It's pretty small bore, but the idea of implementing "pay as you drive" car insurance, as proposed by Jason Bordoff in Democracy and various other worthies elsewhere seems like a no-brainer of a good idea. The basic concept would be that people who drive more would, ceteris paribus, pay more for their car insurance. After all, your odds of getting into an accident have something to do with how much driving you do. This would be a bad deal for half the people, but a good deal for the other half, and should help reduce congestion, air pollution, carbon emissions, etc.
« Modest Proposals | Main | Optimism »
Pay As You Drive
29 Mar 2008 03:53 pm
Comments (59)
Any reform on car insurance seems like a hit to me. Nothing makes me want to run screaming towards anarchy like car insurance. (Ever have an accident where you end up hoping the bill goes higher than $500 so that you actually get some benefit from your insurance policy?)
Some will say that this is already out there, as many (most? all?) insurers vary their rates depending on annual mileage estimates. However, as someone living in a very small town where driving is often unnecessary and going two months between visits to the gas station is not uncommon, let me note that I have a hell of a time convincing my insurer to believe a miles per year estimate under about 8000. Actual miles last year - based on oil-change stickers :p - 1750.
So, personally, I'd really go for the real thing.
Such a proposal is complicated by the fact that 1000 miles on the interstate is less dangerous then 1000 miles driven in congestion. Maybe it should be "pay for accelerating".
Isn't this already done to some extent? I know my insurance uses the distance of a daily commute as one factor when setting rates. I got a nice rate cut last year when I moved closer to work.
If it weren't for the attention to personal privacy historically adhered to in American policy, it would make a huge amount of sense to require a LoJack in every car on the road today. The possible applications are useful and limitless: consumption tax, insurance, tracking vehicles of criminals or felons on probation, triggered responses if vehicles are seen swerving or exceeding the speed limit or driving off of marked roads. If the system were universal and secure (ie require a warrant to actively track the movement of criminal suspects, etc) it would be an amazing way to save money. It would also effectively subsidize the expansion of cellular coverage to every community accessible by road. Pay for that, and you could ax traffic police, toll collection/enforcement, various parole enforcement costs etc, etc, etc. You could even make money if the government charges a fee for access to non-identifying driving statistics for non-academic purposes. The cost of a GPS receiver + cellular transponder would be minimal if it were universally adopted.
The only problem is that it'd require a government that can be trusted to act responsibly, maintain privacy, and operate the system with rare operational faults. Oh well.
In response to Matt's comment that this move by the insurance industry "should help reduce congestion, air pollution, carbon emissions, etc," it would do so in a way that's divorced from measurements of the car's energy and emissions efficiency. A combination of a tax on fuel consumption and a vehicle registration fee adjustment that accounts for vehicles that have lower or higher emissions per gallon of gas consumed would be a much better way to address that problem... and probably just as or more popular than attacking the God-given American right to drive as much as we damn well please for a flat fee with cheap, unlimited gasoline.
Ian Parry at Resources for the Future, the most prominent environmental economics think tank, has written about this issue for some time. See http://www.rff.org/Documents/RFF-DP-05-15.pdf/. Note that, as Zach notes above, pay-as-you-drive insurance does not address negative externalities from the burning of gasoline (i.e., carbon emissions), even though it does address other externalities directly.
Great idea, except for that whole Orwelian GPS tracking thing necessary to cary out the idea. Maybe charge you more for driving in bad parts of town, for accelerating too fast, for stopping at too many fast foot restaurants.
As j.e.b. and k point out, insurance companies already do this to a certain degree by charging you based on your commute distance and by charging you extra for traffic violations or accidents--all of which are influenced by not just how you drive but by how much you drive.
er, or what Zach said.
The Minnesota DOT conducted a study on the market for PAYD. Here is the abstract of a paper presented at the 2007 TRB.
The Minnesota Department of Transportation (Mn/DOT) carried out a pay-as-you-drive (PAYD) demonstration intended to test the idea that if people are given frequent price signals regarding the costs of their automobile use, they may reduce their mileage, and alleviate congestion and air quality problems. The project simulated the conversion of fixed costs of vehicles, such as leasing or insurance to variable costs while providing frequent price signals. An important part of the study, and the focus of this paper, was market research into customer acceptance of such potential products. A survey of a sample of Twin Cities Area residents probed interest in PAYD concepts for leasing and insurance. While 25% of respondents indicated some interest in the concept of PAYD, interest dropped to 18% when given explicit choices. Respondents were most attracted to the concept of paying only for the miles they drive. Reasons cited for the lack of interest were the uncertainty of monthly costs, privacy concerns, and a dislike of leases in general. These findings indicate that PAYD products have appeal to a small, but not insignificant portion, of the population, and that PAYD programs will be most effective when they are designed to focus on specific subsamples of the overall population. The overall impact on vehicle miles traveled and traffic during congested conditions were estimated in the remainder of this study, and are documented elsewhere.. You can find the full paper at Transportation Research Board Annual Meeting 2007 Paper #07-1687.
