|Abstract or Summary
- Management decisions are generally considered to be made
under one of three categories of future knowledge: certainty, risk,
or uncertainty. All three categories occur in forest management.
However, forest management decisions whose outcomes are dependent
upon future levels of timber yields, prices, utilization standards,
or social and legal institutions are made under uncertainty.
Forest managers have always recognized that uncertainty existed;
however, they have not systematically included it in their decision- making
The objectives of the study were to: (1) establish the importance
of systematically considering uncertainty in forest management
decision-making and (2) illustrate and evaluate a model or
procedure, for the systematic consideration of uncertainty in forest
A review of the present status of forest management decision-making
constituted fulfillment of the first objective. Theoretical decision-making models which are currently used in forest management,
e. g., present worth analysis, capital budgeting, financial
maturity, and linear programming, while conceptually capable of
considering uncertainty, imply certainty That is, forestry applications
of these models have employed single-valued expectations.
Fulfillment of the second objective consisted initially of a
review of recent developments in the theory of decision-making
under uncertainty. All decision-making problems have some common
components. These components are: decision-alternatives,
the actions which the decision-maker deems possible to take; states
of nature, the future events which determine the outcome of the actions;
and consequences, the result of taking a specific action and
finding that a particular state occurs. The more popular theoretical
models for decision-making under uncertainty were reviewed: minimax,
minimax regret, Hurwicz index, and Laplace. While useful in
some cases, each of these models has specific disadvantages. In
addition, all the models have one common major disadvantage, they
contain the implicit assumption that the decision-maker is completely
ignorant about the states of nature which influence his problem.
In reality, forest managers and other decision-makers usually
possess some information, although it may be vague, about their
problems. If a decision-maker is not willing to assume complete
ignorance about the occurrence of the states of nature, he cannot apply any of the above models.
There is a theoretical decision-making model which appears
compatible with reality. The model, Bayesian decision theory,
allows the decision-maker to arrive at a solution which is compatible
with his opinions or judgements about the states of nature. Also,
he can combine these opinions or judgements with experimental
data to derive a solution using all available information, both subjective
Fulfillment of the second objective was completed by illustrating
the application of Bayesian decision theory to a hypothetical problem.
The problem, optimal degree of land ownership for an industrial
forestry firm, was defined within the Bayesian model and a
solution derived. Since the problem was hypothetical, the actual
solution is not the primary result of the study. The resulting implications
for actual situations is the primary contribution.
If forest managers are to make decisions which contain uncertainty,
the uncertainty should be systematically recognized in
the decision-making process. The Bayesian model is a logical
procedure for such recognition. By adopting and applying such
models, the efficiency of forest management decision-making will