AN ANALYSIS OF PROPERTY-CASUALTY INSURERS: State-specific Control Variables 3

tax-saving
Sommer (1996) reports a positive relation between the price of insurance and the insurer’s capitalization and a negative relation between price and the insurer’s investment portfolio volatility. He infers from these relations that prices are decreasing in an insurer’s solvency risk. To control for such effects, the regression includes two measures of solvency risk. CAPITAL is an insurer’s statutory surplus divided by its total admitted assets. PORTRISK is the percentage of an insurer’s assets in more volatile investments, namely equity, unoccupied real estate and mortgage loans.10 PRICE is predicted to be increasing in CAPITAL and decreasing in PORTRISK. itat

The final controls are a vector of measures denoting the insurer’s percentage of property-casualty underwriting in accident and health, automotive, credit and surety, liability, medical malpractice, reinsurance, and worker’s compensation. All remaining lines are combined. The lines of business are included in the regression to control for possible price variations across lines. No predictions are advanced for the coefficients on the lines of business.

Data and Sample

Each insurer’s financial data, state of incorporation, and the number of insurers operating in each state are collected for 1993 from the NAIC Property/Casualty Annual Statement Database.11 Each state’s total taxes collected from insurance companies are gathered from the NAIC Insurance Department’s Resources Report for 1993. Information regarding state taxation of insurance companies is gathered from the NAIC publication, Retaliation: A Guide to Retaliatory Taxes, Fees, Deposits, and Other Requirements. Information regarding the state regulation of insurance prices is collected from a 1989 NAIC rate regulation survey. The states with no-fault insurance are identified in Insurance Information Institute’s 1993 Fact Book.