Colin Camerer - US grants

Many decisions made by individuals, firms, consumers, citizens' groups, governments, regulatory bodies and courts involve dealing with potential misfortunes or catastrophes before they occur-- through insurance, risk avoidance behavior, or compensation for risk bearing-- or after they occur--through private or government compensation. These decisions often appear to be inconsistant, inequitable or irrational. Insurers, uncertain about future awards, refuse to offer policies against certain risks because they base their decisions on worst-case scenarios. Individuals react to risks inconsistently, treating small risks as though they were nonexistent (failing to wear seatbelts or to take precautions against AIDS) or so large as to overwhelm all other considerations (opposition to genetic recombination). Governmental attempts to respond to such concerns are often ad- hoc, such as the Delaney amendment, resulting in unanticipated contradictory outcomes. This collaborative interdisciplinary research program examines these puzzling features of behavior within a multi-disciplinary tradition of research on biases and fallacies in decision making. It investigates the question of whether these biases distort the functioning of institutions, markets and societal allocation of resources. The research will develop and evaluate several approaches designed to correct decision biases or protect individuals from their effects, including education, novel ways of presenting information, and mixed strategies coupling incentives and regulations with insurance and compensation. The research plan organizes biases into four categories: reference-point effects, in which preferences for outcomes depend on what they are compared to; contingent weighting, in which choice depends on the way in which a decision is expressed; decision rules, in which absolute priority is given to one attribute (e.g. saving money, avoiding catastrophic consequences); and threshold effects, in which small risks or costs (e.g. the risk of one cigarette or the extra cost of low- deductible insurance) are totally ignored even if they are repeated. The research will examine whether such biases affect decisions about offering or purchasing insurance, setting premiums, compensating victims of misfortune, compensating people for the risk of misfortune beforehand, and engaging in protective behavior. Multiple methods will be used, including questionnaires, laboratory studies, experimental markets, process-tracing, field studies, and microeconomic analysis. The results will bear on theoretical controversies within psychology, economics and decision theory about the relevance of biases to real decision making, on the question of whether market competition reduces the effects of bias, and on questions about the nature of bias itself.

Most game-theoretic applications in economics assume players are rational, mutually rational, and purely self-interested. In experimental studies of games one or more of these assumptions often appear to be violated because people do not play Nash or more refined equilibria (learning sometimes reduces violations but not always). It is usually difficult to tell from strategy choices alone which assumptions a subject is violating. Since game- theoretic solution concepts can be interpreted as algorithms for choosing a strategy, we can test the algorithms directly using a computer system that records what information in a game (e.g., which payoffs) a subject is looking at, and for how long, along with their choices. The information processing data are revealed preferences for information, which can be used to infer an unobservable thinking process. We can also measure the effect of learning on information processing to see how subjects learn, and whether they learn optimal responses or general principles (e.g., solution concepts). The games and ideas that will be studied in this project include: sequencing bargaining, forward induction, Nash vs. subgame perfection, and "almost common knowledge" games.

This is a collaborative proposal with Dr. Ho (SBR-9511137). The PIs propose to study iterated rationality in the context of economic game theory. Iterated rationality refers to the belief that others are rational, that others believe that you are rational, that you believe that others believe that you are rational, etc. Iterated rationality is a central assumption of game-theoretic analyses that have been applied to a large number of phenomena in economics, psychology, and other social sciences. Since the empirical basis for this assumption is weak, the PIs plan to examine it in several ways. First, they will use a numbers game in which the estimate of a winning number by a contestant reveals to what extent that contestant believes in iterated rationality. Second, they will use variations of the numbers game in order to determine which changes in the procedure foster more rational choices. Such procedural changes will provide clues as to subjects assumptions about the rationality of their opponents. Third, the PIs will examine participants with various characteristics, such as group homogeneity or experience with the numbers game, in order to determine what factors influence participants rationality. Finally the PIs will examine whether what a participant learns about rationality in one game transfers to a different game.

In strategic situations, people, firms, or nations care about what others are likely to do. Examples of such situations include bargaining, business decisions which require coordinating various activities, or 'signaling games' in which the actions people take signal something about their abilities or intentions. In the last few decades, a large body of mathematical 'game theory' has developed about how people will make choices in these situations. However, these theories generally assume that people know or can figure out how other people in the strategic situation are likely to behave. In fact, people usually figure out what others are likely to do by learning from experience. Our project proposes a general theory of how this learning occurs. The theory combines two very different forces - 'reinforcement,' which means that successful strategies will be repeated, and 'belief learning,' which means that players keep track of what other people have done to figure out what those people will do in the future, then they choose strategies which will give the biggest payoff if their guesses are right. These different types of learning were thought to be different for about 50 years. In earlier NSF-funded research, we discovered that the two theories are actually special kinds of a single kind of learning, 'experience weighted attraction' (EWA) learning. The current research proposes to extend the EWA theory in three ways -- to incorporate the obvious fact that different people may learn in different ways; to extend the theory to cases where people aren't sure what the payoffs from different choices are (which is of course more realistic); and to allow the possibility that people realize, as they learn about what their opponents do, that their opponents are learning also. When the extensions are incorporated we will have a very general theory of learning which can explain the way people bargain, coordinate, and signal to each other changes over time in response to experience.

Camerer #0078911


Our research addresses how people and organizations learn from experience
in strategic situations like bargaining, coordinating joint actions
(teamwork), choosing prices and features for new products, bidding in
auctions, etc. In previous research we discovered a mathematical formula
which explains how people appear to learn from experience, but the
numerical details of the formula (its "parameters) seem to vary from
situation to situation, as if people are learning in different ways. We
therefore propose to explore why these parameters seem to vary. In
addition, most mathematical theories of strategic learning assume that
people only look back at past experiences. We also propose to extend these
theories to allow for people who realize that other people are learning
from experience, and are able to therefore outguess what others will do
based on what happened in the past. If players are "sophisticated", in
this sense, it pays for them to take actions that are not perfect in the
near-term, to "teach" other players who are learning to take actions which
will benefit the "teachers" in the long-term. This teaching can be
beneficial for the teacher but bad for society (e.g., when firms scare
away innovative competitors by threatening illegal retaliation), or
beneficial for everyone (e.g., when firms reassure others that they can be
trusted). Our research develops a precise mathematical theory of how
sophisticated players behave and when it pays for them to teach. We use
the theory to explain observations from experiments and predicts whether
teaching will occur in new situations.

Organizational culture is a familiar concept to both researchers and
managers. It is frequently cited as the source of a firm's efficiency or as
the cause of problems in firm adaptation or mergers. However, in spite of
its familiarity, it has been difficult to capture exactly what
organizational culture is and how it affects firms. We propose a series of
studies that use an experimental procedure to create something very similar
to culture in the laboratory. The basic elements of culture, based on
common threads in the research literature, are that it is something that
arises through shared history and understanding between members of an
organization, that it depends on the organization's history and can
therefore vary greatly between firms, and that it allows members of a firm
to coordinate activity and therefore perform more efficiently.

Our experiments use a task that allows "firms" consisting of two or more
subjects to develop culture through repeated interaction. In our
experiments, language serves as a metaphor for culture - subjects need to
develop a way to refer to unknown and complex objects using simple, short
phrases in order to perform a task quickly. These "cultures" that subjects
develop end up being based on the group's shared history in performing the
task (what aspects of the objects did they focus on initially?), they end up
being idiosyncratic and varying greatly between groups (as an example: one
group came to refer to an object as "Macarena" while another group referred
to the same object as "coffee cups"), and they end up allowing the groups to
coordinate and improve efficiency (the one word descriptions allowed groups
to jointly identify objects quickly).

Our initial experiments investigate what happens when two groups that have
developed culture independently - and have become efficient at performing
the task - are merged. Not surprisingly, differences in cultures lead to
decreases in the efficiency with which the merged group performs the task.
In addition, subjects are not aware of how difficult it will be to integrate
even these simple cultures, leading them to overvalue the merged firm.
Finally, once culture conflict arises, subjects blame the source of the
failures on incompetence on the part of other subjects rather than on the
difficulty in integrating different cultures. Future experiments will
further explore the last phenomenon by more closely examining the extent to
which subjects place the blame for merger failure on others rather than on
culture incompatibility. One implication is that managers may fire
employees too frequently, blaming low performance on their incompetence
rather than on the need for cultural integration. We will also explore how
cultural integration is related to whether members of an organization have
common or opposed incentives.

The investigators will combine approaches from economics, decision science, and cognitive neuroscience. The goal is to develop a system to measure how well different people reason in strategic social situations, and to explore how specific parts of the brain may be involved in strategic decision making.

The investigators plan to develop a measure to summarize how specific individuals perform in a set of strategic tasks conducted in a laboratory environment. They will administer these tasks -- eg, test "strategic IQ" --to individuals in separate groups with distinct characteristics. Three groups are made up of people who we might expect to be especially good at strategic thinking: mathematically gifted undergraduate students, students trained in the formal analysis of strategic thinking (game theory), and experienced business managers. The fifth group are individuals with specialized brain lesions that have specific and well-documented effects on cognition.

Data from all groups will be combined to test a new theory developed by two of the PIs about how people learn to think strategically. An additional study will involve only members of the fifth group. These individuals will also complete strategic tasks while undergoing fMRI brain imaging scans. The goal is to link evidence about the "games" that these participants play poorly with evidence from brain images, to link specific kinds of errors to specific brain circuits.

This Integrative Graduate Education and Research Traineeship (IGERT) award supports the development of a multidisciplinary graduate training program in Brain, Mind, and Society. Its purpose is to provide students with the analytical foundations and the experimental skills needed to pursue scientific careers at the intersection of neuroscience and the social sciences, who are capable of integrating neural, psychological, and economic approaches to attack basic and applied problems related to valuation, human decision making, and social exchange. Trainees will take a rigorously designed, largely team-taught course sequence, spanning from nervous system organization and function to mathematical models of decision making and social exchange. This coursework will be complemented by equal balance in cross-disciplinary laboratory research, thereby tightly integrating research training with scholarship to create true intellectual hybrids across both disciplines.
The Brain, Mind, and Society program emphasizes the inclusion of highly qualified underrepresented students through a four-tiered outreach program, Science Matters, involving a team-based mentorship program bringing together students from the this program, underrepresented undergraduate students at Cal State University, Los Angeles and underrepresented high school students in Los Angeles' Belmont Schools. The resulting diversity of the program's collaborative teams will reflect the program's broader impact in five key social application areas, which may ultimately provide a new scientifically-enriched discourse to help us understand critical social problems, in economic, therapeutic, educational, philosophical, and business and political applications. IGERT is an NSF-wide program intended to meet the challenges of educating U.S. Ph.D. scientists and engineers with the interdisciplinary background, deep knowledge in a chosen discipline, and the technical, professional, and personal skills needed for the career demands of the future. The program is intended to catalyze a cultural change in graduate education by establishing innovative new models for graduate education and training in a fertile environment for collaborative research that transcends traditional disciplinary boundaries.

"This award is funded under the American Recovery and Reinvestment Act of 2009
(Public Law 111-5)."


Firms, governments, and scientists are constantly striving to figure out what goods and services people want most and quantify their worth. The main technique used to do this in economics is to statistically analyze data on what people have bought before, in order to guess how purchases might change if people have more to spend, if prices rise or fall, or if similar products are introduced. However, this method is of limited use in forecasting the value of brand new products, and goods that are not traded in markets (particularly public goods which benefit everyone, such as clean air.)

When people are choosing, a complex cognitive and biological process underlies those choices. The proposed research uses empirical tools from cognitive neuroscience to measure aspects of these processes when experimental subjects make actual choices of consumer goods. Measures will include brain imaging, eyetracking of visual attention, and speed of responses. The goal is to use these measures to infer what people value, in order to understand the neural foundation of choice and to predict actual choices more accurately than other measures can.

In this project the Principal Investigators will develop methods to rapidly personalize the series of questions that are asked in an experiment or survey, adjusting new questions based on a respondent?s previous answers and the speed of their responses. The methods guarantee, mathematically, that the most informative question is being asked at each step. The software challenge is being able to recompute rapidly on a variety of computer platforms. We will also make software user-friendly and widely available.

In terms of broader impacts, this research will enable scientists to figure out, faster and better, how people are different from each other in what they like and how they behave. Our method will enable creation of short surveys that yield more insight. The method will specifically be used to find out how much risk people accept, how patient they are toward future costs and benefits, and how much they like to cooperate and compete with other people. These methods can help consumers and companies figure out what kind of investments they should make, and help match workers with jobs that are ideal for them.

This project will focus on identifying fundamental attributes of economic decision making, a set of measurements that the researchers call "econographics. The project also will focus on measuring econographics in countries throughout the world and within the U.S. over time, and as well as examining how these attributes relate to popular attitudes and participation in governmental processes. The researchers will conduct an incentivized survey that will measure econographics across thirteen countries, and over three times within the U.S. The countries selected to survey include many different types of governmental and economic environments, including parliamentary democracies with high redistribution, federal systems with a large informal economies, transitional democracies, developing democracies and authoritarian states. These states also have a variety of economic systems. Through their measurement of econographics across a broad spectrum of countries, the researchers will test hypotheses related to the processes through which these fundamental attributes correspond with the political environment and with economic performance. The three survey waves within the U.S. will straddle a national election to examine how major national events, which are known to change popular attitudes, affect econographics.

