Information overload and belief formation
In this paper, Bohren et al. propose a framework that incorporates well-studied limits on attention and memory to derive predictions on how these limits affect beliefs and choices depending on whether information is presented simultaneously or sequentially.
- When learning from simultaneous information, memory plays little role and bottom-up channeled attention leads to the overweighting of unlikely but salient events. This can generate a preference for prospects that are most likely to underperform, but when they do outperform, they do so by a lot. A strict preference for such unlikely outperformers can be interpreted as the overweighting of rare events.
- When learning through sequential sampling, however, imperfect memory recall implies that an infrequent event will be misremembered. This can shift preferences towards prospects that outperform most frequently—behavior that looks like the underweighting of rare events.
Importantly, because the behavioral predictions are derived from basic cognitive processes, the learning environment is expected to change people’s perceptions and mental representations of the risky prospects—learning from information presented simultaneously versus sequentially will lead to different beliefs about the relevant distributions.
A Cognitive Foundation for Perceiving Uncertainty
Authors: J. Aislinn Bohren, Josh Hascher, Alex Imas, Michael Ungeheuer, Martin Weber
From: University of Pennsylvania, University of Chicago, Aalto University School of Business, Universitaet Mannheim
Do people make economic decisions according to their beliefs?
In this paper, Jeffrey Yang argues that in certain settings, decision-makers may not understand how to optimally incorporate beliefs about a given economic quantity into their decision-making, i.e., decision-makers are uncertain about the belief-action map. Yang develops a theoretical framework where he uses rational inattention and cognitive noise to explain the attenuation in the relationship between beliefs and behavior, weak behavioral responses to information interventions, and poorly calibrated beliefs. More specifically, the decision-maker (DM) in Yang’s model is initially uncertain over a decision-relevant core economic quantity and chooses how much to learn about the quantity before taking an action, following standard approaches to modeling rational inattention. The main difference with rational inattention models, though, is that Yang does not assume that the DM perfectly observes the decision weight but instead has access only to a noisy signal of the true decision weight. The model generates two predictions:
- First, uncertainty over the belief-action map attenuates the cross-sectional relationship between the DM’s beliefs about the corresponding quantity and behavior.
- Second, uncertainty over the belief-action map reduces the responsiveness of the DM’s behavior to information about the quantity.
In particular, in an incentivized survey experimenting relating subjects’ stock return expectations to their portfolio choices, this paper documents that uncertainty over the belief-action map attenuates the relationship between subjects’ beliefs and behavior and reduces their behavioral response to information.
On the Decision-Relevance of Subjective Beliefs
Authors: Jeffrey Yang
From: Harvard University
