COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	HOW ARE ALL THESE GOOD THINGS TO COME ABOUT?
C00018 ENDMK
C⊗;
HOW ARE ALL THESE GOOD THINGS TO COME ABOUT?


	The developments discussed in other sections of this book can
come   about  in  a  number   of  ways.    Here   are  some  relevant
considerations:

	1. A lot depends on whether the development is a system  or a
product.  We  will use the term \F1product\F0 for  something that can
be  developed and manufactured  by a company and  sold to individuals
and which has  the additional property  that \F1the first person  who
buys  one will  benefit\F0.   A  \F1system\F0,   on  the other  hand,
requires a  public  decision to  implement  it,   and  requires  many
customers before  any  get benefits.   Often  a system  is a  natural
monopoly  so that a decision  to implement it one  way excludes other
implementations and may exclude competing companies.  Sometimes there
is a choice whether  a given benefit can be obtained  by a product or
system, and,  in that case, I would prefer to see it implemented by a
product,   because competition is  then possible,   and because  each
person can decide for himself  whether the benefit is worth the cost.
Of course, a  case could arise  where the costs  are very much  lower
with a system, and  then it might be preferable.   Within systems, we
can  also distinguish  services like the  telephone that  are natural
monopolies, but where each person can decide how much of  the service
he wants to buy, from services  like public roads, sewers, or defense
where, once  a public decision has been made of how much society will
buy, each person has to pay his assessed share. The former is usually
preferable if  the service  can be organized  that way.   Primary and
secondary education is a service that can be paid for by the consumer
but which  is usually paid  out of taxes.   It seems  to me  that the
usual  reasons for handling  free education  are valid and  may apply
also to most health care.

	However, imagine an artificial heart that was completely
convenient but cost $10 million and was the only one available for
some purely technological reason.  Some authors write as though they
would favor prohibiting it until everyone could have it.  My opinion
is the opposite.  Letting the rich have it is the best way of making
it likely that we will all be able to afford it.  There will be
very strong pressures on the manufacturers to develop models selling
for $1 million and on down to the $100 thousand level at which it
would be purchased by most who needed it.  Yes, we'd buy it at that
price to the dismay of our heirs and at the cost of selling our houses.

	2. The simplest  kind of product  to put into  use is one  in
which the major development expenses  are for a particular design and
preparing  for production in  the factories of  a particular company.
The first  company  to introduce  the  product knows  that  potential
competitors  will  face approximately  the  same  expenses.   Another
favorable case is when the product involves inventions of a type that
can get  good patent  protection. The  worst case  is when there  are
large  proof-of-concept  and market  education expenses  that  do not
result in patent protection. In such  a case, a company that  doesn't
already  have  a  near  monopoly may  find  introducing  the  product
economically  not worthwhile, because other  companies will enter the
market after the  market is  established,  and  the pioneer will  not
recover  its initial  expenses.   This is  particularly acute  when a
small company  considers introducing  a product  and can  see that  a
certain very  large company will  be able to  take 90 percent  of the
market after the concept has been proved.  Many possible improvements
to automobiles have this character,  and an example is  the first smog
control  device. In  the 1950s, a California
law made  smog control  devices mandatory as  soon as two
manufacturers had demonstrated  them.  Two  did,  but the  automobile
manufacturers  made  their  own  instead  of  buying  them  from  the
companies that made the demonstrations.  (Check to make sure this  is
right.) The computer controlled car, as discussed  earlier,  may well
have  htis character  unless it  turns  out to  be accomplished  in a
patentable way which is very doubtful. Another example is that Thomas
Edison gave the first proof-of-concept for supplying electricity as a
public  utility, but the other electric  companies didn't have to pay
him anything for this.

