Skip to main content

Labor Heterogeneity in Search Markets

Final Report Summary - LABORHETEROGENEITY (Labor Heterogeneity in Search Markets)

This project considered heterogeneity of workers and information frictions as the key source of labor market imperfections. If workers have to learn about how good they are they might choose types of work that use too little of their skills or require too much. Those who realize that they cannot meet the requirements would move to occupations where they can actually carry out the tasks, while those who realize that their skills are underused would move to occupations that are more demanding and pay more. This research documents this type of selection by looking at Danish data covering the entire working population. Indeed workers that earn relatively little in their occupation as well as workers that earn relatively much tend to leave the occupation, and the former tend to move to lower-paid occupations while the latter move to higher-paid occupations. Those with intermediate pay levels are more likely to stay. Previous work had mostly presumed the presence of the former group, while it was supposed that those with high earnings are most likely to stay. Our work shows that the well-known fact that those with longer occupational tenure earn higher wages is not just driven by the fact that high-wage workers tend to stay: the selection is rather symmetric between high and low earners. This suggests that there are actual real returns to staying in an occupation.
Occupational mobility might further be limited by lack of information about which occupations currently hold many vacant positions and whether one’s skills would be good enough to obtain one of the jobs. We designed a job search website that integrates information on this: Unemployed job seekers are asked which occupation they are looking for. Using data on other individuals who have already worked in that occupation, we infer related occupations to which these workers subsequently move. We also compile data on related occupations where skill transferability is high. We present this information to a test group of job seekers in Edinburgh, and show them all jobs that are available in these related occupations. This field experimental setup shows that this induces them to consider a broader set of occupations, especially if otherwise they searched over a rather narrow set. For that group it significantly increased job interviews. This is all evaluated relative to a control group. While there are many aspects that call for a larger study, the key innovation is to propose a replicable way to change the search process on the internet by embedding relevant information that was traditionally reserved only to labor market specialists.
Within this field experiment we also investigate which jobs are targeted by job seekers. We find that theories that presume that job seekers are choosy where to apply and where the wage is a main guiding force to attract applications do well relative to the data we see.
On the theory side this project extends our knowledge on how workers of different skills match with occupations/firms of different skill requirements. It opens a new field by considering firms that care both about the type and the number of workers, and to highlight in theory the driving forces that shape whether better workers are higher in better firms, what that means for the wage distribution, and how that affects the size distribution of firms. In a second step this is used to study the impacts of recent changes to economic activity, namely the increased opening up to international trade. This gives new theoretic insights both in wage dispersion across occupations (which had been the focus of well-known prior contributions) as well as within occupations. Finally, an ongoing theory project studies how workers that have to choose amongst different occupations and are screened by firms might get “discouraged” when they do not get jobs and therefore change the occupations they target. The main innovation here is to embed the well-known results when a single worker experiments across multiple possible occupations (the “bandit theory”) to a setting with many workers and many firms, and to analyze how the employment agency in a country might want to intervene to improve the job search outcomes.