14

Advancing the Field of Behavioral Economics

Behavioral economics has grown steadily as a field over the last several decades. Journals and academic departments devoted to it have established themselves, and behavioral economists have studied the application of their ideas to numerous policy challenges, as detailed throughout this report. Returning to the charge that guided this study, we close with our review of the development of the field and its theoretical foundations, as well as our findings about the primary behavioral economics strategies and how they have been applied. In this chapter we synthesize key conclusions from the research to date and offer recommendations to guide the future development of the field and capitalize fully on its potential to help meet critical societal goals, including directions for the research that is needed to support continued growth.

CONCLUSIONS

We draw two primary conclusions from our review: that behavioral knowledge is indispensable to the development of effective policy interventions and that work is needed to support intervention design and broad-scale implementation.

Developing Effective Policy Interventions

Foundational theoretical work that has integrated understanding of cognitive and psychosocial processes with economic analysis has pointed to five core principles that help to explain human decision making: limited

attention and cognition, inaccurate beliefs, present bias, reference dependence and framing, and social preferences and social norms (see Chapter 3). Empirical research has provided strong evidence that these behavioral factors play a major role in human decisions and are therefore key factors to consider in assessing the impact of public policy programs and in designing interventions to modify human behavior.

We acknowledge that our review of the available evidence was not comprehensive and that, for reasons discussed throughout the report, positive findings of effects may be more likely to garner attention than null or negative findings. We hope the field will continue to produce studies that can provide the most robust evidence of effectiveness—or lack of it—such as large-scale studies and meta-analyses.

Nevertheless, the accumulated evidence across the six domains is significant: it shows that decision processes are dynamic, malleable, and context dependent and that understanding these factors helps to explain how and why people behave in ways that appear to be counter to rational calculations. That evidence from research in behavioral economics demonstrates that those behavioral decision processes recur repeatedly and have significant impact on policy-relevant behavior, which in turn has strong implications for policies that differ from those that are suggested by traditional economics.

This report examines this impact in six important domains, but behavioral economics concepts are also being applied in many other contexts, including contexts not generally considered for behavioral economics input. For example, recent reports of the National Academies of Sciences, Engineering, and Medicine that addressed improving fuel economy (2021), reducing consumer food waste (2020), and reducing alcohol-impaired fatalities (2018) all discuss behavioral economics approaches among varied other interventions. Behavioral economics has also had a profound influence in commercial contexts, as we note below. In the future, understanding of behavior has the potential to play a key role in other urgent societal issues we have not addressed, such as combating misinformation; see Box 14-1.

Conclusion 14-1: The very strong evidence that complex cognitive, social, behavioral, and contextual factors influence judgment and decision making means that behavioral economics concepts are indispensable for advancing scientific understanding of policy-relevant human behavior and for designing public policies. Behavioral economics has produced invaluable evidence about why people act in seemingly irrational ways, how they respond to interventions, and how public policy and practice interventions can be designed to modify the habitual and unconscious ways that people act and make decisions.

Supporting Design and Implementation

The evidence for the importance of behavioral ideas is strong, but the refinement of strategies for applying these principles in a systematic way to the design of policies and interventions remains a frontier challenge. There is clear and strong evidence that specific interventions based on these principles have been effective at changing certain targeted behaviors, and it is also clear that behavioral economics research has contributed to a large and expanding set of policy interventions that includes strategies with excellent empirical evidence of effectiveness (e.g., defaults, framing). For other intervention approaches commonly studied, such as behaviorally informed incentives and social proof interventions, the evidence is mixed, nuanced, or still emerging.

Some types of decisions appear to be more amenable to behaviorally informed interventions than others. It is comparatively easy to influence a one-time, up-or-down decision, such as whether or not to opt in to a retirement savings plan, or to make it easier and more likely for a target population to complete a specific action, such as filling out an application form. Interventions that have targeted more complex behaviors, such as programs to reward teachers for increasing their students’ test scores, or

approaches to changing how clinicians approach complex diagnosis and treatment decisions, have not yielded such clear-cut or consistent results.

It is important to acknowledge that research demonstrating positive effects for behavioral economics interventions typically shows modest effect sizes. This is not necessarily a surprising finding, particularly because many of the intervention studies are comparatively low in cost and easy to administer. However, as we emphasize clearly in the work on climate change—a challenge of unmeasurable magnitude—the application of combinations of individually modest interventions can cumulatively bring important changes and benefits for relatively little cost.

Throughout the report we also identify additional persistent challenges in the application of behavioral economics for policy: it is not easy or straightforward to generalize findings from specific contexts, and implementing successful interventions at scale remains elusive. Moreover, deeper understanding of why and when observed effects occur would be valuable. Even when intervention or policy studies demonstrate a positive effect on behavior, there is insufficient knowledge about mechanisms of action and about differential responses to the intervention or policy by different groups. For example, in the context of inaccurate beliefs, understanding of the relevant mechanisms would be enhanced if data on both beliefs and behavior were collected and analyzed. For example, testing whether beliefs held by participants before the intervention were inaccurate and whether the intervention was effective in changing them would be valuable. Our review of six policy domains (in Part II) also suggests that behavioral economists have yet to fully embrace the concept of behavioral design—the systematic process by which interventions and strategies to change human behavior are designed. That is, specific interventions are often tested without sufficient attention to behavioral design principles.

Conclusion 14-2: The field of behavioral economics has made significant advances over the past 20 years, producing evidence about both general principles and specific intervention approaches that address policy challenges in many domains. However, the field has not yet produced generalizable and implementable practice guidance and intervention design strategies for determining what works, when, and for whom. Whether the goal of providing such specific guidance can be achieved, given the importance of context and the unique characteristics of many targets of behavior change, is not clear.

RECOMMENDATIONS

As noted above, behavioral economics research has generated a large volume of promising results. But there is also much more to learn, and there are contributions to be made in new policy domains. The committee

recommends strategies for funders, research and professional organizations, and universities to support future development in the field of behavioral economics; such support would certainly be appropriate in other research fields, but our focus is on work in behavioral economics. We also recommend priorities for research.

Behavioral economics is rooted in the integration of ideas from diverse academic domains, and its future progress will depend on continued interdisciplinary collaboration. Such interaction is not easy and faces well-known obstacles: differences in theoretical perspectives, terminology, research methods and tools, and standards of evidence all bring challenges. Nevertheless, interdisciplinary work, involving researchers from many domains, including cognitive and social psychology and neuroeconomics, can help to pinpoint how the human brain processes and responds to information.

Practical strategies, such as setting up (and funding) multidisciplinary labs or working groups in which teams of researchers can regularly collaborate and identifying funders for methodologies and projects that do not fit established research contexts, can help to address the challenges of implementation (see Chapter 13). Examples include the Behavioral Economics Design Initiative at the University of Pittsburgh, which focuses on behavioral economics and mechanism design, and the Nudge4 Lab at the University of Virginia.¹ In general, establishing stronger incentives (e.g., in the contexts of research support, journal publication, academic advancement) for interdisciplinary research and training for students and younger scholars in interdisciplinary work will yield long-term benefits for the field.

___________________

¹ See https://sites.google.com/view/bedi-university-of-pittsburgh/home and https://nudge4.org

As evidence that an intervention has been effective in a particular context accumulates (in any research field), it is important that larger studies, with larger samples and more variants, be conducted to replicate the findings and test how well the intervention works in practice, across contexts and populations. In some cases, the results will be confirmed; in others, they will not; and others will point to a redesign or a refinement. This phase of the research—scaling studies that involve patience and resources—is essential for behavioral tools to bring meaningful societal benefits.

Many social science disciplines and policy domains emphasize the importance of systems of ongoing evaluation, in which research, design and development, testing of ideas, and evaluation each contribute to a continuous cycle of learning and refinement and improvement of policy interventions (see, e.g., Patton, Sawicki, & Clark, 2012). In the context of behavioral economics interventions, ongoing data collection, pilot testing, and long-term surveillance are of particular importance because of the need to assess the durability and sustainability of effects, as well as differential response to policies by different populations. The committee examined many policy interventions that were designed with behavioral principles in mind but was not able to look systematically, from the perspective of policy makers, at how regularly behavioral factors are currently considered

in the design of policy. It seems likely, however, that the majority of policy recommendations still rely primarily or exclusively on traditional economic modeling.

Recommendation 14-5: Researchers, funders of research, and entities that support or sponsor behavioral units in organizations should prioritize ongoing investigations into the role of behavioral economics in society, with specific attention to the equity implications of behavioral economics policies and interventions; the implications of public attitudes toward the ethics of behavioral economics research and practice, as well as their acceptance by the general public; and possible public policy interest in commercial applications of behavioral economics findings.

It was not part of our charge to identify priorities for the specific policy goals that should be addressed using behavioral economics approaches, but three issues in policy implementation raise broad questions: equity of impact, attitudes about interventions, and commercial application of behaviorally based findings.

Equity of Impact

It is reasonable to ask whether behavioral interventions are reaching those most in need, and whether such interventions in some cases may have the negative unintended consequence of increasing inequalities (Blumenthal-Barby & Burrough, 2012; Lunze & Paasche-Orlow, 2013; Lin, Osman, & Ashcroft, 2017). These are empirical questions and, as far as the committee is aware, there is limited evidence on them.

