2

Development of Behavioral Economics

Economists have long been interested in human behavior, but it was not until the second half of the 20th century that they began to systematically integrate ideas from psychology—particularly growing understanding of cognitive frameworks—into their work. The field that emerged, behavioral economics, challenges some of the assumptions of traditional economics and seeks to use detailed understanding of the social and cognitive aspects of decision making in the design of policy strategies.

The development of behavioral economics as a distinct field is just one of many important developments in economics in recent decades. Rapid gains in computing power and the availability of new kinds of data, along with new methodologies for analyzing data, have contributed to important advances in most social science fields, and economics is no exception.¹ Economists have also increasingly extended the application of their methods beyond traditional areas, such as the analysis of financial markets, to support the development of public policies to address a wide range of issues, including environmental protection, public education, and poverty. Behavioral economics has a complex history, and many different scholars and approaches have contributed to its development. This chapter provides an overview of the history and development of behavioral economics; key ideas, particularly from the domain of psychology, that have been

___________________

¹ Two examples are applied microeconomics, the study of such decisions as people’s choice of one alternative over another, and econometrics, the application of statistical methods to the analysis of economic data.

influential in the field; its growing influence; and critiques and challenges for the future.

ORIGINS OF THE FIELD

Behavioral economics draws on insights from social and cognitive psychology, sociology, and neuroscience to challenge some of the assumptions made in traditional economic analysis (see Box 2-1). The term behavioral economics was first used in the 1940s, though it has never had a precise consensus definition (Svorenčík & Truc, 2022).² The name has its roots in early work by psychologists such as B. F. Skinner, who studied the role of conditioning and reinforcement in shaping human behavior. The idea that psychological and sociological factors are crucial for economics can be traced back to the early days of economic thought, with roots in the work of Adam Smith and other early economists who recognized that human behavior often does not follow the path that rational analysis would predict (Ashraf, Camerer, & Loewenstein, 2005).

Among other precursors of behavioral economics was the work of Herbert Simon, a researcher working in the 1950s and 1960s at the intersection of computer science, organizational science, and economics. He

___________________

² The chapter draws on the paper by Andrej Svorenčík and Alexandre Truc that was commissioned for this study, available at https://nap.nationalacademies.org/resource/26874/BE_history_20221009.pdf

advocated for the concept of satisficing, accepting an option as acceptable for the circumstances, as an alternative to the concept of maximizing, the traditional assumption used by economists. His concept was intended to capture the reality that people’s rationality is limited, or bounded (see, e.g., Simon, 1955). But Simon himself noted that a number of prominent economists, some who worked far before his time—including John R. Commons, George Katona, Joseph Schumpeter, and Thorstein Veblen—had already introduced behavioral economics concepts.

Drawing on Psychological Theory

While the concepts associated with behavioral economics have been influenced by many disciplines, including sociology, research in psychology is the source for the primary ideas in the field. Cognitive psychologists, particularly Daniel Kahneman and Amos Tversky, contributed concepts that have been critical to the development of behavioral economics. They systematically documented how heuristics (short-cut problem-solving strategies) and biases influence people’s perceptions of probability and their decisions (see, e.g., Tversky & Kahneman, 1973). Equally influential was their concept of prospect theory, which proposed an explanation of individual decision making that takes into account comparative judgments, framing, and reference dependence (Kahneman & Tversky, 1979).^3,4 Indeed, such was the influence of prospect theory that the publication in which it was introduced is one of the most cited articles in economics in the past 50 years (Kim, Morse, & Zingales, 2006).

This work laid the foundation for the collaboration of psychologists with economists that created behavioral economics as a distinct field. Economist Richard Thaler worked with Kahneman, Tversky, and others to incorporate insights from cognitive and social psychology into economics. Among other contributions, Thaler (1981) pointed to the importance of evidence about present bias (the tendency to give greater weight to near-term risks or benefits than to more temporally distant ones) from the field

___________________

³ Prospect theory posits that individuals compare the expected outcomes of decisions they have to make with the expected outcomes for an alternative possible decision, called a reference point. Prospect theory assumes that the reference point is the default choice—the one that would seem easiest or most obvious—but reference dependence can refer to any situation where a decision is weighed against a hypothetical alternative choice. The framing of alternative decisions (the way they appear to the decision maker) also plays a role in other areas: framing may determine whether the possible options or outcomes are perceived as an overall loss or gain relative to the reference alternative.

⁴ Reference dependence is the tendency to make decisions by comparing the options to a single possible case, such as considering whether an outcome would be better or worse than the current status quo, rather than objectively assessing a range of possible outcomes.

of psychology. A set of articles that stimulated growth in the field examined the impact of people’s taste for fairness in the setting of prices and wages, and the impact of prospect theory on trading (Kahneman, Knetsch, & Thaler, 1986).⁵ The authors suggested the existence of an endowment effect, the idea that people place a higher value on something when they have it than when they do not (i.e., sellers value goods at higher rates than buyers; Kahneman, Knetsch, & Thaler, 1986).⁶ This early work set the stage for several of the key themes in behavioral economics discussed in Chapter 3.

