8
(production function) and the accuracy and adequacy of the observations used in the estimates of the coefficients.
Linear programming is a normative too] of analysis. Using observed or derived input-output coefficients and prices, the solution obtained is that which best satisfies the objectives and constraints of a contrived problem. The solution indicates the action which should be taken to achieve a stated goal, and does not necessarily describe the practices which are presently being followed in the study area. The output indicated by linear programming is that which represents the optimal utilization of the resources available, combined according to the relationships specified by the coefficients. Thus, linear programming is used to indicate a reorganization of production and resource use in order to realize the given objective. Numerous factors may cause this normative model to differ from actual conditions, and the value of it lies in its use as a guide toward efficiently employing resources in achieving specified ends. A Note of Comparison
A basic difference between regression analysis and linear programming was discussed in the preceding section comparing positive and normative analysis. Briefly, it should be remembered that production function analysis, using coefficients derived by regression analysis, can be used to solve the maximization (minimization) problem under existing conditions, that is, under the existing organization of production. The solution given is in the context of, and in accordance with, the actual structure of the production process. The use of linear programming in the maximization or minimization problem permits