Firm size and growth in Sweden's life insurance market between 1855 and 1947: A test of Gibrat's law

Abstract

Using data for the period from 1855 to 1947 and the two sub-periods, 1855–1902 and 1903–47, the article examines whether the organic growth rates of 38 Swedish life insurance firms are independent of size, as predicted by Gibrat's (1931) Law of Proportionate Effects. Using panel unit root tests and panel Generalised Method of Moments (GMM) regression, the article finds a significant difference between the growth rates of small and large Swedish life insurance firms (with smaller firms tending to grow faster than larger firms), a result that clearly contradicts Gibrat's Law as a long-run tendency in the Swedish life insurance sector. significant influences were also found on firm growth from profitability, organisational form, reinsurance, the real rate of interest and the Swedish regulatory environment.

Keywords::

• Gibrat's law
• insurance
• Sweden

Notes

^1.Santarelli, Klomp, and Thurik, “Gibrat's Law.”

^2. Mazzacuto, “Risk, Variety, and Volatility.”

^3.Geroski, Machin, and Walters, “Corporate Growth and Profitability.”

^4.Audretsch et al., “Gibrat's Law.”

^5.Hardwick and Adams, “Firm Size and Growth”; Choi, “U.S. Property and Liability Industry.”

^6.A'Hearn, ‘Finance-Led Divergence.”

^7. e.g., see Lindmark, Andersson, and Adams, “Evolution and Development”; Lindmark and Anderson, “All Fired up”; Pearson, “Mutuality Tested”; Pearson, “Insuring the Industrial Revolution”; Ryan, “Expansion of the Norwich Union”; Zanjani, “Pricing and Capital Allocation.”

^8.Smith and Stutzer, “Mutual Formation and Moral Hazard”

^9. Lindmark, Andersson, and Adams, “Evolution and Development”; Adams, et al., “Mutuality as a Control for Information Asymmetry.”

^10. For example Swiss Re, “World Insurance in 2006” report that in 2006 gross global insurance premiums written amounted to roughly US4 trillion, while Browne et al. “International Property-Liability Insurance” estimate that approximately 8% of the world’s GDP is spent on insurance products, mostly in European and North American markets.

^11. Dunne, Roberts, and Samuelson, “The Growth and Failure.”

^12. Life and non-life insurers are regulated, taxed and accounted for on a different basis. As a result, we consider that any improved statistical precision possibly emanating from the pooling of both life and non-life insurance firms would be offset by the problem of measurement bias in estimating the linkage between firm size and growth rate patterns, see e.g. Hardwick and Adams, “Firm Size and Growth.”

^13.Skogh, “Returns to Scale.”

^14.Andersson, Eriksson and Lindmark, “Life Insurance and Income Growth.”

^15. See e.g. Waldenström, “Taxing Emerging Stock Markets.”

^16. Lindmark, Andersson, and Adams, “Evolution and Development.”

^17. Andersson, Eriksson, and Lindmark, “Life Insurance and Income Growth.”

^18. Ibid.

^19. Ibid.

^20. Ibid.

^21.Hägg, “Insurance regulation.”

^22. Lindmark, Andersson, and Adams, “Evolution and Development.”

^23.Sutton, “Gibrat's Legacy.”

^24.Gibrat, “Les Inegalites Economiques.”

^25. Hardwick and Adams, “Firm Size and Growth.”

^26.Chesher, “Testing the Law of Proportionate Effect.”

^27. See e.g. Dunne, Roberts, and Samuelson, “Growth and Failure”; Geroski, Machin, and Walters, “Corporate Growth and Profitability”; Hardwick and Adams, “Firm Size and Growth.”; Choi, “U.S. Property and Liability Industry.”

^28.Hardwick, “Measuring Cost Inefficiency.”

^29.Cummins, Tennyson, and Weiss, “Consolidation and Efficiency.”

^30. Hardwick and Adams, “Firm Size and Growth.”

^31. See e.g. Cummins et al., “Consolidation and Efficiency”; Hardwick, “Measuring Cost Inefficiency.”

