Randomized experiments provide policy relevant treatment effects if there are no spillovers between participants and nonparticipants. We show that this assumption is violated for a Danish activation program for unemployed workers. Using a difference-in-difference model we show that the...

We employ a large social experiment combined with register-based data allowing for up to 12-year follow-up to evaluate a long-lasting employer-sponsored health and well-being program. We show that employees at treated worksites receive fewer consultations from their primary care physician and...