A growing number of economists drew on work from psychology as they developed new theories to explain how people make decisions under uncertainty and how biases can influence individual decision making. During this period, concepts and findings from psychology were incorporated by economists into formal theoretical models of decision making and tested using observational and experimental methods. Early examples focused on patterns of behavior in individual decision making under uncertainty and over time, and the ways in which the framing of the available options affects decisions.

A later wave of research focused on field data and extended the work on behavioral biases to two main areas: the study of financial markets (by scholars such as Richard Thaler, Robert Shiller, and Andrei Shleifer; see, e.g., Thaler, 1993) and the study of consumption and savings by David Laibson and others (see Ericson & Laibson, 2019). These empirical projects with field data were very important to the success of behavioral economics. Equally important, at a time when behavioral economics was not taught in most universities, was the establishment of training opportunities for economists interested in these ideas, especially a series of behavioral summer camps initiated in 1994 that offered two weeks of training by leading researchers to interested Ph.D. students from across the globe. The Russell Sage Foundation played an important role in supporting this summer school and other initiatives in the area.

Learning from Laboratory Experiments

While one set of insights for behavioral economics came from cognitive psychology, another source for understanding the significance of human

___________________

⁵ Taste for fairness refers to the fact that “a wealth of experimental data corroborates the everyday observation that people care about, and respond to, fair and unfair treatment. Individuals exhibit both an aversion to arbitrary inequality and a preference for inequality based on desert” (Fennell & McAdams, 2013, p. 1).

⁶ Other contributors to early work in behavioral economics in the 1980s and early 1990s include, among others, George Akerlof, James Andreoni, Colin Camerer, Catherine Eckel, Ernst Fehr, Robert Frank, Elizabeth Hoffman, George Loewenstein, Kevin McCabe, Matthew Rabin, Tom Schelling, Andrew Schotter, Robert Shiller, and Robert Sugden.

psychology for economic analysis emerged from the work of experimental economists. The insights of Vernon Smith, Charles Plott, and Reinhard Selten, among others, as well as Elinor Ostrom, a political scientist who worked on culturally embedded policy solutions, played a key role in these developments (Grether & Plott, 1979; Isaac, McCue, & Plott, 1985; Smith, 1989, 2003; Ostrom, 1990; Selten, 1998). What distinguished these scholars was their interest in how individual behavior produced outcomes in specific institutional contexts.⁷

Experimental economists examining the behavior of individuals and groups in controlled conditions were confronted by the complexity of human motivation in the course of their attempts to test formal game theory in laboratory experiments. Unexpected findings emerged repeatedly and became well documented (e.g., Güth, Schmittberger, & Schwarze, 1982; Isaac, McCue, & Plott, 1985).⁸ That is, subjects’ responses systematically diverged from game-theoretic predictions: they were clearly sensitive not only to their own payoffs but also to payoffs for others. Their behavior diverged from the predictions in several ways, showing that people were generous, trusting of, and trustworthy toward other people; cared about relative earnings; and were responsive to the efficiency effects of their choices. A growing body of evidence demonstrated that economic agents responded differently to incentives and information than was usually predicted by traditional economists’ models.

The junction between this work and the work in psychological theory discussed above led to the modeling of such phenomena as social preferences (the way in which people take account of the utility of others, such as the taste for fairness or inequity aversion; Fehr & Schmidt, 1999; Bolton & Ockenfels, 2000; Charness & Rabin, 2002). This work fostered the development of behavioral game theory, the study of how people make decisions in strategic situations, such as in business negotiations or political elections,

___________________

⁷ Smith argued that individually “rational” behavior emerged when it was most valuable—for example, when individuals are subject to the discipline of the market and other competitive situations—while cooperation is favored when it can increase the size of the resource pie: that is, the key to understanding behavior is an understanding of the institutional environment. Plott focused on nonmarket decision making and the development of institutions for solving collective action problems. Selten stressed the importance of observing regularities in subject behavior in lab settings, such as whether and how groups achieved equilibrium play, and using those to modify theoretical models. Ostrom documented the ways in which local public goods and commons problems were solved with cultural contexts by the development of local institutions.

⁸ Richard Thaler hosted a column in the Journal of Economic Perspectives for many years, titled “Anomalies.” Written with a variety of cohosts, these columns introduced economists to “behavioral” phenomena. While many of the columns focused on heuristics and biases, a number of them featured the cooperative anomalies arising from experimental studies (Dawes & Thaler, 1988; Thaler, 1988; Camerer & Thaler, 1995).

and how these decisions are influenced by such factors as emotions, biases, and social norms (Camerer, 2011). Behavioral game theory thus incorporates the idea that people often make decisions based on factors other than economic self-interest, such as fairness, trust, and reciprocity. These ideas had roots in sociology and earlier economic theory but had not generally been explicitly factored into economic models (Homans, 1961; Blau, 1964).

