total number of hours worked (if some work less than max allowed) Whether or not work is done in monitoring condition (some allowed not to may opt to)

Cited by

No citation data available for this registration yet.

Keywords