Job Search on Digital Platforms: A Large-Scale Experiment

How to cite

Rademakers, Emilie; Goos, Maarten. 2021. "Job Search on Digital Platforms: A Large-Scale Experiment." AEA RCT Registry. AEARCTR-0007130. https://doi.org/10.1257/rct.7130-1.0

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

Emilie Rademakers* , Maarten Goos

* Primary investigator

Registry ID
AEARCTR-0007130
DOI
https://doi.org/10.1257/rct.7130-1.0
First registered
Last updated
Status
In development
Trial dates
2020-04-01 – 2022-10-08
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/7130

Abstract

In this field experiment, we provide unemployed job seekers with additional information about online job search at the start of their unemployment spell. This is done in two stages. First, we provide encouragement to search actively. Second, we integrate a nudge for job seeker towards using the online job portal provided by the public employment service. Job seekers are tracked for up to 12 months. During this time, we will monitor search behavior on the platform as well as administrative records of job seekers and their job finding rates. This will give us greater insight into remaining barriers towards the use of online platforms in job search and how cost-effective nudges in communication may help lower them. Moreover, this information allows us to measure the impact of search online on the job finding rates for the treated job seekers.

Experimental design

Our intervention consists of augmenting the existing email that is sent at the start of the unemployment spell with additional information and motivation. This is done in two steps, which we refer to as Nudge A and Nudge B. First, Nudge A consists of general information on the labour market and the importance of actively searching for a job. Second, Nudge B encourages job seekers to act by referring them directly to an existing search tool.
Job seekers are randomized at the individual level in three groups after registration. The first group is given the existing communication with administrative information only. The second group is given the extended email containing Nudge A. The third group is given the extended email containing Nudge A as well as Nudge B. Randomization occurs immediately after registration since the emails are normally sent out the following day.

Design details

Randomization method
randomization done by a computer
Randomization unit
individual
Clusters
44 000 individuals
Observations
44 000 individuals
Arms
14 670 individuals control, 14 670 individuals in treatment 1: Nudge A, 14 670 individuals in treatment 2: Nudge A and Nudge B.
Primary outcomes
job finding rate

Cited by

No citation data available for this registration yet.

Keywords