Introduction
Behavioral public policy (BPP) draws on insights from various behavioral sciences, with behavioral economics playing a particularly influential role in shaping BPP’s empirical and methodological foundations (Oliver, Reference Oliver2023). Incorporating insights from behavioral economics into policy discussions has arguably improved predictions about the effects of economic incentives, such as retirement savings subsidies and income tax policies (Chetty, Reference Chetty2015). Behavioral insights have also led to new policy tools that address internalities rather than externalities, such as nudges (Thaler and Sunstein, Reference Thaler and Sunstein2008) and sin taxes (Gruber and Koszegi, Reference Gruber and Koszegi2001). At the same time, behavioral insights have made welfare analysis more challenging (Sugden, Reference Sugden2018). For instance, different framing can alter people’s preferences by affecting the salience of choice attributes and shifting the weight individuals assign to them (Bernheim, Reference Bernheim2021). Seemingly irrelevant contextual factors can result in markedly different decisions (Hargreaves Heap Reference Heap2013). Issues of self-control further complicate matters, as time-inconsistent behavior undermines the coherence assumption required for standard preference representations. When preferences are both context-dependent and time-inconsistent, determining which preference, if any, accurately reflects welfare becomes difficult.
Over the past two decades, behavioral economists have developed several approaches aimed at addressing these challenges and enabling welfare analysis ‘in a behavioral world’.Footnote 1 One prominent approach is preference purification. This approach assumes that individuals have ‘true’ preferences, from which they deviate due to heuristics and decision biases (Sugden, Reference Sugden2018, 65). The approach models agency as an ‘inner rational agent,’ whose preferences ‘deep within’ adhere to neoclassical rationality axioms but who struggles to act on them since he is trapped in an error-prone ‘psychological shell.’ In modeling practice, this means that standard models are extended by behavioral elements that represent cognitive biases leading to mistakes. The economist’s task is to identify the choosers’ ‘true preferences’ and potentially correct the decision biases, enabling them to fulfill their true preferences. The logic of this preference purification has led behavioral economists to develop extensive lists of ‘biases’ to explain the divergence between observed choices and true welfare. This approach, widely adopted by behavioral economists, aligns closely with the assumptions of standard welfare economics (Sugden, Reference Sugden2018, 53; Hargreaves Heap, Reference Hargreaves Heap2017, 252). A key assumption is that ‘true’ preferences are consistent and context-independent, because in the absence of this condition, the approach would lack a clear normative standard for evaluating revealed preferences (Sugden, Reference Sugden2018, 62–63).
Despite the influence of this and other behavioral approaches to welfare analysis, there has been a vigorous debate about the conceptual underpinnings of these perspectives. A prominent contributor to this debate is Mario Rizzo, whose work has challenged the conceptual, empirical, and practical foundations of behavioral paternalism in a series of articles (Rizzo, Reference Rizzo and Henderson2018, Reference Rizzo2019; Rizzo and Whitman, Reference Rizzo and Whitman2009a, Reference Rizzo and Whitman2009b; Whitman and Rizzo, Reference Whitman and Rizzo2015), culminating in the book Escaping Paternalism: Rationality, Behavioral Economics, and Public Policy (Rizzo and Whitman, Reference Rizzo and Whitman2020). Rizzo acknowledges that behavioral economics has enhanced economists’ understanding of how humans – rather than hyperrational ‘Econs’ – make decisions. As he and Glen Whitman state, ‘[to] the extent that behavioral economics has exposed the genuine failings of old rational-choice models, it has been a boon to the economics profession’ (Rizzo and Whitman, Reference Rizzo and Whitman2020, 3). However, Rizzo critiques the tendency of some behavioral economists to move too quickly from descriptive analyses to normative conclusions, particularly in support of soft paternalistic policies like nudges.
