American insurers charge reckless rich drivers less than safe poor drivers

The Consumer Federation of America did a mystery shopper review of several auto insurers and found that drivers with at-fault accidents paid lower premiums than drivers with spotless records -- provided that the careless driver was rich and well-educated and the careful driver was a single renter without an advanced degree.

Using two hypothetical characters the group compared premiums offered to two 30-year-old women. Both had driven for 10 years, lived on the same street in a middle-income Zip code and both wanted the minimum insurance required by whichever state the group was researching.

The imaginary woman who wasn’t married, rented a home, didn’t have coverage for 45 days but has never been in an accident or ticketed with a moving violation was compared to a married executive with a master’s degree who owns her home and has always had continuous insurance coverage. But she’d been in an accident (again, hypothetically) that was her fault and caused $800 in damage within the last three years.

The results were somewhat surprising, although there were differences across the five insurers. Farmers, GEICO and Progressive always gave a higher quote to the safer driver than the woman who’d caused an accident. Across all 12 cities in the study, State Farm offered the lowest or second lowest premiums.

“State insurance regulators should require auto insurers to explain why they believe factors such as education and income are better predictors of losses than are at-fault accidents,” said J. Robert Hunter, CFA’s director of insurance and former Texas insurance

Consumer Group: The Rich May Pay Less For Car Insurance Even If They’re Not Safe Drivers [Consumerist/Mary Beth Quirk]



  1. That doesn’t surprise me any. I wonder if it’s because they can “trust” the rich driver more to pay on their own if they screw up badly?

    1. I’ve heard something similar with Credit Scores. Very low credit score gets you high insurance rates, good credit scores get you low insurance rates, perfect scores get you high rates. The thought behind it being people with good credit (think, 700-750) may pay for small claims out of pocket and never even submit to the insurance company, but people with perfect (near 800) credit will always follow rules and submit for even tiny things, making them a higher risk for overall claims.

      I have no idea to the truth to this, but I’d love to see the actuarial modeling behind a lot of it (I work in health insurance, so property fascinates me!)

      1.  Yeah, my understanding of the secret score mechanics is that the “best” credit is for “has enough money to pay us back but doesn’t do it all at once so we get to milk as much interest & such from them as we can.”

      2. People with the lowest 20% of credit scores file 250% more insurance claims than those in the top 20% of credit scores, according to data from Allstate several years ago.

        1. “File more insurance claims” isn’t the same as “being a greater risk for causing accidents”. If paying a higher rate because I’m more likely to simply file a claim (eg, in a no-fault state) is ok, then we should just call them savings plans, not insurance plans.

    1. Exactly, it’s a bad study with one variable too many. A break in coverage ALWAYS brings a huge rate hike, regardless of any other factor. 

      1. Why is that? I can understand if they suggest the receptionist was *driving* without insurance, but there’s no reason to charge a person more simply for not driving. I guess that’s the big question for me. I personally took a couple years off of driving and cancelled my (state farm) insurance, and it was no problem to restart without an inflated rate, in fact my rates have declined because State Farm has a safe driver discount, and I presume because of my age. I’m closer in class to the receptionist than the wealthier woman. Of course, if you know that somebody makes the choice to drive without insurance that they may not be incredibly responsible. Regardless, the government needs to step in. Even if these drivers are riskier it’s more beneficial to society at large to have them easily insured.

      2. ALWAYS? Read the study. It doesn’t ALWAYS. In all twelve markets, State Farm would charge the receptionist less. Geico, Farmers, and Progressive always charge her more or refuse to give her a quote. AllState charges her 38% less in Los Angeles, but more than double in Baltimore.

        1. To be fair, he said a lapse will ALWAYS result in a hike. Not that it ALWAYS means the receptionist will be charged more than the executive because of it. Now, we don’t know for sure from the results of this study whether the lapse does always result in a hike, but we also can’t conclude that it doesn’t. Unless we reapplied for the coverage with all other factors being the same. Which they didn’t, which is why the study is a bit dodgy in the first place.

    2. Why? I moved to a walkable city and ditched my car last year. Does that mean if I move a year from now and get a car again I’ll be paying dramatically more?

  2. Could be the higher income means they’re less likely to miss payments, the education may mean they’ll be less of a hassle to deal with. Both items may be negatively correlated with being a scammer. Basically, they may be attempting to estimate at their costs of doing business with the person apart from their driving history. Or, it could be ‘fuck you, you’re poor’.