This project will provide new metrics to assess fundamentals of economic culture and decision making, and it will facilitate new modes of research to link behavioral economics with other social and behavioral science fields. The project will provide new insights related to the conduct of large-scale, cross-country, incentivized surveys, thereby facilitating future research based on empirical tests on representative samples of the population rather than on subjects in a social science laboratory. The project will provide valuable education and training opportunities for post-doctoral researchers and graduate students, and project results will provide valuable information and insights to assist policy makers and others in determining the most effective ways to attain socially desirable goals. This project is supported through the NSF Interdisciplinary Behavioral and Social Sciences Research (IBSS) competition.

The research team includes researchers at Caltech and Virginia Tech with expertise in Economics and Cognitive Neuroscience. They will study how the human brain works when faced with the tasks that are part of buying and selling in asset markets. They plan to use lab experiments with market designs that encourage the formation of market bubbles, periods where the price paid for an asset is well above the actual value of the asset. They want to determine whether traders who buy at high "bubble" prices are systematically different than other more cautious buyers. They will use brain imaging techniques and analyze the resulting data to test whether neural activity can predict how large a bubble will become and how long it will last. Because market bubbles can have serious consequences on the broader economy, understanding more about the possible causes of bubbles is important for financial market regulation.

The project studies the behavioral ecology of trader types, predictive 'decoding' of when bubbles will form and crash, and neural activity during bubbles. For economics, the project contributes to our scientific understanding of bubble dynamics by helping us understand why and how people participate in bubbles. For cognitive neuroscience, studying asset prices is one way to advance the science of understanding how the brain computes a complex dynamic value that changes over time depending on the reactions of others. The combination of behavioral observation and neural measures will give us data that will help us better understand how emotions such as 'irrational exuberance', social influences like 'herd behavior', and momentum trading all affect asset markets.

During the COVID-19 pandemic, people were forced to try new routine at school, work, and home ? ?forced exploration?. That is, habits used to guide how we work, conduct our daily routine, exercise, eat, go to school, and interact with friends and neighbors, Habits are formed to make the same routines effortless, to save time and mental effort, when routine choices work well. However, while the mental savings from habits are a benefit, when people are choosing habitually, they may not be exploring other options which could be even better? that is the hidden cost of habit. Forced exploration can actually be beneficial if it shows people better ways of school, work, and social interaction. This kind of exploration is like going to your favorite restaurant and finding out they have run out of your favorite dish ? now you have to try something new, which you would not have explored without the disruption. This wave of forced exploration raises important questions: What new habits are formed that will persist? what will be the ?new normal?? Consider, for example, wearing a face-mask outside of the house. This is exactly the kind of ?muscle memory? behavior that usually habitizes? it can be triggered while stepping out of your car, or entering a store, and quickly becomes automatic and effortless. Whether a lot of other people are wearing masks or not can also be a trigger that prompts habit (in either direction).The same question arises across the board: Will people go back to movie theaters (or stay home with streaming)? Will restaurants fully reopen or will home delivery take over? Will knowledge firms switch to more remote ?tele-work?? Will schools find better mixtures of home learning and in-school activity? This project will analyze two different kinds of big data to test whether or not this kind of forced exploration really does result in new habits.

In social sciences, habits are usually modelled mathematically using a simple equation in which the more an activity has been done in the past, the more it is done in the future. This is called a "reduced form" approach because it reduces a biologically complicated mechanism to something much simpler. It is a good starting point but cannot answer questions such as "What if past behavior is disrupted?? This research project uses a new approach to habits based on animal learning and human cognitive neuroscience. The starting point is that habits have developed to save effort ?? both physical and mental. The ?neural autopilot? framework proposed here predicts that individuals develop habits for actions which, after repeated decisions, have proven to be reliably rewarding. Such habitual behavior drains fewer physical and mental resources. At the same time, when people are habitized?- about exercise, eating, or work ? they ignore new goods and activities they would prefer if they actually tried them. While the neural autopilot approach has been tested in many lab studies of animal and human habituation, it has never been systematically explored using a large amount of data about how people actually behave in everyday life. An ideal test of this model is in a field setting where choice sets are artificially truncated, so people resort to new choices; and that is exactly what happened during the ongoing lockdowns. This project will use data from Weibo chat data and Fitbit fitness and sleep tracking. These large sets of data contain fine-grained measurements of behavior. Using this data, we will develop and test a statistical neural autopilot model, to recover values for the model?s main parameters. The parameters are numbers that measure, for each person, how fast habits are formed and the threshold to break out of a habit and explore something that might be better. The estimated parameter values will be used to make predictions about which habits acquired during the pandemic will persist, and which behavior will revert to pre-pandemic patterns.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.