I recently received a proposed total loss settlement for an auto from my insurance company based on an average of three "corrected" values based on local list prices for the same vehicle.

How the insurer arrived at the "corrections" is a separate issue. What I am disputing is the number of vehicles used in the "average".

Essentially, I am arguing that they need to account for confidence limits in their calculation of the average. If they are averaging only three values of, say, $9,000, $10,000, and $8,000, it is possible that if they pick 3 other vehicles with values of $15,000, $20,000, and $10,000, they would get a much different result.

The 95% upper confidence limit on the mean is the sample mean plus around twice the sample standard deviation of the samples presented. For the first example above, the sample average is $9,000 and the sample standard deviation is $817. Thus I would argue that a more accurate estimate of the average value is $9,000 + $1,600 ($10,600), not the straight sample average of $9,000 that they are proposing.

Is there any precedent for this kind of argument in any Texas or Federal Courts?

  • Does your policy documentation (not so clearly) state how they will arrive at the total loss value for your property?
    – Ron Beyer
    Apr 30, 2018 at 21:05
  • 1
    I'm curious on what basis you have chosen to use the 95% upper limit. On its face, it seems arbitrary and tilted in your favor. While you're at it, why not use the 99% upper limit? Do you have evidence that your car was at the 97.5th percentile in value? On the flip side, it seems like the insurer could argue, with equal plausibility, that the 95% lower limit ought to be used. Apr 30, 2018 at 23:57

1 Answer 1


Insurance broker in two Canadian provinces here (since the early 1990s).

This highly depends on jurisdiction. However, in many countries (all of Canada, and based on my reading, most if not all the US), there is a statutory or legal right of appraisal, which is a fancy way of saying that if there is no dispute that a loss occurred and that that loss fits within the insuring agreements (i.e. it is covered), and the dispute is only about dollar amount, then you can go to a dispute mechanism to settle the dollar amount.

Here in Canada, although the mechanism is not labelled as such, it functions much like binding arbitration. The insured and insurer each pick an arbitrator, the two together pick an umpire, and the agreement of any two of the three resolves the dispute. I believe the system works much the same in the U.S. I invite comments from Law.SE users in other countries to comment on how things work in their jurisdiction.

Actual cash value figures for vehicles and other property are an inexact science. It is difficult to be precise. The best friend of these calculations are comparable goods of similar condition and value, recently sold in the same jurisdiction. Whenever there is a gap between this ideal and reality, a compensation factor will have to be applied.

I don't think it is realistic to expect a 95% upper confidence limit on a set of values. With the sample you provided, the mean is probably a reasonable amount. It will be up to you to demonstrate that the upper end of that current sample is more representative of your vehicle's value, by demonstrating that the circumstances in that situation more closely mirror your situation than do the other two.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .