The impact of vocational rehabilitation on employment outcomes: A regression discontinuity approach.

Abstract

Since 2015, Finnish disability pension applicants who are rejected or receive a short-term temporary pension have, under certain conditions, also received a preliminary decision for vocational rehabilitation (VR). A key requirement for eligibility is a certain amount of earnings during the previous five years (€34 910.29 in 2017). We exploit this discontinuity to examine the impact of assignment to VR on labor market outcomes. All new disability pension applicants from 2015 to 2017 were included. Fuzzy regression discontinuity design was used to evaluate the impact of assignment to VR on employment, unemployment and earned income two years later among those close to the threshold (+/- €20 000) providing eligibility for the preliminary decision. Arguably, those just below and just above the earnings limit are similar to each other, allowing causal interpretation of the estimates. For each of the employment outcomes, we found a modest effect in the expected direction at the income threshold, but there is considerable uncertainty in these findings. On average, exceeding the income limit increased the probability of employment by 7.6 percentage points, but the estimate was far from statistical significance. Unemployment became slightly less common and earned income slightly increased, but the estimates were also clearly statistically non-significant. We found no consistent evidence of the impact of assignment to VR on employment outcomes among low-income disability pension applicants. However, given the narrow and specific study population, this should not be taken as evidence of ineffectiveness of VR more generally.