Faculty Bibliography

Whether social media platforms increase well-being can be explored from multiple angles. Three empirical studies raise corresponding puzzles, with implications for valuation, choice, and well-being in general. The first finds that people are willing to pay far less to use social media platforms than they would demand to stop using them. The second finds that people lose welfare from using Facebook, and that Facebook users become more anxious and depressed, but that even after experiencing a good month without Facebook, they would demand a significant amount of money to stop using the platform for an additional month. The third finds that while many people would demand a significant amount of money to stop using Instagram and Tik Tok, they would also be willing to pay to eliminate Instagram and Tik Tok from their community. Each of the three puzzles has a plausible solution, but we do not yet know the ground truth. A reasonable conclusion is that people would demand a lot of money to be excluded from social media networks, which suggests that inclusion confers significant benefits, contingent on their existence – but that if social media networks did not exist, many users would be better off. This conclusion has broad implications; it suggests that people often spend time or money on goods whose existence they deplore. The three puzzles offer broad lessons for choice and welfare, and for how to think about their relationship.

This study explores preferences for algorithmic versus human decision-making across six countries using nationally representative samples. Participants evaluated ten decision scenarios, typically involving serious risks of one or another kind, in which they choose between algorithmic or human decision-makers under varying informational conditions: baseline (no additional information), brief information about the expertise of the human decision-maker, brief information about the algorithm's data-driven foundation, and a combination of both. Across all countries, a strong majority preferred human decision-making. A brief account of the expertise of the human decision maker increased that majority percentage only modestly (by three percentage points). A brief account of the data on which the algorithm relies significantly reduced the size of the majority preferring the human decisionmaker (by eleven percentage points). When information about both the human and the algorithm was provided, the size of the majority preferring the human decisionmaker was also significantly reduced (by eight percentage points). Other variables, above all prior experience with algorithms, were correlated with increases or decreases in the size of the majority favouring human decision-maker or the algorithm. Prior experiences were significantly correlated with preferences, with positive interactions reversing the baseline preference for human decisionmakers when algorithmic information was provided. Methodological robustness was ensured through OLS-, Logit-, and Poisson regression, as well as Random Forest analyses. The findings suggest that informational interventions alone have a relatively modest effect on algorithm acceptance.

At defining points in American history, there have been massive shifts in constitutional understandings, doctrines, and practices. Apparently settled principles, and widely accepted approaches, are discarded as erroneous, even illegitimate, in favor of new principles and approaches. Less momentously, views that were once considered unthinkable do not quite become the law on the ground but instead come to be seen as plausible and part of the mainstream. Relatedly, Americans transform how they talk and think about their Constitution-its core commitments and underlying narratives-and those transformations change our practices. These radical, dizzying changes often trigger what we term a sense of "constitutional vertigo." Our goal is to provide a conceptual map and to describe how and why constitutional vertigo occurs. First, we ask whether theories of interpretation trigger radical change or whether desires for radical change impel people to generate new (or modify old) theories of interpretation. Second, we explore why so many people experience a form of vertigo. Third, we investigate the drivers of radical constitutional change, both the familiar bottom-up pressures from "We the People" and the less-familiar top-down approaches, where legal elites back and impose a new constitutional regime. Given that the Roberts Court's changes have triggered widespread vertigo, it is an apt moment to theorize about the phenomenon writ large.

Many policies take the form of nudges, defined as liberty-preserving approaches that steer people in particular directions, but that also allow them to go their own way Some nudges attempt to correct self-control problems. Some nudges attempt to counteract unrealistic optimism. Some nudges attempt to correct present bias. Some nudges attempt to correct market failures, as when people are nudged not to emit air pollution. For every conventional market failure, there is a potential nudge. For every behavioral bias (optimistic bias, present bias, availability bias, limited attention), there is a responsive nudge. There are many misconceptions about nudges and nudging, and they are a diversion.

Both free markets and government regulators tend to use willingness to pay (WTP) as the measure of value of goods that people do not own, and willingness to accept (WTA) as the measure of value of goods that people do own. The challenge is WTP and WTA are not perfect proxies for the welfare effects of buying or selling goods, especially when people do not have experience with those goods. The reason for use of WTP and WTA is not that they are perfect, but that they seem to be the best and the most administrable method for capturing the relevant welfare effects. At the same time, WTP and WTA might be infected by a lack of information, by behavioral biases, and by hedonic forecasting errors (all of these might be either cured or aggravated by market processes). Challenges also arise when WTP is low because people do not have money and when distributional weights might be necessary to align WTP or WTA with welfare effects. There are questions about how to proceed when WTA is much higher than WTP for the same goods; the WTP-WTA disparity has yet to be fully understood. These questions are especially challenging when valuing environmental amenities, animal welfare, and social media.