This work further made the case that it is important to more explicitly take account of human cognition and behavior in research and, ultimately, policy making. It also led to the investigation of mechanisms to solve public goods problems and other market failures and to the development of new theoretical models of behavior and tests using observational and experimental methods.

Behavioral economics has benefited from the confluence of the research from these two origins: the analysis of heuristics and biases and the experimental analysis of behavior in specific contexts (particularly contexts of interest to policy makers).⁹ While these two branches in behavioral economics also continue to work independently, together they have fostered appreciation of the importance of more psychologically accurate models of decision making. Some researchers primarily consider individual-level behavior and policies, using laboratory experiments, formal theory, observational data, and field experiments; others focus on lab experiments to test mechanisms for fundraising, public goods provision, and market design (including kidney exchange and medical-intern matching “markets”). Behavioral economics has achieved considerable success and influence in economics; the awarding of two Nobel Memorial Prizes for behavioral economics research and experimental economics highlighted the field’s growing influence.¹⁰

By taking the utility-maximizing model typically used by traditional economists as a baseline and building related models that incorporate behavioral elements, and by adopting state-of-the-art empirical methods to obtain field evidence (such as natural experiments and randomized controlled trials), behaviorally inclined economists have generated work that

___________________

⁹ The heuristics and biases research has largely been funded by the Sloan and Russell Sage foundations, while the experimental economics agenda has largely been funded by the National Science Foundation. The National Institutes of Health has supported work on applications of behavioral economics to aging and health. In all these cases, key personnel at the relevant organizations were instrumental in supporting and fostering the emergence and growth of behavioral economics.

¹⁰ In 2002, Daniel Kahneman and Vernon Smith were each awarded a Nobel Memorial Prize: Kahneman for his work on behavioral biases, and Smith for his work on how institutions shape behavior. In 2017, Richard Thaler received the prize for his pioneering and wide-ranging contributions to behavioral economics. In addition, Herbert Simon, one of the early contributors to the field, received the prize in 1978.

has become widely accepted. These scholars succeeded in making psychological research a fully established source of economic knowledge.

INFLUENTIAL BEHAVIORAL IDEAS

A wide range of psychological findings have been influential in the development of behavioral economics. That mass of evidence covers human emotions, cognitive processes, and behavior, as well the complexity of social interactions. In particular, cognitive psychologists and neuroscientists have produced an ever more detailed picture of the complex functioning of the human mind, while social psychologists and sociologists have focused on the complex ways that environment and context influence thinking, emotions, and behavior. By providing empirical evidence that people make judgmental errors, exhibit inconsistent preferences, are often overconfident, have a taste for immediate gratification, and are influenced in complex ways by social norms and contexts, among other findings, these fields have been invaluable to the economists seeking to refine their models and methods for understanding and influencing behavior. The field of economics has been influenced by many developments in the behavioral sciences, such as work on learning and new approaches to surveying both individuals and firms about their actions and thinking, but behavioral economists have been particularly focused on several core ideas. An acquaintance with some of these ideas is an important foundation for understanding the core principles of behavioral economics, which are discussed in Chapter 3.

Research in cognitive psychology and neuroscience has shown that human decision making is influenced by fundamental cognitive processes, such as perception, attention, and memory, and their significant limitations. Research in social psychology and sociology has also contributed to understanding that decisions are highly contextual, shaped by multiple aspects of the circumstances in which a decision is made and past experiences that affect how each individual construes the world. These two core sets of ideas—the multiple cognitive influences on decision making and the influence of context on how decisions are represented—have particular relevance for thinking about decision making and public policy. The descriptions below are designed to provide a flavor of the research and ideas that have been influential.

Cognitive Processes That Affect Decision Making

Theoretical and empirical research in cognitive psychology, neuroscience, and the newer field of neuroeconomics has fueled increasingly detailed insights about human cognition and decision making. A thorough review of this vast research landscape would be beyond the scope of this report, but

three fundamental classes of cognitive processes stand out as particularly relevant: perception, attention, and memory. This work has demonstrated that these three processes have important influences on decision making across individuals and in a wide range of domains.

The Nature of Perception

The way people perceive the world around them involves input from their senses, but this information is filtered and shaped by past experiences, beliefs, motivation, attention, and other factors. That is, people’s day-to-day experiences over the course of a lifetime, along with their attitudes, perspectives, and other factors, shape (or bias) how they respond to information. The word bias has a negative connotation in some contexts, but humans naturally take into account past knowledge and experience when processing new information; in the context of cognitive science, the word refers to the fact that the brain necessarily filters information (National Academies of Sciences, Engineering, and Medicine [National Academies], 2018). Indeed, this capacity to filter is the basis for the development of knowledge and expertise. Yet the influence of past experience on both perceptions and preferences can lead to distortions in perception and even in how neurons encode information. These influences may distort even seemingly straightforward perceptions, such as visual interpretations of an image (Sun & Perona, 1998). And these distortions can lead to inaccurate inferences that affect people’s decisions.

