The mechanisms of nudging

Nudges can be divided into two broad categories: architectural and educative (Sunstein, Reference Sunstein2022). Architectural nudges target the choice architecture, or how choices are presented to people. This can include the placement of options in a list or their physical location in space, the language used to present them, and whether people have to opt in or out of them. For instance, placing healthier foods at eye level in a shop is a form of architectural nudging. Educative nudges involve providing warnings, reminders, and information to guide decision-making. This includes communicating descriptive norms informing people about others’ behaviour (Cialdini, Reference Cialdini2007). For example, a sign in a hotel bathroom saying ‘75% of our guests reuse their towels to protect the environment’ encourages people to follow the majority’s eco-friendly behaviour.

The nudge approach has been described as a form of libertarian paternalism (Thaler and Sunstein, Reference Thaler and Sunstein2003). It is paternalistic because an external authority prompts people to act in ways that the authority has decided is beneficial, individually and/or collectively. It is libertarian because it maintains freedom of choice and does not remove options or force people to act in a particular way.

The appeal of nudging

The nudge approach has proved popular among policymakers, arguably for several reasons. First, nudge policies can be highly cost-effective (Benartzi et al., Reference Benartzi, Beshears, Milkman, Sunstein, Thaler, Shankar, Tucker-Ray, Congdon and Galing2017). Even when they result in very small changes, they do so at low cost and the population-level impact can be significant (Halpern, Reference Halpern2015). Second, because nudging does not involve outright prohibitions or potentially costly support programmes, it appeals to governments that are averse to overt regulation and that advocate for fiscal conservatism and reduced government involvement in societal affairs (Halpern, Reference Halpern2015). Third, unlike other policy options, nudging does not require governments to take on corporate interests. A study of 11 countries that tried to impose sugar taxes on soft drinks, for example, found that all faced energetic lobbying and criticism (Lauber et al., Reference Lauber, Rippin, Wickramasinghe and Gilmore2022). Fourth, the nudge approach can absolve governments from their responsibility to address collective and societal problems. Framing issues like obesity, ill-health, and poverty as failures of individual decision-making diverts attention from systemic drivers of behaviour for which governments are responsible, such as allowing the production and dissemination of harmful products (Nestle, Reference Nestle2015; Kozyreva et al., Reference Kozyreva, Lewandowsky and Hertwig2020; Chater and Loewenstein, Reference Chater and Loewenstein2023). Fifth, nudge arguments align with broader views that politicians often hold about the public. The Hobbesian legitimization of government is that, left to themselves, individuals are incapable of creating a safe and stable society (Hobbes, Reference Hobbes1651/2016). Nudging can be seen as providing a ‘scientific’ justification for this viewpoint and a logical solution for societal problems. The more individuals are portrayed as cognitively and motivationally deficient, the more credible it becomes to blame them for societal issues and to cast nudging as a justifiable solution.

The deficit model of human cognition

Proponents of the nudge approach frequently refer to an extensive body of research – ‘thousands of studies’ (Thaler and Sunstein, Reference Thaler and Sunstein2021, 10) – purportedly showing that human thinking is flawed and biased. Indeed, the heuristics-and-biases programme, started by psychologists Daniel Kahneman and Amos Tversky in the early 1970s (Tversky and Kahneman, Reference Tversky and Kahneman1974), has profoundly changed how many psychologists and behavioural economists perceive the quality of people’s statistical reasoning, intuitions, inferences, and choices, as well as the cognitive mechanisms, or heuristics, that underlie these behaviours.

The main message of the heuristics-and-biases programme is a specific interpretation of Herbert Simon’s (Reference Simon1956) foundational concept of bounded rationality. Simon, a Nobel Prize laureate in economics, was a vocal critic of the rational choice framework – the classical model of rationality – as a descriptive model of human choice. He noted that this model expects individuals to have unlimited knowledge, computational power, and time to make perfectly rational decisions. Such an ‘Olympian model’ (Simon, Reference Simon1983, 19) of rational choice was perhaps suitable for omniscient gods, but unrealistic in the real world. As an alternative, Simon proposed the concept of bounded rationality, which acknowledges that people often approximate rather than optimize when making decisions.

Whereas Simon emphasized that people’s approximate decisions can be adaptive, good enough, and satisficing because the ‘environments to which [organisms] must adapt possess properties that permit further simplifications of its choice mechanisms’ (Simon, Reference Simon1956, 129), the heuristics-and-biases programme has interpreted bounded rationality to mean systematic biases in human judgment and choice. The research objective, as stated by Kahneman, was ‘to obtain a map of bounded rationality, by exploring the systematic biases that separate the beliefs that people have and the choices they make from the optimal beliefs and choices assumed in rational-agent models’ (Kahneman, Reference Kahneman2003, 1449). Richard Thaler, who worked closely with Kahneman and Tversky, famously reinforced this perspective by concluding that ‘mental illusions should be considered the rule rather than the exception’ (Thaler, Reference Thaler1991, 4).

