This study tests whether citizens’ evaluations of the performance of artificial intelligence (AI) in public policies are subject to motivated reasoning. Specifically, we test whether respondents’ preferences for AI regulation or their subjective attitudes toward AI are sources of motivated reasoning across varying use cases, differing in nature, complexity, safety-criticality and normative considerations: AI in municipal services, self-driving cars and recidivism prediction. Experimental results from two preregistered studies conducted among German citizens reveal that subjective attitudes toward AI cause substantial and robust motivated reasoning across all three policy domains. Regulatory preferences are only a selective source for motivated reasoning about AI in public policy. Overall, the results point to the cognitive limitations of strategies that attempt to objectify the benefits of AI without considering the context of the application domain. Politicians and policymakers need to consider these limitations in their attempts to increase citizens’ appreciation of AI in public policy.

Experiments on saving behavior reveal substantial heterogeneity of behavior and performance. We show that this heterogeneity is reliable and examine several potential sources of it, including cognitive ability and personality scales. The strongest predictors of both behavior and performance are two cognitive ability measures. We conclude that complete explanations of heterogeneity in dynamic decision making require attention to complexity and individual differences in cognitive constraints.

The level-k literature classifies subjects into different Lk types by their observed levels of reasoning in games. However, it remains unclear whether the observed level-k behavior is determined by belief or reasoning ability. This study proposes a strategy to identify the ability-bounded Lk subjects, who could not reason more than k steps of iterated best responses and thus have reached the upper bounds of their reasoning capacity. The identification utilizes a combination of simultaneous and sequential ring games. In the sequential games, it requires an extra step to best respond as Lk, and thus the ability-bounded ones would fail the task. Results show that more than half of the L2 and L3 subjects are ability-bounded. Additionally, subjects’ CRT scores, a measure of their cognitive ability, support the separation of the ability-bounded types. The findings suggest that both belief and reasoning ability could determine the observed levels, and thus one must be cautious when trying to infer belief or ability from the existing level-k data.

This paper examines the effects of competition in experimental posted-offer markets where sellers can confuse buyers. I report two studies. In one, the sellers offering heterogeneous goods can obfuscate buyers by means of spurious product differentiation. In the other study, sellers offer identical goods and make their prices unnecessarily complex by having multi-part tariffs. I vary the level of competition by having treatments with two and three- sellers in both studies, and having an additional treatment with five-sellers in one study. The results show that average complexity created by a seller is not different for the treatments with two, three and five sellers. In addition, market prices are highest and buyer surplus is lowest when there are two sellers in a market.

Bargaining and dilemma games have developed in experimental economics as fairly separate literatures. More than a few analysts are now persuaded that the patterns of behavior in these games are closely related, and considerable effort is being put into a search for models that bridge the gap between the two types of games. I focus on a handful of models that, when taken together, outline the conceptual issues, and provide a sense of the progress that has already been made.

This paper experimentally investigates the effect of introducing unavailable alternatives and irrelevant information regarding the alternatives on the optimality of decisions in choice problems. We find that the presence of unavailable alternatives and irrelevant information generates suboptimal decisions with the interaction between the two amplifying this effect. Irrelevant information in any dimension increases the time costs of decisions. We also identify a “preference for simplicity” beyond the desire to make optimal decisions or minimize time spent on a decision problem.

We explore the effects of the provision of an information-processing instrument—payoff tables—on behavior in experimental oligopolies. In one experimental setting, subjects have access to payoff tables whereas in the other setting they have not. It turns out that this minor variation in presentation has non-negligible effects on participants’ behavior, particularly in the initial phase of the experiment. In the presence of payoff tables, subjects tend to be more cooperative. As a consequence, collusive behavior is more likely and quickly to occur.

Previous experimental research suggests that individuals apply rules of thumb to a simplified mental model of the “real” decision problem. We claim that this simplification is obtained either by neglecting the other players’ incentives and beliefs or by taking them into consideration only for a subset of game outcomes. We analyze subjects’ eye movements while playing a series of two-person, 3 × 3 one-shot games in normal form. Games within each class differ by a set of descriptive features (i.e., features that can be changed without altering the game equilibrium properties). Data show that subjects on average perform partial or non-strategic analysis of the payoff matrix, often ignoring the opponent´s payoffs and rarely performing the necessary steps to detect dominance. Our analysis of eye-movements supports the hypothesis that subjects use simple decision rules such as “choose the strategy with the highest average payoff” or “choose the strategy leading to an attractive and symmetric outcome” without (optimally) incorporating knowledge on the opponent’s behavior. Lookup patterns resulted being feature and game invariant, heterogeneous across subjects, but stable within subjects. Using a cluster analysis, we find correlations between eye-movements and choices; however, applying the Cognitive Hierarchy model to our data, we show that only some of the subjects present both information search patterns and choices compatible with a specific cognitive level. We also find a series of correlations between strategic behavior and individual characteristics like risk attitude, short-term memory capacity, and mathematical and logical abilities.

The substantively rational value of the games studied in this paper does not help predict subject performance in the experiment at all. An accurate model must account for the cognitive ability of the people playing the game. This paper investigates whether the variation in measured rationality bounds is correlated with the probability of winning when playing against another person in games that exceed both players’ estimated rationality bound. Does seeing deeper into a game matter when neither player can see to the end of the game? Subjects with higher measured bounds win 63 percent of the time and the larger the difference the more frequently they win.

We run an eye-tracking experiment to investigate whether players change their gaze patterns and choices after they experience alternative models of choice in one-shot games. In phase 1 and 3, participants play 2 × 2 matrix games with a human counterpart; in phase 2, they apply specific decision rules while playing with a computer with known behavior. We classify participants in types based on their gaze patterns in phase 1 and explore attentional shifts in phase 3, after players were exposed to the alternative decision rules. Results show that less sophisticated players, who focus mainly on their own payoffs, change their gaze patterns towards the evaluation of others’ incentives in phase 3. This attentional shift predicts an increase in equilibrium responses in relevant classes of games. Conversely, cooperative players do not change their visual analysis. Our results shed new light on theories of bounded rationality and on theories of social preferences.

This paper explores the role of artificial intelligence (AI) within economic institutions, focusing on bounded rationality as understood by Herbert Simon. Artificial Intelligence can do many things in the economy, such as increasing productivity, enhancing innovation, creating new sectors and jobs, and improving living standards. One of the ways that AI can disrupt the economy is by reducing the problem of bounded rationality. AI can help overcome this problem by processing large amounts of data, finding patterns and insights, and making predictions and recommendations. This insight raises the question: can AI overcome planning problems – could it be that central planning is now a viable option for economic organisation? This paper argues that AI does not make central planning viable at either the nation-state level or the firm level, simply because AI cannot resolve the knowledge problem as described by Ludwig von Mises and Friedrich Hayek.

The outbreak of COVID-19 unleashed a severe crisis in society. The suddenness and speed with which the disease spread into a global pandemic makes it an outstanding case for showing how bureaucracy acts in response to a crisis marked by uncertainty and urgency. This article focuses on the role played by the central government bureaucracy in preparing and enacting the Danish government’s response to the COVID-19 crisis. It is based on full access to internal government files related to crisis management during the winter and spring of 2020. These files include memos, e-mails, decision notes, and draft decisions from key civil servants. The article demonstrates the strength of the theory of bounded rationality when it comes to analyzing the interaction between top civil servants and political executives. Moreover, it shows how administrative and political executives can mold a governmental organization to overcome the inertia inherent in bureaucratic organization and procedures.