This report describes studies in two states – North Carolina and Washington – to test new communication methods to increase Unemployment Insurance (UI) claimants' compliance with various "work search" requirements. The requirements include contacting a certain number of employers per week, recording employer contact information, position sought, and results of the contact, and maintaining records on the work search activities. The communication methods tested included, among others, pop-up alerts and emails, as well as clearly stated information in "tip sheets" about the types of employer contacts that counted and how to document work search activities.

The new approaches were devised after an analysis of the barriers to compliance and identification of possible solutions using behavioral principles, as part of a broader initiative, the Behavioral Interventions in Labor Programs Evaluation project, under the Department of Labor's Chief Evaluation Office. This project also involved collaboration with the Employment and Training Administration, as well as the two states' UI agencies.

Two separate studies, using multi-method, multi-phased approaches, were conducted. In Washington, a randomized control trial (RCT) involving 26,967 claimants compared knowledge and work search behaviors between a treatment group who received the intervention and a control group who did not. Outcomes were assessed using administrative data as well as information from phone interviews with staff. In North Carolina, an RCT involving 24,416 claimants explored the communications’ influence on self-attested work search compliance, while a series of quasi-experimental design (QED) studies explored other outcomes. Outcomes were assessed using administrative data as well as information from phone interviews with staff and in-person focus groups with staff and claimants. However, the COVID-19 pandemic limited some aspects of data collection and prevented the possibility of improving features of the interventions and further testing, as is typical in behavioral trials.

The quantitative data in both states showed no statistically significant impact on claimants’ knowledge of work search requirements, as measured through self-attestation of work search or work search behavior. However, qualitative findings suggested areas for improving the interventions to increase claimants' understanding of certain aspects of the requirements, such as how to record and provide documentation of work search activities – which could inform future research.

The trials demonstrated the feasibility of quickly deploying behaviorally-informed communications and measuring their effectiveness with administrative data. State partners reported the collaborative design process to apply behavioral principles was useful. In North Carolina, state partners have scaled up the use of the behavioral intervention materials from the trial with new claimants and have demonstrated interest in applying behavioral insights to improve operations. In Washington state, staff reported that learning to apply behavioral insights has increased their ability to quickly revise communications going forward.