    1.  I would think the opposite in terms of a rich educated person vs. a poor uneducated person being a hassle.  If you are educated you are generally more likely to question things and if you are rich you have more time and energy to fight back, especially if you can afford a lawyer.  I tend to think it would be easier for the insurance company to scam the poor uneducated customer by not paying out certain claims, (something a rich person would be less likely to let them get away with) etc, than it would be for the poor uneducated person to scam the insurance company with false claims. 

  3. Well no coverage for 45 days was probably a significant factor.  After all if you’re NOT getting into accidents, the main thing for the insurance company to worry about is not getting paid.  One $800 accident is a minor fender these days and probably less than the payments over three years.  I wouldn’t be surprised it it isn’t very predictive of future, significant payouts.

  4. I don’t have a problem with this. Insurance is a very competitive business and it’s highly unlikely that this is the result of some coordinated conspiracy. It’s more likely that rich people insure more expensive cars. Or perhaps rich people are statistically more likely to insure homes and boats and airplanes. 

    1.  If it was so competitive you’d think that the other companies would adjust their pricing to match State Farm, who the study said across states and locations charged much less to the “receptionist” than the other companies.

      1. State Farm is a not-for-profit: it is a mutual insurance company which means any profits go back to the policy holders in the form of rebates and/or lower prices in the future.

        It also really matters what underwriting criteria is used: State Farm may underprice the “gap in coverage” penalty lower than the other companies (just like they underprice their motorcycle insurance by only writing based on displacement).

        1. That is not quite the same as a “non-profit” though. It’s just owned by the policy holders instead of stock holders. It’s a Forbes ranked company that makes profits, it just returns them to policyholders. They should still have the same risks as those other insurance companies when deciding who to insure and for how much, as well as a need to compete with all those other insurers.

          1. Except that since profit == reduced premiums, they effectively act like a nonprofit in terms of cost structure.  

            Since a for profit like Geico earns half a billion (a bad year) to a billion (a good year) in PROFIT to Berkshire/Hathaway, this means that Geico must cost $40-100 more per customer per year than an equally efficient mutual insurance company.

          2. That still does absolutely nothing to explain charging the receptionist more. State Farm has the same risks, no matter how they handle their profits. You are talking about difference in costs, but not explaining the difference in risk. In your argument, State Farm should be charging the receptionist more than the wealthy person too, but they’d both overall have lower premiums. The flip flop in “risk” isn’t explained, or justified in any way at all.

      2. The study does not say that State Farm charged the receptionist less than the affluent woman. It says that they had the lowest or second lowest quotes in every market. In other words, State Farm is just cheaper.

  5. She let her coverage lapse.  That puts you in a penalty box as far as the insurers are concerned.  There is no evidence this is in any way malicious, insurance companies want continuous coverage to get the stats to all work out.  If your coverage lapses the numbers don’t work out for them.

  6. It’s been a long time since I went without insurance (only because I went without a vehicle), but I recall having to pay a significantly higher rate when I signed back up; nothing else of note had changed. That 45-day lapse is not something to ignore.

  7. It could be that there is some nefarious insurance statistical model out there that simple says something to the effect:

    “We can charge poorer less educated customers more because they are less likely to figure out how much we are ripping them off and have less choices or options. Wealthy more education customers are more likely to be mobile and discriminating so we should offer them a better deal to discourage switching insurance.”

    Or as several mentioned, it could just be about continous coverage.

  8. Actuarial statistics (if I have that expression right) deals with people in the aggregate, not individual cases. There will always be outliers who buck the trend; using those outliers as a basis for insuring the average person is a poor way of doing business.

    You can point to the person with a spotless record and say they should be considered a poor risk, but there’s a reason they say “past performance is not an indication of future performance” in the finance industry.

    I agree that past performance should be taken more into account than it is, and you’ll find that some companies either lower your rates or give you a rebate if you go a certain time without an accident.

  9. In the UK there are about 100 different things they rate on and jobs can have a major impact. If you are in a low paid job like bar work you are in double trouble (a) As a low income household house you are more likely to make a false claim in an attempt to claim injury money (they are big payouts for insurance companies) (b) Your car is going to spend a lot of time in bar car parks where there is good chance some drunk will back in to it or it will get stolen.

    Someone with a high income is far less likely to be involved in false claims and when in work their car will be parked in a more secure place.

    Are these factors in the US? A break in cover is not really an issue here in the UK until you get to about 3 years when you could lose your no claims bonus

  10. Petty bias happens a lot on the part of humans… but insurance companies operate on mathematics and statistics to make their money, otherwise they’d tank, right?

