Abstract
This paper describes the development of the revised Workers Compensation Experience Rating Plan. This is based on sound statistical theory, certain modeling assumptions, and thorough empirical testing. It is heartening that the empirically derived parameterization is consistent with most of the assumptions needed to simplify the algebraic foundation.
The paper begins with an heuristic derivation of a general modification formula based on losses split into primary and excess portions. It delineates the assumptions about the components of loss ratio variance leading to the algebraic form of the formulae tested.
Iterative testing is used to parameterize those formulae. A simple preliminary test procedure is described to clarify the basic concepts. The operative test procedure is then specified, and results of iterative testing using that procedure are displayed for the selected formulae.
The parameterized formula finally approved by the National Council on Compensation Insurance (NCCI) was subject to certain adjustments to maintain continuity during the transition from the old to new plans. Credibilities have been scaled to account for differences in state benefit levels and the effect of inflation.
Keywords: Workers Compensation, Credibility, Experience Rating Plans.
Volume
LXXIX
Page
21-56
Year
1992
Categories
Actuarial Applications and Methodologies
Ratemaking
Experience Rating
Business Areas
Workers Compensation
Publications
Proceedings of the Casualty Actuarial Society
Prizes
Woodward-Fondiller Prize