A multiple goal resource allocation model as applied to the board of investments' industrial structure planning
Date
1970-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
One of the more common problems encountered by development planners is the determination of an industrial investment program that optimizes the multiple goals expressed by policy-makers and society in general, subject to the resource and technical constraints of the economy. This is the type of problem posed to economic planners by the Board of Investments.
There exists a multitude of powerful constrained optimization models that may be applied to such problems. However, the main impediment to the application of these models has been the empirical specification of an objective function that is technically sound and truly reflects the wish of policy-makers. Without an objective function, constrained optimization methods cannot proceed.
The source of difficulties is technical as well as psychological. The matter of specifying an objective function reflecting a policy-maker’s judgment on multiple policy goals has always been a difficult job. It involves requiring the policy-maker's judgment on what is the marginal trade-off between goal achievement levels under varying conditions. The information requirement of such an exercise is virtually impossible to obtain and even if available the quality of the judgment must be doubted if the implications of choices made are not known to the policy-maker.
A second source of difficulty is psychological in nature. Policy-makers find it easier to think in terms of target levels for each goal, rather than in terms of marginal trade-off between goals when a choice is made between industrial projects. To illustrate--one finds in the Investment Priorities Plan of the Board of Investments (BOI) a statement as to how much foreign exchange, contribution to GNP,
how many jobs, etc., will be generated by BOI's Investment Plan (i.e., a program of investment). The amount of foreign exchange, GNP and number of jobs generated by BOI's Investment Plan may be taken by the planner as the policy-maker's aspired goal targets for his multiple goal objectives. One
does not find a policy statement as to how much foreign exchange may be substituted for a job opportunity created or for a peso contribution to GNP. Even if marginal rates of substitution between goal achievement levels are implicit in the choice of goal target levels, the policy-maker does
not consciously use these information to justify his choice of goal target levels. In short, it is operationally more convenient for the economic planner to make use of aspired goal target levels of policy-makers, as the starting point of a resource allocation model design.
These observations were taken into consideration in the design of the multiple-goal resource allocation model developed in this paper to solve the BOI problem of determining the investment program which will optimize the achievement of multiple policy goals. The model developed used expressed goal target levels as the starting point of analysis. BOI expects its Investment Priority Plan to generate a certain level of foreign exchange, GNP, and a certain number of jobs if the investment requirements of the plan are fulfilled. Insofar as these goal achievement levels were acceptable to BOI policy-makers, these goal achievement levels were considered in the model as the "official targets," although the Investment Priority Plan itself may not be the optimal program of investment to fulfill these targets.
Using a proto-type of linear programming, called goal programming, under- achievement of the specified goal targets was minimized and where possible, over-achievement maximized. The optimal solution of this problem gives the investment program which best achieves the given set of goal targets. This is then one main result of the paper--given a set of goal targets and a certain stock of resources (and values for other technical parameters), the optimal program of investment is determined. The optimal program of investment is that which leads to the highest achievement levels of the given goals, or alternatively it is that program of investment which minimizes under-achievement of given goal targets or maximizes over-achievement of given goal targets.
It is intuitively clear that the choice of goal targets and the relative priority weights given to each goal will influence the optimal achievement level of goals. Hence goal setting is a critical phase of decision-making. In general, policy-makers do not realize the implications of their chosen targets and therefore the chosen goal targets will most likely not be the wisest nor is the choice truly rational. Shackle
defines rationality in the following manner –
“Choice is rational when it conforms to a
perfect knowledge of all circumstances which
will affect its relevant outcome. Any gap in
his (the decision-maker's) knowledge of those
circumstances destroys the possibility of his
demonstrating to himself a rigid connection bet-
ween his chosen act and his desired state of
mind, and thus leaving his choice arbitrary"*.
*G.L.S. Shackle, A Scheme of Economic Theory, Cambridge
University Press, 1965. p. 12
If the policy-maker knows how his choice of goal targets and relative priority weights affects the achievement levels of his given policy goals, the policy-maker can more rationally and more wisely tell his economic planner the goal targets to be optimized. This paper shows the implication of alternative goal targets other than the official BOI targets, by going through a sensitivity analysis of goals.
Finally a third set of results is the determination of the effect on goal achievement levels of changes in critical technical or resource parameters.
A total of fifty- two linear programming problems were solved in this paper in order to analyze the effects of changing technical and goal parameters. Some of the more important empirical findings of the paper are here enumerated.
1. All BOI official targets for foreign exchange,
GNP and employment generation (for the mining
industries) can be achieved, even if investible
funds available is only 80 per cent of total
investible fund requirements, if the optimal
program of investment (as riven in this model)
is adopted. This means that BOI's investment
programs as given in the Investments Priorities
Plan is not the most efficient investment prog
ram for meeting its given goal targets.
2. The employment and GNP generation targets of
BOI are easily over-achieved, while the foreign
exchange generation target was difficult to
achieve. It is achieved only when investible
funds available is 80 per cent of total requirements.
This indicates that a higher priority weight
should be given to the achievement of the foreign
exchange goal. Experimentation with goal priority
weights showed that the achievement of the foreign
exchange goal should be valued three times greater
than the valuation of the GNP and employment goals.
3. The more labor intensive industries are relatively
more efficient in meeting the BOI policy goal targets
than the capital intensive industries.
4. Experience has shown that investible fund
requirements of BOI's Investment Priorities Plan are
difficult to meet. This paper shows the optimal
program of investment, given the situation that
only 10, 30, 50, and 100 per cent respectively of
investible fund requirements can be met.