We examine how job seekers' subjective beliefs about re-employment wages affect their perceived labor market prospects, job search behavior and labor market outcomes.
Experimental design
We measure respondents' beliefs about re-employment wages and shift them using information calculated from the Danish registers. We then study effects of our intervention on perceived labor market prospects, job search behavior and labor market outcomes.
Design details
Randomization method
Computerized via oTree.
Randomization unit
Individual
Clusters
We invite 83,591 job seekers through their "e-boks" - the official digital mailbox for government services in Denmark used by all Danish residents. This represents all individuals in Denmark that we can potentially use in our experiment, excluding those that we contacted for pilot studies. We will use reminders in the weeks after the initial invitation to increase the overall response rate. Based on previous experience, we expect to achieve an overall response rate of between 10% and 15%, giving us an expected overall sample size between 8,395 and 12,539. Out of these, we expect to screen out about 40% because they cannot be used for our experiment for different reasons: those who have already accepted a job offer, those who exclusively consider part-time jobs, those who have never been employed before (for which the previous wage, which is used to calculate the treatment information, is undefined), and those in small demographic cells (for which the treatment information cannot be reliably calculated).
Observations
We invite 83,591 job seekers through their "e-boks" - the official digital mailbox for government services in Denmark used by all Danish residents. This represents all individuals in Denmark that we can potentially use in our experiment, excluding those that we contacted for pilot studies. We will use reminders in the weeks after the initial invitation to increase the overall response rate. Based on previous experience, we expect to achieve an overall response rate of between 10% and 15%, giving us an expected overall sample size between 8,395 and 12,539. Out of these, we expect to screen out about 40% because they cannot be used for our experiment for different reasons: those who have already accepted a job offer, those who exclusively consider part-time jobs, those who have never been employed before (for which the previous wage, which is used to calculate the treatment information, is undefined), and those in small demographic cells (for which the treatment information cannot be reliably calculated).
Arms
After screening out some participants that we cannot use in our survey, we allocate our respondents randomly into treatment or control group with equal chance. We expect to end up with between 2,500 and 4,000 observations in each group.
Primary outcomes
The main outcomes we are interested in are the following:
- Job seekers' beliefs, e.g., about own re-employment wages, and planned job search behavior measured in the survey, e.g., their search effort, reservation wages and willingness to make concessions.
- Job search beliefs about own re-employment wages, reservation wages, perceived job finding prospects, as well as the intensity of search as elicited in the survey.
- Job seekers' search activities on the online platform of the Danish public employment service (jobnet.dk) measured through click-by-click data, e.g., the frequency of logins, searches for vacancies and the "types" of jobs targeted in the vacancy database.
- Job seekers' realized labor market outcomes, e.g., job finding probabilities, as observed in the administrative records.
Cited by
No citation data available for this registration yet.