People often judge how embarrassing an activity or condition is on the basis of its perceived prevalence. They infer prevalence in part by considering how often they hear other people discussing it. But how often a condition is discussed is a function not only of its prevalence but also of how embarrassing it is. If people fail to take this into account, they will tend to judge embarrassing conditions as being rarer, which will accentuate their embarrassment, and, in turn, further amplify their reluctance to disclose those conditions - a "spiral" of shame and silence. We present results from two studies that support the existence of such a feedback process. The first, a cross-sectional survey study, asked respondents a series of questions about different embarrassing and non-embarrassing conditions. Respondents (1) indicated whether they had the conditions, (2) judged how embarrassing the conditions were, (3) reported whether they had disclosed, or would disclose, having the conditions to others, and (4) estimated what fraction of survey respondents had the conditions. As predicted, reports of disclosure were negatively related to judgments of embarrassment, and when embarrassment was greater, estimates of prevalence were lower, both for conditions that respondents had and for conditions they did not have. The second, an experimental study, manipulated whether people received a high or low estimate of population prevalence for 5 different conditions, and found that receiving a high prevalence estimate reduced embarrassment and increased self-reported willingness to disclose the condition to others, and vice versa.

If you're injuring someone, you should stop—and pay for the damage you've caused. Why, this book asks, does this simple proposition, generally accepted, not apply to climate change? In Climate Justice, a bracing challenge to status quo thinking on the ethics of climate change, renowned author and legal scholar Cass Sunstein clearly frames what's at stake and lays out the moral imperative: When it comes to climate change, everyone must be counted equally, regardless of when they live or where they live—which means that wealthy nations, which have disproportionately benefited from greenhouse gas emissions, are obliged to help future generations and people in poor nations that are particularly vulnerable. Invoking principles of corrective justice and distributive justice, Sunstein argues that rich countries should pay for the harms they have caused and that all of us are obliged to take steps to protect future generations from serious climate-related damage. He shows how “choice engines,” informed by artificial intelligence, can enable people to save money and to reduce the harms they produce. The book casts new light on the “social cost of carbon,”—the most important number in climate change debates—and explains how intergenerational neutrality and international neutrality can help all nations, crucially the United States and China, do what must be done.

A great deal of work in behavioral science emphasizes that statistical predictions often outperform clinical predictions. Formulas tend to do better than people do, and algorithms tend to outperform human beings, including experts. One reason is that algorithms do not show inconsistency or “noise”; another reason is that they are often free from cognitive biases. These points have broad implications for risk assessment in domains that include health, safety, and the environment. Still, there is evidence that many people distrust algorithms and would prefer a human decisionmaker. We offer a set of preliminary findings about how a tested population chooses between a human being and an algorithm. In a simple choice between the two across diverse settings, people are about equally divided in their preference. We also find that that a significant number of people are willing to shift in favor of algorithms when they learn something about them, but also that a significant number of people are unmoved by the relevant information. These findings have implications for current findings about “algorithm aversion” and “algorithm appreciation.”

People are said to show “algorithm aversion” when they prefer human forecasters or decision-makers to algorithms, even though algorithms generally outperform people (in forecasting accuracy and/or optimal decision-making in furtherance of a specified goal). Algorithm aversion also has “softer” forms, as when people prefer human forecasters or decision-makers to algorithms in the abstract, without having clear evidence about comparative performance. Algorithm aversion has strong implications for policy and law; it suggests that those who seek to use algorithms, such as officials in federal agencies, might face serious public resistance. Algorithm aversion is a product of diverse mechanisms, including (1) a desire for agency; (2) a negative moral or emotional reaction to judgment by algorithms; (3) a belief that certain human experts have unique knowledge, unlikely to be held or used by algorithms; (4) ignorance about why algorithms perform well; and (5) asymmetrical forgiveness, or a larger negative reaction to algorithmic error than to human error. An understanding of the various mechanisms provides some clues about how to overcome algorithm aversion, and also of its boundary conditions. These clues bear on the numerous decisions in law and policy, including those of federal agencies (such as the Department of Homeland Security and the Internal Revenue Service) and those involved in the criminal justice system (such as those thinking about using algorithms for bail decisions).

In the current era, product differentiation is increasing; it is often fueled by big data and artificial intelligence (AI). Whereas product differentiation is generally welfare enhancing when consumers are informed and fully rational, such differentiation might reduce welfare when consumers suffer from misperceptions, either because of a lack of information or because of behavioral biases. We show that the positive and normative implications of product differentiation depend on whether consumers over or underestimate the benefits from some products. In particular, overestimation of the benefits is a potential source of significant welfare losses. We also study sellers’ incentives to promote, or combat, misperception. Our analysis can inform policymakers who are debating regulation that can make product differentiation more difficult (or easier), especially when the differentiation is instigated by AI algorithms powered by big data.

