This report uses the goal programming concept to estimate livestock roughage consumption by type of roughage and State. The concept involves multiple conflicting objectives or goals which are evaluated subject to goal and nongoal constraints. The goal programming algorithm develops a solution to a multiple objective problem by minimizing the nonattainment of all goals subject to the restrictions imposed by nongoal constraints. Data for the State of Wisconsin are used to illustrate how the model is set up and run. Input data requirements for applying the model to each of the 48 contiguous States are described, followed by the running of the model for each State.

As the title suggests, the data on roughage consumption by type of livestock, type of roughage, and State as presented in the tables are not official estimates of the U.S. Department of Agriculture. The goal programming procedure (described in this manuscript) used for generating this data is a novel analytical technique. Its application should be viewed as a preliminary test of a potentially useful mathematical model. Hence, the reader should exercise caution in using and interpreting the results.