Rizzo’s critique of behavioral paternalism encompasses several dimensions, particularly the narrow interpretation of rationality within the framework: ‘despite having rejected rationality as a model of how people do behave, the behavioral paternalists still accept rationality as a model of how people ought to behave’ (Rizzo and Whitman, Reference Rizzo and Whitman2020, 16). This represents a peculiar development in the history of ideas, as the restrictive assumptions of neoclassical rationality were originally adopted not for their normative plausibility but for their analytical convenience and mathematical tractability. A central theme in Rizzo’s discussion is the difficulty of externally defining rational behavior and mistakes: ‘[in] the rush to characterize certain “anomalies of choice” as violations of rationality, behavioral paternalists have been insufficiently subjectivist’ (Rizzo and Whitman, Reference Rizzo and Whitman2020, 317). In this context, Rizzo asserts that behavioral paternalists often impose an external set of values by using the neoclassical definition of rationality as a normative standard for ‘good’ decision-making. This is problematic because adherence to neoclassical rationality does not necessarily lead to welfare-improving choices. Experimentation, mistakes, and self-discovery might be essential components of individual welfare, suggesting that inconsistencies between preferences and values can be integral to the pursuit of welfare.
In addition to critiquing the prescriptive interpretation of neoclassical rationality, Rizzo has addressed several practical challenges associated with implementing behavioral paternalism (Rizzo and Whitman 2009a; Whitman and Rizzo, Reference Whitman and Rizzo2015). He emphasizes that it is not straightforward to translate the myriad findings on cognitive biases from highly controlled laboratory experiments into real-world policy applications. Without gathering field data in the contexts where proposed paternalistic policies will be enacted, the magnitude and prevalence of identified biases remain uncertain. Furthermore, Rizzo argues that in many real-world situations, individuals are aware of their cognitive biases and often devise effective self-debiasing strategies, such as seeking advice or engaging in group decision-making. As a result, the impact of these biases in everyday contexts may be less pronounced than in laboratory settings. This creates a complex knowledge problem for policymakers, who may struggle to accurately assess the extent of an individual’s bias or whether that individual has developed an adequate level of self-control.
Successfully implementing paternalistic policies requires policymakers to possess a deep understanding of individuals’ supposedly ‘true’ preferences. By rejecting revealed preference theory, behavioral economists face an epistemological dilemma regarding how to identify these true preferences. Given the evolving nature of preferences, it remains unclear in which choice contexts individual choices or verbal statements should be regarded as normative inputs for policymaking. Moreover, Rizzo warns against the ‘nirvana fallacy’ (Rizzo and Whitman, Reference Rizzo and Whitman2020, 310; Demsetz, Reference Demsetz1969), the notion that any discrepancy between the ideal and the real justifies behavioral paternalistic intervention. While it may be correct to say that individuals fall short of some behavioral ideal, interventions aimed at bridging the gap between the ‘real’ and the ‘ideal’ may prove less effective than the spontaneous adaptations of imperfect individuals.
Several factors contribute to this concern: policymakers themselves may be subject to cognitive biases, such as overconfidence, confirmation bias, or salience bias. Additionally, behavioral paternalistic policies can be vulnerable to manipulation by special interest groups. Rizzo also cautions against the slippery slope of small or moderate paternalistic interventions leading to more intrusive and autonomy-reducing measures in the future. Core concepts of behavioral paternalism – such as welfare, freedom, and autonomy – are often vague and employed in an ad-hoc manner. While discussions initially focus on the identification of individuals’ ‘true’ preferences, the difficulty in conceptualizing and measuring these preferences can shift the focus from the agent’s perspective to experts imposing their own values. For example, in the case of sin taxes, the normative rate of time discounting is often assumed to be the longer-run rate.
