Fig. 2. Longer employment bans worsen employment trajectories of refugees.

(A) Employment trajectories of FRY refugees who arrived in Germany in 1999 (green) and 2000 (red) (n = 1748). The 1999 arrival cohort faced a 13- to 24-month employment ban (depending on their month of arrival), while the 2000 arrival cohort faced a 12-month employment ban. The average difference in the length of the waiting period between the 1999 and 2000 cohorts is 7.1 months. The dots indicate the percentage of respondents who are in paid employment by survey year. The curved regression lines and corresponding 95% confidence intervals are a nonparametric approximation of the employment trajectories using regression B-splines. (B) Results of the first placebo test: Turkish immigrants who arrived in 1999 and 2000 but were not subject to the ban experienced very similar employment trajectories (n = 3712). (C) Results of the second placebo test: FRY refugees who arrived in 2000 and 2001 and were subject to the same 12-month waiting period experienced virtually identical employment trajectories (n = 1067).