-
Objectives
- elimination of unnecessary government regulatory requirements
- removal of obstacles that impede competition and innovation
- streamlining of regulatory procedures that adversely affect business operations
- harmonization of regulatory requirements internationally and all across Canada
- reduce cost of regulatory compliance
- develop regulatory environment which recognizes the fact that P&C industry must be economically viable in order to maximize consumer protection
- recognition that P&C industry is distinct from financial sector
-
Principles
- need for government regulation to reduce insolvencies
- balance between intrusive government regulation and strict reliance on market forces
- cooperation between insurers and regulators
- harmonized approach across nations and provinces
- pragmatic solutions instead of philosophical debates
- cost, both direct and indirect
- control
- inherent limits to solvency monitoring
- disclosure of information
-
Paper summary
- US data have found that regulation of insurance premium affects both the amplitude and length of insurance cycle
- Canadian analysis shows that prior approval rate regulation is reliably associated with greater volatility
-
Types of price and rate regulation
- prior approval: require regulatory approval before use
- flex rating: may be an expedited process if rates fall into range
- both are strict forms of regulations
- file-and-use: file rates, after some period rates are deemed approved
- use-and-file: file rates but may begin to use immediately
- both are competitive forms of regulation
-
Types of regulation in Canada
- ON has strict prior approval system since 80's added flex in 2000
- AB & Atlantics use variations of variations of the file-and-use
- BC, MA, SK have monopolistic government run insurers for mandatory basic auto coverage, and don't regulate competitively delivered optional auto coverage
- QC has a use-and-file system
- overall, auto insurance is the only line of insurance where rates are regulated in Canada
-
Conclusion
- prior approval regulation does not necessarily produce lower insurance rates without consideration of product features
- strict price regulation has been found to limit competition, reduce availability of coverage and increase volatility
- regulatory lags in prior approval process could produce volatility by weakening the link between expected claim and premium
- regulatory build-ups occur when insurer holds off filing smaller , more frequent rate increases in favour of larger ones that justify the cost of assembling the detailed actuarial filing requirements
-
Empirical Model
- test the effect of rate regulation on volatility in the unexplained growth rate of average premium
- we analyze growth in auto premium that is not predicted by growth in claim costs, accident frequency, and other variables expected to contribute to the cost of insurance
- step 1: regress average premium on set of explanatory variables
- step 2: use volatility of premium as proxy for volatility of insurance rates and regress on a second set of explanatory variables to test whether regulation possesses any power in explaining the unexplained volatility
- set of variables 1: average claim, UW profit, Herfindahl, CPI, accident frequency
-
Empirical Model Results
- severity and frequency are the primary determinants of average insurance premiums
- UW profit margin is significant and positive
- CPI level also contributes
- index for regulation is positive and significant, suggesting that the primary difference in premium volatility between jurisdictions and over time, after accounting for other input costs and product changes, is the regulation of prices
- lack of significance of Herfindahl index
- regulation is not significant for TPL, COLL and COMP but is significant and positive for AB
- change in competitive environment is significant for COLL
-
Diagnostic Tests
- Hausman specification test
- in general models with fixed coefficient are more appropriate
- but for collision a random effects model is more appropriate
- results suggested by the test used in the model
- Breusch-Pagan test
- confirms that a fixed effects model is appropriate
- where test suggests that individual specific effects should not be rejected, a random effects model was used
- Durbin-Watson test
- positive autocorrelation found in all coverages except collision
- corrected by including a first lag of the dependent variable; correction affects primarily the Herfindahl variable
- Variance inflation factors (VIF) approach
- measures the effect of the intercorrelation of the regression on the variances of the least-squares coefficient estimators
- since no variable has value > 4.5, existence of multicollinearity is rejected
-
ON, Illinois, South Carolina
- since 1989, ON and SC have repeatedly reformed their auto insurance rate regulation system, whereas Illinois has been free of such legislative activity
- we notice price instability in SC and ON from 1989 to 1999
- premium volatility in IL was relatively stable over the same period
-
South Carolina
- originally introduced a prior-approval system
- by 1997, suffered a significant availability crisis
- 93 to 98, claim cost increased faster than premium; residual market accounted for 40% of drivers
- 99 reform replaced prior-approval system with competitive market rating, and UW restrictions were eased; uniform classification, merit rating and territories were abolished
- results are positive so far: more insurers, more refined risk classification, alternative policy options
-
Illinois
- since 1971, only US jurisdiction not regulating auto rates
- rates are filed but not subject to regulatory approval
- once of the healthiest market in the US with few affordability or availability issues
- premiums are lower and less volatile than US average; residual market size stable at around 10%; higher number of insurer
|
|