Rizzo’s work goes beyond merely deconstructing the foundations of behavioral paternalism but also engages with an alternative approach to policymaking that he refers to as the paternalism-resisting framework. He critiques the tendency within behavioral economics to frame policy issues not as ‘whether or not paternalism is desirable,’ but rather as ‘what form of paternalism shall we have?’ (Rizzo and Whitman, Reference Rizzo and Whitman2020, 392). This framing is of course not the only way to address behavioral problems. A central tenet of his alternative framework is the concept of inclusive rationality, which he and Whitman define as ‘purposeful behavior based on subjective preferences and beliefs, in the presence of both environmental and cognitive constraints.’ They emphasize that inclusive rationality does not impose a normative structure on preferences and beliefs a priori; instead, it accommodates a diverse range of possibilities regarding how individuals set goals, form and revise beliefs, structure decisions, and conceptualize the world (Rizzo and Whitman, Reference Rizzo and Whitman2020, 26).
This special issue reflects on Rizzo’s contributions to the conceptual, empirical, and practical foundations of behavioral paternalism. It delves into his recent work on behavioral economics as well as his earlier research in law and economics, both of which have significant implications for advancing the field of BPP. In this context, the issue aims to revisit foundational concepts in BPP, including rationality and well-being. The contributors – comprising psychologists, behavioral economists, legal scholars, historians of economic thought, and philosophers – are united in their efforts to revitalize discussions on the (im)possibility of welfare analysis and process-based approaches to rationality in BPP.
Summary of arguments
In the first contribution to this issue, Gerd Gigerenzer provides a reflection on what he terms the ‘rationality wars’ among three distinct approaches within the behavioral sciences in the 20th century. He notes that during the Cold War, logical rationality – grounded in consistency axioms, subjective expected utility maximization, and Bayesian probability updating – became the cornerstone of economics and other social sciences. However, in the 1970s, this framework faced significant criticism from the heuristics-and-biases program spearheaded by Daniel Kahneman and Amos Tverksy. This program interpreted logical rationality as a standard for how individuals ought to make decisions. Gigerenzer notes that such an interpretation is absent in much of the canonical and foundational work in decision theory by John von Neumann and Oskar Morgenstern or L. J. Savage. According to the heuristics-and-biases program, deviations from logical rationality in people’s judgments were thought to reveal stable cognitive biases, which were subsequently viewed as underlying social problems and justifying governmental paternalism. In the 1990s, the ecological rationality program emerged, inspired by the work of Herbert Simon. This approach moves beyond the restrictive boundaries of logical rationality and examines how individuals and institutions make decisions under conditions of uncertainty and complexity. Gigerenzer argues that this broader perspective reveals that many of the supposed cognitive biases are markers of intelligence rather than indicators of irrationality, and that heuristics serve as essential guides in navigating a world filled with uncertainty.
The second article in this issue, authored by Adam Oliver, explores a specific aspect of neoclassical rationality: choice consistency. Oliver aligns with a significant claim in Rizzo’s work, asserting that choice inconsistencies are legitimate facets of individual decision-making. Oliver highlights that a core normative assumption of welfare economics is that individuals should maximize expected utility and, consequently, maintain consistency in their choices. While behavioral economists have documented systematic choice inconsistencies, they often respond by attributing these behaviors to individual errors rather than relaxing the normative assumption of expected utility maximization. Oliver argues that this attribution is itself an error – what he calls an ‘error error.’ In reality, planners cannot fully understand the diverse desires that influence an individual’s choices, making it very difficult to determine which choice in an inconsistent set is erroneous. Furthermore, Oliver points out that individuals who exhibit inconsistencies may not view either of their choices as erroneous if the context meaningfully interacts with their valuations of outcomes. According to Sugden (Reference Sugden2018), planners should refrain from intervening in market mechanisms to correct behavioral inconsistencies, advocating instead for the free market as the best means for individuals to arrive at mutually agreeable exchanges. Notwithstanding the compatibility of these points with Rizzo’s general perspective, Oliver nevertheless posits that policymakers have a legitimate role in enhancing individuals’ agentic capabilities. The most effective way to achieve this is by investing in human capital and creating institutions that are widely regarded as foundational to personal agency. Additionally, he acknowledges the need for ‘boosts’ to help people avoid make costly decision errors. Oliver argues that ‘budges’ are justified, i.e., government regulations against market actors who use behavioral insights in a self-interested manner to manipulate others.