Another paper looks at the benefits of mileage based vehicle insurance, taxes and fees and here are its conclusions:
The analysis in this paper has demonstrated that strategies that convert fixed vehicle user costs to variable user costs can be an efficient
solution to address congestion, make better use of expensive highway resources, manage the operation of the highway system, and postpone the need for highway expansion. The analysis suggests that a nationwide variabilization policy adding 10 cents per mile to variable user monetary costs (without increasing total user monetary costs) could produce a 20-year stream of congestion reduction
benefits conservatively estimated at over $44 billion.. The source for this is the Transportation Research Record 1812, pp. 171-178.
This is a stupid idea. You want lower insurance? Don't get any points on your license, and get a liability-only policy with high deductibles from GEICO. You'll even get a discount if you own any Berkshire Hathaway.
On the insurance kick, what would be outstanding for nerds like me is if insurance companies were more open about the statistics that drive rate choices. I know I can't get a good rate w/out disclosing my credit rating, but I get the same rate regardless of what I pull out of my ass for an estimated annual mileage (grad student... literally no clue whether I'll drive 1500 or 15000 miles in any given year). I changed addresses in Baltimore to what I'm sure is a safer zip code but a less safe actual address and my rate went down -- does that mean car companies stupidly adjust by zip codes instead of more finely sifting the available data? And, if I can get a lower rate by disclosing my SS#, why can't I share my race/health conditions/right-handedness/musical-preferences or something else that might strongly correlate with being a good driver to get an even better rate?
It's hard to think of a more regressive tax than this. People in rural areas are much poorer than people in urban areas. People in poorer areas produce much less pollution than people in urban areas.
Yet you want to tax people in rural areas much more stringently than people in urban areas? No. This is politically suicidal.
when Matt says this is a bad idea for about half the people, he neglects to mention that it's the poorer half, and the half he does not belong to.
Hard not to take that into account.
@ Zach,
We know a couple of reasons why insurance companies won't reveal their calculation methods. One is that the accuracy of their methods in predicting, on average, how much a customer will make in claims is the key to success in the insurance business. Telling other companies exactly how you do that puts you at a competitive disadvantage, because they can copy your methods without the development costs and they can undercut your pricing very precisely knowing what your basis is.
The other reason is that the public would be appalled at the methods insurers use to calculate risk and it is much easier to get away with those methods in secret.
The idea that half of people will benefit is a false one. Without proper statistics we don't know how many people would benefit, but more importantly Matthew is presuming that the insurance companies will pass along a savings--which is not necessarily the case.
Aren't soullite's claims that (a) rural people pollute less, and (b) rural people would pay more under a scheme that requires higher insurance premiums based on miles driven at odds? The extra driving causes extra pollution, right? Is that pollution offset by per capita consumption or something?
Also, how is this a "tax"?
1000 miles on the interstate is less dangerous then 1000 miles driven in congestion.
Is that by number of accidents per mile driven or amount of damage done per miles driven, which is what insurance cares about? I was in a fender-bender on the interstate that did a lot more damage than I hope I'd ever do driving in the city. (No people or animals were hurt, only cars.)
On the other hand this was on I-95, proving that "on the interstate" and "in congestion" are not mutually exclusive.
And I think Zach is right; if you want to reduce the other externalities from driving lots of miles, tax gas. With a corresponding progressive reform of the tax system so even poor people who drive a lot come out winners.
Really, what is the liberal wonk's obsession with breaking everything down to money? The winner in each and every one of these equations is the wealthy and the losers are the poor and those with marginal incomes.
Can you find another discriminating factor other than ready access to lots of cash?
Zach, you want a nice break on your insurance, get a Master's degree and tell the insurance company you don't work for a living.
No shit.
"Really, what is the liberal wonk's obsession with breaking everything down to money? The winner in each and every one of these equations is the wealthy and the losers are the poor and those with marginal incomes."
Well, no, actually. Progressive, as opposed to regressive, taxes can mean "win win." People who are able to afford it pay more and the poor pay less. Where as, say, a tax on dried beans will affect the rich virtually not at all but the poor very much.
The only logical metric for an insurance company to use is Risk. For instance, let's say that their statistics determine that you are going to cost them $500 per year on average in claims. If they charge less than $500 for insurance, they lose money. If they charge dramatically more than $500, they'll be undercut by another insurance company that's willing to accept a smaller profit margin.
In other words, they have a huge incentive to keep the price of their service just slightly above the amounts of their payouts. Trying to get them to set the price to anything else - be that miles or whatever - is a losing battle.
This idea sounds positively German (if not germane) in its potential for complexity. And sure enough, the Germans already monitor freight trucks along every inch of autobahn to make sure that everyone pays for road usage as accurately as possible.
So aside from earlier regulations to keep truckers in the right lane, to limit their speed to 20kmph below other traffic, and to ensure that all truckers pull off for regularly scheduled nap time, they found this fantastically ornate way to hand a fat taxpayer subsidy to Telekom and Siemens to install the road use monitoring system nationwide -- but by golly it's fair! And I'm sure many government bureaucrats got bigger offices with snazzier high-tech gear, an assigned parking place closer to the door, a pay-grade enhancement, etc.
One question that those involved in government in Germany never, ever ask (well, perhaps with the exception of the Free Democratic Party) is what the cost-benefit estimates look like. No, in Germany it's the bureaucrat's Field of Dreams: Build it, and they will pay...
