We propose to conduct an experiment on a major international online freelancing labor market platform to study the impact of greater flexibility in choosing work hours on productivity and worker benefit. This document describes the analysis plan for the natural field experiment. This plan outlines the design of the study, the outcomes of interest, and the econometric approach.
Experimental design
The study aims to understand the effect of providing flexibility in choosing the time of working on productivity and labour welfare. The study will use a natural field experiment to causally estimate the effect. We will post identical jobs on a major online freelance labour market platform that only differ in flexibility. These will be data entry jobs that are to be completed in a continuous two hour period and will have both a fixed wage component and a bonus component as well. A randomly chosen set of freelancers will be invited to apply to these jobs and all the applicants will be hired. At the end of the task, candidates will be asked to complete a survey. Our targeted sample size is 750 freelancers across all treatment arms. The target sample for the inflexible job is 300 freelancers (finishing the task) and for the committed hours is 300 freelancers (finishing the task) and 150 freelancers (finishing the task) for flexible jobs.
Design details
Randomization method
Randomization done in office by a computer
Randomization unit
Individuals
Clusters
Not Clustered
Observations
750 Individuals
Arms
150 individuals per treatment arm
Primary outcomes
Number of acceptances, Number of cells entered, Number of complete cells entered, Number of correct cells entered, Number of incorrect cells entered, bonus earned, Uninterrupted time spent on the web portal, number of cells correctly entered/time spent actively working and number of cells correctly entered/time spent actively working, Number of complete rows entered, Number of correct rows entered, Uninterrupted time spent on the web portal, number of rows correctly entered/time spent actively working and number of rows correctly entered/time spent actively working, Self assessment of productivity, Adjustments made to schedule, difficulty in finding time for the task, room for oneself during the task, stopping for caring responsibility, self assessment of work experience.
Secondary outcomes
Cover letter length, work samples, experience measures (total prior contracts, total prior contracted hours, total prior earnings), counteroffer, underbids, overbids and hourly wage, Age, married, number of children, number of young children, spouse employment, full time status on the platform, fixed work hour schedule, responsibilities during the day and educational qualification. completion of the task in the first attempt (compliance) and completion of the survey.
Our primary interest of heterogeneity is by gender. We will look at heterogeneous effects by gender for all the variables. We will also look at heterogeneity by region and stated hourly wage level.
Other variables of interest are stated preference for flexibility, stated time preference, acceptance of an inflexible job, chose to answer the optional survey and employer niceness. For continuous outcome variables we will also use log transformations of the outcome variables.
Cited by
No citation data available for this registration yet.