An example of this distortion comes from neuroscientists who measure the activity of neurons. They have proposed the efficient coding hypothesis, which posits that neural representations adapt to the statistics of sensory input available in the environment so that perception is more precise for stimuli that occur more frequently (Barlow, 1961; Laughlin, 1981). The idea of efficient coding is also relevant for how individuals learn and make decisions. In both learning and decision making, a critical problem is how agents internally represent information. In the domain of learning, there is evidence that individuals adopt a hierarchical scheme for encoding actions, similar to an efficient coding strategy (Botvinick et al., 2015). Researchers have also suggested that efficient coding extends to the domain of risky choice and could explain why risk-taking behavior varies across different environments (Frydman & Jin, 2022).

The key idea is that efficient coding results in a subjective value function (an individual’s estimation of the value of possible rewards in a given situation) that is more sensitive to frequently occurring payoffs and less sensitive for infrequent payoffs: that is, the function is steeper for frequently occurring payoffs and flatter for infrequent payoffs. Similar arguments about perceptual biases have been used to explain small-stakes risk aversion (Khaw, Li, & Woodford, 2021). Perceptual biases can be either positive or

negative. People must and do learn over time to distinguish what is most important among the many stimuli they encounter. In the context of the kinds of decision making that behavioral economists study, it is important to understand that people’s existing knowledge and experience lead them to recognize and process information and apply interpretations to what they perceive in unique ways that may differ from what might be expected in a given circumstance (National Academies, 2018).

The Nature of Attention

In making decisions, people need to process multiple pieces of information. For example, making a binary decision between two options in a gambling situation requires attention to the potential value of each option as well as the probability of each possible outcome. When people are faced with multiple pieces of information, they cannot allocate attention to all of them at the same time. Researchers who study attention—using such methods as tracking the eye movements of participants as they view stimuli—have suggested that people allocate attention to subsets of the available information and develop preferences for the aspects of the information to which they are attending (e.g., Krajbich, Armel, & Rangel, 2010); see Box 2-2. As they deliberate, people shift this “window of attention” to evaluate each of the alternatives they face. In this way, attention acts as a spotlight, enhancing processing within a certain area (similar to a zoom lens; Posner, 1980; Carrasco, 2011; Logan et al., 2021).

In decision making, people tend to choose options that they have looked at longer, a phenomenon known as attention bias (Fiedler & Glöckner, 2012; Gluth et al., 2020). Gaze (a common measure of visual attention) is hypothesized to both reflect and influence preferences (Shimojo et al., 2003). This phenomenon has been demonstrated in varied settings, such as food decisions, risky monetary decisions, and social decisions (e.g., Krajbich, Armel, & Rangel, 2010; Smith & Krajbich, 2018, 2019).

Numerous factors may influence how people allocate their attention. A decision maker may prioritize processing of information that is of highest relevance to particular goals, such as gaining the most reward in the least amount of time. For example, one study has shown that visual fixations and choices in simple choice tasks approximate an optimal information sampling policy to maximize rewards and minimize costs when taking account of environmental constraints (Callaway, Rangel, & Griffiths, 2021). In this study, the researchers assumed that values are uncertain and need to be estimated through a sampling process and that the goal of the decision maker is to select the highest-value item within a reasonable amount of time. This idea is closely related to research on rational inattention in economics, which examines the hypothesis that as people allocate effort to

extracting pieces of information they factor in the costs of acquiring that information (Sims, 2003; Caplin & Dean, 2015). If some information is costly (in time or money) to obtain, then people may not consider all available information. This can result in decisions that might appear irrational by some standards but are optimal if the costs of information are taken into account. Rational inattention models are discussed further in Chapter 3.

Contextual factors can also affect attention and, consequently, decision making (Trueblood, 2022). For example, decisions involving multiple options that have different features (e.g., consumer goods that vary in both price and quality) are often sensitive to the context created by the choice set (Tversky, 1972; Huber, Payne, & Puto, 1982; Simonson, 1989). This is best illustrated by the attraction effect, which was first discovered in research on marketing (Huber, Payne, & Puto, 1982, 2014) and subsequently studied in marketing and other domains including psychology, neuroscience, and economics (Busemeyer et al., 2019). This effect occurs when a clearly inferior (decoy) option is added to a choice set and boosts preference for a similar,

but slightly superior (target) option over other alternatives. Eye-tracking studies have shown that this boost in preference is related to increased visual attention toward the decoy and target options (Marini, Ansani, & Paglieri, 2020; Trueblood, 2022). The attention bias is significant because it reveals how simply attending to particular information can influence a decision.

The Nature of Memory

Memory—the capacity to store and retrieve knowledge and information—also plays an important role in decision making. But memory is a complex process. It is not the case that stable memories are stored and can be retrieved in identical form whenever they are needed. Rather, people reconstruct past experiences each time they are remembered and identify new connections among them and with new experiences (National Academies, 2018). That is, many factors, including a person’s emotional state, may influence which memories are recalled in a given moment. In addition, memory has subsequent consequences on decision making.

