I study how limited abilities to process information affect choice behavior. I model the decision-making process by an automaton, and measure the complexity of a specific choice rule by the minimal number of states an automaton implementing the rule uses to process information. I establish that any choice rule that is less complicated than utility maximization displays framing effects. I then prove that choice rules that result from an optimal trade-off between maximizing utility and minimizing complexity are history-dependent satisficing procedures that display primacy and recency effects.

Panel and experimental data are used to analyse the economic outcomes in the extended warranty market. We establish that the strong demand and high profits in this market are driven by consumers distorting the failure probability of the insured product, rather than standard risk aversion or sellers’ market power. Providing information to consumers about failure probabilities significantly reduces their willingness to pay for warranties, indicating the important role of information, or lack of, in driving consumers’ purchase behaviour. Such information provision is shown to be more effective in enhancing consumer welfare than additional market competition.

We develop a framework for modelling choice in the presence of framing effects. An extended choice function assigns a chosen element to every pair (A, f) where A is a set of alternatives, and f is a frame. A frame includes observable information that is irrelevant in the rational assessment of the alternatives, but nonetheless affects choice. We relate the new framework to the classical model of choice correspondence. Conditions are identified under which there exists either a transitive or a transitive and complete binary relation R such that an alternative x is chosen in some (A, f) iff x is R-maximal in the set A. We then demonstrate that the framework of choice correspondence misses information, which is essential to economic modelling, and which is incorporated in the extended choice function.

We develop a satisficing model of choice in which the available alternatives differ in their inherent complexity. We assume—and experimentally validate—that complexity leads to errors in the perception of alternatives’ values. The model yields sharp predictions about the effect of complexity on choice probabilities, some of which qualitatively contrast with those of maximization-based choice models. We confirm the predictions of the satisficing model—and thus reject maximization—in a novel data set with information on hundreds of millions of real-world chess moves by highly experienced players. Looking beyond chess, our work offers a blueprint for incorporating complexity at the level of individual objects into models of choice and for detecting satisficing outside of the laboratory.