Productivity Growth in the European Insurance Industry: Evidence from Life and Non-Life Companies

Dev Vencappa*, Paul Fenn, Stephen Diacon

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


This paper estimates and decomposes productivity growth for a sample of European insurance companies using Standard and Poor's Eurothesys panel data set of company accounts. We focus on a period (1995-2008) where substantial deregulation took place, as well as significant shocks to global capital markets and unforeseen climatic and terrorism-related events. We estimate parametric stochastic production frontiers for life and non-life insurance from which we decompose productivity growth using a derivative-based approach. In order to capture the inherent variability of productivity in the insurance sector, we explore the consequences of adopting three different proxy measures for insurance output. In each case, we observe temporal variations in overall productivity growth for both life and non-life insurance, corresponding to systematic shocks to the market. The decomposition of these fluctuations provides important lessons for the measurement of long-term productivity growth in insurance markets and more generally in sectors which are exposed to stochastic and market-wide systematic shocks to performance. Our findings cast doubt on many previous studies of productivity growth in the financial service sector where outputs are inevitably difficult to measure due to their contingent nature. The choice of output proxies appears to be critical, particularly when exploring long-term productivity trends.

Original languageEnglish
Pages (from-to)281-305
Number of pages25
JournalInternational Journal of the Economics of Business
Issue number2
Publication statusPublished - Jul 2013


  • Insurance
  • Stochastic Frontier Analysis
  • Total Factor Productivity Growth


Dive into the research topics of 'Productivity Growth in the European Insurance Industry: Evidence from Life and Non-Life Companies'. Together they form a unique fingerprint.

Cite this