Do people like financial nudges? To answer that question we conducted a pre-registered survey presenting people with 36 hypothetical scenarios describing financial interventions. We varied levels of transparency (i.e., explaining how the interventions worked), framing (interventions framed in terms of spending, or saving), and ‘System’ (interventions could target either System 1 or System 2). Participants were a random sample of 2,100 people drawn from a representative Australian population. All financial interventions were tested across six dependent variables: approval, benefit, ethics, manipulation, the likelihood of use, as well as the likelihood of use if the intervention were to be proposed by a bank. Results indicate that people generally approve of financial interventions, rating them as neutral to positive across all dependent variables (except for manipulation, which was reverse coded). We find effects of framing and System. People have strong and significant preferences for System 2 interventions, and interventions framed in terms of savings. Transparency was not found to have a significant impact on how people rate financial interventions. Financial interventions continue to be rated positive, regardless of the messenger. Looking at demographics, we find that participants who were female, younger, living in metro areas and earning higher incomes were most likely to favor financial interventions, and this effect is especially strong for those aged under 45. We discuss the implications for these results as applied to the financial sector.

Sometimes legal systems are “noisy”; they show unjustified variability or “scatter,” which means that what emerges is likely to be unpredictable and unfair. Potential examples include assessment of whether a risk is “significant” or “unreasonable,” fines for environmental harms or safety violations, compensatory damage awards for libel, pain and suffering, hedonic losses, and sexual harassment, and punitive damage awards for corporate wrongdoing. To understand why and when law is noisy, it is useful to note that psychologists commonly distinguish between two kinds of scales: category scales and magnitude scales. Category scales are bounded and anchored in verbal descriptions at specified points. By contrast, magnitude scales are unbounded and defined by a meaningful zero point. In some settings, money might operate as a magnitude scale. For purposes of policy and law, here are the two key psychological findings. First, judgments on magnitude scales are often highly variable, or noisy, when there is no “modulus” to define the various points. The variability occurs even when there is no reason to believe that people actually disagree about anything meaningful. Second, distributions of judgments are “positively skewed,” with a long right tail. People involved in law and policy often use magnitude scales, above all money. High levels of noise, and susceptibility to bias (especially from anchors), are likely results. This is the problem of “noisy law,” an insufficiently explored area of behavioral public policy. Theories of optimal deterrence might help to reduce noise, but many policymakers, and many people involved in law and policy, do not accept those theories.

Suppose that historians establish that the original public meaning of the First Amendment leads to a much less protective system of free speech than the one to which we have become accustomed, or that the due process clause does not protect much at all. Or suppose, far more dramatically and consistent with the work of Stanford's Jonathan Gienapp, that historians establish that contemporary lawyers and judges have mangled the founding, in the sense that they have fundamentally misunderstood what the founding generation established. What then? Nonoriginalists need not much struggle with that question, but originalists might have to bite some hard bullets. They might have to call for a system of constitutional law that contemporary judges, lawyers, politicians, and citizens would not recognize or might even deplore. Alternatively, they might defend public meaning originalism on the ground that it protects the rule of law and related values, even if it does not really channel the founding, and even if it produces a constitutional order that the founding generation would not recognize and would in fact deplore. But most originalists are unlikely to want to defend their approach on that ground; for better or for worse, they seek to maintain continuity with the founding era. The affective pull of originalism lies in a claim of continuity, even though the strongest arguments on behalf of originalism have exactly nothing to do with that affective pull. In the end, any theory of constitutional interpretation must be justified, not on the ground that it will preserve some kind of continuity with the distant past, but on the ground that it will produce a constitutional order that deserves general support.

Could AI predict the outcome of a coin flip? Could AI have predicted in (say) 2006 that Barack Hussein Obama would be elected president of the United States in 2008? Could AI have predicted in (say) 2014 that Donald Trump would be elected president of the United States in both 2016 and 2024? Could AI have predicted in (say) 2005 that Taylor Swift would become a worldwide sensation? The answer to all of these questions is "No." AI could not have predicted those things (and no human being could have predicted those things, either). There are some prediction problems on which AI will not do well; the reason lies not in randomness, but in an absence of adequate data. There are disparate challenges here, but all of them are closely connected to the knowledge problem, and in particular to the unfathomably large number of factors that account for some kinds of outcomes and the critical importance of social interactions. In important respects, the Socialist Calculation Debate and the AI Calculation Debate are the same thing.

The current method used by the United States Government to calculate benefits and costs substantially underestimates the monetary value of some regulations. The problem is that the method does not recognize the possibility that individual valuations, reflecting judgments in an isolated, uncoordinated situation, might be significantly lower than individual valuations in a situation of coordination. For example, people might be willing to pay $X to protect members of an endangered species in their individual capacity, but far more than $X for the same purpose, assuming that many others are paying as well; one reason may be that an individual expenditure seems futile. So too, people might be willing to pay $X for a good, supposing that other people have that good, but might be willing to pay $Y to abolish that good, supposing that no one will have that good. We sketch, identify, and explain this unmeasured value, which we define as coordination value, meant as an umbrella concept to cover several categories of cases in which individual valuation might be inadequate. Changing the methodology of benefit-cost analysis to include coordination value, where it is relevant, would present serious empirical challenges, but would eliminate the undervaluation.

