By identifying well-being with preference satisfaction, mainstream normative economists were able to leave the determination of which specific things make people better or worse off to the individuals themselves. The findings of behavioral economics undermine the possibility of deferring in this way to individual preference. One response to this challenge to welfare economics is to distinguish the true preferences of individuals from their manifest preferences and to take true preferences to guide policy. In The Community of Advantage, Robert Sugden criticizes this strategy and proposes that economists appraise policies, institutions and outcomes by the opportunities they provide rather than by their contributions to welfare. This paper criticizes Sugden's view and argues for a modest solution that makes cautious use of preferences as indicators of well-being.

A behavioural public policy movement has flourished within the global policy realm. While this movement has been deemed interdisciplinary, incorporating behavioural science theories and methods in a neoclassical economics-governed policy process, this paper analyses the bounded form of interdisciplinarity that characterizes it. We claim that an engagement is missing with the broader sweep of social sciences, which share similar concerns but deploy different analytical perspectives from those of behavioural public policy. Focusing on two central concepts (context and evidence), we aim to show how behavioural public policy's bounded interdisciplinarity implies constrained understandings of context and evidence, thereby limiting its complex problem-solving abilities. At the same time, we highlight some alternative examples of behavioural public policy practice that do explore new critical interdisciplinary horizons.

Defaults may not directly get people to behave as intended, such as saving more, eating healthy food or donating to charity. Rather, defaults often only put people on the ‘right’ path, such as joining a savings plan, buying healthy food or pledging money to charity. This an issue because getting more people to take those first steps does not necessarily motivate them to go on with further steps. Indeed, the default does little to help them understand the benefit of doing so. This can greatly reduce the impact of the default. We test this idea in a charitable giving experiment where people first can promise to give to charity (‘pledge’) and then can go on to donate. We find that participants pledge more often when that is the default, but those who pledge in that case are less likely to take further steps to donate than those who pledge when pledging is against the default. We interpret this in terms of motivation and transaction costs. Some people pledge only to avoid the psychological costs of going against the default. Those people are closest to indifference between donating or not and are therefore less motivated to go on to donate. We also show that the intrinsic motivation of pledgers is lower when pledging is the default and that making pledges the default does not change attitudes to charities.

Behavioral economics research has helped with understanding charitable behavior and has shown that charities can encourage donations by carefully designing their pledges. However, there is still scope to extend current research on who gives, what drives the decision to donate and at what levels, especially when behavioral insights are applied in context. In cooperation with a major Italian charity for cancer research, this study implements a natural direct mail field experiment, with over 150,000 letters sent to donors. By exploring the behavioral responses to different donation anchors, evidence was found that, within the given framework, including donation menus significantly increased the average amount donated without affecting the likelihood of donation. Furthermore, introducing additional explanations of how to make a payment significantly increased overall returns. Lastly, individual heterogeneity (high- and low-frequency donors, as well as senior and junior donors) had a direct effect on donations.

We explore whether simple behavioural insights techniques can be successful for addressing a policy issue within one of society's more complex and difficult sectors: child protection. Child protection reporting practices in New South Wales, Australia, reveal that the public's primary response is to report to the statutory authority, who only deal with cases of the highest risk. As a result, a large volume of statutory resources are spent processing lower-risk reports that lead to no benefits for lower-risk families and slow down response times for families that require a statutory response most. Our goal was to reduce lower-risk reporting by encouraging alternative responses to these situations. To do this, we altered report feedback for cases deemed lower risk in order to make alternative responses more salient and we added a persuasive message framed as a gain or loss. We then examined subsequent reporting accuracy. We found that our trial was linked to a modest improvement in reporting accuracy, though the results may have been diluted by a spill-over effect. We discuss how facilitating a greater behavioural change likely requires multi-organization collaborations, extending the range of insights drawn from behavioural science and/or addressing issues from multiple angles.

Under what circumstances will members of the public hold positive attitudes toward new or higher taxes? While some scholars have posited that the practice of “earmarking” – designating tax revenues for a particular purpose – can increase support for taxes, the existing literature has not identified the conditions under which earmarking will prove effective in this regard. Here, we draw upon previous research on consumer behavior to hypothesize that support for earmarked taxes will be stronger when such taxes satisfy the criterion of “source–use alignment” (i.e., when the connection between the revenue source and the use for which those revenues are earmarked accords with familiar consumer fairness norms). Evidence in support of this hypothesis comes from two experiments on a sample of US residents matched to Census data, in which subjects were randomly assigned to read descriptions of hypothetical earmarked taxes with varying levels of alignment. Individuals consistently expressed stronger support for earmarked taxes that achieved source–use alignment as compared to earmarked taxes that did not satisfy the source–use alignment criterion. Our theory and results not only help to explain why some earmarked taxes are more popular than others, but also suggest a means for increasing public support for taxes.

This paper compares the effects of two types of behaviorally informed policy – nudges and boosts – that are designed to increase consumer demand for insurance against low-probability, high-consequence events. Using previous findings in the behavioral sciences literature, this paper constructs and implements two nudges (an ‘informational’ and an ‘affective’ nudge) and a statistical numeracy boost and then elicits individual risk beliefs and demand for flood insurance using a contingent valuation survey of 331 participants recruited from an online labor pool. Using a two-limit Tobit model to estimate willingness to pay (WTP) for flood insurance, this paper finds that the affective and informational nudges result in increases in WTP for flood insurance of roughly $21/month and $11/month relative to the boost, respectively. Taken together, the findings of this paper suggest that nudges are the more effective behaviorally informed policy in this setting, particularly when the nudge design targets the affect and availability heuristics; however, additional research is necessary to establish sufficient conditions for this conclusion.