    Pretty sure the chosen variables simply correlate to a higher cost / lower profit index for the companies, and they provided rates accordingly.  Is it fair to the individual?  Hell no, nothing based on statistics is.  But insurance pools are all about statistics for now, with incremental changes towards individualizing that many of us are quite wary of: “let us install a GPS tracker and then we’ll give you a lower rate maybe!”

    1. I’m skeptical of the pure, emotionless statistics argument. Poor people were better with their mortgages than the rich were during the housing crisis, yet they were often preyed upon with more expensive rates. Beyond that, there was rampant racial discrimination with banks charging higher rates and fees to blacks, and Latinos than to whites with the exact same credit rating/risk. Why the insurance industry would be thought to be upholding a gold standard is beyond me.

      Frankly, I’m kind of blown away to see this much trust (especially on Boing Boing) in the notoriously ruthless ( and corrupt) “FIRE” industry at large.

      1. Pure emotionless statistics can still account for rampant racial discrimination, if you assume utter heartless evil on the part of mortgage lenders.Tracking and correlating race with things like ability to fight back against predatory practices, ability to get a fair rate from a competitor, etc., could well support a policy of racial discrimination on the basis of pure, “rational” profit-seeking – as long as reluctance to break the law and/or damn yourself to Hell are no obstacles.

      2. I’m kind of blown away to see this much trust

        Sniffing the taints of the rich and powerful never goes out of style.

    2. That’s what you would think, but then why did State Farm consistently quote lower for the mid income/clean driving record customer, while the others typically quoted lower for the high income / minor accident customer?

      Hypotheses I can see:
      – The different companies have significantly different statistical models of expected costs & revenues per customer.  This would suggest that the statistical models are much less accurate than one would expect – their management may think they’re being scientific, but they may be blind to significant flaws in their statistical modeling.

      – Some of the companies are not, in fact, working on pure math – biases are consciously included into their calculations, e.g. “We want to position ourselves as an insurance company for the wealthy”, “We want to help ensure that the less wealthy have insurance” (not that that’s necessarily a bad thing).

      – There is significant self-selection by customers between insurance companies, such that:
          – among State Farm customers, the first profile typically goes with a lower-cost customer
          – among Geico (etc.) customers, the first profile typically goes with a higher-cost customer

      What could possibly account for the last option (the only one where all of the insurance companies are acting rationally on the basis of pure math, and possess the good statistical models you give them credit for)?  The best I can do, and it seems like a pretty feeble hypothesis, is that the fact of State Farm being customer-owned tends to attract collectivists, and something about a collectivist mentality correlates differently with the differing aspects of the two fake customers, so a collectivist-minded single receptionist is less likely to submit claims than an individualist-minded single receptionist.

      Pretty feeble, right?

      1. I went with State Farm because they had offices in my neighborhood, and I could conveniently see an agent from my neighborhood who was happy to talk and work with me any time I had issues, or questions. I don’t see any collectivist motivations. In the case of State Farm, they have offices in regular city neighborhoods, unlike a lot of the other companies.

        In general I believe the more desperate drivers go with really fly-by-night companies. The kind that have notoriously (so bad they’re good) commercials on tv for SR-22 etc:

      2. To add to your list, a big decision factor in which model to apply is probably based on how much risk a company wants to take on in favour of a potentially higher return. Low risk models might rate the same customer much higher/lower than an equally correct high risk model.

    3. I disagree that it is just base on math. Not when they use credit reporting as part of their decision on what to charge. If you examine the practices of credit reporting agencies (the big 3) with their customer relationship on both sides of the fence (consumers who have been made to fear the power of one bad item on their credit report and the bank who they also sell their services to) you’ll see they are playing both sides against each other. How do you make the MOST money at the most acceptable risk if you are a bank? You get people with really good jobs and no real credit issues, who have never sued anyone or shown any inclination towards fraud, then you get them to have ONE dispute with a cell phone company or their cable company over a bill. At some point they will just send you to collections. Good luck with that. One collection on your report for 7 years, with no taking into account that it was a dispute, or how much it was for ($30, $40?). There’s 40 points off your credit score. But that’s great for the institutions that use credit reporting as it operates now, because you WANT someone with a 20 year history of never being late on their mortgage (or any other payment for that matter) with an artificially low credit score. Low risk with high penalty for the consumer, and if the insurance oversight people in you state government ask why they just point to your credit score and say “look, they are high risk!”. And if that’s your model as a credit agency, a bank is more likely to buy your services because your scores will get them more money from the customer. So that’s what credit reporting agencies are competing against each other to do, give the banks (and other users) the numbers they want to see. Just one of the reasons these agencies keep their methods and formulas an “industry secret”.Until credit reporting is based on an open and scientific model, overseen by government, it will always me a scam to get the bank or insurance company a higher rate.

