My reading is the lower initial price is a incentive to allow the company to collect data. The data can then be used as evidence of poor driving, which then allows the company to raise prices. Although I've never used such a tracker, my understanding is they use proxy metrics for safety (accelerometer thresholds vs. accident record) that can facilitate this.

If your theory is that they can raise prices arbitrarily for profit, then you have to explain why they have not already done so.

Insurance companies complete on actuarial accuracy. I'd rather bad drivers pay more, and be in incentivized to drive better, than young drivers and all drivers pay more.

> then you have to explain why they have not already done so

That may be easy: there is a lot of competition in the market. If one company raises their rates too high above others, it is very easy to switch. Lots of people switch every few renewals, chasing a cheaper price.

With the individualized data, they can selectively raise their rates and either take more in from those customers, or get them (and just them) to flee. All without having to wait until _after_ an accident.

I know someone who's an actuary, and the reality is much simpler: people who consent to tracking tend to be safer drivers because they aren't afraid of the insurance company seeing their driving habits.

It gives the insurance company a way to identify a lower risk group of drivers so they can charge that group lower prices.

> people who consent to tracking tend to be safer drivers because they aren't afraid of the insurance company seeing their driving habits

Is that true? The handful people I know who have/had these devices are definitely wouldn’t fit that category. And most removed it after a while because their premiums went back up (because of their driving). They just chose it because it offered cheaper rates and they all said the companies promised it would not raise rates above what they were already paying. So it was a “what do I have to lose” decision.

Obviously anecdotes aren’t data. I think it makes sense that the drivers that are egregiously unsafe (and know it) will avoid them. But it would take seeing actual data to make it clear to me that the majority of people who consent are safer drivers.

The problem with this tracking nonsense is that it doesn't change anything. The business model is still the same. Unless the insurance company pays out less claims there won't be any savings.

For the members, probably not. But if the insurance companies can more accurate determine who not to cover, they can see huge savings in payouts. That’s why they already have teams of actuaries determine rates based on the data they do have. They just want more specific data.

That seems to support the point you replied to. You point to a group of people who opted out of monitoring and exhibited risky driving behavior, leaving leaving lower risk people in the monitored group

They all consented, initially. The insurance companies got the data they need.

The people who stay in the plan are the low risk group, obviously there will be some percentage of churn in the first year.

IDK. I've declined the tracking with State Farm, and I drive pretty conservatively. I might speed by 5mph or so but I really try to watch traffic and minimize my need to accelerate/brake as it does result in better fuel economy (and saves wear on the brakes). The MPG on my cars absolutely drops by 2-3 MPG when my kids are driving. I just reject the tracking on principle. There's too much incentive for the companies to monitize this data once they have it, and most of us know that really anonymizing data can be quite difficult.

I'm the same.

But it's not hard to see how nearly 100% of people who drive recklessly would avoid such programs, but less than 100% of good drivers would.

The statistical skew makes sense to me. I doubt that insurance companies are assuming that every driver who doesn't take part in such program are bad drivers, but insurance is purely a game of statistics.

Having worked in the insurance industry, and spent a fair amount of time chatting to the underwriting team about what goes into their models, they do on the whole use this data in a straight forward way - if you drive well, you'll get a lower rate. If you drive badly they're going to charge you for the privilege.

Don't get me wrong, there's definitely a bit of "but we can charge more" baked into the pricing models, but that tends to go in as a completely separate "we can charge more" line item within the internal model. This if nothing else is because in the UK the Financial Conduct Authority have an extensive list of things you can not price on, regardless of whether your models show they're indicative of a greater risk. The FCA don't care if you're charging more than you have to, but they will tear you a new one if you're shown to be breaking the rules.

It's what they insinuated might be possible. You COULD save $X with this shiny new surveillance device! Im sure the executive who proposed this scheme didn't pitch it to the rest of the staff as "Ive found a way we can cut our revenues by 30%".

"I've found a way to reduce payouts with data collected by this device that our payers don't get to directly see except through our filtered interface"

"that if it ever got out we were doing that would turn out to be majorly illegal and also would probably cause us to bleed customers fairly quick!"

"I've found a way to reduce payouts and reducing adjuster staffing, by helping drivers avoid crashes"

I’ve got the phone driving app from Allstate that monitors my driving via an iPhone app. I’ve seen my insurance go down as a result.