"In other words, they have a huge incentive to keep the price of their service just slightly above the amounts of their payouts. Trying to get them to set the price to anything else - be that miles or whatever - is a losing battle."
Except we also say for the good of society that you may not use certain factors that may correlate with risk, even though that may harm your business model. For instance, you are not allowed to use race as a risk factor for car insurance, even if your actuaries determine a strong correlative risk. Likewise, we can tell insurance companies that they must use overall mileage as a factor in insurance rates, and they will have to adjust rates so that the cost of insurance will cover the cost of the pool. Insurers won't loose money, but if they loose some of the granularity of risk adjustment people will be thrown into a more diverse risk pool. Thus some people will benefit and some will loose relative to their previous risk pool, ie some people's rates will go up and some will go down.
Insurance is about discrimination based on produces against you, not based on you as an individual, thus it is entirely appropriate that government regulate the factors by which insurance companies discriminate.
People who are able to afford it pay more and the poor pay less.
But the POINT of making whatever more expensive is to make it prohibitive to use/enjoy. So this is automatically going to favor those with lots of money and discriminate against those with marginal incomes.
If you respond by making it progressively easier for the poor to use/enjoy then you aren't accomplishing your goal of restricting access.
errata: "produces against you" should be "prejudices against you"
@Face The Music
I was speaking generally about regressive taxes, not about this instance specifically.
Wow! Guess Mr. Bordoff doesn't know anybody who lives in a rural area... or the mountain west. My carpenter neighbor here in the Sierra drives a hundred miles daily in his pickup for a job that pays 25 bucks an hour. That's the way employment works in the country. Plus, he's a lot less of a threat on the road than the city driver going two miles to the office talking on the cell phone.
So yes, Mr. Bordoff's plan is a way of taxing the working class while feeling righteous.
It is a great idea. It should be combined with extra charges for driving into cities, like London has. That also would reduce the differential between remote rural residents and urban residents.
Hate to throw a wrench into your enthusiasm but, when I had a car, the only times it was hit were while parked on the street. Once, as the innocent victim of two colliding vehicles; the second, as a truck maneuvered in or out of an alley.
The only logical metric for an insurance company to use is Risk.
Correct. Risk that I will bill my insurance company for a moving vehicle accident when I am not driving: zero. Risk that I will be in a moving vehicle in my car when I am driving: small, but finite. For those that drive only a few times per month, the idea of buying short term (as in days-at-a-time) car insurance is a good one.
With this and the congestion pricing issue, I think a lot of people have to learn to accept something: the rich can afford a certain level of convenience that the poor can't. Any time you charge for something, it will be easier for the rich to pay for than the poor. This self-evident fact is not a reason we should oppose every sort of pay-for-use idea. The number of rich is small, and I'm certainly willing to accept a few "sore winners" paying-their-way into a life of easy-access in order to subsidize the rest of us and create a more efficient distribution of services for most everyone else.
"Correct. Risk that I will bill my insurance company for a moving vehicle accident when I am not driving: zero"
As Judy points out, that is false. Plenty of accidents happen to parked cars. However, you seem to also be implying that if such an accident happens you won't make a claim. While you may elect to do that, that is not the case for all people, so it is still false to imply that cars have zero risk for claims if they are not currently being driven.
A good sign of whether you are reading a liberal blog is whether you can scan 31 comments in a thread on car insurance reform before seeing any mention of the effect of lawsuits on insurance rates.
A good sign of whether you are reading a liberal blog is whether you can scan 31 comments in a thread on car insurance reform before seeing any mention of the effect of lawsuits on insurance rates.
Go ahead, enlghten us. What is the effect of lawsuits on car insurance rates and how is that relevant to the topic at hand?
As an actuary (someone who prices insurance for a living), I think I might be able to contribute something to this. The idea has been batted around in insurance circles for a while. The company that has done the most on this front is Progressive. I believe they even have a pilot program set up. They also hold a business methods patent on at least some form of pay as you drive insurance (someone please step in if they know the details) which other companies claim inhibits their ability to innovate in this area. I also think consumer advocates tend to oppose measures that raise premiums for any subset of drivers, even when they lead to lower premiums overall.
I think it would be best to refocus the discussion a bit. Lower pollution and congestion would be nice things, but they can't be the objective of pay as you drive insurance. Insurance pricing (by law and by actuarial principles) is meant to reflect the expected costs of providing insurance coverage without being unfairly discriminatory. So the questions we should be asking hear are i) does pay-as-you-drive better reflect insurance costs? and ii) is it unfairly discriminatory?*
i) A big benefit of pay as you drive insurance is that it makes premiums better reflect exposure to loss. Car accidents kill over 40,000 people a year, injure and maim tens of thousands more and cause billions in property damage. Car insurance spreads those costs across society and attempts to charge more to those groups who contribute disproportionately to that death and destruction (teenage boys, say). The principal social benefit of car insurance (aside from the personal benefits that individual drivers get from its protection) is that it encourages those most likely to cause accidents to stay off the road (perhaps not getting a car until after high school). Each extra mile you drive make you more likely to contribute to the death and destruction that cars cause. If you had to pay more to drive more, you'd probably contribute less.
ii) The unfairly discriminatory question is a lot thornier. Discrimination is allowed in some ways in car insurance (eg age, sex) but not others (eg race, sexual orientation--though I won't claim to know whether any races or SOs would pay more as a result of that). People here seem to be making the argument that either lower income people drive more or that their lower means make it more difficult for them to pay higher premiums. I think it would help to focus the discussion by weighing the death and destruction that more accurate pricing prevents against the social good created by affordable car insurance for lower income people.