  11. I don’t see the issue with understanding why one was charged more than the other. Insurance is a GAME and you have to know* the rules** in order to win:

    1. If you are in an accident, you have to pay more to the insurance companies at the time of the accident, and after that because they raise your rates. INSURANCE WIN. CUSTOMER LOSE.

    2. If you never get into an accident, you don’t pay the insurance company anything but your monthly payment and they can’t raise your rates because you are a ‘good driver’. INSURANCE LOSE. CUSTOMER WIN.

    * In which you have bits of paper which explain the current terms and regulations, but as you will see in the next point…

    ** They can change them at any time, for any reason, and only notify you of the changes. They don’t have to ask your consent to do so. So you are playing their game which they make the rules and they can change at any time.

    1. #2 seems like insurance win. They keep getting your money and don’t have to pay anything. Thanks for the money, safe driver!

      1. From Ambrose Bierce:
        INSURANCE, n. An ingenious modern game of chance in which the player is permitted to enjoy the comfortable conviction that he is beating the man who keeps the table.

        That was a century ago, though, and permitting that conviction seems to be less important these days.

      2. They lose because if you never do a claim you can call in and lower your rates–but I was trying to keep it simple and not introduce variables.

    2. Also re. 2, if there is a general jump in insurance claims, not even necessarily related to car insurance, then you can expect your premiums to rise as well, even if your record is clean. This is partly because of the re-insurance market (insurers insure themselves against excessive claims) and partly because of greed.

      1. The issue with this is that we (the consumers/people) do not have access to this data that they use to calculate this number–so it can either be:

        1. Based in reality such as higher claims happening in your area, the average payout of each incident would be higher and therefore they would ‘need’ to charge more for your premiums.


        2. They make up data and they twist numbers in their favor in order to bleed more money out of a stone.

        1 takes work, and 2 is virtually impossible to argue out of legally and is easy. Which one do you think they pick…

  12. Insurance Industry Guy Here.  The folks like me who set the rates are typically highly educated, nerdy number geeks.  But more importantly, we also tend to be progressive / liberal.  So hold off before you typecast us as evil industry minions.

    The inconvenient truth is that the data don’t lie.  Income level and credit score are VERY highly correlated with loss propensity.  Which means, at a basic level, that rich people tend to be better auto risks, and poor people less so.  It’s one of those situations where our progressive impulses, and the hard cold numbers, diverge.  Such is life.

    Interestingly – states like California expressly outlaw insurance companies from using variables like income and credit in setting their rates … no matter how highly predictive those variables may be.

    More importantly though, the auto insurance industry is very competitive.  Carriers are at war trying to poach each other’s market share.  So if you think that all the insurance carriers are colluding in a grand conspiracy to defraud the poor working class people, then you’re misinformed.  The carriers are simply looking at the same (huge, statistically significant) pool of data, and their actuaries are reaching the same conclusions.

    Sometimes, these conclusions are difficult to grapple with.

    1. OK, a tame insurance guy, excellent. Can you answer the question that I, and apparently other’s here, have about insurance lapses? Why does it matter if you don’t have continuous insurance coverage? 

      1. Coverage lapses are highly correlated with increased loss cost.  Conversely, continuous coverage is linked to decreased cost.

        Why?  Who knows.  My personal hypothesis is that people often drop coverage in a financial bind, and keep driving even without it. That’s a ballsy (risk-prone) thing to do.  Risk prone drivers = costly drivers.

        But enough with my unscientific guessing games.  The data is what matters here.  And the data says: people with lapses cost more overall.

        1. It sounds like there would need to be a break down of why someone had a lapse. There are those that might lapse because they are in a financial bind and they are trying to cut corners and continue to participate in “risky” behavior. There are also those that don’t have insurance because they don’t have cars, sometimes because their finances dictate it as being out of reach, but not all of the time.They might just have other priorities.

          I personally don’t have car insurance, nor have I had it for over 4 years because I don’t have a car. I don’t have a car because I found myself barely using the thing and I still had to worry about the base costs of insurance, registration, parking, etc. even if I didn’t use it. Turns out it is possible to live without a car in Los Angeles. It’s not that I was in a finincial bind, I just have no need for a car right now. I suspect as time goes on, there will be more people like me in cities.

