This paper is concerned with the problem of balancing an optimization model consisting of two submodels. The submodels are represented by separate linear programming problems and are linked by dependence on common resources, or by the presence of the same variables in both of them. 

The method for coordinating the activities of submodels, in order to reach an overall optimum, is based on the direct exchange of proposals between submodels. Computational improvements in comparison with the conventional master-subproblems scheme are shown.