no outside trading partners, and maybe the Soviet Union could too. If both
countries want to improve their living standards substantially, however, they
should take advantage of their potential abilities to buy and sell products from
each other. The Soviet Union is one of the leading grain buyers of the world
and could be a leading buyer in almost any field, given the size of its population.
The Soviet Union is one of the leading producers of oil and gold, and it could
become a leading producer in other fields if it would concentrate its resources
on what it can do relatively well, the way Japan has. The United States could
sell grain and other products to the Soviet Union in return for gold, oil, and
other products, thereby easing the international deficit, which can be paid in
gold, and the U.S. energy problems.

Everything that is shown in Table 17-6 can be subjected to a computerized
what-if analysis. Such an analysis enables one to determine what it would take
to bring any of those alternatives that are tied for second place up to first place.
It also enables one to determine the effects of adding additional goals, changing
the alternatives, or changing any of the inputs. This kind of what-if capability
may be the most important purpose that is served by working with decision-
aiding software. Some of the benefits can also be obtained by working with a
spreadsheet matrix (like that of Table 17-6 with alternatives, goals, relations,
and total scores) even if the matrix is not computerized. One type of sensitivity
analysis that is often especially helpful is to ask what would be the winning
alternative if we just concentrate on the conservative goals or give them extra
weight, which in this case would be having a higher GNP with full employment
and lots of consumer goods. One exciting characteristic of super-optimum so-
lutions is that they win even when one uses only conservative weights or only
liberal weights. The reason here and often in other SOS situations is that the
SOS alternative does better than the other alternatives on every individual goal,
not just better on the overall total. There are ways of systematically arriving at
super-optimum solutions. That, however, is partly a subject for Chapter 19 on
"Broadening the Applicability of Multi-Criteria Dispute Resolution" and for
my next book on Super-Optimum Solutions in Public Controversies.

-210-