        2. I under stand that predictive correlations aren’t necessarily causal, but still make good business sense.

          I made a comment above but maybe you’re able to answer it. I moved to a walkable city, returned my car at the end of it’s lease, then didn’t need insurance anymore. If I get a car again in a few years, will my rates be higher because of the lapse?

        1. The actuaries don’t care about causation one way or the other. They care about making sure at the end of the year they have paid out less in claims than they took in in premiums. Correlation is adequate for their needs, and causation is irrelevant. 

    2. The whole point, regardless of the anomalies of statistics is to pool risk. Everybody wins. In Europe now it is illegal for insurance companies to factor in sex as a variable in determining premiums. Here in the UK until very recently men’s (especially young men’s premiums) were considerably higher then women’s. Women’s premiums have now increased considerably and men’s premiums have been reduced minimally in response to legislation. Governments can legislate against ‘discrimination’ but that doesn’t necessarily produce a fairer outcome or a truer picture of reality.

    3. You lost my confidence when you said “credit report” because credit reporting is a for-profit business, which rules it out of any chance if it remaining un-biased in our capitalist business model. For years credit reporting was something that went on at the bank, and the consumer wasn’t really involved. It was supposed to be an actuarial process that gave a “true” representation of your risk to a creditor. But they now sell to the consumer and the bank. Their business model is to make money for their investors. One might argue they would make the most money by giving the most accurate risk assessment possible. but that’s not true. They make the most money when creditors realize that the credit reporting company can find ways to push lower risk consumers into a slightly higher risk category. THAT’S the credit reporting model that for-profit bank/insurance company really wants. And once one CRA does it the others in the industry will follow suit or lose customers. Just follow the money in the credit reporting industry and you’ll see why it works. CRA’s are not nearly as regulated as banks and insurance, so that’s where the unethical practice is most likely to take place. Plus it allows their customers (their REAL customers, the banks) to hold up their hands and say “it’s not that WE want to charge you more, but the number says you’re a higher risk”. The state needs to regulate the CRAs as tightly as the banks and insurance. Ask yourself this: Is a consumer who shops around for the best home loan rate and does lots of research a better credit risk than someone who just walks into a lender and takes what ever they give them? Particularly before the mortgage lending scam-and-crash of the century, you would get TAKEN by unethical mortgage brokers if you didn’t shop around. But if you go to 10 different lenders and get a  rate quote you will see “score lowered due to number of inquiries on your credit” right there in black and white. The CRA’s true customer does not want you shopping around. I can believe that a non-profit model of insurance would want the truest representation of risk, but the for-profit model profits more from a slightly skewed representation of risk. The actuarial eggheads don’t make the final decision on business practice, the MBA eggheads do.  

  13. loaded data, to produce loaded result to get uneducated people to rage against insurers

    the “poor” driver let their insurance lapse for 45 days

    insurance lives and dies on the law of large numbers.

    there is obviously a higher correlation to claims pay outs of people who let their insurance lapse then people having at fault $800 accidents

    1.  insurance lives and dies on the law of large numbers.

      Obviously not, if some of the others are charging much lower rates to the “poor” (why’d you put that in scare quotes?) driver.

  14. The difference between the two drivers wasn’t income, it was continuous vs non-continuous coverage!  You are mischaracterizing the difference between the drivers.  

    The insurance company prefers to insure the woman with continuous coverage b/c they will more likely be able to milk her for money for several years, while the person with gaps is only going to give them money for a short time so they want more per month to make it worth the overhead of signing them up etc.

    p.s. Cory: Don’t let your political beliefs cause you to forget your skepticism!

  15. This is a flawed study.   Continuous vrs non continuous coverage is a huge rating factor.   What type of vehicle you drive is relevant to rating even if you carry only minimum liability because of loss history (actuarial results) of the vehicle types.  Married people often get lower rates than unmarried people (again this is based on loss statistics).  Income is a inaccurate way to phrase it, it is a lot more likely that financial stability ( in other words credit history) which isn’t based on income at all was used.  Education is not a factor with most carriers and when it is, it isn’t a large one. 
    Point is that driving record is one piece of how you get rated because a *lot* of factors come into statistical relevance when determining what kind of risk pool you fall into.   

    1. You’re missing the most interesting part of the study – one company, State Farm, consistently quoted lower for the lower-income / clean record / gap in coverage customer, while the others mostly quoted higher.

      Why is that?  Is State Farm using flawed actuarial stats?  Are the other companies?