*for those actuaries reading, I'll admit that I'm simplifying a bit. There are other concerns in using deciding rating variables (eg privacy, efficiency), but not room to discuss them here.
Insurance prices already depend on how much you drive. Duh. It would be stupid of companies not to charge higher premiums to people who drive more - more mileage equals more risk of an accident.
If you want a government policy to discourage people from driving, how about a "gas tax"...
Plenty of accidents happen to parked cars.
Having made a claim to an insurance company when a tree fell on my parked car, I am aware of the risk that I could have to make a claim on a non-driving day, but there are plenty of situations where I'd be willing to assume the risk of damage to my car on a non-driving day or be willing to have some kind of limited coverage to cover these contingencies... plus liability is a larger part of your insurance bill than theft/collision (especially on an older car), in any case, so a "pay-as-you-go" liability insurance might not be a bad idea.
In the first comment on this thread, blah suggests that anyone who endorses pay-as-you-drive should also endorse pay-as-you-drink, pay-as-you-smoke, and pay-as-you-gorge on trans fats (through small taxes). I'll let the cigarette industry and trans fat fanatics speak for themselves, but on the question of drinking blah has gone too far. As a dedicated pedestrian and advocate of public transportation (I live in Chicago, I hate suburbs, and I originally come from Iowa, so there's no converting me), I think pay-as-you-drive merits serious consideration. But as a Twenty-first Amendment extremist, I object to the levying of additional taxes on my beer. To adapt a phrase from a different Amendment's extremist: "* * * from my cold, dead hands."
Very few people here seem to have bothered to read the link. Matt is wrong to say this would be a good deal for half the people; Bordoff specifically says the distribution of miles driven suggests it would be a good deal for about two thirds. Furthermore, the idea that this is on the whole a regressive tax is squarely at odds with Bordoff's piece:
there is a very strong correlation between income and driving, most low-income people would see their rates come down
Now maybe he's full of shit and the editors of Democracy didn't bother to check, but I see no particular reason to assume that. If you're sure he's wrong about this, give me a reason to believe it beyond evidence-free assertions.
Drew, thanks for the injection of expertise.
Andrew Tobias has in the past backed no-fault auto insurance financed by a tax on gasoline. As I recall the lawyers (led by Ralph Nader) killed it. Something like 2/3 of auto insurance premiums goes to pay lawyers.
Drew's comment above is very very good. I'll just toss down a few related thoughts here:
The point of "pay-as-you-drive" is not to charge people in rural areas more than people in urban areas, but to charge people more precisely according to their expected contribution to the insurance company's claims. This can be defended pretty strongly on both equity grounds (drivers more likely to get into accidents should pay more for their insurance) and business sense grounds; again, see Drew's comment.
Insurance companies already use rural vs. urban as a rating variable. They try to use estimated miles driven as a rating variable as well, but obviously putting a device in your car that tracks how much/when/where/how you drive would lead to greater accuracy than, say, using distance between home and work. As noted above, the big problem is not the idea in theory, but the creepy Big Brother vibes.
As Drew notes, another benefit of this sort of plan is that it increases the incentives to be a better driver (and in a fair way, not in a "death penalty for speeding" way).
Scote's concern about whether or not insurance companies will pass on any cost savings to consumers seems misguided. I believe the car insurance market is already rather competitive, as it is rather easy to get competing quotes. More accurate rating systems increase competition -- if you are getting overcharged, your insurer's competitors will be better able to steal you away with lower rates while still making their required profit. If you are getting undercharged, well better pricing means you will have to pay more, but that only seems fair.
Finally, our resident conservative Fred is supposedly here to explain the facts of business and life to liberals with a blinkered view of the world. Yet he doesn't grasp any of the economics basics at all, dismissing more accurate pricing as a "stupid idea" while proposing a magic plan to lower your insurance rates (buy less insurance) and complaining that no one else is complaining about lawsuits.
As already pointed out, it isn't nearly as simple as a linear relationship between miles driven and prob of an accident. People who are much better at the statistics than I have spent a lot of time analyzing it... since the insurance industry potentially can make a lot of money by assessing their risks more accurately. Additionally, the government put quite a bit of research into it too... trying to figure out how the policies like 55mph speed limit work out in reality.
I was exposed to a nice stack of the white-papers and reports when I took Econometrics (which I failed BTW). It was one of the profs favorite real-world example problems.
All that said, the fixed costs of ownership of a car are too high IMO. A more ideal situation would be where people could own their own cars to use when it makes sense, and use other ways of getting around when those make more sense. As it stands, the marginal costs are pretty low so it really doesn't make much sense not to use your car. (I calculated that it was cheaper for me to drive than to take the bus, even at a student rate!)
""Insurance pricing (by law and by actuarial principles) is meant to reflect the expected costs of providing insurance coverage without being unfairly discriminatory."