The third article, by Roberto Fumagalli, can be read as a response to some of Gigerenzer’s and Oliver’s arguments, as it critically assesses Rizzo’s concept of inclusive rationality and defends standard rationality as a foundation for BPP. The paper aims to contribute to the ongoing debate about the connection between rationality and welfare. At the core of Fumagalli’s argument is a qualified defense of the consistency conditions of standard rationality against Rizzo and Whitman’s main criticisms. Fumagalli posits that reported violations of consistency conditions decrease significantly when experienced decision-makers are involved, particularly when individuals have the time and incentives to learn about the choice problems they face. Additionally, Fumagalli provides a qualified normative defense of consistency conditions. He argues, for instance, that completeness requires individuals only to specify preferences over relevant alternatives in a given decision problem and that transitivity can be defended by highlighting the losses incurred from violations and the tendency for individuals to revise intransitive choices once they recognize their inconsistencies. Furthermore, Fumagalli suggests that many reported descriptive violations of consistency can be reconciled by adjusting the descriptions of choice options in counterexamples to the axioms of standard rationality. While Fumagalli acknowledges the valuable critical insights Rizzo’s work provides regarding the descriptive and normative validity of standard rationality in evaluating public policies, he notes that Rizzo’s account of inclusive rationality remains vulnerable to objections.
The contribution by Richard Epstein critiques the choice theories of both neoclassical and behavioral economics, arguing that despite their differences, these two rival theories share several unfortunate similarities. Standard rational choice theory posits that individuals employ rational methods to choose goals that satisfy their self-interested preferences. In contrast, modern behavioral economics highlights systematic deviations from these principles. However, both approaches assume that all preferences are based solely on individual choice. Epstein invokes the evolutionary principle of inclusive fitness, which suggests that in familial contexts, the welfare of the family unit, rather than that of any single individual, accounts for the natural love and affection that emerge from interdependence and the redistribution of wealth. Furthermore, both standard theories overlook or underemphasize variations in tastes and competence levels, which can generate gains from trade. Epstein reinterprets the treatment of nudges and various legal doctrines related to disabilities, product liability, and organizational structure, arguing that the standard assumptions of uniform behavior fail to capture the salient features of human behavior and social interactions. He acknowledges Rizzo’s achievement in exposing the challenges inherent in behavioral models that portray deviations from standard rational choice as fatal flaws in economic theory. However, Epstein contends that merely undermining the case for paternalism is insufficient. His broader narrative recognizes that while some individuals may require the guidance of parents or guardians, significant variations in preferences and competencies do not warrant state paternalism.
In the next article, Shaun Hargreaves Heap explores a significant theme in Mario Rizzo’s work, viz., the argument that rationality should be conceived more broadly than the conventional rational choice model. Following Rizzo, Hargreaves Heap argues that rationality encompasses the ability to adjust behavior based on experience and the mistakes that arise from it, emphasizing the concept of learning-by-doing. Hargreaves Heap posits that learning-by-doing should be recognized as a central insight in economics, on par with widely acknowledged concepts like the gains from trade and the significance of unintended consequences. To substantiate this claim, he references discussions by Adam Smith and Albert Hirschman on learning-by-doing. Hargreaves Heap then examines the determinants of learning-by-doing processes within teams, noting that Rizzo’s work often emphasizes that social learning occurs in teams within organizations. For instance, the success of entrepreneurship in society heavily relies on the effectiveness of team learning. Hargreaves Heap identifies a key challenge for teams: how to harness the diversity essential for learning-by-doing while preserving the cooperation necessary for effective collaboration. Drawing inspiration from political theory, he addresses the broader societal issue of designing rules that encourage individuals with fundamentally different perspectives to engage in positive-sum activities rather than zero-sum ones. Hargreaves Heap advocates for checks and balances to prevent any individual or faction from pursuing zero-sum behaviors, framing this as a critical concern for public policy. His proposals suggest that teams adopt an egalitarian structure reminiscent of de Tocqueville’s principles of democracy to foster cooperation and maximize learning.