For example, it is hypothesized that during deliberation, people sample arguments for various alternatives from memory (Johnson, Häubl, & Keinan, 2007; Appelt, Hardisty, & Weber, 2011; Weber & Johnson, 2011). This memory sampling process can be biased so that arguments supporting some options are sampled before thoughts supporting alternative options, thus biasing the deliberation process and ultimately choices. Empirical studies have also shown that decision makers prefer options they can remember better (Kraemer et al., 2022). In some cases, people are willing to accept relatively bad options when they remember them well (Gluth et al., 2015). Thus, the memories that are brought to bear when an individual is making a decision might or might not be the ones that would best support a purely rational decision-making process.

This brief overview of the role of perception, attention, and memory in decision making cannot do justice to the sophistication of ongoing work in psychology, neuroscience, and neuroeconomics, but the research highlighted here is especially relevant for understanding of decision making; see Box 2-2 for a discussion of modeling approaches.

The Relevance of Context

Social psychologists and sociologists have explored the concentric rings of influence on how people think, learn, behave, and feel about their world, from family and friends, to school and work environments, to geographical location, and outward to the sociopolitical context. This work has produced a significant body of evidence that the context in which decisions

are made and how people interact is critical. Research on the influence of context and culture on cognition and behavior has spanned decades and disciplines and yielded an increasingly nuanced view of the influence of the physical and social environment on development, cognition, behavior, and even biological systems (National Academies, 2018). Research on learning has demonstrated that an individual’s cultural and social contexts interact with cognitive and biological processes throughout the lifespan to shape thinking, memory, and learning—all of which have implications for decision making. This important work from other disciplines has had a major impact on behavioral economics.

For understanding behavioral economics, there are two key points: (1) any time an individual makes a decision, cognitive processes are influenced not only by the cognitive biases discussed above, but also by factors that reflect cultural and other influences from the decision maker’s environment; and (2) the specific context in which the decision is made may have a strong influence on the decision maker’s choices. The context for a decision includes the specific combination of options in a choice set, whether the context is a lab experiment, a supermarket shelf, a place (a bank, workplace, school, or doctor’s office), or another circumstance of daily life. Contexts might also include societal norms and expectations; temporal or financial concerns; the salience of particular personal identities; the structure of relationships among people; and the cultural, linguistic, and institutional setting of the decision maker. It is the incorporation of the interaction of cognitive and noncognitive forces with the surrounding environment, and the attention individuals pay to the detailed context in which they are making decisions (called framing), that make behavioral economics models different from those usually developed in other disciplines.

Myriad Contextual Influences

Behavioral research has shown that humans are deeply influenced by their immediate contexts, including the time and place, the social and economic environment, and the relationships they have with other people at the time. This context can shape many aspects of life, including perceptions, emotions, reasoning, choices, and judgments. For example, preferences are highly malleable and can be shaped by context in ways that people may not even be aware of. As a result, seemingly minor contextual factors can have a significant impact on behavior, even when it does not seem rationally justified. This is often seen in framing effects, where small changes in the way options are presented can lead to widely different perceptions, reactions, and choices (see below).

Other contextual factors that affect decisions include culturally transmitted concern about fairness and societal norms, the tendency to defer

difficult decisions, and even attitudes toward gains and losses. Fairness and social norms have appeared extensively in the work of economists working in traditional frameworks and in the work of scholars in other disciplines, but the interaction of those factors with difficulty in individual decision making and attitudes toward gains and losses has not been emphasized. All of these considerations may be, at least in part, conscious and intentional. Others, however, like those based on automatic or implicit processing, happen entirely out of awareness. For example, job candidates interviewed on rainy days tend to receive lower ratings than people interviewed on sunny days (Schwarz & Clore, 1983). This bias extends to admission interviews at large medical schools, where the difference in scores accounted for by the weather was equivalent to approximately a 10 percent lower total score on the Medical College Admission Test (Redelmeier & Baxter, 2009).

Personal identity (both one’s own and one’s perception of others’ identities) is another aspect of context that has strong influence. For example, the concept of “mother” evokes strikingly different values and ideals from those evoked by “CEO.” Identity salience has been shown to affect various behaviors, including resistance to persuasion, reactions to advertisements, and ratings of consumer products (e.g., Forehand, Deshpandé, & Reed, 2002; Reed, 2004; LeBoeuf, Shafir, & Belyavsky Bayuk, 2010).

Evoked identities are those that are brought to the forefront of a person’s thoughts and behavior by a specific context or situation, and these identities often have associated tastes, values, and priorities that can shape decision making. Although people might be expected to make consistent choices, based on well-ordered preferences, evoked identities can influence which concepts and priorities are activated at the time of a specific decision, along with individual tastes and values. The result can be inconsistent, context-dependent preferences associated with the salient identity (e.g., Higgins, Rholes, & Jones, 1977; Bargh, Chen, & Burrows, 1996). One study that illustrates this concept explored the effects of ethnic stereotypes on decisions about saving. The experiment made ethnic identities salient by asking participants to first fill out a background questionnaire that asked which languages were spoken by their family and how many generations their family had lived in the United States. The researchers found that those who filled out the questionnaire before making a decision tended to make choices that aligned with ethnic stereotypes about their own family group (Benjamin, Choi, & Strickland, 2010).

