Skills mismatch - defined as the discrepancy between worker’s abilities and job skills requirements - it is ubiquitous and associated with long lasting wage penalties. What are the sources of skills mismatch and how to mitigate its adverse consequences on workers? This project seeks to advance our understanding of these two questions in two complementary strands.

The first strand focuses on skills mismatch at the onset of workers’ careers. The research will identify and compare the relative importance of various sources of mismatching at the micro level (uncertainties about abilities, job characteristics and psychological  constraints), so that early interventions on individuals may help reduce the unfavorable consequences of mismatches. The analysis will apply the combined approaches of randomized controlled trials (RCT), reduced form, and structural modeling techniques to (i) jointly analyze the roles and processes of beliefs about abilities, pecuniary and non-pecuniary aspects of jobs on occupational choices, and (ii) identify the relative importance of the sources and mechanisms affecting mismatching at the early stages of careers.

The second strand concentrates on analyzing aggregate factors at the macro level - such as changes in the distribution of job offers - which could affect worker-job allocations at any career stages. The growing technological transformation such as AI calls for a deeper assessment of the issues as they are likely to cause drastic changes in job offer distributions (characteristics and match quality) and result in unequal impacts across skill and experience distributions. Using the RCT, online job postings, and matched employer-employee data, I will study (i) the roles of beliefs on mismatching, and (ii) the drivers of the unequal impacts of changes in technologies on worker-job allocations.

Co-investigators:
Ilse Lindenlaub (Yale University)
Matthew Merkel (UCL)
Rune Vejlin (Aarhus University)
Warn n. Lekfuangfu (Universidad Carlos III de Madrid)
Weerachart Kilenthong (Research Institute for Policy Evaluation and Design)

Co-authors:
Arnaud Maurel (Duke University)
Igansi Merediz Sola (UCL)
Rachel Tan (Singapore Management University)

Research Assistants:
Vrinda Anand (UCL)

The RCT intervention has been completed and endline data collection is currently in progress.