New Study Shows Poverty, Not Age, the Key Factor in Teen Crashes

[KrazyHorse]

I lost a lot of faith in actuaries when I realised that rates from different insurance companies vary by ~20%

This wasn't just a constant decrease either (which could be explained by one company having lower margin and lower overhead than another).

In my old zip code (21218) Progressive was ~15% higher than Geico. In my new zip code (21231) Progressive is 20% lower than Geico. Equivalent coverage in all cases.

screw you, otbot

[rah]

Actuaries assess risks. Each company weighs factors differently, and some use different factors so differences of up to 20% shouldn't be considered that abnormal.

[OzzyKP]

Originally posted by lord of the mark
youre pretty much only looking at fatals. Do you even have enough fatals in relevant subcategories, (age by econ group, for ex) to be able to make statistically significant judgements?

Insurance companies have millions of dollars riding on their decisions, and I suspect do a deeper job of analysis than your group. Dont they always charge a teen premium?

Well the sample size was 35,000, that is a fairly large group to draw from. That includes every fatal accident in the areas he was studying.

When AAA put out a study a few months ago that the media went nuts over, they just looked at fatal crashes. No one thought to question it. Furthermore their study looked at all fatal crashes teens were INVOLVED in. So even if the teen wasn't at fault, they were lumped in together to paint a terribly skewed picture of teen driving.

No one questioned that either.

Except us, but no one wants to hear from us because we dare to challenge the existing erroneous assumptions about teens.

[KrazyHorse]

Originally posted by rah
Actuaries assess risks. Each company weighs factors differently, and some use different factors so differences of up to 20% shouldn't be considered that abnormal.

I understand the actuarial process.

I don't think that having that much variability is normal.

Something is off.

[KrazyHorse]

If different insurance companies show a mean variation of 20% then the error on a single measurement of true risk should be ~14% (20%/sqrt2).

That seems to me to be very high. It is, for instance, much larger than the margin insurance companies operate on (insurance companies tend to break even in nominal terms, and use their float to turn a profit of ~4-5%)

[rah]

Having worked at a couple of different insurance companies I can attest that there is that much variability.
One company was trying to incorporate credit rating in the calculation for auto insurance. (until it was discovered and people started calling it racist)
Redlining is still quite common despite the illegality of it. It's just done more descretely based on other factors that correlate to the ones they're not supposed to use. So it's harder to prove when the regulators show up.
How companies translate demographic information will have a huge impact on rates from zip code to another. (as has been pointed out)
If they take it down to the census track level you can see huge swings inside a zip code.
And another big difference is experience. Most companies will reweight factors in the equation based on their companies experience. There experiences could be considerably different from other companies based on the lengh of time they've been in a market or their targeted customers. There is considerable demographic differences between online customers and those serviced by an agent.

When you're talking numbers this large a single percentage point in loss ratio translates into huge dollars.

And yes, most insurance companies make their money in investment income not premiums. It's always great PR for them when they can make those broad statements after diasters that they've paid out more in claims then all the premiums collected there for the last 20 years combined, but neglect to mention how much they made off the investment of those premium dollars.

[Dauphin]

KH, not to be funny, but did you switch age brackets in that time?

[KrazyHorse]

Again, I understand that the different insurance companies measure and weight factors differently.

It seems to me that consumers would very quickly exploit these differences by switching from one company to another. With standard error of close to 15%, such movement would destroy the profitability of the insurance industry as a whole.

[KrazyHorse]

Originally posted by Dauphin
KH, not to be funny, but did you switch age brackets in that time?

Not unless the age bracket was between 25 years 2 months and 25 years 8 months.

Even if I had, I contacted both companies for an updated quote, so they both had the same age to work with.

[Kontiki]

I think you'd find few people who are even aware of the differences, let alone people who would be willing to switch companies. There's generally a pretty high entrenchment rate for financial services.

[KrazyHorse]

If something saves me 40$ a month I'll do it in a heartbeat.

(yes, that does mean I was previously paying close to 200$ a month for the minimum legal coverage. **** all the bad drivers in ****ing Baltimore)

[rah]

Yes, you're correct to a certain degree.
Most consumers are not that savy. But may be becoming more so. On-line quotes are helping to change that. But for a good majority of the people they never change companies (unless they're dropped) They get their first polices from an agent or their dad's agent and stay with them for life. Especially when they have multiple policies with the same company and their home insurance is paid through an escrow company. They're getting a multi policy discount that makes it
more "confusing" to compare. But it accomplishes what the companies want.
It's silly to change companies on whole or universal type life insurance policies. A hassle if your escrow company is paying your home insurance. So if you're getting a discount on your car insurance based on your other policies, you're highly probable to never change it either.


When I was on the customer service side of it they used to produce those charts of what each customer was worth to us based on how many different policies they had over the life of the policies. If they had life, auto and home we were expected to really kiss their ass.

[Japher]

I have all my insurance through one company. I like it that way. I only have to deal with one person for all of it, and I only have to write one check.

[rah]

I do to.
I play golf with my agent on occasion.

And since I worked for the company that insures me, I'm more savy when it comes to those main rating catagories and where the cut off lines are. And my agent keeps me informed when they change so we fill out the forms together making sure that when they ask questions like, how far do drive to work/school and the cuttoff is 5 miles, I can answer 4.5. Or annual milage driven is always 7400 since 7500 translates to a higher catagory. Each company is different and I like having a company where I'm familiar with them.
The dedicated personalized service that I get from this one guy offsets the small savings I might get if I shopped around.

[Japher]

I worked 20 miles from job
somehow I onle had 80 miles per week on my truck?!