  16. Gap in coverage, likelihood of their ability to pay premiums, and one other thing not being mentioned: Married people get a big discount.  It is interpreted (rightly or wrongly, but statistically) as reliability, maturity, etc.
    Income level is rarely relevant, and I used to rate insurance for home, auto and others.

    1. Back in the late 90s, when I was a young man in Alberta, I was surprised to learn that my 3000/year car insurance would only have been 2400/year had I been married.  Not quite worth getting married for, but I was surprised at how much of a difference it would have made.

      My understanding, though, is that it is a huge factor for young males.  Less significant for females, and less significant as the person gets older

      1. I don’t know if that controls for the number of cars in the family, but my wife and I pay less as a couple than the total amount we’d pay in individual policies separately. But we also share a car, so that only seems fair to me.

  17. I have nothing to say about the study but I think the BB headline writer (Cory?) needs to break out the dictionary and look up “reckless”.

  18. This thread serves as a good reminder that going carless was one of the best choices I ever made. Any extra money I have to spend toward transit and cab fare is easily offset by gas, insurance, and trips to the mechanic.

    ALSO, it’s a big reminder that it’s expensive to simply just be poor and working class. The hidden costs that the lower classes have to bear, from groceries to inflated/dubious insurance charges are numerous, and often dismissed or disbelieved by those that have never had to struggle to get by. If you wonder why it can be so hard for people to get ahead…..well here you have a real example.

    1. The everyday cost of living for the poor is generally higher in both relative and absolute terms. A greater percentage of a smaller income is taken up by increased unit costs. This is just a marketing ploy. With car insurance who decides the criteria for defining and comparing variables? Should a reckless rich woman be compared to other reckless drivers (regardless of income) or other wealthy drivers (regardless of history) in determining risk. It is impossible not to be biased. Individual circumstances can never be made to fit universal criteria. They are trying to justify discounting, not cost.

      1. With car insurance who decides the criteria for defining and comparing variables

        I’m thinking it needs to be something that receives some new scrutiny.

        Reading up a bit, it seems that agents often don’t have to tell the driver that they qualify for the cheapest rates, if they think they can roll the dice on getting them to bite at a more expensive plan.

        It is impossible not to be biased

        It is not impossible to try to limit it.

  19. Also does a single $800 accident strongly establish that this person is an expensive customer?

    That’s a very cheap accident, like a shallow scratch or small dent.  Whats the deductible on this policy, would the insurance company even have paid out anything?

    The economic status info probably pretty strongly predicts that the more educated person is likely to eventually become a more lucrative customer. Not hard to believe it outweighs a single very minor accident.

    Besides, want to see a no-accidents driver who pays a lot for insurance?  Look at any single male under 25.  I think I was paying $800 per six months minimum liability only when I got my first car with separate insurance at 22 or so after 4 years of spotless driving (not even a speeding ticket driving to highschool  20 miles daily for two of those years).  At around 25 that just fell off a cliff (and I still wasn’t married then).


  20.  State Farm’s goals are different. State Farm seeks to be (and is) the largest provider of home and auto insurance. Its agents are rewarded on the basis of volume of policies, not the profitability of those policies. This is possible because State Farm is a mutual, owned by own policyholders, and because State Farm is interested in establishing relationships with people whom they can sell the full suite of State Farm products — home, auto, life, banking.

    Geico and Progressive ONLY sell auto insurance, and both are part of publicly traded companies (Geico is part of Warren Buffett’s Berkshire Hathaway, Progressive is traded as itself.) For them, earning underwriting profits on every risk matter a lot more than they do for State Farm.

    1.  Its agents are rewarded on the basis of volume of policies, not the profitability of those policies.

      This is unusual at other agencies? Can you cite?
      Still don’t see how this accounts for the utter flip flop in risk.

      1.  State Farm uses “captive” agents, who work directly for the company and only the company. This used to be common but it is today relatively rare. Most property and casualty insurance is sold by independent agents and brokers who sell products from multiple companies, and yes it is common for agents to receive bonuses based on how profitable the business is.

        In the case of Geico and Progressive, Geico doesn’t use agents at all. Progressive does use independent agents, but most of its business is acquired (like Geico’s) online. And that’s actually relevant to this discussion as well. The use of credit, in particular, kicked off the “automated underwriting” craze. When you can boil down whether or not to take a risk, and how much to charge them, to a relatively simple formula, that makes it much much easier to use the Internet as your sales force, automatically quoting and binding business. State Farm doesn’t have an online sales tool — all of its policies are sold by individual agents, and the risks are reviewed by human underwriters.

        1. Most property and casualty insurance is sold by independent agents and brokers who sell products from multiple companies, and yes it is common for agents to receive bonuses based on how profitable the business is.

          If the goal is simply profits, then shouldn’t all these companies be using the same formulas? Certainly that makes one of these methods wrong.

          In the case of Geico and Progressive, Geico doesn’t use agents at all.

          They actually do have agents, but most of their business is handled online and by phone, true.

          Farmers, who are also in the study have 50,000 agents. Allstate, who are second next to State Farm is agent based too.

          I’m still not seeing any kind of compelling argument for why State Farm would do the opposite.

          1. The goal of State Farm isn’t merely profits. Profits is one of the goals, for sure, but there are different motivations when you are a mutual company (where profits are simply going to be paid back to policyholders as dividends) than when you are a public company (where profits are going to be paid out to shareholders who are not policyholders.) Moreover, as I said, State Farm has a broad array of products that they look to sell to their customers, whereas Geico and Progressive do not. Farmers and Allstate do have a few products, but Allstate’s actually gotten out of the life insurance and banking businesses, and it’s been shrinking, rather than growing, its homeowners business for a long time now. Allstate is likewise a public company, and Farmers is controlled by Zurich, which is a public company in Switzerland.

          2.  Help me understand why this matters?  In the study they were only asking for basic auto policy quotes, not home insurance or anything else. Not sure how this makes any difference…..

  21. Conclusion: lie on the form when you sign up. That is assuming that not one of these insurance companies figured out that both these people were fake.
    Do enough of these and you could craft the optimal profile for the lowest rates for each area of the country, insurance company, etc. I bet there’s an (underground) market for that.

    1.  Except that, when it comes time to make a claim, lying on your application forms would be cause to invalidate your policy. So, yes, you’ll have enjoyed cheaper insurance rates on a policy that will be useless to you when it matters.

  22. This is a pretty terrible article and a very poorly conducted experiment. 
    First off, there is a major difference between the “poor” and “rich” driver depicted in the article which does not consider their driving history or wealth, such as the poor driver’s lack of insurance history vs. the rich driver’s continuous coverage, which plays a major factor in the rates offered.
    Second, in order to rate based on factors such as education and job title, an insurer must first extensively prove the correlation between these factors and risk. Thus insurers are already proving to regulators how these factors are better predictors of losses than at-fault accidents.I’m fairly certain these are blanket regulations, that are required nationally, but they do often vary by state as well.
    So the same factors and weights utilized in Texas do not apply to New York and otherwise. I should add that my knowledge-base is limited to much of the Northeast so I could be ill-advised of what is required elsewhere.
    Although the article is not necessarily wrong in calling bullshit on this tactic, it’s “science” is heavily flawed.
    While these are not factors that are outwardly divulged to consumers, in most places, an insurer is obligated or required to advise how these rates are determined behind the scenes, if asked.
    In fact, one of the biggest factors often used to determine rates, which is not mentioned in this article but does apply in most states, is the driver’s credit history. Driver’s with better credit history tend to be wealthier and better educated than those with poor credit, and thus pay lower premiums. Had the article taken this into consideration or made their argument based on this fact (although impossible without an actual driver with a SSN) they’d have a point to make. Unfortunately, they did no such research.

    1. such as the poor driver’s lack of insurance history vs. the rich driver’s continuous coverage, which plays a major factor in the rates offered.

      The piece says that the receptionist drove for 10 years, just like the rich lady, but just hadn’t had coverage for the 6 weeks prior to trying to get coverage, which could mean any number of things. A change in vehicle and shopping around for a better policy, or yes, god forbid deciding it was better to save the bill for a few weeks and look for a cheaper provider.

      1. My post is not in defense of insurance companies underwriting practice. I’m just trying to provide some context. Most companies consider any lapse in insurance coverage to be a negative factor when determining rates. 
        If it’s a situation where you own a car and are in a lapse of coverage, it is often weighed more heavily against the driver. The amount of days (usually under 30 or over 30) has a proportionate impact as well.
        I believe they do often consider situations where you’re buying a new car as well – but a driver with continuous coverage is still a preferred risk.

    2.  “I’m fairly certain these are blanket regulations, that are required nationally, but they do often vary by state as well.”

      They vary quite a bit state to state. With the relatively new exception of HHS acting in same cases like a health insurance regulator under the ACA, there is no federal insurance regulator. It is an entirely state regulated business

      1. Right  – I guess what I meant is that each state has to approve the factors which any auto insurer intends to use for underwriting.