In their contribution, Agnès Festré and Stein Østbye highlight an important point in Rizzo’s work, viz., the epistemological divide between contemporary mainstream behavioral economics – particularly its libertarian paternalistic variant as represented by Richard Thaler and Cass Sunstein – and Hayek’s complexity and evolutionary perspectives on rationality and learning. Central to their argument is Hayek’s knowledge problem, which refers to the inherent limitations in policy-making due to the dispersed nature of knowledge in society. Knowledge is decentralized, fragmented, and often includes the knowledge of ‘the particular circumstances of time and place.’ This individual knowledge encompasses tacit knowledge, a concept elaborated by Hayek’s contemporary Michael Polanyi. Festré and Østbye point out that while Hayek’s theory of tacit knowledge shares similarities with Polanyi’s, the latter’s theory diverges by emphasizing the tacit, personal, and perceptual aspects of knowing. Polanyi’s ultimately rejects any attempt to categorize different types of tacit knowledge. He even questions the existence of general knowledge independent of tacit understanding. Festré and Østbye contend that tacit knowledge warrants greater attention to the limitations of the nudging approach in BPP. They argue that determinism is intrinsic to the nudging strategy, yet this stance conflicts with both Hayek’s and Polanyi’s emphasis on indeterminacy. In this context, Festré and Østbye posit that tacit knowledge should be foregrounded in the debate, advocating for indeterminacy as a compelling argument against nudging and in favor of prioritizing education and learning.
Daniel Connolly, George Loewenstein, and Nick Chater expand on earlier insights by Chater and Loewenstein (Reference Chater and Loewenstein2023), who argue that behavioral scientists often focus on individualistic (i-frame) solutions to policy issues that stem from systemic (s-frame) causes. Typically, i-frame policies begin with the assumption that adverse outcomes like obesity and retirement insecurity arise from inherent human frailties, such as present bias. This perspective aligns with the idea that public policy can modify the choice architecture by means of nudging to encourage better decision-making. For example, to address insecure retirement, policymakers may change the default settings in retirement savings accounts, while combating obesity may involve providing calorie information to help consumers make healthier food choices. However, Connolly et al. argue that by concentrating on mitigating individual cognitive biases and behavioral errors, the i-frame approach neglects and distracts from the systemic issues (s-frame) which significantly contribute to the problems these interventions seek to address. While i-frame policies aim to assist individuals in making ‘better’ choices within existing frameworks, s-frame public policies focus on transforming the societal rules and institutions that shape the choices available to individuals. This emphasis on institutional change aligns with Rizzo’s work (e.g., Rizzo, Reference Rizzo1980, Reference Rizzo2021), which suggests that the s-frame is where meaningful societal progress can be made. The authors contend that shifting BPP’s focus toward the s-frame and creating beneficial systemic change will require the behavioral research community to confront a new set of questions, necessitating a different mix of research methodologies. They advocate for a transition from field experiments, which are well-suited for testing i-frame interventions, to quasi-experimental and qualitative observational approaches that examine policy changes across countries and over time.
In his contribution, Cass Sunstein engages with Mario Rizzo’s recent work on radical uncertainty (Dold and Rizzo Reference Dold and Rizzo2021; Rizzo and Dold, Reference Rizzo and Dold2021) and its implications for institutional analysis at both the organizational and public policy level. Sunstein’s argument draws on the distinction made by John Maynard Keynes and Frank Knight between uncertainty and risk. He explains that Knightian uncertainty arises when individuals cannot assign probabilities to conceivable outcomes. For instance, people may be aware that a particular course of action could lead to adverse outcomes, but lack information about the likelihood of each occurring. Challenging the standard view in economics, Sunstein asserts that Knightian uncertainty is a widespread phenomenon that presents complex and unresolved issues for decision theory and regulatory practice. He points out that it affects various public policy problems, including those related to the uncertainty around the consequences of (un)regulated artificial intelligence. Sunstein argues that while it may be tempting to eliminate worst-case scenarios by adopting the maximin rule, this approach can lead to significant drawbacks. According to Sunstein, such drawbacks include (1) imposing high risks and costs, (2) sacrificing substantial benefits or potential breakthroughs, and (3) generating uncertain risks. Consequently, there are good arguments for why individuals or societies would not necessarily opt for maximin in situations of uncertainty. Much depends on the trade-offs involved in eliminating the worst-case scenario, including the potential for increased resilience. In many cases, available information allows for informed judgments about these trade-offs, at least in general terms. However, Sunstein recognizes that in the most challenging and intriguing cases, no simple rule can be defended. Instead, some form of judgment must be made, which may be inherently political and thus susceptible to the public choice considerations emphasized in Rizzo’s work.