^32.Fields, “Expense Preference Behavior.”

^33. Westall, “Invisible, Visible and ‘Direct’ Hands.”

^34. Hardwick, “Measuring Cost Inefficiency.”

^35. Hardwick and Adams, “Firm Size and Growth.”

^36. Ibid.

^37. Geroski et al., “Corporate Growth and Profitability.”

^38. Zanjani, “Pricing and Capital Allocation.”

^39.Jaffee and Russell, “Catastrophe Insurance.”

^40.Santomero and Babbel, “Financial Risk Management.”

^41. See e.g. Hardwick and Adams, “Firm Size and Growth.”; Choi, “U.S. Property and Liability Industry.”

^42.Jovanovic, “Selection and Evolution of Industry.”

^43. Choi, “U.S. Property and Liability Industry.”

^44. (Adams et al., “Mutuality as a Control for Information Asymmetry.”

^45. Ibid.

^46. Andersson et al., “Life Insurance and Income Growth.”

^47. Kader et al., “Determinants of Reinsurance.”

^48.Doherty and Smetters, “Moral Hazard in Reinsurance Markets”; Choi, “U.S. Property and Liability Industry.”

^49. Kader et al., “Determinants of Reinsurance.”

^50.Adams et al., “Competing Models of Organizational Form.”

^51.Kader et al., “Determinants of Reinsurance.”

^52.Zanjani, “Regulation, Capital, and the Evolution of Organizational Form.”

^53. Lindmark, Andersson, and Adams, “Evolution and Development.”

^54. See e.g. Hardwick and Adams, “Firm Size and Growth.”

^55. The 1948 Insurance Act introduced a suite of new and radical changes that affected the structure of the Swedish insurance market right up to the mid-1980s. These measures included, amongst other things, tighter solvency requirements and investment controls, premium tariffs, and tougher licensing rules (e.g., see Lindmark et al., “Evolution and Development.”

^56.Sveriges Officiella Statistik, Enskilda Försäkringsanstalter.

^57.Waldenström, “Svenska Aktiekurser, Aktieavkastningar”; Edvinsson, Jacobson, and Waldenström, “Historical Monetary and Financial Statistics”; Krantz and Schön, “Historical National Accounts.”

^58. Hardwick and Adams, “Firm Size and Growth.”

^59.Heckman, “Sample Selection Bias.”

^60.Dunne and Hughes, “Age, Size and Survival”; Hart and Oulton, “Size and Growth of Firms.”

^61. Goddard, Wilson, and Blandon, “Panel Tests of Gibrat's Law.”

^62.Goddard, Wilson, and Blandon, “Panel Tests of Gibrat's Law”; Oliveira and Fortunato, “Testing Gibrat's Law.”

^63.Levin, Lin, and Chu, “Unit Root Tests in Panel Data”; Im, Pesaran, and Shin, “Testing for Unit Roots in Heterogeneous Panels.”

^64.Mazzucato, “Risk, Variety and Volatility.”

^65. The variance-inflation factor for the ith independent variable is calculated as: VIF_I = 1/(1 −  ), where R_i is the multiple correlation coefficient when the ith independent variable is predicted from the other independent variables in the model. As a rule-of-thumb, variables can be regarded as highly collinear if a VIF exceeds 10 (see Gujarati, “Basic Econometrics,” 339). The results were: ln(SIZE)_t − 1: 2.70; IC: 1.72; PR: 2.86; AGE: 4.35; ORG: 2.13; REINS: 1.30; INT: 1.02 and ln(GDP): 3.45.

^66.Westall, “Invisible, Visible and ‘Direct’ Hands.”

^67. Ibid.

^68. Pearson, “Insuring the Industrial Revolution.”

^69. Lindmark, Andersson, and Adams, “Evolution and Development.”

^70. Kader et al., “Determinants of Reinsurance.”

^71. Waldenström, “Taxing Emerging Stock Markets.”

^72. See e.g. Zanjani, “Regulation, Capital, and the Evolution of Organizational Form.”