Following the generic problem-solving process listed above requires at
some point knowing what outcomes or impacts are going to be used to
rank policy alternatives. Presumably, this means knowing the objectives
or goals of the policy. If the problem is, say, educational performance,
then policy alternatives can be ranked by their estimated impact on educational
performance indicators. Unfortunately, in practice, the problems
public policies are called upon to solve rarely have a universally agreed
upon, precisely defined, easily measurable, single goal. Educational performance,
for example, is a pretty vague goal