Framing

Psychologists use the concept of construal to understand the influence of context: it is the notion that people’s perceptions are the product of their mental representation or interpretation of themselves, others, or aspects of

the world around them. That is, people react to the world around them in a way that is filtered through their internal representations of the people, situations, and things they observe—and those representations can vary significantly from one person to another. Psychologists describe these different representations as framing. For example, one person might look at a situation in terms of losses versus gains, while another looks at it in terms of lives saved versus lives lost. They are looking at the same outcomes but describing and interpreting them differently.

Framing can profoundly affect decisions. For example, people may be more favorably inclined toward actions that prevent disease when their chances of success (i.e., a gain frame) are highlighted, while highlighting the avoidance of a negative outcome (i.e., a loss frame) may be more effective at promoting screening or detection procedures (Rothman et al., 2006). Similarly, in a very different context, ground beef, which can be described as 75 percent lean or 25 percent fat, tends to be evaluated more favorably when the percentage of lean is provided (Levin, 1987; see also Levin, Schnittjer, & Thee, 1988). And a community may allocate more police resources if it is described as having a 3.7 percent crime rate than if it is described as being 96.3 percent crime free (Quattrone & Tversky, 1988).

This phenomenon has clear relevance to decision making. Traditional economic models generally portray individual decision making as a function of probabilities—the likelihood that different future states of the world will occur if they take different actions. Those probabilities are typically assumed to arise from rational evaluations of past events (learning) and from the information available to an individual. The behavioral economics framework views the problem through a different lens, positing that decisions are not directly the product of rational calculations of objective probabilities of states of the world but rather the product of assessment of the world as the person has mentally represented it. That is, individuals choose not between things in the world but between those things as they are mentally represented. These two views of decision making are not inconsistent—behavioral economists do not discount the role of rational analysis—but they portray the behavior in different terms.

Researchers have found that the way a choice is framed can significantly influence the outcome of the decision. This phenomenon, known as a framing effect, has been studied in a variety of settings. For example, in one field experiment, clients of a money lender in South Africa were sent letters offering short-term loans (averaging one-third of borrowers’ gross monthly income) at randomly determined interest rates. Not surprisingly, those offered higher rates were less likely to take up a loan than those offered lower rates. In addition, however, various peripheral features of the offer letter also had a large effect on take-up. The picture of a smiling woman

randomly placed in the corner of the offer letter had the same positive effect on take-up among male clients as dropping the monthly interest by 4.0 percentage points (Bertrand et al., 2010). Similarly, showing one example of a possible loan (easy to process) rather than four examples (harder to process) had the same positive effect on take-up as dropping the monthly interest by more than two percentage points. In a similar vein, Huberman, Loch, & Önçüler (2004) found that employees’ overall participation in 401(k) plans drops as the number of fund options proposed by their employer increases.

Other studies have examined the effects of weighting the various elements of a set of options. Because people are often uncertain about the relative importance of various dimensions of a choice, the weights assigned to those dimensions are often influenced by relatively immaterial changes in the task, the description, and the nature of the options under consideration. For example, changing the weighting of relevant attributes (e.g., emphasizing the payoff of a gamble rather than the risk) will yield changes in people’s expressed preferences (Slovic, Griffin, & Tversky, 1990). Similarly, attributes that are difficult to evaluate in isolation get weighted more heavily when they can be directly compared (Hsee, 1996; Hsee et al., 1999).

Behavior is the outcome of a variety of social, cognitive, and affective processes that are partly influenced by their context. But local contexts change, and with them the cues and features that affect people’s behaviors also change. As a result of this malleability, people’s decisions and behaviors fluctuate and frequently diverge from what might be expected. As this sampling of the large landscape of work on human cognition and behavior suggests, there is a firm empirical base of understanding about the ways humans process and act on information that can be integrated into the work of behavioral economists.

The research discussed in Chapters 5–10 highlights insights from research in psychology that have been incorporated in economic models and how they have been applied in a variety of policy domains. This body of work can best be understood in the context of a few additional notes about the field’s growing influence and ongoing challenges it faces.

APPLICATIONS AND INFLUENCE

In recent years, behavioral economics has been applied in a variety of fields, including economic development, finance, marketing, and public policy. It is also increasingly represented in university undergraduate and graduate curricula. While only a few colleges and universities taught behavioral economics, experimental economics, or behavioral game theory in the 1980s and 1990s, most major universities now offer such coursework. Behavioral economics approaches have also gained popularity as a tool for policy makers, and the consultants who advise them, to use in designing policies that take account of what is known about human cognition and behavior.