Attitudes About Behavioral Interventions

The use of behavioral interventions, such as nudges, can be perceived as paternalistic (as can many government interventions). That is, behavioral economics reveals that people’s behaviors and actions cannot be assumed to accurately reveal their preferences or reflect their conscious thinking about a decision. Thus, policy makers using behaviorally informed interventions are making decisions for people about what would be optimal for them. Thus, it is important to give close scrutiny to how the target of change is chosen and how certain policy makers are that the change will actually benefit the people who are induced to make it.

It is also important to consider how researchers and policy makers know that a particular outcome, which may appear to be desirable from a societal perspective, can be construed as beneficial to an individual, and

that, in effect, manipulating a person to change their behavior is an unmitigated good (Liscow & Markovits, 2022). Another important question is whether people’s attitudes about interventions affect their responses to them—that is, whether the perception that nudges and other behaviorally based interventions may be paternalistic might influence the way people respond when they recognize that they are the target of one. These issues may arise with any sort of incentive, but the possibility that some people may respond when they recognize behaviorally based strategies—such as an energy bill that compares one’s energy use with that of one’s neighbors—by determining to resist the desired behavior merits investigation.

Commercial Application of Behavioral Economics Knowledge

Ideas grounded in behavioral research are used in commercial contexts, usually to induce people to buy things. While it is not the job of behavioral economics researchers to police the applications of their ideas, they might profitably explore such understudied questions as how behavioral ideas are being used, how consumer responses may vary across business and policy domains, and when marketing that takes advantage of knowledge of biases crosses an ethical line.

All human decision making is influenced by the context in which those decisions are made. The influence of context on decisions often occurs in ways that consumers are aware of, through design, detectable nudges, and so on. But many factors have been shown to influence people’s decisions in ways that they are often unaware of and cannot control; moreover, even hypothetical questions can influence people’s subsequent decisions (Fitzsimons & Shiv, 2001).

Advertising, of course, is a hugely profitable and effective means of shaping preferences, often in ways consumers are unaware of and unable to avoid. In a study offering loans to low-income borrowers, various features of the offer letter that had no economic implications were found to have a large effect on take-up. There is compelling evidence that television advertising influences children’s food and beverage preferences, requests, and consumption habits. One study found that a majority of Australian children ages 9–10 believed that Ronald McDonald knew best what children should eat (Olfman, 2005). Another study found that the number of hours of television watched by children ages 3–8 correlated with their caloric intake and their requests for—and their parents’ subsequent purchases of—foods they see on television (Taras et al., 1989). These peripheral and undetected effects on people’s decision behavior have been observed among novices and experts, in situations of intentional manipulation, in cases where people are not aware of the manipulation, and sometimes when the manipulating factors are altered without any human involvement.

Although researchers in the field of behavioral industrial organization, in particular, have examined some of these questions, there are no straightforward rules or answers (for a review, see, e.g., Heidhues & Kőszegi, 2018). It is important that researchers and those who apply and use research keep these questions in mind when making decisions, including in the design of research studies and selection of study populations. Empirical investigation of how behavioral economics findings are used by both public and private entities to influence consumer attitudes and behaviors could contribute to public policy thinking about the possibility of using regulatory mechanisms to limit actions that cross ethical boundaries.

Recommendation 14-6: Funders and university leaders should foster the development and application of behavioral economics by supporting training opportunities for public policy professionals. They should also support learning about practices for research transparency.

The successful development and implementation of policies that incorporate knowledge from behavioral economics depend in part on people who have been trained in behavioral principles and understand their application in policy making. Many policy schools are now including coursework on behavioral public policy; further progress in this area will be valuable for the field and for policy makers.

CONCLUDING OBSERVATION

We close with a few reflections on the future of behavioral economics. As ongoing questions about whether research discussed in this report is more properly regarded as a domain of economics or a domain of behavioral science suggest, the field’s boundaries are not precise. From the committee’s perspective, this was less important than the benefits behaviorally based approaches can bring in the development of policy. It is also perhaps a reason to think about future aspirations for application of these ideas, however they are categorized.

Most of the research we identified investigated ways in which behavioral biases interfere with a desired policy outcome and how behaviorally based interventions counteract those biases. This analysis can be applied beyond the context of the individual behaviors that are the focus of most behavioral economics research to help explain nonrational responses to complex regulatory structures, for example. That is, there may be behavioral solutions to problems that are not primarily the result of individual behavioral biases, such as the externalities that are such important considerations in combating climate change.

Thus, it is important not only to consider a broader range of solutions to behavioral biases but also to consider applying behavioral solutions even when there is no clear problem of cognitive bias. It is likely that ideas not explicitly identified as coming from behavioral economics research, but that nevertheless take advantage of behavioral insights, have already influenced the development of policy. All of these are reasons to be optimistic about the future contributions of the field.

REFERENCES

This page intentionally left blank.