An alternative can be defined as a project or a (complex)
of policy measures with which the gap between the desired
situation and the zero-situation can be (fully or partly)
be closed. In other words, it is (a combination of) measures
which solve the earlier defined problem sufficiently.
Two sub steps can be identified; generation and pre-selection.
In addition, the alternatives are restricted by a set
Generation of alternatives
In general, the generation of alternatives is not a controlled
process. Possible solutions come already on the surface
during the problem analysis. It can even be the case that
the proposal of a certain alternative forms the start
of a policy analysis.
In most cases, the initiator of the policy analysis will
come up with alternatives. In addition, alternatives can
be proposed by individuals or groups that are not directly
involved in the study (like pressure groups). The initiator
can also demand that certain alternatives are not included
in the study.
In certain studies, both the number and the content of
alternatives can be determined on forehand. In these cases,
it is needed to know the background of these decisions
(like time limit: a decision is needed in short term).
It can also be the case that the study initiator likes
to prevent that incorporation of new alternatives will
lead to the situation that 'his' alternative comes in
danger. A further analysis of the alternatives of the
study initiator can reveal the implicit boundary conditions
of the policy study (like financial restrictions).
In other cases, the analyst has the freedom to determine
both number and content of the alternatives within given
the list with points of attention for the generation
Pre-selection of alternatives
It is very well possible that in this phase of the project
study, there are a large number of alternatives, even
a too big number of alternatives. It is impractical and
very costly to work out all these alternatives at sufficient
detail, to present them and to include them in the final
decision making process. A further reason to limit the
number of alternatives might be the evaluation method
that will be applied, as some methods can handle only
a limited number.
In those cases it is necessary to make a pre-selection.
The large number of alternatives is reduced to a limited
number of promising solutions. A first reduction can already
be obtained through clustering: alternatives with a certain
similarity are clustered and considered as one alternative.
In a later stage, such clusters can be split up into variants,
which feasibility can be further analysed.
It might be needed that during pre-selection, a global
inventory is made of possible effects.