Artificial intelligence (“AI”), including generative AI, is not human, but restrictions on the activity or use of AI, or on the dissemination of material by or from AI, might raise serious First Amendment issues if those restrictions (1) apply to or affect human speakers and writers, or (2) apply to or affect human viewers, listeners, and readers. Here, as elsewhere, it is essential to distinguish among viewpoint-based restrictions, content-based but viewpoint-neutral restrictions, and content-neutral restrictions. Much of free speech law, as applied to AI, is in the nature of “the law of the horse”: established principles applied to a novel context. But imaginable cases raise unanswered questions, including (1) whether AI as such has constitutional rights, (2) whether and which person or persons might be a named defendant if AI is acting in some sense autonomously, and (3) whether and in what sense AI has a right to be free from—for example—viewpoint-based restrictions, or whether it would be better, and correct, to say that human viewers, listeners, and readers have the relevant rights, even if no human being is speaking. Most broadly, it remains an unanswered question whether the First Amendment protects the rights of human viewers, listeners, and readers seeking to see, hear, or read something from AI.

In 1921, John Maynard Keynes and Frank Knight independently insisted on the importance of making a distinction between uncertainty and risk. Keynes referred to matters about which ‘there is no scientific basis on which to form any calculable probability whatever’. Knight claimed that ‘Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated’. Knightian uncertainty exists when people cannot assign probabilities to imaginable outcomes. People might know that a course of action might produce bad outcomes A, B, C, D and E, without knowing much or anything about the probability of each. Contrary to a standard view in economics, Knightian uncertainty is real, and it poses challenging and unresolved issues for decision theory and regulatory practice. It bears on many problems, potentially including those raised by artificial intelligence. It is tempting to seek to eliminate the worst-case scenario, and thus to adopt the maximin rule, which might seem to be the appropriate approach under Knightian uncertainty. But serious problems arise if eliminating the worst-case scenario would (1) impose high risks and costs, (2) eliminate large benefits or potential ‘miracles’ or (3) create uncertain risks.

Food consumption and production contribute significantly to global greenhouse gas emissions, making them crucial entry points for mitigating climate change and maintaining a liveable planet. Over the past two decades, food policy initiatives have explored interventions to reshape production and consumption patterns, focusing on reducing food waste and curbing ruminant meat consumption. While the evidence of "what works" improves, evaluating which policies are appropriate and effective in specific contexts remains difficult due to external validity challenges. This paper demonstrates that a fine-tuned large language model (LLM) can accurately predict the direction of outcomes in approximately 80% of empirical studies measuring dietary-based impacts (e.g. food choices, sales, waste) resulting from behavioral interventions and policies. Approximately 75 prompts were required to achieve optimal results, with performance showing signs of catastrophic loss beyond this point. Our findings indicate that greater input detail enhances predictive accuracy, although the model still faces challenges with unseen studies, underscoring the importance of a representative training sample. As LLMs continue to improve and diversify, they hold promise for advancing data-driven, evidence-based policymaking.

Many moral judgments are rooted in the outrage heuristic. In making such judgments about certain personal injury cases, people’s judgments are both predictable and widely shared. With respect to outrage (on a bounded scale of one to six) and punitive intent (also on a bounded scale of one to six), the judgments of one group of six people, or 12 people, nicely predict the judgements of other groups of six people, or 12 people. Moreover, outrage judgments are highly predictive of punitive intentions. Because of their use of the outrage heuristic, people are intuitive retributivists. People care about deterrence, but they do not think in terms of optimal deterrence. Because outrage is category-specific, those who use the outrage heuristic are likely to produce patterns that they would themselves reject, if only they were to see them. Because people are intuitive retributivists, they reject some of the most common and central understandings in economic and utilitarian theory. To the extent that a system of criminal justice depends on the moral psychology of ordinary people, it is likely to operate on the basis of the outrage heuristic and will, from the utilitarian point of view, end up making serious and systematic errors.

For many decades, administrative law has been clouded, or perhaps haunted, by a Grand Narrative. According to that narrative, the Supreme Court has abdicated. It has allowed the modern administrative state to breach the safeguards established by Article I, Article II, and Article III. The Court permitted the breach of Article I by authorizing Congress to delegate broad discretionary authority to agencies (and thus to become legislators). The Court permitted the breach of Article II by authorizing Congress to create independent agencies, immunized from presidential control. The Court permitted the breach of Article III in two ways: (1) by giving Congress broad authority to allow administrative agencies to engage in adjudication, unprotected by the Constitution's tenure and salary provisions and (2) by granting interpretive authority to such agencies. In recent years, the Court has acted as if the Grand Narrative is essentially right. Thus it has sharply cabined Congress' power to create independent agencies; imposed new constraints on Congress' power to allow agencies to adjudicate; signaled the vitality of the nondelegation doctrine; insisted on independent judicial interpretation of law; and invoked the separation of powers, through the major questions doctrine, to limit the exercise of discretionary power by agencies. The Grand Narrative also affects other areas of administrative law, including "arbitrary or capricious" review. There are other grand narratives about administrative law (originalist, Burkean, Thayerian, and pragmatic), and they might well be more compelling; but in the current era, they are not nearly as grand, or as influential, as the Grand Narrative. Law has multiple equilibria, and the current equilibrium, if it can be called that, is one in which the Grand Narrative is on the ascendency.

