The (in)visible hand: do workers discriminate against employers?

Last registered on May 04, 2020


Trial Information

General Information

The (in)visible hand: do workers discriminate against employers?
Initial registration date
May 04, 2020
Last updated
May 04, 2020, 1:58 PM EDT



Primary Investigator

Hofstra University

Other Primary Investigator(s)

PI Affiliation
University of Mannheim
PI Affiliation
Indiana University

Additional Trial Information

Start date
End date
Secondary IDs
Do employees discriminate against potential employers? We implement a random- ized experiment in an online labor market that is pro-typical for a set-up where employees have the opportunity to discriminate against employers to answer this research question. In our experiment, workers make labor-supply decisions after we randomly expose them to signals about the race and sex of the potential em- ployer. Investigating average effects for the full sample of workers, we find evidence of discrimination on the quantity and quality of work (conditional on working), but not on the general willingness to work in our labor task. Upon completing the experiment, workers fill out a survey in which they report whether they discerned the signals regarding race and sex of the employer. The survey responses reveal an interesting pattern of heterogeneity: subjects who discerned the treatment sig- nals were more likely to work for the minority employer, while those who did not discern the signal were less likely to work for the minority employer. The latter result suggests that discriminating employees try to conceal their behavior ex-post. An additional survey with randomized components among online workers suggests that our results are not driven by statistical discrimination.
External Link(s)

Registration Citation

Doerrenberg, Philipp, Denvil Duncan and Danyang Li. 2020. "The (in)visible hand: do workers discriminate against employers?." AEA RCT Registry. May 04.
Experimental Details


We recruited subjects from Amazon Mechanical Turk (mTurk) to complete a survey about a topic unrelated to our discrimination study. Upon completion of this survey, subjects are randomly assigned to one of five groups and offered an opportunity to complete an additional bonus task where they are paid a piece rate to transcribe information from gasoline receipts. The announcement of the additional bonus task features a photograph of a hand holding a gasoline receipt. We signal the employer’s race and sex by randomly varying the presence and characteristics (gender and race) of the hand in the photograph across groups. After being exposed to the treatment photograph, subjects are asked if they wish to complete the bonus task. Subjects who respond yes are allowed to transcribe up to 40 gasoline receipts.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
1. Subjects' decision to work for the bonus task or not
2. The number of receipts subjects choose to transcribe
3. The share of accurate entries across all transcribed receipts
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We signal race and sex of the potential employer by showing subjects a picture of a hand holding a receipt. We have five treatment groups:
i) Black-Female (BF): Subjects see receipts held by a female hand with black skin and nail polish.
ii) Black-Male (BM): Subjects see receipts held by a male hand with black skin (no nail polish).
iii) White-Female (WF): Subjects see receipts held by a female hand with white skin and nail polish.
iv) White-Male (WM): Subjects see receipts held by a male hand with white skin (no nail polish).
v) Control: Subjects see receipts that do not include a hand.
Experimental Design Details
Randomization Method
The randomization process is computerized through Qualtrics. Subjects who opted in the bonus task were randomly assigned to one of the five treatment groups.
Randomization Unit
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
2500 individuals
Sample size: planned number of observations
2500 individuals
Sample size (or number of clusters) by treatment arms
500 individuals per treatment group
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
Hofstra Unversity
IRB Approval Date
IRB Approval Number
IRB Name
Indiana University/IU Health
IRB Approval Date
IRB Approval Number


Post Trial Information

Study Withdrawal

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information


Is the intervention completed?
Intervention Completion Date
February 20, 2019, 12:00 +00:00
Data Collection Complete
Data Collection Completion Date
February 20, 2019, 12:00 +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
2500 individuals
Final Sample Size (or Number of Clusters) by Treatment Arms
500 individuals per treatment group
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials