Interim Report: No Proof Insurers Labeled Katrina Wind Claims as Flood

August 16, 2007

  • August 16, 2007 at 10:38 am
    Anon says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Are you kidding me? The sample group was 98 claims? USA Today does a better job with research on their polls that this. What’s the margin of error? Something like +/- 10,000?

    I’m not implying anything about insurance companies (I’ll leave that to Joe) but this investigation was a joke!

  • August 16, 2007 at 2:01 am
    Flood Guy says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    The sample was selected from Hancock, Harrison, and Jackson counties only, and represented a good snapshot of properties subjected to surge and wind damage.

  • August 16, 2007 at 2:30 am
    Statistics Dude says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    A ‘snapshot’ is not a statistically valid sample size. How many claims were there in total? Statistically, they needed to look at least 15% of the claims from MS to TX. If there were 750 flood claims in total, that would be OK. I think the sample size was possibly/probably too small.

  • August 16, 2007 at 2:47 am
    Not a Stats Dude says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Satatistics Dude,
    It’s been a while since I took stats in college but if a valid sample size is 15% then how can all of these wonderful political polls be valid when they contact maybe a couple of thousand “voters” and project the attitude of the country?

  • August 16, 2007 at 3:13 am
    Ratemaker says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    The political polls always publish a sampling error – a “dumbed-down” term for a confidence interval – which is a function of the observed proportions and the number of observations.

    From a couple of thousand observations (if properly and truly randomly selected), you can believe the results when the answers are about 20-80%. Take the results from 5-20% and 80-95% with a grain of salt, and answers more extreme than that need a lot more data to determine a true proportion.

    From 98 observations, if zero instances of wind passed off as flood are observed, statistically it is strong evidence that if it happened, it was not a widespread practice. However, I cannot say with 95% confidence that the proportion of times this may have happened is less than 3% – which is still too high in my opinion.

  • August 16, 2007 at 4:24 am
    New Orleans agent says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    I had multiple insureds come in claiming that neighbors and relatives were told by their Direct writer adjuster that they would pay them full policy limits on flood if they dropped their wind claim.

    My insureds were as you can imagine put off by the fact that this was happening while their company was adjusting the claim correctly.

    Now try and find one of those insureds that was paid full policy limits on a partially flooded house to come and testify to their fraud. Not going to happen.

    For investigation purposes they should look at those claims that paid policy limits and did not file or collect under the wind policy.

    Only then will you catch the companies that frauded all of the american taxpayers and hurt the independent agents reputations down here.



Add a Comment

Your email address will not be published. Required fields are marked *

*