    3. This is a pretty terrible article and a very poorly conducted experiment.

      I would consider taking you seriously if any part of your comment didn’t repeat things that other commenters have already said.

  23. After reading the comments, the moral of the story seems to be that insurance companies are evil even when they’re not trying.

    Yeah, yeah, “continuous coverage”, blah blah.  If someone can’t afford continuous coverage in the first place then they’re going to have extra trouble affording higher premiums in those periods when they can afford insurance.

    A lot of people seem to think it’s OK because no one’s trying to screw poor people.  No, the problem is that poor people are getting screwed.  It doesn’t matter if anyone intended that outcome or not.  If folks already having trouble covering rent and food also have to pay more for insurance that’s a serious problem.

    1. So what is the solution then? Have richer, statistically less risky people pay higher premiums to offset those of the risky poorer people? I mean, I might even be in favour of such a shift, but I thought Americans hated socialism.

      1. No, Americans hate the word “socialism”.  They love socialism itself but the poor dears do not actually know what the word means.

        And yes, since auto insurance is a basic necessity for most of the working poor in the USA it should be subsidized somehow.  As it stands, someone who can’t afford auto insurance has the choice between being unemployed and driving uninsured.  That’s not only bad for that person.  If the person remains unemployed society loses out on the useful work that person could do.  Otherwise, if the person gets in an accident someone else will have to pay for the damage.

        If you don’t think it’s fair to give insurance to people who can’t afford it try getting in an accident with someone who doesn’t have insurance and see how fair that is.

        1. I have low rates because I’m 55. I feel that the overall quality of my life would be improved by just socializing it. My rates would be somewhat higher, but it would get more people insured. I know a lot of people who have been hit by uninsured drivers and it makes an accident a much worse experience.

  24. The CFA’s study is fatally flawed because it allowed for more variables between the two woman than just education and income.  Notably the lower income woman was not insured for 45 days which is a high risk factor.  The study should have had that fact be the same for both.  In fact the data sets for both women should have been EXACTLY the same except for one factor like income or education, then you could make the claim the CFA is making.  And, sadly, Bob Hunter is a former actuary and KNOWS this.  So clearly this was just an attempt to get publicity using flawed data to try to attack the insurance industry.
    In fact if you look into the survey they used (and most reporters won’t spend the intellectual capital to do so), you’ll see that in 30 cases the executive got a better rate.  But not always.  In a full 20 cases the receptionist got a better deal.  So how do you say that is improperly discriminatory?
    What is good about the CFA’s press release is it calls attention to the fact that it is a very wise consumer tip to ALWAYS shop your insurance coverage at each renewal because the market is very competitive.

  25. I wonder what kind of damage you can do to (a couple of) cars these days that only results in $800 in repairs.  Not much, I bet…

        1. No, that’s the front end of one Acura and the rear of one Prius.  I’m some kind of magnet for getting rear-ended and/or hit-and-run’d.  As for the cars that hit mine, I neither know nor care what their repairs cost were.

  26. Rich drivers are more likely to pay their bills on time. Why would you alienate the guy who pays you every month, in favor of the guy who pays late, if at all? Once you drop the guy for late payments, do you think he’s coming back? The rich customer will see the low rates and continue paying them for years and years. Insert quote about how much it costs to get one new customer etc etc.

  27. Lessons learned…
    – Rich people get better things for being rich.
    – Regular people get less nice things.
    – Survey data can be skewed by factors not considered.

    A better question would be what would the rates of the 2 test cases have been in 3 years of continued coverage.  Betcha the regular persons rate wouldn’t change much.  I remember when I was a little coward, I was looking forward to getting past the magic number age when my rates would drop, because then I’d be a real grownup.  They spiked my rate before that birthday, and then cut my rate by almost as much as they spiked it by.  No accidents, no tickets, continuous coverage… but rates jumped so I would see a “savings” for becoming a mature driver.

  28. “In the study they were only asking for basic auto policy quotes, not home insurance or anything else. Not sure how this makes any difference…..”

    No room to reply above, so I’ll continue here. Unlike Geico and Progressive, State Farm has other products to cross-sell. This, in addition to the fact that they are a mutual rather than a public company, informs their interest in volume of policies as being a higher priority than attracting only the best, most profitable risks. They are interested in having more people establish relationships with the company and its captive agents, and then using that relationship to encourage policyholders to buy other products.

Comments are closed.