Using DEA and survival analysis for measuring performance of branches in New Zealand's Accident Compensation Corporation

Abstract

This paper describes an approach to solving the problem of assessing relative branch performance in the Accident Compensation Corporation (ACC), the New Zealand state-owned, no-fault, personal injury compensation insurance company. The methodology described in this paper is innovative in that it assesses relative performance via a modified data envelopment analysis (DEA) process. Performance is assessed and can be examined in terms that are meaningful to business goals. Performance evaluation has generally aligned well with management preconceptions of performance. DEA results are clustered to find common performance themes that are reviewed for business acceptability and the successful themes are used to measure performance against a Farrell frontier. The methodology is applicable to other situations where there are multiple instances of a unit performing the same or similar functions. Technical analysis as used in this paper is based on survival analysis, DEA, clustering and knowledge of the organisation's business operations and objectives.