	The automatic delivery system is a system par excellence.  It
requires  digging up  the  streets, it  needs lots  of  customers for
success, and  the  benefits to  an  individual subscriber  depend  on
stores, etc. being subscribers.  One way it could come about is for a
local government to decide that it is a good thing and issue bonds to
install it.  Another is for a private company to get a  franchise for
a certain area  from the government and raise the  money itself.  The
latter has  the advantage that the company risks its investors' money
on the proposition  that the service  will prove popular, whereas  if
thhe investment is  made with public money, many  more people have to
be convinced.  Naturally,  the  investors will expect to be  rewarded
for their gamble by higher rates than the city  might charge.  If the
proof-of-concept  and proof-of-market expenses  are very high,   as I
expect they would be  for the delivery  system,  the franchise  would
have  to  cover a  wide  area  in order  to  be  worth the  necessary
investment - a wider area than could immediately be promised service.
An attempt  by the  government to  optimise  by giving  away a  small
franchise and the expecting to put other franchises up for bids might
fail if potential investors thought they would be developing the system for
the  benefit of  free-loaders.    An  alternative would  be  for  the
government  to  develop  a  demonstration  system  and then  put  the
franchises  up for  bids.    Even  this  might  require  a  different
political  atmosphere  than  exists  now -  an  atmosphere  in  which
legislatures  would support something that  promises a benefit rather
than confining  themselves to  supporting  projects whose  supporters
predict disaster unless the project is undertaken.

	All this suggests another way of encouraging development. The
idea is that the  developer of a  system should be  able to expect  a
financial  reward for  his  pioneering (if  it  is successful)  under
circumstances  in which the  patent system is  not good  enough.  Why
isn't the patent system good enough?  

	First and simplest,  if it takes  the public  a long time  to
decide they like a product or system and the seller loses money while
this  is going on, when the public  finally is attracted, there
is no patent protection for the result of the  market development. If
the original developer is a  very large company, it may hope to get a
large share of the market and  will accept the free-loading. This  is
what happened with RCA's development of color TV. A small company may
simply decide that it is impractical to undertake the development.

	Secondly, patent  protection may be unobtainable, because the
basic ideas  are in the  public domain,  and the  realization may  be
carried out in  a variety of ways.   Moreover, most of the  cost in a
development may be in going from a device that sometimes works to one
that is  reliable.    Besides all  that,  the concept  of  patentable
invention as  distinguished from a  natural development of  the prior
art has been whittled away by innumerable court decisions.

	One idea for relieving  the situation is  to make post  facto
awards of royalties.  The idea is that  after an development has been
put  into use  and  has proved  valuable, a  legal  proceeding should
determine who  has contributed  what to  the  development and  divide
royalties over  a period of  years accordingly among  individuals and
firms.   There  would have to  develop a body  of case  law before it
became feasible for a company to predict what it might get back if it
put  money into  a  development that  would subsequently  be  used by
others.  The payment should include compensation for risk as well  as
for direct  costs.   Perhaps  a company  contemplating a  development
could  announce  it and  announce  the risk  they  thought  they were
taking. Another firm  who thought the  risk was over-estimated  could
put up some of its own money.  Otherwise, it would have to accept the
estimate put  forth by the developer. A feasible scheme might require
considerable development, but  the general idea  is to encourage  new
development by increasing the  probability that success would lead to
profit.\.

	The Apollo Project cost $25 billion.  The research phase of
the Strategic Defense Initiative is also budgeted at about $25 billion.
Developing personal.
flying machines or computer controlled cars might be comparably
expensive.  Since the resulting industry in either case would have
sales of many tens of billions per year, the development cost isn't
excessive.  However, at present private industry doesn't raise money
for such expensive research and development projects.  Major airplanes
cost about a billion to develop, but the company usually gets orders
for the airplane before undertaking the main part of the expenses.
Investments in the development of major oil fields are
in the billions.  In both cases the costs and the resulting market
are calculated in advance in a manner acceptable to the investors.
Even these calculations are often mistaken.

	So how are the billions required for the development
of personal flying machines, computer-controlled cars or automatic
delivery systems to be raised?