Anyone who’s ever searched for an auto insurance company knows by the questions they’re asked—plentiful and at times personal—that a lot of factors go into determining individual rates: our credit history, zip code, age, marital status, previous driving and claim records, and more all help insurance companies estimate the approximate risk in taking us on as customers. It’s also true that each state has different laws about what types of data insurance companies can use to determine premiums. Though there’s some information insurers cannot legally use in any state (income, religion, and race), consumer groups assert that questions about education level and occupation (which are legal) are being used to guestimate income levels.
All of this may or may not surprise you, depending on how deep your claws are in the insurance game. But The New York Times recently reported something that shocked even us: a big car insurance company registered a patent for “a computerized system for analyzing voter registration and voter history as a means of determining risk.” Because, the company argued, voters aren’t as likely to file a claim as non-voters. A frightening prospect, especially for those of us who generally prefer to fly under the radar.
But before you install your intercom and color code your wardrobe, let’s look at the facts. Just what can insurance companies use—and what can’t they?
Who regulates insurance companies?
Not the federal government. Instead, each state legislates their own oversights and laws regarding fair pricing and adequate coverage. Consumer groups also investigate and report their discoveries of how closely auto insurers hew to the law. Quoted wrote about consumer groups’ accusations of the insurance industry’s potential use of price optimization, and we’ve heard news stories of neighboring zip codes with wildly different rates for the same people. Consumers themselves of course also have a hand in regulating insurers, using their wallets. Shopping around is now quite simple, and consumers are entitled to jump ship whenever a better opportunity comes along, which theoretically keeps insurers honest (or, at least as honest as the next place).
How much do voting records predict insurance claims?
NASDAQ reports that if a person has voted even once, they are less likely to file a claim, and if they’ve voted in the last twelve months, they are even less likely to file a claim. Though voting records are publically available, insurers of course cannot see how a person votes, so the data now is about yes’s and no’s, not donkeys and elephants.
Is it already happening?
It isn’t currently legal to use voting records when determining premiums. As we’ve reported, the insurance industry party line is and has been that insurance rates are fundamentally based on risk. However, if insurers maintain data that proves voting records do in fact predict the number of claims a person is likely to make, that’s risk. Consumer groups often claim that the insurance industry finds loopholes around regulations, and using voter records could certainly be one of them. And if it isn’t legal yet, it could be soon.
But hey, until the auto insurance companies, consumer groups and state governments sort it all out, that insurers know voters file fewer claims is the most unique reason to get out there and do our civic duty we’ve heard in a while.