A conditional analytic technique for evaluating interdependent decisions, based on Bellman's 'Principle of Optimality' is proposed. The procedure computes both the present and future impacts of a decision and permits: (1) each decision to be evaluated independently; (2) an accurate and more informative evaluation of such decisions relative to existing evaluative schemes; and (3) important insights regarding the decision process, such as the employed planning horizon, the degree of choice conservatism, and whether and to what degree learning occurs. A nontime-staged assignment problem and time-staged production planning problem are used to illustrate its potential utility and generalizeability. The implications of the technique for evaluating non-programmable (e.g. managers) and programmable (e.g. heuristic) decisions as well as for designing organizational planning and control systems are discussed.
