Psychology and Economics

Date
Wed, Aug 17 2022, 9:30am - Fri, Aug 19 2022, 1:45pm PDT
Location
SIEPR Koret-Taube Conference Center, Room 130
366 Galvez Street, Stanford
[In-person session]
Organized by
  • B. Douglas Bernheim, Stanford University

  • John Beshears, Harvard Business School
  • Vincent Crawford, University of Oxford and University of California, San Diego
  • David Laibson, Harvard University
  • Ulrike Malmendier, University of California, Berkeley

This workshop brings together researchers working on issues at the intersection of psychology and economics. The segment will focus on evidence of and explanations for non-standard choice patterns, as well as the positive and normative implications of those patterns in a wide range of economic decision-making contexts, such as lifecycle consumption and savings, workplace productivity, health, and prosocial behavior. The presentations will frequently build upon insights from other disciplines, including psychology and sociology. Theoretical, empirical, and experimental studies will be included.

In This Session

Wednesday, August 17, 2022

Aug 17

9:00 am - 9:30 am PDT

Check-In and Breakfast

Aug 17

9:30 am - 10:05 am PDT

Tailored Stories

Presented by: Chiara Aina (University of Zurich)

This paper studies the problem of persuading a boundedly rational agent by proposing a menu of models (i.e., interpretations) of possible future events. For example, a political campaign could suggest ahead of a vote that it might be fair, but could also be rigged; or a stock-market analyst could suggest several ways of looking at future earnings announcements. Following the release of data, the agent adopts the model that best fits the data given her prior beliefs, and takes an action that maximizes her posterior expected utility under the adopted model. Anticipating this, a persuader can strategically communicate models to manipulate how the receiver interprets the different signals.I characterize the extent and limits of belief manipulability in this setting. In particular, I show that the agent may hold inconsistent beliefs across signal realizations — posterior beliefs across signal realizations do not average to her prior — because they trigger her to adopt different models. But, while persuasion can mislead the receiver, its impact is bounded. When agents differing in initial beliefs are exposed to conflicting models, polarization occurs

Aug 17

10:05 am - 10:40 am PDT

Shared Models in Networks, Organizations, and Groups

Presented by: Joshua Schwartzstein (Harvard Business School)
Co-author(s): Adi Sunderam (Harvard Business School)

Why did the market rise yesterday? What are the implications of the latest school shooting? Why did a particular employee get promoted? To answer such questions, we often exchange models, stories, narratives, and interpretations with others. This paper provides a framework for thinking about such social exchanges of models. The key assumption, following Schwartzstein and Sunderam (2021), is that when people are exposed to multiple interpretations they adopt the one that best explains the data. A key implication is that within a network interpretations evolve. This evolution driven by social learning hardens reactions to data that are open to interpretation: following the exchange of models, people are more convinced they are able to explain the data. For certain network structures, we show that social learning also mutes reactions to data that are open to interpretation: the exchange of models leaves beliefs closer to priors than they were before. Beyond studying fixed networks, we also consider how firm managers, politicians, and other agents are able to influence patterns of communication to their advantage. Agents who benefit from the status quo or broadly shared understandings will encourage a robust exchange of interpretations; agents who instead want new data to change behaviors will try to limit the exchange of interpretations, especially interpretations that suggest the data are not surprising. We apply the framework to consider the goal and structure of meetings in organizations, as well as the evolution and persistence of myths in social networks.

Aug 17

10:40 am - 11:15 am PDT

Coffee Break

Aug 17

11:15 am - 11:50 am PDT

Is Attention Produced Optimally?

Presented by: Dmitry Taubinsky (University of California, Berkeley)
Co-author(s): Erin T. Bronchetti (Swarthmore College), Judd B. Kessler (University of Pennsylvania), Ellen B. Magenheim (Swarthmore College), and Eric Zwick (University of Chicago Booth)

This paper investigates whether people know their attention cost functions and allocate attention optimally. We characterize how demand for instruments that reduce attention costs varies with the pecuniary incentives to be attentive, under the null hypothesis of correct perceptions and optimal choice. We examine if the optimality conditions are satisfied in three experiments. The first is a field experiment (n = 1373) with an online education platform, in which we randomize incentives to complete course modules and incentives to utilize a plan-making tool to complete the modules. In the second experiment (n = 2306), participants must complete a survey in the future. We randomize survey-completion incentives and how long participants must wait to complete the survey, and we elicit willingness to pay for reminders. The third experiment (n = 1465) involves a psychometric task in which participants must identify whether there are more correct or incorrect mathematical equations in an image. We vary incentives for accuracy, elicit willingness to pay to reduce task difficulty, and examine the impact of learning and feedback. In all experiments, demand for reducing attention costs increases as incentives for accurate task completion increase. However, in all experiments—and across all conditions—our tests imply that this increase in demand is too small relative to the null of correct perceptions. These results suggest that people may be uncertain or systematically biased about their attention cost functions, and that experience and feedback do not necessarily eliminate mistakes.

Aug 17

11:50 am - 12:25 pm PDT

Disposed to be Overconfident

Presented by: Katrin Godker (Masstricht University)
Co-author(s): Terrance Odean (University of California, Berkeley)

Overconfidence, the notion that investors tend to systematically believe that their own investment abilities are better than they are, is a key concept used in behavioral finance to explain the observed high trading volume in markets. Most studies in financial economics treat investor overconfidence as a static personal trait; they do not examine the processes through which investors become more—or less—overconfident. In this paper we show that the disposition effect, a well documented pattern in investor behavior, can be a source of investor overconfidence. We provide experimental evidence for a biased learning process through which the disposition effect leads to investor overconfidence. We show that investors update beliefs about their own investment ability based on realized gains and losses rather than the overall performance of their portfolio. We formalize this learning process in a theoretical model in which the disposition effect leads to overconfidence, excessive trading, and lower investment performance.

Aug 17

12:25 pm - 1:45 pm PDT

Lunch

Aug 17

1:45 pm - 2:20 pm PDT

On the Origin and Persistence of Identity-Driven Choice Behavior

Presented by: Caroline W. Liqui Lung (Paris School of Economics)

A recent literature shows that a priori identical individuals belonging to different social groups tend to make different occupational and educational choices. This paper provides a novel explanation for this identity-driven choice behavior that does not rely on builtin differences in preferences, discrimination or social pressures, but works through the optimal biasing of beliefs. Agents observe social identity cues that stem from the prevalence of their subgroup among the successful individuals in the social context, and find it optimal to let their choices be driven by these social identity cues, even when these cues do not directly affect utility and are informationally irrelevant in a Bayesian sense. The influence on choice differs across types, but is not driven by ability differences. I show the existence of a stable population equilibrium in which the task allocation and the use of social identity cues differ between a priori identical subgroups

Aug 17

2:20 pm - 2:55 pm PDT

Information Resonance

Presented by: Ulrike Malmendier (University of California, Berkeley)
Co-author(s): Laura Veldkamp (Columbia University)

People process the same information differently depending on who delivers it. Information resonates with recipients when they identify with the person who communicates it or whose personal experience it reflects. Resonant information imprints itself in our memory and comes to mind when making relevant decisions. We describe and formalize the phenomenon of resonant information being more heavily weighted in the process of belief updating. The model delivers an empirical tool to identify which social characteristics determine reso-nance. We then work through the model implications for social transitions in response to COVID-type shocks, for the influence of role models on behavior change such as vaccine adoption, and why social media changes the influence of experts. Evidence on occupational choices within geography-ethnicity-gender groups, and their responses to labor market outcomes inside and outside their respective groups, suggest that resonance may be a powerful force

Aug 17

2:55 pm - 3:30 pm PDT

Coffee Break

Aug 17

3:30 pm - 4:05 pm PDT

The Supply of Motivated Beliefs

Presented by: Michael Thaler (Princeton University)

When people choose what messages to send to others, they often consider how others will interpret the messages. In many environments, particularly in politics, people are motivated to hold particular beliefs and distort how they process information in directions that favor their motivated beliefs. This paper uses two experiments to study how message senders are affected by receivers’ motivated beliefs. Experiment 1, conducted using an online sample of social media users, analyzes the effect of incentivizing senders to be perceived as truthful. These incentives cause senders to send less truthful messages. When incentivized, senders send more false information when it aligns with receivers’ politically-motivated beliefs, controlling for receivers’ current beliefs. However, receivers do not anticipate the adverse effects of senders’ incentives. Experiment 2 further isolates the role that information processing plays by analyzing an environment in which receivers assess the truthfulness of messages from a computer and senders choose one of the computer’s messages to determine their earnings. Senders predict that receivers distort information processing in the direction of their politics, demand information about receivers’ political preferences, and condition on the receivers’ politics to strategically choose less truthful computer messages.

Aug 17

4:05 pm - 4:40 pm PDT

Strategic Thinking and Media Bias: Evidence from Chinese Microblog Users

Presented by: Yihong Wang (Harvard University)
Co-author(s): Juanjuan Meng (Peking University) and Xi Weng (Peking University)

It has been shown that media bias contributes to belief polarization from the supply side. This paper provides a novel perspective from the demand side by testing whether news consumers are strategic enough to detect media bias. We propose a theoretical framework which shows that the tendency of reposting news with political inclination inconsistent with the media outlet’s general ideology is an indicator of strategic thinking. Combining both field data from Chinese Microblog users and an online experiment that exogenously varies whether news sources are revealed, we show both observational and causal evidence that Microblog users are more likely to repost inconsistent news, an indication of general strategic reaction. We also show that simply reminding people of media bias can make their decisions more strategic.

Aug 17

6:30 pm - 9:30 pm PDT

Dinner

Thursday, August 18, 2022

Aug 18

9:00 am - 9:30 am PDT

Check-In and Breakfast

Aug 18

9:30 am - 10:05 am PDT

Social Comparisons in Peer Effects

Presented by: Seung-Keun Martinez (University of Nottingham)
Co-author(s): Jason Bigenho (Amazon.com, Inc.)

Numerous studies document peer effects in individual decision making, even absent material incentives to coordinate. We extend Benabou and Tirole (2006) to propose a self-signaling theory of social comparisons that produces peer effects in choices when individuals use others’ behavior as useful information for judging their own actions. This model also makes the novel testable predictions that (1) individuals will make more costly self-esteem improving choices in anticipation of private social information; and (2) introducing extrinsic incentives to engage in an otherwise intrinsically motivated task will mitigate this anticipation effect. We test these predictions in a series of incentivized charity experiments. We show that (1) individuals choose to complete more tasks for a charity in anticipation of learning how many tasks previous participants chose to complete; and (2) this anticipation effect disappears if task completion incurs not only a donation to charity but also a payment to the participants themselves.

Aug 18

10:05 am - 10:40 am PDT

Bottlenecks for Evidence Adoption

Presented by: Woojin Kim (University of California, Berkeley)
Co-author(s): Stefano DellaVigna (University of California, Berkeley) and Elizabeth Linos (University of California, Berkeley)

Governments increasingly use RCTs to test innovations before scale up. Yet, we know little about whether and how they incorporate the results of the experiments into policy-making. We follow up with 67 U.S. city departments which collectively ran 73 RCTs in collaboration with a national Nudge Unit. Compared to most contexts, the barriers to adoption are low. Yet, city departments adopt a nudge treatment in follow-on communication in just 27% of cases. As potential determinants of adoption we consider (i) the strength of the evidence, as determined by the RCT itself, (ii) features of the organization, such as “organizational capacity” of the city and whether the city staff member working on the RCT has been retained, and (iii) the experimental design, such as whether the RCT was implemented as part of pre-existing communication. We find (i) a limited impact of strength of the evidence and (ii) some impact of city features, especially the retention of the original staff member. By far, the largest predictor of adoption is (iii) whether the communication was pre-existing, as opposed to a new communication. We consider two main interpretations of this finding: organizational inertia, in that changes to pre-existing communications are more naturally folded into year-to-year city processes, and costs, since new communications may require additional funding. We find the same pattern for electronic communications, with zero marginal costs, supporting the organizational inertia explanation. The pattern of results differs from the predictions of both experts and practitioners, who over-estimate the extent of evidence-based adoption. Our results underline the importance of considering the barriers to evidence adoption, beginning at the stage of experimental design and continuing after the RCT completion.

Aug 18

10:40 am - 11:15 am PDT

Coffee Break

Aug 18

11:15 am - 11:50 am PDT

Liquidity Constraints and the Value of Insurance

Presented by: Keith Marzilli Ericson (Boston University)
Co-author(s): Justin Sydnor (University of Wisconsin)

We explore how liquidity constraints affect the value of insurance using a model incorporating limits to borrowing and saving and insurance contracts that span multiple consumption periods. Two insights emerge. First, the timing of expenditures matters, which generates new results about optimal contract design and the importance of when premiums are paid. Second, insurance provides a consumption-smoothing benefit beyond its traditional risk-protection benefit. This model can explain puzzling patterns of choice, and can explain the purchase of insurance even when premiums are so high as to be dominated in the standard model. Rational liquidity-constrained individuals have a different normative benchmark of insurance demand than the standard model: they may not buy full insurance even at actuarially-fair prices, can benefit from costsharing designs that are more complex than straight-deductibles, and will not treat all expected-out-of-pocket costs similarly

Aug 18

11:50 am - 12:25 pm PDT

Present Bias Amplifies the Household Balance-Sheet Channels of Macroeconomic Policy