Thanks for the voice of reason, Drew Drytellar. I am all for encouraging people to drive less, smarter, use public transport, etc. Auto insurance is complicated enough, in addition to utilizing residence, driving habits, age, gender, employment status - many states now allow companies to utilize drivers FICO scores in assessing rates.
Something like 2/3 of auto insurance premiums goes to pay lawyers.
Dude, get a bullshit detector. WTF?
Maybe this is true for something like medical malpractice insurance. Maybe.
I too am an actuary who does insurance pricing (personal auto, homeowners, etc.) and I'd like to echo what Drew said above. This is not a new idea at all, and I think there are a number of reasons why it hasn't been introduced here yet.
First, almost every company does include some sort of commute distance/annual miles driven factor, but the data is of limited use because it's notoriously inaccurate. So the biggest one is probably just the cost involved in collecting accurate data, but I think it's getting to the point where it's pretty cheap (and getting cheaper) to deploy gps technology and collect the data, so companies like progressive are starting to introduce this. I think you'll see more and more companies offer this as an option going forward. It would obviously be attractive to certain people (those who don't drive their car very much!), and you could capture this lucrative market, especially if you were the first in an area to introduce it, or you provided an easier/less intrusive feeling service. Everyone else is going to have to do it, or get selected against.
I am aware of another company besides progressive, milemeter, which is planning on introducing this in texas. Their model is that you simply pay up front for a certain amount of miles, and if you are over that, your insurance isn't valid. The mileage is printed on your insurance card. It remains to be seen which payment model will work better- milemeter's system might work better for non-standard risks, or something like that. It's a lot like cell phone plans. You might have a plan where you pay for a certain amount, then get surcharged next time if you go over, etc.
Another reason there was some resistance is that I think there was a feeling among some in the industry that people wouldn't like having their privacy invaded, but I think it's obvious by now that at least some people would be willing to put up with it for cheaper insurance. I think attitudes toward the intrusiveness of this have probably changed, now that we've all become accustomed to all kinds of other surveillance that technology has made possible.
Also, it's not clear that regulations in many states will let you do this. Lots of states are pretty restrictive about how you can charge people for insurance. This has also changed recently as I believe several states (texas, oregon, I know there are others) have changed their regulations to allow this, and if it proves popular, probably other states will too.
Also, to reply to Zach at 4:54: Yes, we often stupidly adjust by ZIP code instead of more finely sifting the data. There's a balance between sifting the data finer and finer, and being able to have any sort of confidence in your rating, and even if you didn't care about that, lots of states require that you have a certain amount of statistical validity to your rating. Some states won't even let you get as fine as ZIP code with your territory rating. Also, technology for mapping and aggregating much more data about the geographic location of a risk has improved a lot in the last 10 years and people are starting to look at how to use all this to get much more accurate, detailed adjustment factors for your garaging address. Of course, you wouldn't be able to use this in a state that mandated county as the geographic rating variable.
The reason we don't use the other stuff is that either the state would find it unfairly discriminatory, and/or it would be intrusive to collect from every policyholder, and/or there is no reason to think it would correlate well to risk so why bother with the expense of collecting/storing/analyzing it. Or maybe it does correlate well to risk, we are wrong, and someone will figure it out and have a temporary opportunity to capture the highly profitable left-handed musician market.
Finally, to reply to scote above @5:02: Both of your reasons are wrong, as insurers' rating plans, including all the factor tables and how they are applied are publicly available. Almost every company does, in fact, dissect their competitors rating plan to see what they are doing and how they can undercut them. Some companies actually do just straight up copy the plans of their competitors (you can't do this in every state though). There are companies (Insight is one) that exist that will actually do this for you, that is they build a tool that lets you enter a risk, and see what factors your competitors are using and what premiums they are charging. There are other companies, that for a fee will go the the DOI in the state you are interested in and copy the filing of your competitor for you. Also, I doubt anyone would be surprised at how the rating plan is done- it's pretty much what you would expect. I know GEICO uses education level to rate on, which might piss some people off, and most companies use credit models which is somewhat controversial, but I really can't think of anything that would shock anyone in most plans.
By the same token, shouldn't there be a health insurance surcharge every time you drink alchohol, smoke a cigarette, or eat at a fast food restaurant?
Posted by blah
"Sin" consumption items have little effect on health risk, unless abused. Studies show moderate consumption of alchohol has health benefits compared to teetollers or heavy drinkers. Taxes that discourage moderate use discourage the health benefit. Smoking, too, curiously enough, costs society less than long-lived elderly - the State and insurers, pension benefit programs pay out less lifetime on average for nonsmokers than smokers. And studies have shown the little ghetto porkers and their Type II diabetes is primarily a function of all the free, easy to make "nutritious" carb-heavy food their mommas make up by the boatload. And a food industry and FDA that has touted things like "lots and lots of pasta" and high fructose syrup "nutritious" fruit juice are good - McDonalds is Eviiiiil! - mentality.
Of course for Lefties, the so-called "fast food obesity epidemic" followed on the heels of their claim of hungry, starving children so weakened by lack of food that the 230-lb 6th graders were getting bad grades...
