You're a good driver but no one knows why
The topic of car insurance comes up in a conversation with friends. You learn that you all have the same insurance company and coverage, but they pay hundreds of dollars more each year. You suspect that your friends are serial car-crashers - but you'd probably be wrong.
Forces unknown to you -- and even to insurance regulators - are likely picking the lucky ones in the car insurance lottery. Your premiums might have as much to do with a utility bill as safe driving.
How does it work? Auto insurers predict the likelihood of how much you'll cost them before they determine your annual premium. To do this, they need to know as much as possible about you: specifically, whether you're likely to experience an accident, theft or fire. And they use numerous predictive indicators to figure it out.
You might think much of this data has little to do with your driving skills, but insurers say they have found correlations between data and the likelihood you'll file a claim.
FICO is old news
Many old-school rate-setting techniques are still used. Reaching age 65 may cost you more. Belonging to an affinity grouplike AARP could get you a betterpremium. Women are often quoted lower rates than men; married people usually get a better price than singles.
Another traditional data source for insurers is FICO, an acronym for publicly held Fair Isaac Corp., which provides models used in credit scoring, including insurance scores used by insurers.
"All our models know is: If you assumed credit responsibilities, how are you managing those responsibilities," says Lamont Boyd, the insurance market director of analytics for FICO. "Each insurance company decides if and how they want to use these scores as part of their overall pricing."
The director of insurance for the Consumer Federation of America (CFA), Robert Hunter, says otherwise, charging that FICO's models inherently discriminate against the poor and minorities. But 46 out of 50 state insurance regulators allow insurers to use FICO models, and only California and Hawaii totally ban their use. One reason: FICO is an accepted yardstick.
"State regulators fully understand FICO's role in the industry," says Boyd.
But probing your life has gotten a lot more sophisticated than simply using credit history. Loads of consumer data, not ordinarily reported to credit agencies, now get harvested by "data miners" including Scorelogix, L2C and the LexisNexis unit ChoicePoint.
What these data miners know about you -- and share with insurers -- can seem very invasive. The Federal Insurance Office (FIO), formed as part of The Dodd-Frank Wall Street Reform and Consumer Protection Act, recently suggested that regulators not only watch the insurers, but also keep a close eye on the vendors that provide this information to the insurers to set rates.
Scorelogix provides a Job Security Score (JSS) that rates how likely a person is to keep his or her job. If you believe the JSS formula, then a university professor with tenure is far more insurable, and at a lower rate, than an independent contractor. And the JSS even factors in the local economy, according to Bloomberg.
The website fatwallet.com says that L2C checks your cell phone, rent and utility bills, as well as subscription payment data to evaluate credit risk.
Some data miners use publicly available sources, such as courthouse records, to find out anything about you: birthplace, age, marital status and current residence, and may even monitor your magazine subscriptions. Since some of this data could be part of your insurance application, it allows insurers to check for accuracy.
Catastrophe modeling provides insurers with an additional layer of data, showing how often an area is likely to experience extreme weather.
Some insurers may even factor in how likely you are to switch insurers if they raise your rates, says the CFA's Hunter.
"We see insurers using auto theft rates, catastrophe modeling and a lot of information from third parties," says Nancy Kincaid, a spokesperson for the California Department of Insurance.
But you'll never know what cocktail of data has gone into setting your own rate. Every insurer keeps its own algorithm private, for competitive reasons.
No one says data use is consistent. For example, since gay marriages are not recognized in about 30 states, same-sex partners won't get the better "married" prices. The director of the FIO, Michael McRaith, points out that same-sex couples could pay up to 20 percent more in those states.
The Insurance Information Institute (III), which represents property insurers, notes that another factor in determining rates is what the III refers to as "judicial hellholes."
"New Jersey has the most lawyers of any state," says III spokesperson Loretta Worters. "Couple that with large medicals costs and it's not surprising that it has some of the highest insurance rates in the country," she adds. (See Insure.com's state rankings of car insurance rates.)
Challenging the status quo
Now some groups want to expose the rate-setting process.
Among them is the National Conference of Insurance Legislators or NCOIL, state representatives whose committees keep an eye on insurance regulators. NCOIL supports a "model act" that requires insurance companies to, among other things, give consumers the right to challenge their credit information based on extraordinary circumstances like divorce, serious illness, identity theft, job loss or military deployment. It would be up to each state whether to implement the model act.
The FIO is challenging the use of credit history. In December 2013 the office pointed out that a driver with a poor credit report could pay up to 40 percent more in premiums. While the FIO admits that insurance scoring is an "important tool," it says regulators should be aware of the dangers of data mining.
In an ominous warning to state regulators, who value their independence, the FIO says it will "monitor state regulatory activity and move for federal involvement" if it sees discrimination against the poor, minorities or gays.
Kincaid feels that this extra layer of regulation is unnecessary. "Simply regulating insurers gives us what we need to protect consumers," she says.
Insurers argue that looking at your credit, your utility bill, your job security or anything else is a viable predictor of your future claims. They're also making in-roads to take a closer look at your actual driving, with telematics programs that deliver a vast amount of information on how and when you use your car. Pretty soon they might actually know why you're a good driver.
More from Ed Leefeldt here