http://www.article-3.com/why-you-should ... als-913891
Why You Shouldn’t Trust Those Cute Car Insurance Commercials
By Sam Uretsky | May 13, 2014
First it was just Santa Claus and the NSA. But now the Geico gecko, Flo from Progressive, and a bunch of strangers are turning into snoops too. While The Most Interesting Man In The World may sell beer, the most interested people are the ones pushing auto insurance. The more they know about you, the more they can get away with, and that can be a great deal.
Watching their ads, you might think the auto insurance companies are in a race to the bottom, but nobody really knows. That’s why, on April 10th, the Treasury Department posted a notice in the Federal Register headed “Monitoring Availability and Affordability of Auto Insurance.” It begins: “The Dodd-Frank Wall Street Reform and Consumer Protection Act provides the Federal Insurance Office with a number of authorities including monitoring the extent to which traditionally underserved communities and consumers, minorities, and low- and moderate-income persons have access to affordable insurance products regarding all lines of insurance, except health insurance.” The notice was a request for comments from consumer groups, insurance companies, and other interested parties about what to measure and how to perform the measurements. This followed a press release from the New York Public Interest Research Group claiming that the top New York insurers would charge higher rates to a high school graduate with a blue collar job than a college graduate with a professional or managerial job.
State insurance departments normally limit the factors an insurance company may use in setting rates, including ZIP code, age, driving record, car make and model, marital status, and gender. California, Massachusetts, and Hawaii ban the use of credit scores in setting rates, and California also prohibits the factoring in of occupation or education. While the insurers argue that all the factors they consider are closely linked to driving patterns, the Consumer Federation of America has claimed that the insurers are using data mining for “price optimization” – a technique of charging the highest rates a company can get away with before a client jumps to another insurer.
The NYPIRG study was simple – they made a number of requests for insurance quotes using the same basic data regarding age, driving record, and vehicle, but using wildcard variables of education and occupation. While State Farm charged the same price for all models, the other companies charged the least to a college-educated bank executive, and the most to a retail cashier with a GED. The researchers started with a profile of a 30-year old single woman living in moderate income areas (c. $30,000 – $40,000 median income), who has been driving for 14 years, driving a 2008 Honda Civic 7,500 miles annually. They repeated the experiment in different sectors of the state, and the pattern repeated itself. In other variations, a college education and an executive job was offered a lower rate than a cashier, even when the cashier was represented as having a spotless driving record, while the executive was presented as having caused a crash, or had two speeding tickets, or drove twice as many miles each year.
The explanation for the price difference may have less to do with driving records and actuarial studies than with a method of pricing that uses sophisticated software to perform data mining, and determine not the likelihood of a crash, but how much the insurer can charge before the client will start shopping other companies. The Consumer Federation of America showed one model in which the price optimized rate was nearly 25 percent higher than the rate based on traditional actuarial principles. The optimization profile would be used to determine “price elasticity” – meaning how high the price can be raised before the client snaps.
Good news for politically active individuals: While traditional price setting might be based on driving record, make and model of car, and ZIP code, optimization factors in many other things. Credit score is one example. In the NYPIRG review, Allstate refused to provide pricing information until the shopper had authorized a credit check; Progressive’s web site states that its price quote factors in the shopper’s credit history. In reality, the insurers may go beyond that. In October 2012, patent number 8,285,618 was issued for “System, program product and method for segmenting and underwriting using voting status.“ The inventors claim that people who vote regularly are less likely to file an auto insurance claim, or as they put it “…the risk parameters of the first group indicate a relatively lower predicted insurance claim frequency or insurance claim severity than the risk parameters for the second group.” They don’t care whom you vote for as long as you get to the polls. This might actually be useful in increasing voter turn-out, particularly in off-year elections. In a 2012 comparison, NYPIRG created a model buyer, and by manipulating non-driving factors, marital status, education, and occupation, got an 86 percent reduction in the price of the insurance.
In earlier studies, California had the lowest insurance rates, probably because California limits insurers to using driving-related factors in setting rates. While insurance companies defend their methods of pricing as providing the fairest price, optimization methods are a perfect example of rent seeking, which is pretty much defined as trying to get more than a fair price, or a higher price without providing additional value. While some people may get discounts, it discriminates against those who aren’t in a position to shop around due to various factors such as geography, time constraints, or limited resources. These are the people who pay for the discounts offered to those who can afford to pay a bit more. And apparently, while a college education may not teach anything useful, it does teach you not to assume that your insurance company is charging you the best price.