Comment on "Discriminatory Bias in Rates Charged by the Canadian Automobile Insurance Industry": Rejoinder.

The paper focuses on the article written by the author on C.U.A. rate-making procedure, which is biased against high risk categories because of the double-counting which results from treating dependent risk classes as if they were independent. An analyst has presented some data on industry loss ratios, which seem to deny the author's conclusion when applied to the specific case of driving record classes. The data does not refute the author's conclusion that C.U.A. rates are biased upwards for high risk classes, but it does seem to deny the C.U.A. assumption that persons should be classified as high risks simply because they have had an accident in one of the previous three years.