The best solution might be to cut back on food stamp generosity so mammies won't overfeed, and make food stamps good at fast food restaurants so the little porkers might blow the stamps on a 3.40 cent happy meal at 1,450 calories and avoid the 3.00 "nutritious" meal of 3700 calories consisting of two fried turkey thighs, 3 pounds of mac&cheese washed down with a 1/2 quart of "healthy" fruit juice and a quart of ice cream.
And leave us normal people free to grab a whopper or a Popeye's value meal without Nanny taxes for something fast food is not really responsible for - ghetto ignorance is.
***************
Miles only add to existing risk factors based on driver record, type of driving, high crime neighborhood or neighborhood with stats showing a pile of bad reckless, drug abusing drivers.
In some cases, like rural law-abiding areas, or highway travel for most miles, high milage is a negligable risk contributor.
******************
tax gas. With a corresponding progressive reform of the tax system so even poor people who drive a lot come out winners.
Posted by Matt Weiner
Lots of luck trying to set up gas stations charging differently taxed gas prices for different socioeconomic levels. Or "reforming" the tax system so that middle class that drive a lot are punished to discourage excess milage while "poor people who drive a lot come out as winner".
*****************
Same with slapping Joe Public with a huge tax on his 7 mpg RV that he only drives 1200 miles a year while milking taxpayers so Senior Partner Sue Shapiro can drive her Prius on a 200 mile a day commute in her 30 mpg ave car and uses twice as much gas for her needs as the RV owner.
And of course, watch - as such "soak the low mpg vehicle owners" laws would hammer middleclass America and SUV resale purchasers but oddly have all these earmarks inserted exempting 7 mpg Lear Jet owners and the Global Warming celebrities urgently jetting about to "warn about lesser people needing to change their Petro-consumption."
Barbar asked "Dude, get a bullshit detector. WTF?".
Oops. Checking gives a revised estimate of 11%. This is for liability and comprehensive combined, liability alone will be higher but nowhere near 2/3. Fraud appears to be a bigger cost especially in NY.
I can't believe that Chris 'The Klansman' Ford thinks that our food stamp system is generous.
Yeah I'll just bet it's easy to feed a family of 4 on fifty dollars a week.
What a scumbag. Of course this is the guy who still thinks we should have dropped more napalm on Vietnamese peasants to make the world safe for American capitalism, or something like that.
"Finally, to reply to scote above @5:02: Both of your reasons are wrong, as insurers' rating plans, including all the factor tables and how they are applied are publicly available. Almost every company does, in fact, dissect their competitors rating plan to see what they are doing and how they can undercut them. Some"
Interesting. Thanks for the correction.
Isn't this already done to some extent? I know my insurance uses the distance of a daily commute as one factor when setting rates. I got a nice rate cut last year when I moved closer to work.
Every year GEICO asks me whether I use my car to get to work every day, and I don't think they're just making conversation. I knew that Yglesias opines from a pretty narrow background, but this is ridiculous.
McArdle, Sullivan, increasingly boy wonder Yglesias: Do the editors of the Atlantic really think it's wise to aspire to being the next Slate?
"Isn't this already done to some extent? I know my insurance uses the distance of a daily commute as one factor when setting rates."
Uh huh. And there's usually good rea$on to peg your older, safe and less valuable car as your primary one.
We keep hearing that Democrats want to be the party of the West. It's going to be a long time before public transit takes me anywhere near the middle fork of the Yuba River. In fact it's going to be awhile before I can get to the really interesting places (all but one crossing) without a 4 wheel drive vehicle that has significant clearance and at least 185 horsepower.
I've got quite a different proposal: require insurance companies not to discriminate based on any factor. Everyone pays the same premium for the same coverage, whether they're a 16-year-old male or a 35-year-old female.
Why? Because insurance is about pooling risk, not charging people for their claims.
If my insurance company could look into a crystal ball and know that I'd end up making $1200 worth of claims in the next year, what would they do? They'd charge me $1200 for the next year's coverage, plus a little more to cover their profit. I'd be better off putting that money in the bank, paying the claims out of my own pocket, and earning interest instead of paying for overhead. Their having so much knowledge about my behavior would defeat the purpose of buying insurance in the first place, right?
Well, the more risk factors they add into their models, the closer they get to having that crystal ball. The more they know about me, the more they can adjust their rates to reflect the amount they'll end up paying out, and the less utility I get out of being insured.
Perhaps someone can explain why this is a bad idea, but so far it seems to me that we should have more pooling of risk, not less.
Your point rests on the premise that driving fewer miles is directly correlated with less accident prone opportunities. Granted that might be true at face value but you don't account for skill, response time and sheer ability which actually determines the likelihood of accidents to a greater degree.
Who is really more likely to get into an accident - the seasoned driver that is constantly on the road, the 'professional driver', the 'occasional driver' or the low mile driver?
Assessing risk is usually a very multifaceted enterprise, and the mathematical models that map out the relative risks complex.
Your endorsement of pay as you drive, your assertions about the program and what it would mean seem to suggest you have not bordered to educate yourself on the matter.
In my humble opinion, pay as you go would transfer the premium from person to miles driven. And since there are fewer accidents per miles driven than per subscriber, even if the premium per mile is set low - I see only a net monetary gain for Insurance companies.
