Implementation Effects of Alternative Performance Measurement Models in a Multivariable Context.

Abstract The article estimates the economic effect of varying certain information practices in a specific firm under a specific set of circumstances. The major features of the simulated decision context are a large number of decision variables, over twenty that are centrally determined with a global, but imperfect, optimization model, and implemented by a number of semiautonomous individual decision makers. These individual decision makers, in turn, have access to certain local information and can marginally influence implementation of the centrally determined decision variables. While the production aspects of the linear program model focus on determining the optimum mix to produce a specified number of each main product, the marketing aspects focus on how many of each main product should be produced and sold. Implementation effects may be viewed in terms of resultant variations in the parameters in the central linear program model and in the levels of the decision variables that the individuals are instructed to implement. Such variation may be controllable, and may be desirable.