This resource summarizes the findings from a random assignment study that tested the impacts of a behavioral intervention–email communication to Unemployment Insurance recipients–to prompt or motivate the respondents’ participation in a Reemployment and Eligibility Assessment program in Southwest Michigan in 2015.

Finding a job after becoming unemployed can be challenging for many individuals. Even as the unemployment rate has decreased during the recovery from the 2007-2008 crisis, the average duration of regular unemployment insurance (UI) benefits remains high.

One effective tool for helping unemployed workers find new employment faster, and shorten their duration of UI benefit receipt, is the Reemployment and Eligibility Assessment (REA) program.

Michigan was one of 44 states that received part of $80 million in REA grants awarded by DOL in 2015…At that time, Michigan began its REA program with plans to serve 9,000 UI claimants across five workforce areas between January and September 2015. Michigan Works! Southwest, which serves Branch, Calhoun, Kalamazoo, and St. Joseph Counties, was one of the five implementing workforce agencies.

The Department of Labor (DOL) Behavioral Interventions project was launched to explore how insights from behavioral science can be used to improve the performance of outcomes of DOL programs. The project team has designed, implemented, and rigorously tested three behavioral trials in selected Labor programs. DOL was especially interested in testing behavioral interventions that would allow for rapid evaluation and analysis of short-term outcomes and would be easy to replicate and/or scale if found to be effective.

This report presents findings on the effects of emails designed to encourage UI claimants to (1) schedule and attend REA sessions with Michigan Works! Southwest and (2) persist in efforts that will help them succeed in their job search efforts following REA program completion.

The authors conducted a randomized control trial to test the effectiveness of the series of emails, comparing the attendance rates across 372 UI claimants who received emails and 375 UI claimants who received only the Michigan UIA notification letter.

The trial was designed to answer four main research questions:

  1. Did the emails improve the initial response/scheduling rate for the UIA notification letter relative to the status quo?
  2. Did the emails improve the attendance rate for the REA sessions?
  3. Did the emails improve the completion rate for the REA program?
  4. Did the effects of the emails substantially differ across key subgroups?

The authors collected data directly from the Michigan Works! Southwest administrative data.


Major Findings & Recommendations

Results for the trial provide strong evidence that the email interventions encouraged more individuals to sign up for, attend, and complete the REA program. Specifically:

  • Low-cost behaviorally informed emails increased the proportion of UI claimants who participated in the first REA session by 15 percentage points and increased participation in remaining sessions.
  • The emails increased the proportion of UI claimants who completed the REA program by 14 percentage points.
  • The study did not find any statistically significant variation in impacts based on the age and education levels.

The authors offered the following discussion of the results:

  • Impacts on REA program participation could translate into meaningful savings over the long term.
  • Program effects were strongest after the first two emails. The implication is that many claimants may have intended to respond to the first email or the Michigan UIA notification letter but failed to do so. In such cases, reminders are an effective method to increase the probability of response.
  • The intervention has broad relevance and scalability results are especially promising because of the intervention’s low-touch, low-cost nature.
  • Email interventions may be effective even if many people do not read or open the email. The recipient’s awareness of a message sent by a recognized sender on a relevant subject may have produced the desired result.
  • Strategies to get people ‘in the door’ can have a long-lasting effect. The email messages tested in the trial substantially increased the number of treatment group claimants who scheduled their first REA session.
  • The REA emails helped to soften language that could turn off job seekers. It was more customer friendly, less formal. People who received the email were more willing to ask questions, and felt more comfortable interacting with the staff.

Finally, the authors provide implementation lessons for those interested in conducting behavioral trials. For example, in-person brainstorming can speed the development of intervention designs and implementation plans. A two-day site visit allowed the authors to conduct a ‘deep dive’ into the REA program and understand how it operated at Michigan Works! Southwest, diagnose possible causes for poor engagement, identify operational constraints, and propose intervention options.