The never-ending friction between common sense and political meddling has metastasized into needless automobile insurance regulation.  The New York State Department of Financial Services has started probing the legitimacy of considering occupation as a factor in setting automobile insurance rates.  In order to get the right prices, i.e. prices that reflect anticipated claims cost, insurance companies charge more for customers who are in categories that demonstrate higher claims costs.  That reduces rates for customers who are in lower claims categories.  If an insurance company uses lower rates to attract low-cost customers, how does that insult the public interest?

Today, ”insurance companies use age, gender, driving history, vehicle type and estimated miles to be driven, and often occupation and education” to establish which risk / cost category is most appropriate for the potential insurance customer.  We might assume that novice drivers are more of an insurance risk than those with 20 years of experience, and it makes sense that a wannabe professional stock car racer is a higher driving risk than a hearse driver.  The New York DMV says “Certain careers are statistically associated with higher accident risk, while others are associated with lower risk.”  The DMV lists some high-risk occupations (e.g., physician or lawyer) and some considered low risk (e.g., K-12 teacher).  However, the right price factors for different levels of risk requires empirical data from the marketplace, not merely a good guess or a political bias.

We need actual claims experience that unambiguously distinguishes the costs expected from different attribute sets. For example, costs for the attribute set of “middle aged, male, muscle car-owning, rock band musician” could be different from costs for the attribute set of “middle aged, male, muscle car-owning, partner in a CPA firm.”  The two attribute sets might reveal costs that are statistically the same or they may be different, but empirical data is the only fair way to establish disparate risk and cost factors, and the only way to be fair in pricing for the two types of customer.

The National Association of Insurance Commissioners drafted a survey of insurers asking them about the many factors that could be relevant to their rate setting process.  Intuition is a useful preliminary guide on what to research, but actual experience is the definitive guide for insurance companies and the public.  Actual experience should also be heavily persuasive for insurance commissioners who are responsible for assuring that customers are treated fairly, are offered justified prices, and are served by companies that are actuarially sound.  So, if we want to establish the most appropriate pricing factors for rock-band musicians and for accountants, we should look to the costs they cause in property damage.  There is no justification for fixating on how cool they are or how ethnically diverse they are.

The database of costs for the many attribute sets that describe a marketplace is valuable proprietary information for an insurance company.  The database of an insurer’s cost information should not be revealed to competitors nor should it be controverted through anything less than empirical data of similar pedigree.  It is unclear what the New York insurance commissioners hope to accomplish with their probe into occupation as a legitimate pricing factor.

The database of costs for the many customer attributes (including occupation), along with the average administrative cost per customer, give insurers a foundation for sound financial thinking, while serving its marketplace and offering each potential customer a fair price, one that neither overcharges nor undercharges.