Experimenting with Caseworker Direction

Abstract

We experimentally study the intensity of caseworker direction given to eligible voucher recipients selecting a job-training program. Requiring counseling led some to forgo training, but those assigned to the most directive approach had higher survey-reported earnings seven years after study entry, with targeting implications: dislocated workers did no worse with optional counseling, while disadvantaged workers earned less. Although impacts on earnings were inconsistent for dislocated workers when using administrative data, they were consistent for disadvantaged workers. Finally, using simulations, we find that changing caseworker direction can have impacts that are similar in magnitude to participating in any job training.