De Economist (2009) 157:315–335 © Springer 2009 DOI 10.1007/s10645-009-9125-6 DE ECONOMIST 157, NO. 3, 2009 EXPERIENCE RATING AND THE INFLOW INTO DISABILITY INSURANCE ∗∗ BY PIERRE KONING ∗ Summary This paper examines the effects of experience rating on the inflow into disability insurance (DI) in the Netherlands, using unique longitudinal administrative data from the social benefit admin- istration. We follow a difference-in-differences approach to identify the impact of changes in DI premiums. Due to unawareness of the experience rating system, employers seem to have been triggered to increase preventative activities, once they have experienced increases in DI premium rates. We find this impact to be substantial, amounting to a 15% reduction of the DI inflow. Key words: experience rating, disability insurance, panel data JEL Code(s): H22, I12, C23 1 INTRODUCTION As of 1998, disability insurance (DI) in the Netherlands is financed by premiums that are experience rated. This means that, in principle, employers bear the costs of the first 5 years of DI benefits. Initially, the experience rat- ing system did not cause substantial controversy among employers and pol- icy makers. It appears this has been due to the nature of the incentive system itself: each year, a new cohort of disability benefit costs has been added to the disability premium. In 2003 the experience rating incentive had reached its maximum impact. At the same time, criticism against the experience rat- ing has grown steadily, as an increasing group of employers had been con- fronted with substantial increases in their premiums. The major argument of the opponents against experience rating – both employer organisations and politicians – is that employers, in particular those with small employer size, ∗ Corresponding author: Pierre Koning, CPB Netherlands Bureau for Economic Policy Analysis, Van Stolkweg 14, P.O. Box 80510, 2508GM, The Hague, The Netherlands; email: pwck@cpb.nl ∗∗ UWV is gratefully acknowledged for giving access to the data and providing facilities that were needed to carry out research activities. The suggestions and comments of Rob Alessie, Wolter Hassink, Maarten Lindeboom, Richard Nahuis, Peter Rijnsburger and David Wittenburg, as well as two anonymous referees have improved the paper substantially.