Many people prize agency; they want to make their own choices. Many people also prize second-order agency, by which they decide whether and when to exercise first-order agency. First-order agency can be an extraordinary benefit or an immense burden. When it is an extraordinary benefit, people might reject any kind of interference, or might welcome a nudge, or might seek some kind of boost, designed to increase their capacities. When first-order agency is an immense burden, people might also welcome a nudge or might make some kind of delegation (say, to an employer, a doctor, an algorithm, or a regulator). These points suggests that the line between active choosing and paternalism can be illusory. When private or public institutions override people's desire not to exercise first-order agency, and thus reject people's exercise of second-order agency, they are behaving paternalistically, through a form of choice-requiring paternalism. Choice-requiring paternalism may compromise second-order agency. It might not be very nice to do that.

One way to evaluate various legal interventions in people's lives is to ask whether they make choosers better off, "as judged by themselves." This criterion can be understood to borrow from the liberal political tradition insofar as it makes the judgments of choosers authoritative. For lawyers, judges, and policymakers to give ultimate authority to choosers might be taken to respect their own judgments and to promote their welfare (insofar as people are uniquely situated to know whether choices make them better off). But for certain decisions, the "as judged by themselves" criterion is indeterminate. In these situations, involving life-changing, transformative experiences, the criterion does not offer a unique solution; people might be happy either way. It is possible that welfarist criteria will resolve the indeterminacy, despite serious questions about incommensurability.

Both Republican and Democratic administrations make regulatory and funding decisions with close reference to benefit–cost analysis (BCA). With respect to regulation, there has been a great deal of academic discussion of BCA and its limits, but almost no attention has been paid to the role of BCA in government funding. That is a serious gap, not least in connection with climate-related risks, such as wildfire, drought, extreme heat, and flooding. Office of Management and Budget (OMB) Circular A-94 sets out guidelines for the BCA required when people are applying to many federal discretionary grant programs. Through Circular A-94, OMB has long required applicants to demonstrate that the benefits of their projects would exceed the costs. But under Circular A-94 as it stood for many years, efficiency-based BCA could produce results that fail to maximize welfare and that are also highly inequitable. The 2023 revision of Circular A-94 focuses more directly on welfare and equity, which are now – not uncontroversially – being brought directly into policy. At the same time, the new Circular A-94 raises fresh questions about how best to promote welfare, and to consider equity, in practice. This article explains the economic foundations for promoting welfare through distributional weighting – and how the old BCA guidance fell short. It then offers recommendations on how to operationalize distributional weighting on the ground specifically for government spending programs – and for BCA more broadly.

“Choice Engines,” powered by Artificial Intelligence (AI) and authorized or required by law, might produce significant increases in human welfare. A key reason is that they can simultaneously (1) preserve autonomy and (2) help consumers to overcome inadequate information and behavioral biases, which can produce internalities, understood as costs that people impose on their future selves. Importantly, AI-powered Choice Engines might also take account of externalities, and they might nudge or require consumers to do so as well. Different consumers care about different things, of course, which is a reason to insist on a high degree of freedom of choice, even in the presence of internalities and externalities. Nonetheless, AI-powered Choice Engines might show behavioral biases, perhaps the same ones that human beings are known to show, perhaps others that have not been named yet, or perhaps new ones, not shown by human beings, that cannot be anticipated. It is also important to emphasize that AI-powered Choice Engines might be enlisted by insufficiently informed or self-interested actors, who might exploit inadequate information or behavioral biases, and thus reduce consumer welfare. AI-powered Choice Engines might also be deceptive or manipulative, and legal safeguards are necessary to reduce the relevant risks.

Before Trump v. United States, issues of presidential immunity were generally analyzed by reference to the following question: Would subjecting the President to a legal proceeding compromise his ability to perform his constitutional functions? After Trump v. United States, the immunity issue is analyzed, in significant part, by reference to an altogether different question: Does some provision of the Constitution give the President unquestionable power, such that neither Congress nor courts may intrude on it? It is imaginable that the second question is the right one, but the Court's analysis is brisk and conclusory; it does not seriously engage with the standard sources of constitutional interpretation. For example, there is no real effort to show that the Court's various conclusions follow from text or the original public meaning. That is a puzzle. Part of the explanation for the new framework may lie in a single sentence, pointing to "the prospect of an Executive Branch that cannibalizes itself, with each successive President free to prosecute his predecessors, yet unable to boldly and fearlessly carry out his duties for fear that he may be next." This pragmatic concern might provide the motivation for some of the Court's momentous constitutional conclusions. It links Trump v. United States with a variety of other decisions, including Bush v. Gore and Trump v. Colorado, that also seem concerned with the problem of democratic disorder-a kind of Representation Reinforcement 2.0 (or is it 1.0?). For those who insist on fidelity to legal sources, the most charitable verdict on Trump v. United States is Scottish: Not proven.

Loper Bright, overruling Chevron, is likely to produce a shift in authority from agencies to courts, by significantly increasing the likelihood that courts will strike down agency interpretations of law. But will it? And what counts as a significant increase? The answers depend, of course, on the meaning of both Chevron and Loper Bright. Under Chevron, courts hardly gave a blank check to agencies; on the contrary, they frequently invalidated agency interpretations of law. How much will invalidation rates rise? We do not know, in part (1) because Loper Bright retains Skidmore (which calls for respectful attention to agency interpretations); in part (2) because Loper Bright recognizes that Congress sometimes explicitly or implicitly delegates interpretive authority to agencies; and in part (3) because (and these must be counted as some of its effects) Loper Bright will (a) increase litigants' incentive to attack agency interpretations and (b) reduce agencies' incentive to adopt adventurous interpretations. Any numerical projection would be hazardous, but Loper Bright gives a clear signal, a green light to federal courts where Chevron gave a yellow light-which means that it is reasonable to predict a nontrivial increase in judicial invalidations (other things being equal). Still, current evidence is consistent with the following proposition: If the goal is to predict whether a court of appeals will strike down an agency's interpretation of law, it may be more important to know whether the panel consists of Republican or Democratic appointees, and whether the agency's interpretation leans left or right, than to know whether Chevron or Loper Bright is the governing law. It is safe to predict that in the near future, the combination of Loper Bright with increasing judicial skepticism about the administrative state will result in a nontrivial increase in invalidation of regulations designed to protect health, safety, and the environment. It is also safe to predict that in the near future, Loper Bright will lead to a significant increase in ideological divisions in the lower courts.

Many consumers suffer from inadequate information and behavioral biases, which can produce internalities, understood as costs that people impose on their future selves. In these circumstances, “Choice Engines,” powered by Artificial Intelligence (AI), might produce significant savings in terms of money, health, safety, or time. Different consumers care about different things, of course, which is a reason to insist on a high degree of freedom of choice, and a high degree of personalization. Nonetheless, it is important to emphasize that Choice Engines and AI might be enlisted by self-interested actors, who might exploit inadequate information or behavioral biases, and thus reduce consumer welfare. It is also important to emphasize that Choice Engines and AI might show behavioral biases, perhaps the same ones that human beings are known to show, perhaps others that have not been named yet, or perhaps new ones, not shown by human beings, that cannot be anticipated.

Judges and lawyers refer to "the" separation of powers, but the term is an umbrella concept, referring to six different propositions, or six separations of powers. (1) The legislature may not exercise the executive power. (2) The legislature may not exercise the judicial power. (3) The executive may not exercise the legislative power. (4) The executive may not exercise the judicial power. (5) The courts may not exercise the legislative power. (6) The courts may not exercise the executive power. None of these propositions is without ambiguity and all of them must be qualified, but each can be understood to have a core of both meaning and truth. If the goal is to protect liberty or self-government, every one of the six propositions can be strongly defended, but they raise different considerations, and they must be analyzed separately. None of them is a logical truth; all of them rest on empirical judgments, involving the likely capacities and performance of various institutions, that are more than plausible but that may or may not be correct.

At points in American history, there have been significant, even massive shifts in constitutional understandings, doctrines, and practices. Apparently settled principles, and widely accepted approaches, are discarded as erroneous, even illegitimate, in favor of new principles and approaches. Less momentously, views that were once considered unthinkable do not quite become the law on the ground but instead come to be seen as plausible and part of the mainstream. Relatedly, Americans transform how they talk and think about their Constitution – its core commitments and underlying narratives – and those transformations change our practices. Our goal is to provide a conceptual map of radical constitutional change and to describe how and why such change occurs. First, we ask whether theories of interpretation trigger radical change or whether desires for radical change impel people to generate new (or modify old) theories of interpretation. Second, we explore why so many people are baffled or outraged by constitutional paradigm shifts. Third, we investigate the drivers of radical constitutional change, both the familiar bottom-up pressures from “We the People” and the less-familiar top-down approaches, where legal elites back and impose a new constitutional regime. Given that we might well be in the early innings of a radical constitutional change, it is an apt moment to theorize about the phenomenon writ large.

Many consumers suffer from inadequate information and behavioral biases, which can produce internalities, understood as costs that people impose on their future selves. In these circumstances, “Choice Engines,” powered by Artificial Intelligence (AI), might produce significant savings in terms of money, health, safety, or time. Consider, for example, choices among motor vehicles or appliances. AI-powered Choice Engines might also take account of externalities, and they might nudge or require consumers to do so as well. Different consumers care about different things, of course, which is a reason to insist on a high degree of freedom of choice, even in the presence of internalities and (to some extent) externalities. But it is important to emphasize that AI might be enlisted by insufficiently informed or self-interested actors, who might exploit inadequate information or behavioral biases, and thus reduce consumer welfare. AI might increase internalities or externalities. It is also important to emphasize that AI might show behavioral biases, perhaps the same ones that human beings are known to show, perhaps others that have not been named yet, or perhaps new ones, not shown by human beings, that cannot be anticipated.

