Everyone, Out of the Pool?

Commenting on last week's LA Times story-torial (previously cited here), Jon Coppelman of Workers Comp Insider wonders whether increasingly sophisticated computer modeling and data mining will mark "The End of Insurance" as we know it.  Says he:

We'll see.  The insurance industry is undergoing a paradigm shift of enormous importance.  The new model will probably generate some hefty profits, but the party may not last very long.  There will be howls of protest from the millions who, for one reason or another, have been deemed to be 'bad risks.'  Then the rhetoric will explode: 'Bad risks of the world, Unite!  You have nothing to lose but your (coverage) chains!'

Jon's item is one of an interesting dozen recent posts highlighted as part of Cavalcade of Risk #14, hosted on the Cato-at-liberty weblog by Michael Cannon, who adds his own commentary:

Insurance is a tool for dealing with uncertainty (i.e., by subsidizing uncertain losses).  How do we know that?  Because people generally don’t buy actuarially fair insurance to pay for certainties.  When additional information moves a potential loss from the 'uncertainty' to the 'certainty' end of the spectrum, people understandably decry the loss of that subsidy.  But I find it bewildering when some call that 'the end of insurance' or a market failure.  First, unless we’re close to eliminating uncertainty, we will always have insurance.  Second, in cases where uncertainty is reduced, insurance markets are doing exactly what they should: replacing the subsidy with some very valuable information.  Finally, just because the insurance subsidy is gone, that does not prevent society from subsidizing those losses in other ways.

Coincidentally, I gave a presentation today to a group of regulators from the California Department of Insurance -- no pressure there, eh? -- talking about the Times article and other recent instances of what might be seen as result-driven analysis.  (The Louisiana flood exclusion case -- ably discussed in posts by Martin Grace at RiskProf and David Rossmiller on his Insurance Coverage Blog -- was another example I cited.)  In post-talk questions and discussion, while no one indicated knowledge of anything specific on the immediate horizon, several attendees seemed to expect that public pressure will inevitably grow on regulators at some point to "do something" if the "one on one" model of information-intensive underwriting is perceived to be supplanting the traditional "throw them all into an amorphous mashup" model.

    ~~~

UPDATE [120706]:  David Rossmiller also weighs in today on the Los Angeles Times story.  His particular contribution to the discussion turns on this well-taken and well-stated point:

[T]he purpose of insurance is not a transfer of wealth to achieve social leveling, it is a transfer of risk from a present 'you' or group of 'you's' with similarly classed risks to a future you or a future group of people like you.