View this article online: https://www.insurancejournal.com/magazines/mag-features/2014/01/13/316338.htm
In 1988 California voters enacted Proposition 103, in one stroke changing the state’s law regulating insurance rates from one of the least restrictive to one of the most restrictive in the nation. In the 25 years since the passage of Prop. 103 debate has continued over whether this change was beneficial or harmful.
The most recent venture into this debate is a document produced in November 2013 by the Consumer Federation of America, which concluded that in California there has been over “$100 billion in savings for motorists as a result of lower auto insurance rates driven by … Proposition 103.”
However, the analytical methodology of the CFA report is so faulty that its conclusions are nothing more than unsupported allegations and political rhetoric. Although the report purports to provide statistical analysis of the issue, in fact it does nothing of the sort.
The essential argument in the CFA study is that California auto insurance rates have declined since the enactment of Prop. 103, therefore Prop. 103 caused the rates to decline. This is a clear example of the invalid logical argument form of post hoc ergo propter hoc – the second event followed the first event, therefore the first even caused the second event. In terms of statistical analysis, this conclusion violates the fundamental statistical rule that correlation does not imply causation.
Many factors influence auto insurance rates. If we accept the basic statistics in the CFA report showing that California auto insurance rates have declined since 1988, there is still no logical or statistical basis for concluding that the reduction was caused by Prop. 103. Only a rigorous analysis of the impact of these other factors would allow valid conclusions to be reached as to what caused the reduction. The CFA study provides no such analysis.
The most glaring omission in the CFA study is any substantial analysis of the impact of other changes in California law that may have affected the cost of auto insurance.
In 1979 the California Supreme Court decided Royal Globe v. Superior Court, 23 Cal.3d 880. Royal Globe allowed a person with a liability insurance claim – a 3rd party claimant – to bring a tort action against the liability insurer for failure to handle the claim properly. Under this rule a claim on an auto policy that had a policy limit of $15,000 could result in an insurance loss that was potentially unlimited. It is not surprising that under the Royal Globe rule auto insurance liability costs rose dramatically.
In 1988, less than three months before Prop. 103 was enacted, the California Supreme Court overturned Royal Globe in Moradi-Shalal v. Fireman’s Fund, 46 Cal. 3d 287. Suddenly insurers no longer faced losses on Royal Globe claims, and no longer needed the premiums that they had been collecting to pay for these losses. It is unsurprising that liability rates subsequently declined following this change in the law – a decline that had nothing to do with Prop. 103.
The CFA study deals with this only by saying that it cannot have been a significant factor in auto rate decreases because auto comprehensive coverage, which is unrelated to liability exposure, also declined following 1988. In logical terms this is called a non sequitur – an irrelevant fact. There is no necessary connection between the cost factors in liability insurance and those in comprehensive insurance. If fact, the statistics in the CFA study itself shows that the decline in auto liability rates was much greater than that in comprehensive insurance, suggesting that different factors were involved.
So how much of the decline in auto rates was the result of the overturning of Royal Globe? We don’t know. In 2001 the Rand Institute for Civil Justice issued a study far more statistically rigorous than the CFA study, which concluded that auto liability rates in 1997 were 35 percent lower than they would have been had Royal Globe remained in effect.
The CFA study ignores this. It also ignores the fact that in 1989 California significantly strengthened its laws against driving while intoxicated. It also ignores the fact that in 1996 California significantly restricted damages that could be collected from intoxicated drivers or from drivers committing felonies.
How did these and other changes in law affect insurance costs? The CFA study is silent on this specific question, but clearly these changes would only impact insurance by causing rates to decline.
So did Prop. 103 cause auto rates to decrease? Maybe yes. Maybe no. We will have to await a logically and statistically valid study to determine this.
But it is clear that the November 2013 report from CFA is not such a report. In terms of showing the impact of Prop. 103 on auto insurance rates, CFA’s study is statistically and logically worthless.