The United States has an Economic Constitution, governing federal regulation, and explaining how to conduct regulatory impact analysis, with reference to quantification and monetization of the costs and benefits of proposed and final regulations. Known as OMB Circular A-4, the Economic Constitution of the United States was thoroughly revised in 2023, with new directions on behavioral economics and nudging; on discount rates and effects on future generations; on distributional effects and how to account for them; and on benefits and costs that are hard or impossible to quantify. The revised document leaves numerous open questions, involving (for example) the valuation of human life, the valuation of morbidity effects, and the value of the lives of children.

For fans of Thinking Fast and Slow and The Power of Habit, a groundbreaking new study of how disrupting our well-worn routines, both good and bad, can rejuvenate our days and reset our brains to allow us to live happier and more fulfilling lives.Have you ever noticed that what is thrilling on Monday tends to become boring on Friday? Even exciting relationships, stimulating jobs, and breathtaking works of art lose their sparkle after a while. People stop noticing what is most wonderful in their own lives. They also stop noticing what is terrible. They get used to dirty air. They stay in abusive relationships. People grow to accept authoritarianism and take foolish risks. They become unconcerned by their own misconduct, blind to inequality, and are more liable to believe misinformation than ever before. But what if we could find a way to see everything anew? What if you could regain sensitivity, not only to the great things in your life, but also to the terrible things you stopped noticing and so don’t try to change? Now, neuroscience professor Tali Sharot and Harvard law professor (and presidential advisor) Cass R. Sunstein investigate why we stop noticing both the great and not-so-great things around us and how to “dishabituate” at the office, in the bedroom, at the store, on social media, and in the voting booth. This groundbreaking work, based on decades of research in the psychological and biological sciences, illuminates how we can reignite the sparks of joy, innovate, and recognize where improvements urgently need to be made. The key to this disruption—to seeing, feeling, and noticing again—is change. By temporarily changing your environment, changing the rules, changing the people you interact with—or even just stepping back and imagining change—you regain sensitivity, allowing you to more clearly identify the bad and more deeply appreciate the good.

This is the original version of the Tanner Lectures on Human Values, delivered at Harvard University in 1994. The central question is this: How is law possible in a heterogeneous society, composed of people who sharply disagree about basic values? Such disagreements involve the most important issues of social life: the distribution of wealth, the role of race and gender, the nature of free speech and private property. Much of the answer to this puzzle lies in an appreciation of how people who disagree on fundamental issues can achieve incompletely theorized agreements on particular cases. Lecture I sets out the basic idea of incompletely theorized agreements and argues that such agreements have many virtues. Lecture II opposes rules to rulelessness. Its principal goal is to point the way toward a more refined understanding of the ideal of the rule of law, one that sees a degree of particularity, and a degree of lawmaking at the point of application, as an important part of that ideal. The lectures end with the suggestion that incompletely theorized agreements on particular outcomes play a large role not only in law, but also in many other sectors of social life, prominently including democratic discussion.

The Court’s opinion in Dobbs v. Jackson Women’s Health Organization embraces a form of due process traditionalism. With Burkean and Thayerian arguments at work, the Court offers an understanding of the Due Process Clause which allows substantive protection of rights only if they are vindicated by tradition and essential to “ordered liberty.” Within the opinion’s logic, the major challenge is to accept due process traditionalism without simultaneously throwing a variety of nontraditional or antitraditionalist substantive due process cases into doubt. There are uneasy relationships between the traditionalist thrust of the Court’s opinion and the Court’s nontraditional or antitraditionalist jurisprudence in other areas of constitutional law. A central reason must be that the Court believes in some forms of moral progress. The central weakness of the Dobbs opinion is its rejection of the idea that moral progress can and should play a role in the understanding of constitutional rights.

Scientific evidence regularly guides policy decisions, with behavioural science increasingly part of this process. In April 2020, an influential paper proposed 19 policy recommendations (‘claims’) detailing how evidence from behavioural science could contribute to efforts to reduce impacts and end the COVID-19 pandemic. Here we assess 747 pandemic-related research articles that empirically investigated those claims. We report the scale of evidence and whether evidence supports them to indicate applicability for policymaking. Two independent teams, involving 72 reviewers, found evidence for 18 of 19 claims, with both teams finding evidence supporting 16 (89%) of those 18 claims. The strongest evidence supported claims that anticipated culture, polarization and misinformation would be associated with policy effectiveness. Claims suggesting trusted leaders and positive social norms increased adherence to behavioural interventions also had strong empirical support, as did appealing to social consensus or bipartisan agreement. Targeted language in messaging yielded mixed effects and there were no effects for highlighting individual benefits or protecting others. No available evidence existed to assess any distinct differences in effects between using the terms ‘physical distancing’ and ‘social distancing’. Analysis of 463 papers containing data showed generally large samples; 418 involved human participants with a mean of 16,848 (median of 1,699). That statistical power underscored improved suitability of behavioural science research for informing policy decisions. Furthermore, by implementing a standardized approach to evidence selection and synthesis, we amplify broader implications for advancing scientific evidence in policy formulation and prioritization.

