Long-Term Effects of Supported Employment for Young Adults Receiving Disability Pension: A Swedish Randomized Controlled Trial

Purpose Young adults receiving temporary disability benefits face a high risk of long-term labor-market exclusion. We examined the five-year effects of a Swedish randomized trial comparing Supported Employment (SE), Case Management (CM), and Regular Rehabilitation (RR). Methods Young adults aged 19–29 receiving disability benefits were randomized in 2014–2015 to SE, CM, or RR. Outcomes were measured in administrative registers up to five years after randomization. We estimated intention-to-treat effects on mean annual labor earnings, earnings above one statutory price-base amount, mean annual disability-benefit amount, and receipt of disability benefits. Results SE increased mean annual labor earnings by about SEK 10,500 and the proportion with earnings above one price-base amount by 6 percentage points in years 1–2 relative to RR. In years 3–5, the estimated differences in labor earnings and earnings above one price-base amount were smaller than in years 1–2 and were not statistically significant. The short-term reductions in disability benefits did not persist; the long-term estimates were positive but imprecise. Outcomes in RR gradually converged with those in SE over the follow-up period. CM did not produce comparable short-term gains, and long-term estimates indicated higher disability benefit receipt relative to RR. Direct comparisons showed more favorable short-term outcomes for SE than for CM. In years 3–5, the estimated differences were smaller in magnitude and statistically imprecise, and none were significantly different from zero. Conclusions SE accelerated early labor-market entry among young adults receiving disability benefits, but the average differences relative to RR diminished over five years. The results are consistent with the possibility that a strong focus on job entry and workplace support is important, although the relative roles of intervention content and intensity cannot be determined conclusively. The long-term estimates were imprecise and do not rule out modest effects in either direction. Trial registration Registry: AEA RCT Registry; trial registration number: AEARCTR-0003794; date of registration: January 25, 2019; retrospectively registered.