Pay as you go insurance also evenly distributes risk to every driver. As such, a driver with ten previous accidents, a sixteen year old first time driver and yours truly would pay per mile they drive - as if that were the sole source of risk. OK, let me give you the benefit of the doubt and say the premium per mile was calculated relative to the risk presented by the driver - an amalgam of both systems. In which case the true advantage of the pay as you go system is defeated, which is ease of risk calculation - the simpler mathematical model.
Lastly, your theory seems to miss the point that we are currently paying a premium that takes into account the miles you drive (or an estimate thereof). To the extent that your consistent driving means experience that diminishes your risk, it won't adversely affect your premium until the miles driven exceeds a certain threshold where mathematically your risk becomes significant. Not to mention that the accident risk involved in driving a mile in a city is not comparable to a mile driven on an open deserted road or to a mile of ice sheet covered roads - a fact reflected in current pricing models.
Your blatant oversimplification of a complex matter is why policy gets bungled in America. Most things are not simple enough to fit into a blog post. Uninformed opining and the influence of such simplistic lingua on the masses is the root of misinformation.
I don't give a fuck about cars' carbon footprint (whatever we here in the US would do, would make a shadow's worth of difference worldwide, get the cows to stop farting, I say) but I would like to stay alive. Presently, some 1% will leave this world thanks to that lovely hunk of metal we see advertised so beautifully wherever we might turn our eyes. That says nothing about the great great great many of us who will be crippled for life, blinded and otherwise horrifically mangled thanks to our auto-intoxication (as Bergen Evans so beautifully put it).
That being the case, methinks that car insurance in general is a murderous joke and that we ought to do away with it entirely. Thinking about how we can fix our cars after a potential accident (that you likely will slowly bleed to death in anyway!) or how we can get healthcare should we manage to survive (that happens to be free in practically every other civilized country) is entirely the wrong way to think about accidents. Having insurance makes nobody drive safer (quite the opposite in fact - when I briefly was without insurance there was no safer driver on the road than I) and it does nothing to protect the lives of innocent others.
Car Insurance is fucked and I would do away with it entirely. Publicly humiliate, excoriate, flog-naked and imprison people at fault for accidents (and lower the speed-limit, enforce sane safe-driving laws {not those that punish the poor in their jalopies while the hard-carred wealthies zip around flippantly}, require safer cars {thanks Ralph!}, etc.) and you're more likely to live happily and healthily into your old age. Require car-insurance for all and you've done nothing but bring a bunch of suited Wall-Street dicks to orgasm.
mnuez
www.mnuez.blogspot.com
Re: Every year GEICO asks me whether I use my car to get to work every day, and I don't think they're just making conversation.
Every year? I only asked this (by Allstate) if I move and give them my new address.
Jesse:
Well, the more risk factors they add into their models, the closer they get to having that crystal ball. The more they know about me, the more they can adjust their rates to reflect the amount they'll end up paying out, and the less utility I get out of being insured.
Perhaps someone can explain why this is a bad idea, but so far it seems to me that we should have more pooling of risk, not less.
There are better drivers and worse drivers. If you are a good driver, "better pooling of risk" reduces the utility you get of insurance, because in the long run you pay more than you would without insurance.
You are also neglecting the utility you get from having losses spread out predictably over time (ie, paying premiums) rather than having potentially large hits to your finances occur at random times.
For health insurance, I think there is a better argument to be made that there should be cross-subsidization; not so much for car insurance though.
Re Scote @ 10:42: No problem. To be fair, it's public information, but it's not exactly easy to get ahold of. You usually have to go to the DOI to get it in person. Maybe someday it will all be online or something though.
Re Jesse @ 1:41: I think that the general idea of risk rating is that it gives an incentive for safer driving and avoids moral hazard. A big big factor in your insurance price is going to be behavior based- number of speeding tickets, DUIs, at-fault accidents, etc. These are factors you can reduce by consciously driving safer, and be rewarded for it. You would think everyone would do this anyway, given that their personal safety is at stake, but they don't. A financial incentive has an effect. Part of the thinking behind per-mile insurance is that it would be an even more powerful incentive to reduce uneccessary driving, especially if you were in a high-risk class, since you will be paying more per mile, and are aware of that, much like you are aware of your mileage and economize driving when gas prices spike. Whereas with the current situation, you pay once and then, well, you've already paid and there's no reason not to drive as much as you want.
Plus, a lot of people feel they shouldn't subsidize the bad habits of others, especially when those bad habits put others at risk. Of course there are factors- age, gender, etc.- that are out of your control, and I think it's worth having the societal conversation about whether or not it is fair to use that sort of information. People and their legislators seem ok with age and gender for the most part. Race, religion and other factors you could think of are considered (rightly) unfair to use, and so they aren't. Socioeconomic status is generally considered unfair, but it creeps into the rating through geographic location, education level, credit and other factors. Some states are better than others about tempering or forbidding the use of those factors, though they are predictive, because they are seen as unfairly discriminatory. If you think your state should forbid the use of that kind of information, political action can change that, as state legislatures have control over how insurance companies are regulated. Trying to convince companies not to do something probably won't work, because if something works, and one company does it, everyone else has to follow suit or be adversely selected against.
