Originally Posted by runningshoes
However, if you're going to try and define that risk, either relative to other issues or to other locations, then use a meaningful model.
In the absence of the presentation of a better model, supplying better, concrete real-world answers which can be tested such that they demonstrate, in effect, better predictive value: the model that exists is more meaningful than the model that does not.
That is, better some theory/model that can be tested than no theory/model at all. After all, how do you test the untestable? If we can't, then we don't have a scientifically-based theory/model at all. Even a "weak" theory is better than no theory at all; for the former can be tested and improved upon, the latter simply doesn't exist to be tested and worked upon.
Originally Posted by runningshoes
The issue of insurance raises good points and here's what I have found about risk premiums over the last few years, both with personal and business (key man). Travel to certain countries, including Egypt, India, Indonesia, and Israel trigger significant increases in your premiuim due to the increased risk in those countries to both tourists and business people by terror attacks as defined by the insurance industry. Europe does not trigger this problem, nor does China, Japan, and a host of other countries I typically travel to. I am sure there are other places that I don't travel to that do trigger the same rider. None of these increased costs are triggered by the increased risk you associate with vehicle accidents.
The pricing for most insurance premiums -- including the ones mentioned -- are not based merely on the likelihood of an incident occurring, and certainly not on just the incidence/risk of terrorism. The pricing structure is heavily based on the payout (expected outcome) resulting from incidence (i.e., generally more than just one type of incident). The cost to resolve an incident (i.e., for mitigating payout amounts) and/or the payout (i.e., expected outcome) itself increases more with certain categories of incidents rather than other categories of incidents for reasons often completely independent of actual historical damages and actual risk of incidence. For example, if I lose a colleague to a car accident on a trip to Idaho, where he was driving himself, the likelihood that the employer (and/or the employer's insurance company) would have to pay out large sum X for this type of incident is lower than if a like event occurred while he was in Egypt driving himself. Why is this? For the same basic reason that people fear being killed by a stranger (despite the low risk of such) -- namely, irrationality, in that a belief that the "unknown" and "foreign" is necessarily more dangerous and thus deserving of more employer responsibility (and liability).
Yes, the roads in Egypt are less safe and that comes into play, but then again the risk of dying in a road accident is much greater than that of dying in a terrorist attack.
This is to say that insurance premiums may be priced to both: 1) absorb the risk of things besides just the actual risk of one type of incident (i.e., terrorism); and 2) absorb the risk of secondary costs that may have little to do with actual damages from just the incident itself yet are expected to occur anyway due significantly to perception of being "exposed" to the "other"/"foreign" element. In other words, insurance premiums may not be a perfect proxy to measure risk of incidence, for the same kind of reasons you noted in prior posts and for reasons I noted too.