In the final article, Peter Boettke expands the discussion of the previous articles by exploring how BPP could benefit from a deeper engagement with the history of economic thought, particularly Austrian Economics. Boettke highlights Austrian Economics as a school of thought with deep and important insights into market processes, learning, and subjectivism. Boettke argues that insights from Austrian Economics have significant implications for the potential and limits of reforms within political and legal systems. According to Boettke, a core tenet of Austrian economics is that it is not only individual decision rules that help people cope with ignorance but also the institutional safeguards provided by private property and the freedom of contract rooted in the rule of law. Boettke says that Rizzo’s contributions highlight both the stability and dynamic nature of law. A relatively ‘static’ legal framework enhances predictability, allowing individuals to pursue their own goals and shape their own lives. The central question then becomes: which legal framework creates a more stable environment for individuals to pursue their ends in harmony with one another? Building on Rizzo’s insights (Rizzo, Reference Rizzo1980), Boettke claims the answer may lie in the ‘antiquated’ and static system of strict liability, particularly in a dynamic world where the information required for ‘fine-tuning’ is often unavailable. Nevertheless, the law must also adapt to new technological and societal developments. Like the market, the law is in a continuous state of evolution – not erratic or arbitrary change, but gradual adaptation and adjustment. Hence, the governing dynamics for addressing the complexities of a world in flux – both in law and the market – share strong affinities: the ‘fine-tuning’ paradigm is illusory; the only valid basis for comparing the ‘efficiency’ of systems lies at a fundamental institutional level.
Conclusion
This special issue is dedicated to Professor Mario J. Rizzo (NYU), whose influential work at the intersection of economics, ethics, psychology, and law has consistently engaged with foundational questions central to behavioral economics and BPP. In honor of his seventy-fifth birthday, it is a fitting occasion to highlight the importance of investigating key concepts that lie at the heart of BPP but have rarely been subject to deep, systematic reflection, such as the roles of time, uncertainty, and tacit knowledge in individual choice; the significance of dynamic processes like preference construction and learning compared to static notions of preference satisfaction; the challenges that cognitive and behavioral biases present to neoclassical standards of rationality; and the conceptual and normative foundations of economic welfare as a construct of analysis. We believe this special issue represents a significant contribution to advancing the discussion of these key concepts in BPP.
The timing of this issue is particularly appropriate as this journal launches a new section titled ‘Political Economy of Behavioral Public Policy,’ edited by Malte Dold (Pomona College). The editorial board recognized the need for this section, given the rapid development of BPP and the growing demand for more thorough and systematic reflection on its methodological and normative underpinnings. While much of the current literature empirically evaluates tools like nudges, boosts, and budges, there remains a shortage of rigorous analysis of their foundational assumptions. Moreover, while the field has focused primarily on bounded rationality in consumer behavior, it has given less attention to the behavioral dimension of political actors, who frequently face complex knowledge and incentive problems.
This new section aims to bring these issues to the forefront, emphasizing the political economy of BPP. The journal particularly encourages submissions that engage with past or current approaches exploring the complex interplay between theoretical frameworks and behavioral policy interventions. The focus of this new section of BPP aligns closely with the work of Professor Rizzo, whose research has illuminated many of the methodological and practical challenges at the core of BPP.
Open access