Many people associate behavioral economics with the concept of a “nudge.” Though not precise (and possibly overused), the term nudge has come to be a shorthand for many types of interventions based on behavioral economics.¹¹ One well-known definition comes from a best-selling and widely influential book by Thaler and Sunstein (2008, p. 6):

This concept has been particularly influential through the proliferation of nudge units—centers dedicated to identifying nudges that could help a government agency or institution use behavioral science evidence to meet important policy goals. The United Kingdom’s Behavioural Insights Team is widely recognized as the first government-based nudge unit: indeed, the term was coined in reference to the UK group. Since its founding in 2010, many more nudge units have been created around the world at all levels of government, as well as in corporate entities, nongovernmental and multinational organizations, and health systems. Countries that have a nudge unit include Germany, the Netherlands, Qatar, Greece, Australia, Singapore, Peru, Japan, and Canada, as well as the United States. Several nonprofit organizations provide behavioral analysis on a contract basis. Many focus outside of what are characterized as the western, educated, industrialized,

___________________

¹¹ One frequently cited example of a nudge comes from the study of participation in retirement savings plans. If the default is “no,” so that an employee has to make a change to opt in to an offered benefit, fewer people sign up than if the default is “yes.” In traditional economic analysis, there is no difference in the two situations; behavioral economics has shown that this assumption is not correct. The research on this is the subject of Chapter 6.

rich, and democratic (WEIRD) countries: the Busara Center for Behavioral Economics, the Abdul Latif Jameel Poverty Action Lab, Innovations for Poverty Action, and ideas42, among others, pursue research to improve the generalizability of behavioral economics findings (Henrich, Heine, & Norenzayan, 2010; Berge et al., 2020). Questions about the balance of evidence from countries at different stages of economic development are among the challenges for the field as its influence grows.

Despite the influence of nudge-style policies, many policies motivated by behavioral economics are much more than nudges. As we discuss in Chapter 6, for example, the use of automatic enrollment in pension plans clearly addresses behavioral biases but is not a nudge. Rather, it is a redesign that constrains people’s choices by making enrollment the default so that people must actively decline if they do not want to enroll. Another non-nudge intervention is the U.S. Department of Agriculture’s simplification of application procedures for its food stamp program, which reduced the bureaucratic complexity of the program to make it easier for low-income families to participate. Another example of a non-nudge behaviorally informed intervention is the redesign of tax forms based on findings about salience, which suggests ways to make important elements stand out to individuals (see Chapter 3).

CRITIQUES AND CHALLENGES

The committee was mindful of challenges to and critiques of the field. We kept them in mind as we reviewed the literature and developed our recommendations. We note here the responses behavioral economists have offered to several frequently raised concerns:

• behaviorally based interventions may in some cases reflect an unexamined paternalistic attitude toward the targets of behaviorally based interventions;
• many behavioral economics studies showing positive effects have comparatively small effect sizes;
• interventions that focus on individual-level behavioral change may distract from or undermine policy attention to the need for system-level change;
• behavioral insights may be exploited by nongovernmental private agents, such as in the realms of political influencing and commercial marketing;
• behavioral interventions may have the unintended negative consequence of exacerbating disparities in society; and
• the representativeness in the samples used in behavioral economics research is often lacking.

Paternalism

Critics of behavioral economics have suggested that it is implicitly based on, or conveys, a paternalistic vision of human behavior, in which interactions and choices need to be “corrected” (e.g., Wright & Ginsburg, 2012; Gigerenzer, 2015; McMahon, 2015). They make the point that some behaviors operate below the level of consciousness and that interventions that affect them could have a negative impact on the well-being of people targeted by such interventions.¹² A particular concern for behavioral economics is that decisions based on factors other than reasoning may in many cases be valid and that the idea that people need to be nudged or otherwise induced to make more seemingly rational decisions reflects a narrow understanding of human rationality.

Critics have raised the concern that the use of behavioral economics tools in a public policy context could be regarded as an example of government agencies deciding what people should do instead of letting them decide for themselves. A government agency doing this, it is argued, violates the concept of “consumer sovereignty” in economics, the idea that consumers know what is best for themselves. Moreover, many people, perhaps particularly those who are most vulnerable in terms of education and income level, have a strong trust in government and authority. This circumstance heightens the responsibility of those who develop policies to consider them carefully from an ethical point of view—specifically considering possible outcomes and whether people may be induced to make a choice that does not contribute to their well-being.

Behavioral economists respond in multiple ways to this critique. One frequent response is that all government policies are paternalistic in some sense. Even though traditional economic theory recommends government action only in specified circumstances (monopolies, market failures, pollution), there has always been a rationale for what are called “merit goods,” outcomes that society deems to have merit for their own sake. For example, many of the goals that behavioral economists target for interventions, such as high application rates to college, higher take-up of medicines, or higher retirement savings, are broadly supported and generally not controversial. Behavioral economists have suggested that paternalistic actions to improve these outcomes would have a high rate of social approval. Behavioral economists also argue that a core presumption in their approach is that some behavior reflects biased judgments that reduce people’s well-being (e.g., Bernheim & Taubinsky, 2018). They acknowledge that identifying when judgments are biased is difficult—and, in fact, the heart of the issue. But some behavioral economists argue that there are many behaviors that are

___________________

¹² We use “well-being” as equivalent to the term “welfare” that is used in economics.

transparently not the result of intentional behavior, and that the hypothesis that they may have subtle benefits is implausible. For example, individuals who lose tens or hundreds of thousands of dollars by not enrolling in retirement plans offered to them cannot be plausibly understood to be making a rational choice in doing so.

