Testing for Reciprocity of Work Flexibility

Last registered on July 19, 2019

Pre-Trial

Trial Information

General Information

Title
Testing for Reciprocity of Work Flexibility
RCT ID
AEARCTR-0004279
Initial registration date
July 17, 2019

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
July 19, 2019, 11:54 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
National Taiwan University

Other Primary Investigator(s)

PI Affiliation
National Taiwan University

Additional Trial Information

Status
In development
Start date
2019-07-22
End date
2020-01-31
Secondary IDs
Abstract
We plan to conduct an online field experiment exploring whether workers value and reciprocate work flexibility on an online platform. In our study, the additional flexibility takes the form of an extended deadline, allowing the online workers to have additional time to complete the job task. Although recent research has found that some subjects do value work flexibility, the average willingness-to-pay for the work flexibility may be quite low, with most of the gains coming from the tails of the distribution (Mas and Pallais 2017). However, even if this is true, it may still be in the employer's best interest to offer additional flexibility if workers are willing to reciprocate the gift of additional flexibility by working harder, akin to Kube et al. (2012). In addition, this experiment allows us to gather further information on potential differential preferences for males and females regarding time flexibility and higher wages (Goldin 2014).
External Link(s)

Registration Citation

Citation
DeJarnette, Patrick and Charles Po-Cheng Huang. 2019. "Testing for Reciprocity of Work Flexibility." AEA RCT Registry. July 19. https://doi.org/10.1257/rct.4279-1.0
Former Citation
DeJarnette, Patrick and Charles Po-Cheng Huang. 2019. "Testing for Reciprocity of Work Flexibility." AEA RCT Registry. July 19. https://www.socialscienceregistry.org/trials/4279/history/50372
Experimental Details

Interventions

Intervention(s)
We run this experiment on a large online platform, and as a result, we would like to ask for researchers to wait until completion date to see the full experimental design (currently hidden) or contact the authors individually, in case the online workers visit AEA RCT Registry.
Intervention Start Date
2019-07-22
Intervention End Date
2020-01-31

Primary Outcomes

Primary Outcomes (end points)
As the duration of work (3 hours) is fixed, the primary outcome of interest is the number of characters correctly entered in that 3 hour time period.
Primary Outcomes (explanation)
Workers are paid a flat hourly wage and work for a fixed duration of hours, so the number of characters entered is the primary measure of how much effort they put into the task.

Secondary Outcomes

Secondary Outcomes (end points)
A. Additional (unpaid) time spent and characters entered. B. Number and duration of breaks
Secondary Outcomes (explanation)
Although the 3 hours is the most they are getting paid for on Upwork, some workers may put in "extra" time due to their strong feelings of reciprocity. Likewise, although all workers work for 3 hours on Upwork, some may take extended breaks, which might be viewed as an alternative measure of (negative) reciprocity.

Experimental Design

Experimental Design
We run this experiment on a large online platform, and as a result, we would like to ask for researchers to wait until completion date to see the full experimental design (currently hidden) or contact the authors individually, in case the online workers visit AEA RCT Registry.
Experimental Design Details
We will run our experiment on Upwork, an online-hiring platform, and we will hire freelancers as the subjects to participate in our experiment. The experiment lasts for 3 hours for each worker and no worker is hired a second time.

The task is to copy the first paragraph of the first chapter from a list of books available on Google Books. The workers transcribe this paragraph into an online website we create. This task allows us to observe the number of characters correctly transcribed, and we use this as the measure of the workers' productivity. (As time spent working is fixed.)

There are baseline treatment and three additional treatment arms in this experiment. For the baseline treatment, we will give them a 2-day deadline to finish the data-entry work and the payment is $6 per hour. For the first treatment arm, we name it as “Time flexibility bonus”, we will extend the deadline to either 3 days or 5 days, and the payment remains at $6 per hour. For the second treatment arm, we call it “Money bonus”, we will increase the hourly wages to either $6.5 or $8 with the deadline remains two days. For the last treatment arm, we denote as “Choice bonus”, subjects get to choose between a 5-day deadline with $6 per hour or a 2-day deadline with $6.5 per hour.
Randomization Method
Randomization done in advance with a computer using the sample pool spreadsheet.
Randomization Unit
Individual worker
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
270 Workers

Details:
We use the criteria setting from Upwork. We collect those workers who are in the Philippines because Mas and Pallais (2014) found that the Philippines are common on this website, and the time zone between Taiwan and the Philippines are close (0~2 hours difference).

The "Earned Amount" is set as $1 earned; the "Job Success" is set as 80% & up; the "Hourly Rate" is set as $10 and below; the "Hours billed" is set as any hours.

We choose the "Category" as Data Entry; "English level" is fluent; "Talent Type" is freelancers since we don't want them to coordinate and communicate about this job. Finally, The "Last Activity" is set as within 2 weeks.

Under these criteria, we collect 4809 workers. However, some workers appear more than once, so we drop those additional observations. The actual workers become 4048.

We drop those whose posting wage is higher than 6.5 and lower than 4.5, then this gives us 1448 subjects. These 1448 workers serve as our sample pool, as we will randomly assign them to one treatment
Sample size: planned number of observations
270 Observations
Sample size (or number of clusters) by treatment arms
270 subjects in total. According to (List, Sadoff, Wagner 2010) to maximize power for comparisons of interest:

75 subjects in baseline,
30 subjects for 3 day treatment,
45 subjects for 5 day treatment,
45 subjects for 6.5 dollar treatment,
30 subjects for 8 dollar treatment,
45 subjects for Choice between treatments
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
Research Ethics Office
IRB Approval Date
2018-11-20
IRB Approval Number
201810HS026

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials