Boundedly Rational Decision-Making under Certainty and Uncertainty: Some Reflections on Herbert Simon

The sixth in a series of excerpts from Minds, Models and Milieux: Commemorating the Centennial of the Birth of Herbert Simon.

Mark Pingle


Our collective rationality became more bounded on February 9, 2001. Herbert Simon emphasized we humans are cognitively constrained, and those constraints impact our decisions. Yet, Herbert Simon’s mind was less constrained than most of our minds. Because of his exceptional thinking and writing, the constraints binding many disciplines have been relaxed. Consequently, those disciplines have become more rational, and less. The purpose of this essay is to recognize how our collective rationality has been enhanced by the work of Herbert Simon, and related work, on decision making.

Decision-Making as a Processes

Our rationality is bounded by our limited cognitive capacities. This readily recognizable fact should make a theorist uncomfortable about assuming unbounded rationality. It made Herbert Simon uncomfortable. “The expressed purpose of Friedman’s principle of unreality (or as-if hypothesis),” Herbert Simon said, “is to save Classical theory in the face of the patent invalidity of the assumption that people have the cognitive capacity to find a maximum” (Archibald, Simon, and Samuelson, 1963). “The unreality of premises,” Simon continued, “is not a virtue in scientific theory but a necessary evil—a concession to the finite computing capacity of the scientist.”

Ignoring the bounds to rationality is convenient because it allows a decision problem to be specified as a mathematical optimization problem. Environmental factors can be parameterized using variables, so a change in one of the environmental variables will change the set of alternatives available to the decision maker. One can then delineate cause and effect relationships between elements of the decision environment and the optimal decision.

Herbert Simon’s critique of unbounded rationality was twofold. First, while all theories are abstractions, the explanatory power of a theory will tend to decrease as the premises of the theory are less representative of reality. Second, we should not just care about the ability of a theory to predict. The assumptions which underlie our theory are part of our explanation of how the world works, so our explanation lacks credibility to the extent that our assumptions lack realism. Simon (1955, p. 99) sought to “replace the global rationality of economic man with a kind of rational behavior that is compatible with the access to information and computational capacities that are actually possessed”

One of his important insights was that real world decision processes will tend to involve sequential search and adaptation. Search theory (e.g., Wilde, 1964) informs us that sequential search has a fitness advantage over the simultaneous sampling of alternatives because the knowledge obtained from the current search choice can be used to more effectively select the location of the next alternative. Simon’s (1955) bounded rationality model combines a sequential examination of alternatives with a predetermined “satisficing” goal for deciding when to stop incurring deliberation cost and accept an alternative as a choice.

If decision makers understand cognitive limitations imply deliberation costs bind them from achieving what Simon (1976) called “substantive rationality” (i.e., optimality), then they should also understand that no particular decision procedure will be best for all contexts. “Procedurally rationality” (Simon, 1976) involves coping with cognitive limitations and their implied deliberation costs, within a specific environment, by applying reason in some way. What is reasonable, or procedurally rational, can vary by decision maker and by context.

Under bounded rationality, then, understanding choice is not just a matter of relating changes in the decision environment to changes in the location of the optimal choice. It is involves relating changes in the decision environment to the decision maker’s cognitive abilities and to the decision maker’s available set of decision heuristics or processes. Because the choice is the outcome of the decision process selected, procedural rationality is reasoning applied at the level of selecting the decision method more so than reasoning applied to making the choice itself (Barros, 2010).