Are Auto Insurance Companies Redlining Poor Urban Drivers?
It has long been government policy to root out overt discrimination based on race, sex, age, religious and, increasingly, sexual orientation. But in America, it remains politically correct and (in most states) legally permissible to profile based on postal code. They are routinely used by local governments to apportion tax dollars for public education – and by banks to deny or charge extra for loans to households in lower-income or working-class neighborhoods.
Nowhere is zip code profiling more obvious, and in no area does it more obviously violate basic notions of merit, than in auto insurance rating pricing. Where one lives – rather than how one drives – is in fact the primary determinant for how much you can save (or will be legally forced to spend) on auto insurance.
Insurers' most common defense is that pricing by zip code is just a simple way to take into account secondary risk factors of neighborhood – including accident and crime location statistics. Yet, research by UCLA sociologists Michael Stoll and Paul Ong shows that, even when such factors are explicitly taken into consideration, price differences by zip code are still not fully explained.
Instead, good old-fashioned redlining – when a specific geographic area, usually an urban or minority neighborhood, is either excluded or charged higher rates for a loan, insurance or another financial service – explains more of the gap. Postal Code Profiling, it seems, may well be one of the last vestiges of acceptable discrimination in America.
Take the case of Brendan Mulholland, a 40-something geologist with a spotless driving record. He once moved two blocks in Oakland, triggering a 20% increase in his car insurance hike. "The whole premise of basing auto insurance premiums on locality as opposed to individuality is wrong", Mulholland says
His case is hardly unique.
The most egregiously unfair cases of zip code-based pricing may well be in Detroit, where residents pay an additional $1,200 for car insurance a year on average, as compared to car owners in the suburbs. As a result, half of Detroit drivers simply operate cars with no insurance – compared with roughly one in 10 Michiganders who did so. High rate premiums in Detroit are not simply the result of redlining. But redlining appears to exacerbate the city's free rider problem.
With many drivers simply priced out of the market, the result is a skyrocketing number of uninsured in urban areas, and the shrinking risk pool in one area means ever-higher premiums for others in the same area. That serves to further widen the wealth gap between insured, urban motorists and insured upper-middle-class drivers in suburban areas, where insurance is considerably cheaper. Unaffordable insurance "is the most common complaint that I hear from my constituents", said Morris Hood III, a Michigan state representative from Detroit, who served as ranking member of Michigan's House Insurance Committee in 2005.
Because insurers base rates on zip codes rather than individuals' driving records – calling it a "territorial rating system" (TRS) – the higher rates are often borne by the very drivers least capable of paying them throughout the country. For example, an urban driver in a populous northeast or mid-Atlantic states may pay as much as $400 more each year than a nearby suburban driver with the same driving record and same car. Over 50 years – the average time people spend as licensed drivers – lower-income urban motorists will pay $20,000 more for car insurance, in effect subsidizing wealthier suburban motorists.
The current rating system also has gender ramifications. Because more women than men tend to live in poverty – particularly single mothers — more women are impacted by (and stand to disproportionately benefit from the end of) postal code profiling.
The territorial rating system, for all its ubiquity, is hardly transparent. Like the formula for Coca-Cola, the insurance industry jealously guards exactly how – and why – zip codes affect insurance rates, contending that these are industrial trade secrets. But their zeal to protect the formula has raised questions about the veracity of their methods.
The system is so intractable, and the insurers so unwilling to change, that in Pennsylvania, for example, one Pittsburgh neighborhood petitioned to be given an entirely new zip code because it was easier than demanding that the rating system be changed. Meanwhile, one group of Coloradans – concerned about the possibility of price hikes in auto and health insurance – fought against having their zip code folded into a neighboring one in a less-desirable part of northwestern Denver.
The problem of zip code profiling is widespread because, even in urban areas, car ownership is ubiquitous. More than 90% of US households own at least one vehicle, which means that the typical American is far more likely to have a car than a mortgage. And by law in almost every state, every car owner must carry insurance. That gives auto insurers a unique and protected status within our economy: they offer a service that people are required to purchase, and one that is only available on a for-profit basis.
Abolishing zip code profiling in auto insurance should be low-hanging policy fruit. California, for instance, passed a referendum in 1988 to ban the practice, and the net result was a savings of more than $1bn for California drivers since the mid-2000s. Insurers, meanwhile, have spent nearly $100mn between lobbying, campaign contributions and ballot initiatives to undo the law and keep zip code profiling in place.
Americans are often told that we live in a meritocracy, and that how we behave will determine our outcomes – moreso than any social, identity or environmental factors, like one's economic class or neighborhood. But we continue to carve out an exception to this principle continues for auto insurers and wealthy, suburban motorists without even asking why.
It's either time for the insurance industry to prove to its locked-in customer base that zip codes are strongly correlated with risk, or, better yet, start evaluating their customers on their merits, not their neighborhoods.