Presented by: Peter Maxted (University of California, Berkeley)
Co-author(s): David Laibson (Harvard University) and Benjamin Moll (London School of Economics)

We study the effects of monetary and fiscal policy in a heterogeneous-agent model where households have present-biased time preferences and naive beliefs. The model features a liquid asset and illiquid home equity, which households can use as collateral for borrowing. Because present bias substantially increases households' marginal propensity to consume (MPC), present bias increases the impact of fiscal policy. Present bias also amplifies the effect of monetary policy but, at the same time, slows down the speed of monetary transmission. Interest rate cuts incentivize households to conduct cash-out refinances, which become targeted liquidity-injections to high-MPC households. But present bias also introduces a motive for households to procrastinate refinancing their mortgages, which slows down the speed with which this monetary channel operates.

Aug 18

12:25 pm - 1:45 pm PDT

Lunch

Aug 18

1:45 pm - 2:20 pm PDT

Flexibility through Delay

Presented by: Hunt Allcott (Microsoft Research)
Co-author(s): Matthew Gentzkow (Stanford University) and Lena Song (New York University)

People who think they are time inconsistent would like to commit to future actions but also want flexibility to respond to uncertainty. We study one partial solution to this tradeoff, flexibility through delay, in which people can relax a previous commitment, but only if there is a delay before consumption. We show theoretically how flexibility through delay can increase long-run welfare, especially if temptation decays quickly and/or people are partially naive. We demonstrate the idea using a field experiment where we allow smartphone users to set screen time limits that can be relaxed, but consumption can begin again only after a randomly assigned delay. The results provide textbook evidence of commitment-flexibility tradeoffs: consumers assigned to shorter delays set tighter limits and relax the limits more. We use the data to estimate the parameters of a dynamic temptation model, which suggests that 37 percent of the temptation to use social media decays within two minutes. Both the model and additional survey evidence suggest that flexibility through delay would benefit consumers relative to the screen time limits currently offered by both iOS and Android smartphones, which allow full flexibility with zero delay.

Aug 18

2:20 pm - 2:55 pm PDT

Managing Dynamic Inconsistency: The Role of Early Childhood Education

Presented by: Amanda Chuan (Michigan State University)

We conduct an experiment with nearly 400 children to investigate the role of early education on helping children delay gratification. At the ages of 3-4, children were randomized to a preschool program or a control group. Years after the intervention, we conduct an experiment on these same children in which we invite them to allocate rewards that they can receive at earlier and later times. We find that preschool did not make children more patient or less impulsive. Rather, as they aged, treated children who behaved impulsively in the past exhibited greater use of commitment devices to prevent impulsive behavior. Our evidence suggests that preschool helps children regulate executive function, including the use of external tools to overcome impulsive behavior. In contrast to other gains which subside over time, such as IQ, we find that preschool increases use of commitment devices against impulsive behavior as students age.

Aug 18

2:55 pm - 3:30 pm PDT

Coffee Break

Aug 18

3:30 pm - 4:05 pm PDT

Cognitive Endurance as Human Capital

Presented by: Heather Schofield (University of Pennsylvania)
Co-author(s): Christina Brown (University of Chicago), Supreet Kaur (University of California, Berkeley), and Geeta Kingdon (University College London)

Schooling may build human capital not only by teaching academic skills, but by expanding the capacity for cognition itself. We focus specifically on cognitive endurance: the ability to sustain effortful mental activity over a continuous stretch of time. As motivation, we document that globally and in the US, the poor exhibit cognitive fatigue more quickly than the rich across field settings; they also attend schools that offer fewer opportunities to practice thinking for continuous stretches. Using a field experiment with 1,600 Indian primary school students, we randomly increase the amount of time students spend in sustained cognitive activity during the school day—using either math problems (mimicking good schooling) or non-academic games (providing a pure test of our mechanism). Each approach markedly improves cognitive endurance: students show 22% less decline in performance over time when engaged in intellectual activities—listening comprehension, academic problems, or IQ tests. They also exhibit increased attentiveness in the classroom and score higher on psychological measures of sustained attention. Moreover, each treatment improves students’ school performance by 0.09 standard deviations. This indicates that the experience of effortful thinking itself—even when devoid of any subject content—increases the ability to accumulate traditional human capital. Finally, we complement these results with quasi experimental variation indicating that an additional year of schooling improves cognitive endurance, but only in higher-quality schools. Our findings suggest that schooling disparities may further disadvantage poor children by hampering the development of a core mental capacity.

