Automated Occupational Advice for Long-Term Unemployed During Online Job Search, 2019-2021

The data covers job search activities and employment outcomes for participants in an online study on the provision of occupational recommendations to job seekers. Providing job search assistance to job seekers in a cost effective manner is a challenging goal. Interventions aimed at providing tailored advice typically involve large personnel costs that often dissipate the benefits. However, the advances in information technologies and the shift of formal job search to online platforms over the last 20 years offer new opportunities for providing advice at very low-cost. In this study we examine the potential for providing on-line advice to a population of hard-to-place job seekers. In a randomized field experiment, we provided suggestions about suitable alternative occupations to long-term unemployed job seekers. The suggestions were automatically generated, integrated in an online job search platform, and fed into actual search queries. Effects on the primary pre-registered outcomes of "finding a stable job" and "reaching a cumulative earnings threshold" are positive, large, and are more pronounced for those who are longer unemployed. Treated individuals include more occupations in their search and find more jobs in recommended occupations.