[SystemSafety] How far does insurance economic incentive push safety?

[Eric Marsden]

Concerning your point 1, it's worth noting that some insurers are now able directly to measure risky driving behaviour (for example via the huge amounts of information that Telsa collects from each car, or by placing a GPS device or ODB snooping device in the vehicle) and offer differentiated premiums depending on your driving style. With this very comprehensive (and instrusive) level of monitoring, insurance companies are in a better position to control the driver-related component (and also, in a modern car with sensors everywhere, many vehicle-level "roadworthiness" aspects) of road safety than are the traffic police and mandatory vehicle inspection systems, which links back to Phil Koopman's question.

Your point 2 is a well-known effect called "moral hazard". It's the reason that insurers include a deductible on any claim, to ensure that (concerning purely financial consequences) you retain some incentive to take care.

Eric


________________________________
De : Steve Tockey <steve.tockey at construx.com>
Envoyé : lundi 12 avril 2021 18:23
À : Eric Marsden <eric.marsden at foncsi.org>; Phil Koopman <koopman.cmu at gmail.com>; The System Safety List <systemsafety at techfak.uni-bielefeld.de>
Objet : Re: [SystemSafety] How far does insurance economic incentive push safety?


Phil,
Two comments that might not be going in the direction you’re intending, but . . .

1) Could you use the vast amounts of data that exist in the auto insurance business as a proxy? Do people drive more safely with an intent to keep their auto insurance premiums lower?  If this kind of effect is observable in the auto insurance industry then it could be reasonable to suggest that it carries over. On the other hand, if the effect is not visible where there is the most data, then it could be an indicator that the effect may not apply in safety critical situations either.

2) Your comparison seems to only care about higher premiums vs. lower premiums. The data that I saw—again out of the auto insurance industry—shows a different thing that might be considered more important. Namely, the availability of insurance at all. In places where auto insurance was not (at the time) legally required, the studies showed that those who did have insurance tended to be worse drivers than those who did not. Simply, having insurance reduces the insured driver’s financial exposure so they were less concerned about driving safely. They tended to drive more aggressively and got into more accidents because of it. When the driver was uninsured, they unavoidably suffered the full financial exposure of of all damage. So to avoid having to pay much greater amounts, they were generally better drivers. Could the same thing happen in safety-critical? If there were no insurance, the organization might be a lot more careful because they would be liable for 100% of the damages. With insurance, the liability is vastly reduced, so less financial incentive to be as careful?


Cheers,

— steve


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/pipermail/systemsafety/attachments/20210412/d47f386a/attachment-0001.html>