Aug 18

4:05 pm - 4:40 pm PDT

Managing Emotions: The Effects of Online Mindfulness Meditation on Mental Health

Presented by: Advik Shreekumar (MIT)
Co-author(s): Pierre-Luc Vautrey (MIT)

Emotions and worries can reduce individuals’ available attention and affect economic decisions. In a four-week experiment with 2,384 US adults, offering free access to a popular mindfulness meditation app that costs $13 per month improves mental health, productivity and decisionmaking. First, it causes a 0.44 standard deviation reduction in symptoms of stress, anxiety, and depression, comparable to the impacts of expensive in-person therapy, with improvements even among participants with minimal or mild symptoms at baseline. Second, it increases earnings on a proofreading task by 1.9 percent. Third, it makes decision-making more stable across emotional states, reducing the interference of personal worries with risk choices. Overall, our results demonstrate the potential of affordable mindfulness meditation apps to improve mental health, productivity, and the impact of emotions on economic decisions.

Aug 18

6:30 pm - 9:30 pm PDT

Dinner

Friday, August 19, 2022

Aug 19

9:00 am - 9:30 am PDT

Check-In & Breakfast

Aug 19

9:30 am - 10:05 am PDT

Choice by Associations

Presented by: Jose Apesteguia (ICREA)
Co-author(s): Francesco Cerigioni (Universitat Pompeu Fabra)

It has been amply shown that behavior is context-dependent. Evidence from cognitive science suggests that such dependency might be driven by implicit associations. In this paper, we propose and study a choice-based model of contextual associations. We start by formalizing contexts by the set of concepts it contains. We then introduce associations by way of the implicit relationship between alternatives and concepts, which directly impacts the utility evaluation of alternatives. We study the empirical content of the model, its comparative statics, and several extensions. We argue that the model could be instrumental in understanding a wide range of phenomena, and derive new predictions.

Aug 19

10:05 am - 10:40 am PDT

Forced Experimentation During Pandemic Lockdown: A Neural Autopilot Analysis of Social Media

Presented by: Yi Xin (California Institute of Technology)
Co-author(s): Lawrence Jin (California Institute of Technology), Jessica Fong ?(2University of Michigan), Matthew Shum (California Institute of Technology), and Colin Camerer (California Institute of Technology)

This paper describes and estimates a novel “neural autopilot” model of habit formation using
individual-level data on posting behavior from a Chinese social media platform around the 2020
lockdown period. The model produces interpretable parameter estimates about habit formation;
it shows that once habit is neuroscientifically formalized, changes in preferences are no longer
required to explain the observed behavior change. We find that forced experimentation does
not lead to persistent habitual postings. However, counterfactually reducing the volatility in
posting rewards significantly increases habitual postings, suggesting that higher moments of the
reward process may play an important role in creating habit.

Aug 19

10:40 am - 11:15 am PDT

Coffee Break

Aug 19

11:15 am - 11:50 am PDT

Magic Mirror on the Wall, Who Is the Smartest One of All?

Presented by: Yoram Halevy (University of Toronto)
Co-author(s): Johannes Hoelzemann (University of Toronto) and Terri Kneeland (University College London)

In the canonical model of bounded rationality each player best-responds to their belief that other players reason to some finite level. We propose a novel behavior that reflects the player’s belief that while other players may be rational, the player cannot model and hence predict the behavior of others. This encompasses a situation where a player believes that their opponent can reason to a higher level than they do. We propose an identification strategy for such behavior, and evaluate it experimentally.

Aug 19

11:50 am - 12:25 pm PDT

Cognitive Distortions in Complex Decisions: Evidence from Centralized College Admission

Presented by: Ao Wang (University of California, Berkeley)
Co-author(s): Shaoda Wang (University of Chicago) and Xiaoyang Ye (Brown University)

Constructing optimal rank-order lists in centralized matching systems often entails sophisticated risk-taking consideration. We empirically study an admission system that employs a constrained Deferred Acceptance Algorithm to understand how students construct their lists. Students appear overly cautious with their top choices and most of them do not always put safer choices at a lower-ranked spot on the list. We propose that the Model of Directed Cognition could explain such choices. Applicants using the model myopically focus on the spot they are contemplating and neglect its impact on the rest of the list. To differentiate from alternative hypotheses, we deploy an in-field experiment that pinpoints a core prediction of our model concerning framing effects and find clear evidence of it. Structural estimation suggests that 45%55% of the sample are better described by our model and that this boundedly rational decision rule explains 83% of outcome inequality across socioeconomic groups.

Aug 19

12:25 pm - 1:45 pm PDT

Lunch