Objective To determine whether step-downs, which cut the rate of compensation paid to injured workers after they have been on benefits for several months, are effective as a return to work incentive.

Methods We aggregated administrative claims data from seven Australian workers’ compensation systems to calculate weekly scheme exit rates, a proxy for return to work. Jurisdictions were further subdivided into four injury subgroups: fractures, musculoskeletal, mental health and other trauma. The effect of step-downs on scheme exit was tested using a regression discontinuity design. Results were pooled into meta-analyses to calculate combined effects and the proportion of variance attributable to heterogeneity.

Results The combined effect of step-downs was a 0.86 percentage point (95% CI −1.45 to −0.27) reduction in the exit rate, with significant heterogeneity between jurisdictions (I ²=68%, p=0.003). Neither timing nor magnitude of step-downs was a significant moderator of effects. Within injury subgroups, only fractures had a significant combined effect (−0.84, 95% CI −1.61 to −0.07). Sensitivity analysis indicated potential effects within mental health and musculoskeletal conditions as well.

Conclusions The results suggest some workers’ compensation recipients anticipate step-downs and exit the system early to avoid the reduction in income. However, the effects were small and suggest step-downs have marginal practical significance. We conclude that step-downs are generally ineffective as a return to work policy initiative.

Postprint link: https://www.medrxiv.org/content/10.1101/19012286