Nonetheless, there are unquestionably grey areas in making these distinctions. It is important that designers of behavioral interventions, particularly ones that might fall uncomfortably close to constraining individuals’ autonomy, have strong evidence that the people whose behavior they hope to modify are affected by biases and would consider themselves better off if the intervention had its intended effects.

Small Effect Sizes

A second critique is that many, though certainly not all, behavioral economics studies showing positive effects have comparatively small effect sizes. There are examples where extremely large gains have been documented, such as for automatic pension deductions. But in many cases, researchers expect small effect sizes because the intervention itself is modest. Behavioral economists emphasize that the most relevant benchmark is the ratio of the effect size to the cost, and that the ratio can be large even if the effect size is small. This point is discussed further in Chapter 14.

Individual Focus

Another critique of behavioral economics is that it focuses on changing individual behavior rather than on reforms that regulate firms or change government program designs to pursue the same goals (Chater & Loewenstein, 2022). This criticism is mostly focused on the small interventions represented by many nudges that have low effect sizes and do not radically change behavior. In response, behavioral economists point to the many behaviorally motivated interventions that are much more than nudges and that have major effects. But a more fundamental response is that the behavioral patterns identified by behavioral economists can be used to design governments’ regulatory, tax, and transfer policies: that is, for systemic changes. Behavioral economists argue that those policies are more likely to achieve their aims if the behavioral responses of the people targeted are taken into account in the design of the policy. Behavioral scientists argue that most government regulatory efforts to reduce suboptimal behaviors (consumption of sugary cereals, for example) are based on behavioral research even if they address the actions of firms, not individuals. Indeed,

behavioral economists would argue that regulations and behaviorally informed interventions are complementary, not in opposition to one other.

Political and Commercial Use

Others have noted that it is easy to observe—perhaps all too difficult to ignore—the many ways behavioral insights are applied in daily life in the service of marketing efforts and efforts to advance political and other ideas. Knowledge of behavioral bias is not restricted to academics and policy makers advised by academics. That knowledge can also be used in an adversarial fashion to influence people’s thinking and decision making or to frustrate their efforts to make optimal choices, such as through counter-nudges, or “sludge,” that can make it more difficult for people to understand potential advantages, notice hidden costs, or claim benefits. The committee was not able to explore these issues, which concern ways behavioral economics research is used rather than the nature of the research and analysis produced by the field. However, we note that there is a growing body of research that will be of value as policy makers consider ways to address them (see, e.g., Petticrew et al., 2020; Shahab & Lades, 2021; Mills, 2023).

Disparities

It has also been suggested that behavioral economists may not have paid enough attention to the effects of the tools they design on economic and social disparities. One concern is that behavioral biases may tend to affect populations with varying levels of education differently, particularly in the contexts studied by economists, and the result could be that interventions have unintended disparate impacts. A related concern is that interventions designed to bring benefits for certain segments of a population may indirectly disadvantage those not targeted. Some behavioral economists argue that these critiques are refuted in part by the large body of evidence on interventions designed explicitly to close equity gaps (such as unequal distributions of government benefits) and that many behavioral policies are directed to assist disadvantaged groups, at least in the United States (e.g., behavioral redesigns of financial aid forms for low-income college students [see Chapter 9]). Despite this argument, it is important that researchers evaluate the effects of behavioral interventions and nudges on disparities more broadly than they have in the past. Improvements to standard research practices may also help to address this issue, as we discuss in Chapter 14.¹³

___________________

¹³ It is also worth noting concerns about methods used in the conduct of behavioral research (see, e.g., Meyer et al., 2019; Heck et al., 2020; Mislavsky, Dietvorst, & Simonsohn, 2020).

Representativeness

Finally, like researchers in most social science domains, behavioral economists are subject to the critique that they have not paid adequate attention to the importance of ensuring that their research is representative of diverse populations. In the past 20 years a substantial body of behavioral economics research has been conducted in the context of developing countries—including work that has been recognized in Nobel Prize awards—and behavioral development economics is one of the most active areas of the field (see, e.g., Duflo & Kremer, 2004; Banerjee & Duflo, 2009; Banerjee, Duflo, & Kremer, 2020). This work has, thus far, yielded findings in lower-income societies that largely mirror those in the higher-income ones (Kremer, Rao, & Schilbach, 2020). Nevertheless, the bulk of behavioral economics research conducted by U.S. researchers has been carried out in this country, and thus the body of evidence about many behavioral challenges and intervention approaches so far rests primarily on evidence collected in WEIRD contexts; this is a challenge the field will need to continue to address.

REFERENCES