In this retrospective honoring the exemplary psychologist Daniel Kahneman (1934–2024), the authors present a curated selection of quotes from the academic community reflecting on his ideas. These submissions, gathered from a wide range of scholars, highlight Kahneman’s contributions to fields spanning attention, judgment, decision-making, and well-being. From his exploration of cognitive biases to his groundbreaking work on prospect theory, Kahneman’s research revolutionized researchers’ understanding of human behavior and decision-making. Beyond his research, many quotes also emphasize Kahneman’s thoughts on what it means to be a behavioral scientist—focusing on a commitment to criticism, transparency, and adversarial collaboration; showcasing the dynamic nature of scientific inquiry across disciplinary divides; and highlighting his dedication to advancing the greater good. Together, these reflections paint a portrait of a visionary thinker whose theoretical and meta-scientific contributions have left an indelible mark on psychology and other social sciences.

Does the Constitution guarantee free speech on university campuses? The Supreme Court has essentially said so, at least insofar as we are speaking of public universities. Some private universities are interested in adhering to the First Amendment, even though it does not bind them. But what do First Amendment protections entail or require? It is clear that in general, universities may not discriminate on the basis of viewpoint. It is also clear that as educational institutions, universities may sometimes regulate speech to protect their essential mission – by, for example, forbidding “true threats,” prohibiting plagiarism, protecting speakers from being shouted down, forbidding students from taking over buildings, and ensuring that students and teachers focus on the topics of their courses. By exploring thirty-seven scenarios, it is possible to concretize these general propositions, and to see which questions are easy and which questions are hard. The broadest conclusion is that to the extent that universities seek to comply with the First Amendment, they must permit a great deal of speech that is offensive, hateful, and even horrifying.

The foundations of modern administrative law were laid in 1980, with the disparate opinions of a sharply divided Court in Industrial Union Department, AFL-CIO v. American Petroleum Institute (commonly referred to as the "Benzene Case"). Consider four points. (1) The Benzene Case is now understood to be the first contemporary appearance of the Major Questions Doctrine. (2) The Benzene Case marked the return of the nondelegation doctrine, signaled most plainly by then-Justice William Rehnquist's elaborate concurring opinion but also by a favorable reference in the plurality opinion by Justice John Paul Stevens and an open-minded sentence from Justice Lewis Powell. (3) The Benzene Case is the origin of contemporary cost-benefit default principles, permitting the requiring agencies to exempt de minimis risks, to consider costs, and to engage in some form of cost-benefit balancing, unless Congress has squarely said otherwise. (4) The Benzene Case essentially defined "significant risk," with a precise numerical definition (one in one thousand) that persists at the Department of Labor to this day. At the same time, a close analysis of the plurality opinion in the Benzene Case shows that it is best understood as a specification, above all, of the Absurdity Canon--a Church of the Holy Trinity v. United States for the modern administrative state--with the specific purpose of ensuring against the imposition of high costs for small benefits, and thus of requiring a kind of proportionality between costs and benefits. So understood, the Benzene Case had, and continues to have, an important and salutary effect on regulatory programs. Its significant and much broader current role, more than four decades after the opinions were issued, is an intriguing case study in doctrinal development, and in particular, how Supreme Court decisions can plant small seeds that become big trees.

Liberals see human beings as subjects, not objects. They reject despotism in its many forms. With Justice Robert Jackson, liberals believe that “compulsory unification of opinion achieves only the unanimity of the graveyard.” Liberal authoritarianism is an oxymoron. Illiberal democracy is illiberal, and liberals oppose it for that reason. Liberals are puzzled by many of those, on the left and the right, who describe themselves as “antiliberal” or “postliberal.” With respect to some claims of “antiliberals” or “postliberals,” liberals agree with Amos Tversky and Daniel Kahneman: “The refutation of a caricature can be no more than a caricature of refutation.” With respect to those claims of “antiliberals” or “postliberals” that do not amount to a caricature, liberals insist on the importance of freedom of thought and action and deliberative democracy, and on the need to respect reasonable pluralism.

Are administrative agencies illegitimate? A threat to democracy? A threat to liberty? To human welfare? To defining constitutional commitments? Many people think so. But an understanding of the actual operation of the administrative state, seen from the inside, makes it difficult to object to “rule by unelected bureaucrats” or “an unelected fourth branch of government.” Such an understanding casts a new light on some large objections from the standpoint of democracy, liberty, and welfare; indeed, it makes those objections seem coarse and largely uninformed. What is needed is more conceptual and empirical work on welfare and distributive justice, and on how regulators can increase both.