There's also an argument that per-mile rating would lessen the impact of factors like gender (which is why NOW supports it) and factors like credit that correlate somewhat with socioeconomic situation. I'd have to really see the data, but it doesn't sound unreasonable to me.
Also, no matter how much information the insurance company has about you, they will never be able to exactly predict what is going to happen to you in the next year. There is a significant random component that will never go away. Let me put it this way. The $1200 per year you are paying isn't exactly what the company thinks they are going to be paying out in claims to you, it's the expected cost, which is the sum of the cost of all the potential things that could happen to you multiplied by the probability they will happen to you, as near as can be determined (to put it another way, risk rating is about more accurately determining the distribution- or at least some relevant statistics of that distribution- of potential costs for a given class, not predicting exactly what is going to happen. At least in theory). If you put the $1200 into savings, you would be completely SOL if you got in an accident that caused $50,000 of damage to you or to someone else (the "someone else" being why the state requires you to carry insurance). There might be a very small chance of that happening, but there is a chance nonetheless. You buy insurance, maybe even pay 35% above your expected cost due to 30% expenses and 5% profit for the insurance company, to protect yourself against that. Even a $3000 repair bill (not unusual at all), at the wrong time, could be a disaster for someone, whereas a monthly or whatever bill, though annoying, can be factored into your budget. I would say that risk pooling is how the insurance company is able to do this and not go out of business, but it's not the purpose of the insurance.
Re Alice AN @1:46: I would be shocked beyond belief if, for a given driver, miles driven did not correlate extremely strongly with accidents. Also, I don't think anyone is talking about just charging everyone the same flat rate per mile. I think the idea is that you would use miles driven as the exposure unit- there would be a base rate per mile driven, which would then be multiplied by all the usual factors such as age, location, rate of previous incidents per mile driven, etc. to arrive at the final rate charged per mile.
As it stands now, the standard exposure unit is car-years, which simply doesn't correspond all that accurately with true exposure, but has the advantage of being easy to obtain accurately, work with and understand. I don't doubt switching to per mile exposures would be complicated. Everything is based on the way exposures are currently calculated, you would basically have to revamp all your processes. It would be a huge headache, but potentially worth it. Not to mention, if Progressive is doing this and eating your lunch, you probably need to follow suit or go out of business anyway.
It's true that miles driven is currently a factor in rating, but it's very crudely done, and generally VERY inaccurate. As a consequence, this information is not at all used to the degree that it could be.
The point that not all mileage is created equal- that say city driving is different than highway driving and carries a different risk- is a good one. To some degree, this can be accounted for by using ZIP code or some other location identifier (which is pretty much how it is accounted for in current models), but that is not a super accurate way of measuring that, as obviously even people in the same neighborhood could have very different driving habits. Something like a GPS system that could keep track of highway miles vs. city miles would be useful here. After a company had collected enough data, they might even be able to classify individual roads according to risk, and then apply a factor to account for the percentage of driving you do on high-risk vs. low risk roads, or something like that. You could even imagine varying the risk factor of the road by time of day, and accounting for that as well. That sort of analysis would really get one much closer to the actual risk being taken, and premiums could more accurately reflect that, which would incentivize safer habits even more.
The main objection I would have to a system like that is that it seems kind of creepy to know how closely your insurance company is monitoring you, and people might be worried about that information falling into the wrong hands (although a lot of people would probably consider their insurance company already as "the wrong hands":) ) I suppose the information could be collected in such a way that might alleviate this- only keeping track of the "high-risk" vs. "low-risk" factor, and not actually storing the actual location- but people will probably still be wary. However, some people might be comfortable enough with it, and attracted to the savings, that they decide it makes sense for them.
I don't think it should be law that all car insurance has to be this way, but I think it should be allowed as an option. If it's successful, it might eventually replace the old system on its own. This is because all the low-risk drivers who weren't accurately identified as such under the old rating system would get a quote per-mile, do that math and realize they would save by switching. Those people were secretly subsidizing the high-risk drivers before, but now they aren't, so the rates on the high-risk drivers would increase, setting off an adverse selection spiral, and eventually almost all drivers would switch to the new system. This only works if the new system truly does more accurately charge for risk however. Also, people might switch because they figure they can drive less and save, if money is tight. Per-mile gives you somewhat more direct control over your premium.
The first companies that do this, if successful, will definitely see great profits. However, if it is successful, other companies will follow suit, and competition will thin out the profit margins eventually. It's a very, very competitive business.
I don't know for sure that it would be more accurate- I strongly suspect it would be, but I can't be completely sure. Companies are going to start offering this over the next 5-10 years, so we'll see if it works, I guess.
All that said, the fixed costs of ownership of a car are too high IMO. A more ideal situation would be where people could own their own cars to use when it makes sense, and use other ways of getting around when those make more sense.
That's already been invented, though you rent rather than own. It's called car sharing. Go to zipcar.com.
Comments closed April 12, 2008.
| 
Heads in the Sand
By the same token, shouldn't there be a health insurance surcharge every time you drink alchohol, smoke a cigarette, or eat at a fast food restaurant?
Posted by blah | March 29, 2008 4:19 PM