Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84 exptools; site ihlpg.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxn!ihnp4!ihlpg!tan
From: tan@ihlpg.UUCP (Bill Tanenbaum)
Newsgroups: net.motss,net.med
Subject: Re: Politics of AIDS, of Foster Care
Message-ID: <1315@ihlpg.UUCP>
Date: Sat, 28-Sep-85 13:08:07 EDT
Article-I.D.: ihlpg.1315
Posted: Sat Sep 28 13:08:07 1985
Date-Received: Sun, 29-Sep-85 08:09:42 EDT
References: <858@burl.UUCP> <1554@bbncca.ARPA> <865@burl.UUCP> <2034@amdahl.UUCP> <1290@ihlpg.UUCP> <2057@amdahl.UUCP>
Organization: AT&T Bell Laboratories
Lines: 70
Xref: watmath net.motss:2110 net.med:2475

> > > [E. Michael Smith]
> > > The problem:  Insurance is a form of socialism.  The purpose is to
> > > spread the costs generated by one individual over the whole group.
> > > ANY attempt to select out ANY higher risk subgroup is in conflict
> > > with the basic purpose of insurance.  The inevitable result is a
> > > reduction in the cost sharing and a lessening of the 'insurance'.
> > > (Yes, I know there are differential rates based on various
> > > tables, charts, etc.  The conflict still remains.)
> > ---------
> > [Bill Tanenbaum]
> > Wrong.  You are correct only if the higher risk subgroup is either
> > denied coverage completely or assigned to a separate insurance pool.
> > Differential rates (based on risk factors) within the same insurance
> > pool in no way lessens the effects of cost sharing.  An insurance
> > company with a million customers could use so many risk factors
> > that no two customers pay the same rate.  Please explain to me
> > how this adversely affects cost sharing.   ...
---------
> [E. Michael Smith]
> The purpose of
> insurance is to share cost, on that I think we have agreed.  If all
> costs were summed, then shared equally, we would have perfect sharing.
> If all costs were summed, then divided equally by the number of
> people, then some people had costs deducted while others had costs added
> based on some 'risk factors', we would have imperfect sharing.
> (So far, so good, I hope )  At this point I claim that the imperfect
> sharing is a reduction of the 'insurance' by the amount extra that
> an individual must pay due to risk factors.  Admitedly a very small amount
> compared to the potential magnetude of individual costs if one is
> totaly uninsured.
> 
> The only way that I can see where this would not be true would be
> the following.  If the definition of insurance is not taken to be the
> sharing of costs, but rather the substitution of *some known* fixed cost
> for some unknown variable cost.  Then it would not matter that the
> costs were unequally shared.  The only important point would be that
> there was a known rather than unknown cost.  The known cost could vary
> by individual yet still be 'insurance'.  However, the underlying
> mechanism by which insurance companies work depends on the sharing of
> costs over a large group, so I would hold the cost sharing definition
> to be the best one.
-------
I misundersood your use of the term "cost sharing" in your first
posting.  You use it to mean "sharing the costs equally".  To
avoid misunderstanding, I will use "cost sharing" in quotes to
mean what you take it to mean.  
It is NOT TRUE that "cost sharing", in your usage of the term, is
the purpose of insurance.  The purpose of insurance for the customer
is the substitution of a fixed cost for a variable cost.  The underlying
mechanism of insurance does NOT depend on "cost sharing".  It depends
on laws of statistics.  If there is an insurance pool with fixed assets
with one million customers, the costs can be prorated on the basis of
risk, with great assurance that the total cost in benefits to the
company will be predictable.  As a matter of fact, this predictability
deteriorates, rather than improves, if the "costs" are shared equally
among the customers.  An insurance company achieves the most
predictability in benefit payout when it assigns its rates in accordance
with the probability it will have to pay out.  Imagine if you will
a life insurance company run according to your philosophy.  Everyone
pays the same rate, regardless of risk.  All 90 year olds, terminal
cancer patients, AIDS victims, etc. sign up.  Deny them coverage, you
say.  O. K., all 80 year olds, heart patients, diabetes victims, etc.
sign up, etc.  Your insurance company would go down the tubes unless
it limited itself to covering people of the same age and general health.
Even then, because of your reduced number of customers, the
predictability of total benefit payout would be reduced compared to my
company, which insures almost everybody, but with premiums proportional
to perceived risk.
-- 
Bill Tanenbaum - AT&T Bell Labs - Naperville IL  ihnp4!ihlpg!tan