Safety Impacts of Experience Rating in Workers' Compensation

Construction worker on building frame


Experience rating is a fundamental component of all workers' compensation insurance systems in the United States and Canada. Experience rating is intended to perform two functions:

  1. the public policy directive of reducing workplace injuries and illnesses through economic incentives on employers; and
  2. assisting insurance companies in pricing heterogeneous risks among employer in similar industries.

The focus of these two functions differs, safety versus pricing. But experience rating is designed and calculated by insurance companies through insurer controlled "rating bureaus" whose primary concern is pricing. Consequently, the public policy imperative of improving safety through employer incentives may not be efficiently implemented.


This study examines two areas where the application of experience rating could potentially be modified to improve the incentive for safety without reducing its effectiveness as a tool in underwriting risks.

The first area is the experience rating of smaller employers. Only about 20% of employers, the largest in terms of premium, are experience rated. Insurers choose to exclude the smaller employers on the grounds that their experience lacks sufficient credibility when measured against resulting variation in premiums. However, this calculus does not include the potential positive impact of injury and illness reduction. This part of the study measures the safety impact of expanding the experience modification process to include more small employers.

The second area considers the degree to which employers' premiums are sensitive to past experience. The experience modification formula includes several factors meant to reduce the sensitivity of the modifier to changes to claims experience. The insurers' purpose in dampening employers' sensitivity to experience is to reduce the experience-driven variation in premiums when the credibility of past experience is lower (smaller employers) and increase the sensitivity when credibility is higher. But, again, this calculus ignores the question of whether greater sensitivity to past experience would increase employers' attention to, and investment in, safety. We exploit a natural experiment to determine whether similar employers with the same experience, but bigger swings in experience modifiers, respond by reducing injury rates more than employers whose modifiers are less effected.

Project Team

Frank Neuhauser, Principal Investigator
Amelia Haviland
John Mendeloff
Seth Seabury
Susan Straus
Christopher McLaren, PRGS Student