Insurance by/at the pound?
It has never been more of a dog’s world. Consider, 45 percent of households own dogs.
With such a high rate of cohabitation, insurers have seen dog-related claims rise. Today, according to the Centers for Disease Control and Prevention (CDC), there are approximately 4.5 million Americans bitten by dogs each year. Of that 4.5 million, 885,000 people are wounded seriously enough to require medical attention.
It is undeniable that dogs injure an enormous number of Americans. People with dogs in their homes are associated with a higher likelihood of being bitten by a dog and people with two or more dogs in their homes are five times more likely to be bitten than those without any dogs.
The Insurance Information Institute reports that one dollar of every three paid out in homeowners’ claims was a result of a dog-related claim and that the average cost paid out for such a claim was $27,862.
Whether or not the associative risks or costs of dog ownership are recognized, the popularity of dogs persists.
Insurers, as their business demands, assess the cost of the risk that dogs represent, and have compiled information related to dog-related claims. From this information, some insurers have chosen to adjust their underwriting practices to account for the risk profiles of different breeds. This activity sometimes leads to a policy showdown.
Two states, Pennsylvania and Michigan, have chosen to prevent insurers from distinguishing between the risks that different breeds represent. To forestall what proponents of such bans describe as “breed discrimination”, both states have prohibited underwriting practices that are sensitive to breed. There are two rationales behind such bans, one is emotive and the other is policy-based.
First, such bans are, to a dog-loving populous, intuitively attractive. Let’s face it, though legally property, dogs are so much more to those who love them. For those who love dogs, the term “breed discrimination” holds a great deal of rhetorical power. Semantically, “breed discrimination” sounds odious. And politically, fighting against discrimination is almost always good … right?
Well, not really. Risk classification of all types is, in a literal sense, the “quality or power of finely distinguishing.” AKA, legal discrimination. In spite of the legality and inevitability of risk classification, the issue remains sensitive.
Second, even without access to the proprietary data that insurers may have, underwriting according to breed – in the strict sense – is problematic. Though the CDC attempts to measure the health risks posed by one breed versus another, they confess the shortcomings of such an approach. Data about attacks is self-reported and prone to inaccuracy. Further, the majority of dogs in U.S. households are not pure in breed. The existence of mutts and customized cross-breeds (for instance, any dog known as a, “fill in the blank”-doodle) complicates easy classification.
From the perspective of insurers, who are interested in pricing their products competitively so that clients posing a low risk pay a lower premium, forbidding the use of dog breed data is problematic. In every state, insurers are required to underwrite on an “actuarially justified” basis. This means that only legitimate cost factors may be taken into account. Dog breed data, though imprecise, meets this threshold because it is a proxy for indicia of associated risks that are statistically correlative.
More specifically, it is known that male dogs bite more frequently than female dogs; that non-neutered dogs are more likely to bite than neutered dogs; and, that chained dogs are more likely to bite than unchained dogs. Further, it is known that larger dogs are capable of causing greater injury than are smaller dogs. Thus, to the extent that some breeds are more likely to possess any number of the enumerated characteristics, some insurers have come to believe that there is a meaningful correlation between that breed and heightened claim risk.
To be certain, insurers are not underwriting based on breed guided by a normative judgment.
Still, many insurance customers will find such reasoning unsatisfactory because, without access to the data that undergirds it, the correlation will, in their view, not rise to a sufficient level of statistical relevance. Fortunately, there is recourse available to those customers and it lies in the realm of the free market.
Some companies, like State Farm, forego breed-sensitive underwriting in favor of adjusting premiums after a dog has demonstrated itself to be a risk. It is not inconceivable that State Farm, by accepting a certain number of losses that it may have otherwise avoided by employing breed sensitive underwriting, has gained a reputational advantage that could well offset the costs and increase market share and profits. State legislatures, instead of succumbing to the temptation to ban breed-sensitive underwriting, should recognize that the market has a solution to unpopular underwriting practices.
With 885,000 people a year wounded seriously enough by dogs to require medical attention insurers are justified to classify dog owners of all types, as a higher risk than those who do not own dogs.
If legislatures must pass laws to save dogs and their owners from the effects of breed discrimination, perhaps they would allow insurers a rating exception as they pass their laws. The exception would be a straightforward approach to assessing the risk that a dog embodies by simply weighing the dog. We know that the larger the dog the more damage it can do. Can scales predict what breed is not allowed to?
