Summary : Some simple-minded observations on the role of optimization in public systems decision making

Liebman, J., 1976, “Some simple-minded observations on the role of optimization in public systems decision making,” Interfaces, v6, n4, pp. 102-108.

Previous optimization has been successfully applied to resolve private sector problem, but in public sector problems there are many defects to resolve.

In the past, using linear programming method was so useful and successful for optimization in private sector matter and the optimization skills also have grown in many ways, even though there was not so much understanding between decision-maker and analyst. However, when analysts have attempted to resolve public matters, the results of the optimization have showed many failing cases due to some reasons. First, public and social system is vague and difficult to find clear interconnection between cause and effect. Second, there is no perfect agreement in public due to the difference of human tastes. Third, analyst cannot fully comprehend problem of decision-makers, which easily results in failing optimization. This public problems call “wicked problem”.

Therefore, a suggestion is here that optimization should not be used to resolve conflicts which are diversity of problem perceptions, objectives, goals, measures of effectiveness, and constraints.

Optimization tools seem that we cannot use the tools to get the perfect result of optimization in public matters due to the too huge range of elements, which have to be considered by analyst. However, the optimization can still play pivot role in decision. In large degree, analyst cannot provide the best answer but can offer various alternatives for decision-makers to be able to select one.

a.       Why is the paper interesting or significant? I think people who used optimization to resolve public sector problem at that time could not know what was wrong when the modeling and optimization failed, because the results of optimization were usually right. However, this article showed that what was wrong and promoted further study.

b.      What are the faults or limitation of this work? It is too conceptual and mostly about problems.

c.       What is the possible work extending from this work? If this were your research, what would be your next steps to fix the work, apply ideas to other applications, or start new work from these ideas? I would like to select several good examples about failing optimization, and I will analysis them and try to figure out which elements are significant in public sector cases. Then, I will try to compare prior models and new models until I will get some good results.