Feedback complexity, bounded rationality, and market dynamics

Abstract

In a set of experimental markets we investigate how the dynamic structure of an economy interacts with different price-setting institutions to determine market performance, stability and learning. We test the standard assumptions of optimality and rational expectations against a behavioral hypothesis in which subjects ignore critical elements of the feedback structure in which they operate (‘misperceptions of feedback’). We find first that the introduction of even modest levels of dynamic complexity to the structure of the experimental economy significantly and substantially degrades performance relative to optimal, supporting the behavioral hypotheses. In the dynamically complex conditions, subjects generate costly, systematic fluctuations in prices and quantities. The institutional structure of the market also has a strong effect on performance relative to optimal. Performance is improved, but remains significantly below, optimal in the presence of market institutions, even in the case of a perfectly functioning market with computer-mediated market clearing. Markets moderate but do not eliminate the negative impact of bounded rationality and poor mental models. We then analyze the decisions of individual subjects by fitting them to various models of decision making. The estimates reject the hypothesis of rationality at the individual level. We then simulate the markets using the estimated decision rules. The simulations reproduce the most salient features of the dynamics, and variations between markets are related to differences the parameters values of the decision rules which characterize the degree to which subjects misunderstand the feedback structure of the market. In this manner, the analysis links microstructure to macrobehavior, coupling observation of individual decision making behavior and the genesis of aggregate market outcomes. Decision timing data and verbal protocols provide evidence that increasing complexity of the decision-making task leads subjects to ignore important aspects of the decision environment, particularly strategic interactions. In the face of dynamic complexity, subjects revert to simple, reactive decision rules, rules which prove to be dysfunctional. Moreover, subjects tend to attribute the oscillations and other systematic dynamics they experience to exogenous shocks rather than to their own interactions with the system. This latter phenomenon may have important implications for the ability of agents to learn and improve their behavior over time.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 19 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.