Subject: Re: Support as insurance
Date: Mon, 29 Nov 1999 09:33:43 -0500

> writes:
>  > Russell Nelson writes:
>  > > Insurance is a bet.  You're betting that the disaster *is* going to
>  > > happen, and the insurance company is betting that it isn't.  The
>  > > reason they make money is because they're in a better position to
>  > > evaluate the risks than you are.
>  >
>  > Wrong.
> No, I'm not wrong; we're both right.

Well I know that I am right... :-)

Judging from your response to Ben Laurie I think we can agree on the
following statement:

"The potential for an insurance company to make a healthy profit is
dependent upon the fact that they are in a position to diversify the risk
of your disaster over many insurers and you are not.  The fact that
insurance companies do consistently make healthy profits depends upon their
using available information better than their competitors do."

>  > The reason that they can make money is that they can better afford a
>  > disaster than you can.  They can afford to average over many people and
>  > many accidents.  You typically cannot.  The agreement is mutually
>  > beneficial.  They benefit on the odds, you benefit on the guarantee that
>  > you do not personally have the resources to handle.
> You're not getting my point.  In the case of free software support as
> insurance, the purchaser is able to affect the frequency of the
> disaster.  And because of that, unless you take steps to prevent it,
> the only people who will purchase the insurance are those who can
> predict that they will need it.  Then, it's not insurance and if you
> price it as such, you're going to sink like a submarine with a screen
> door.

I did get that point, and I gave a specific example where support on free
software support was provided as a warranty in a model that I think could
work.  Allow me to repeat the description.

A contractor came in and did work.  That contractor left with a warranty
covering their work and the tools they used.  The insurance is here
provided as part of a package deal by the contractor.  It is not "sold
separately" but it is insurance.  Since the contractor is providing the
guarantee on the contractor's own work, the contractor is in a position to
fairly judge the risks.  The specific case that I heard of involved Perl
and Y2K.  I have heard of other contractors giving somewhat similar terms
with systems based on free software.  The service of providing custom work
under warranty is something that customers like to receive as part of a
package, and there is a lot of precedent for this kind of warranty.

If this kind of model proves popular then there is a potential for
establishing a reinsurance market, probably linked with a certification
service.  In this model a small contractor would get certified by the
company as producing quality work, and would then be authorized to provide
guarantees as above and reinsure the risk with the certification provider.
Questions about providing certification for free software come up from time
to time, as well as questions of how valuable it is.  Well this is one
model for how you can do it and make the certification truly meaningful.

> That's one of the reasons why medical insurance (the real thing, not
> the ersatz stuff that employers provide) doesn't cover pre-existing
> conditions.  And if you seen an insurance plan that does, it's not
> insurance, it's socialized medicine.

And socialized medicine is bad?  That is an argument that does *NOT* belong
on this list (I think that my discussion above about warranties does) but I
grew up with socialized medical insurance and I really miss it.  Believe it
or not, the big thing that I miss is the reduction in paperwork.  (A
government bureaucracy can produce less paperwork than private companies?
Yes, by several orders of magnitude!)  Anyways, back to topic.

What you illustrate very well is the role of equal access to information.
In a free market an insurance company that is allowed to discriminate on
significant risk factors whether or not they are "fair" (eg pre-existing
conditions, gender, race), will consistently win over one that does not.
Unless government steps in and regulates what factors may be considered,
the result is a disconnect between social notions of "fairness" and
economic considerations.

The social message has nothing to do with this list, but the business
principle applies to trying to provide warranties for free software as
well.  If you want to try to make a business out of it, you must control
whatever risk factors you can.  (As the certification-reinsurance model
does above.)