The deficit view extends beyond cognitive deficits to include motivational deficiencies: the idea that people’s motivations are inherently maladaptive and that they cannot be trusted to do the right thing. Take the example of the UK government’s response to the COVID-19 pandemic. In early March 2020, the UK’s Chief Medical Officer, Chris Whitty, expressed concern that the public would be unable or unwilling to adhere to virus containment policies for long: ‘[T]here is a risk if we go too early, people will understandably get fatigued and it will be difficult to sustain … over time’ (The Telegraph, 2020). Similarly, the government feared that ‘anything too onerous suggested by the government … might be adopted enthusiastically for a few weeks but then people get bored and leave their homes just as the peak of the illness hits’ (Proctor, Reference Proctor2020). As senior cabinet minister Michael Gove testified to the UK COVID-19 Inquiry, there was a ‘broad view at the time that … the public … would not endure [lockdown measures] for long’ (UK Covid-19 Inquiry, 2023). Consequently, stay-at-home measures were only implemented on 23 March 2020, a delay that resulted in possibly tens of thousands of unnecessary deaths (Stewart and Sample, Reference Stewart and Sample2020).

Another example of the deficit view in the public discourse concerns democratic processes. A strand of libertarian scholars has challenged the quality and, by extension, the legitimacy of democratic decision-making. To support their ‘skepticism about democracy, these libertarians appeal to findings in cognitive and social psychology and political behavior to claim that decision making by ordinary citizens is unlikely to be rational or well grounded in evidence’ (Farrell et al., Reference Farrell, Mercier and Schwartzberg2023, 767). Voters are characterized as irrational (Caplan, Reference Caplan2008) and ‘systematically incompetent’ (Brennan, Reference Brennan2016, 201). Based on such arguments it has been asserted that ‘rule by demagogues … is the natural condition of democracy’ (Caplan, Reference Caplan2008, 19). There is no mincing of words when it is suggested that some individuals ‘ought not have the right to vote, or ought to have weaker voting rights than others’ (Brennan, Reference Brennan2016, viii). Additionally, it has been claimed that ‘widespread public ignorance is a type of pollution’ and that ‘democracy might function better if its powers were more tightly limited’ (Somin, Reference Somin2016, 6, 9).

The focus on individuals

Nudge theory focuses on individual decision-making. Yet this is often too narrow a focus, as illustrated by the ‘behavioural fatigue’ line adopted in the UK in the first wave of COVID-19. The assumption that people would not stick to stay-at-home measures for long was proven wrong; levels of adherence remained high despite psychological and economic hardships (Duffy and Allington, Reference Duffy and Allington2020). It soon became clear that a sense of shared identity and community was a critical factor in maintaining adherence (Jackson et al., Reference Jackson, Posch, Bradford, Hobson, Kyprianides and Yesberg2020), a finding subsequently corroborated by multiple studies (Stevenson et al., Reference Stevenson, Wakefield, Felsner, Drury and Costa2021; Van Bavel et al., Reference Van Bavel, Cichocka, Capraro, Sjåstad, Nezlek, Pavlović, Alfano, Gelfand, Azevedo, Birtel, Cislak, Lockwood, Ross, Abts, Agadullina, Aruta, Besharati, Bor, Choma, Crabtree, Cunningham, De, Ejaz, Elbaek and Boggio2022). This comes as no surprise to those who study mass behaviour in emergencies. Research has consistently shown that people tend not to panic or act destructively in crises (Johnson, Reference Johnson1987; Drury et al., Reference Drury, Novelli and Stott2013). Rather, a sense of common fate engenders a feeling of shared identity which, in turn, leads people to support and expect support from others (Drury, Reference Drury2018). This shared identity helps to sustain adherence to demanding measures in hard times.

Moreover, focusing solely on individual limitations neglects the crucial role of trust in behaviour change. Governments basing policy on an individual deficit model which assumes that people are neither willing nor able to make good choices are unlikely to use strategies known to build shared identity and trust, such as listening to the public, engaging with them, co-producing policy, and showing respect (Tyler and Blader, Reference Tyler and Blader2003; Bonell et al., Reference Bonell, Michie, Reicher, West, Bear, Yardley, Curtis, Amlôt and Rubin2020). Trust was vital to the success of public health responses during the COVID-19 pandemic (Bollyky et al., Reference Bollyky, Angelino, Wigley and Dieleman2022; Lenton et al., Reference Lenton, Boulton and Scheffer2022) and played a significant role in vaccine uptake, both internationally and in the UK (Paul et al., Reference Paul, Fancourt and Razai2022; Viskupič et al., Reference Viskupič, Wiltse and Meyer2022; Allington et al., Reference Allington, McAndrew, Moxham-Hall and Duffy2023). In autumn 2021, while over 90% of White Britons had been vaccinated, the figure for Black Britons was only around 60% (Dolby et al., Reference Dolby, Finning, Baker, Fowler-Dowd, Khunti, Razieh, Yates and Nafilyan2022). A lack of trust contributed to this disparity; 60% of Black Britons felt that health services were less concerned with their issues than with those of White people (Joint Committee on Human Rights, 2020). Addressing this mistrust through a process of engagement (Burgess et al., Reference Burgess, Osborne, Yongabi, Greenhalgh, Gurdasani and Kang2021) – by going to Black communities, working through community representatives, and listening and responding to concerns – proved highly effective (Halvorsrud et al., Reference Halvorsrud, Shand, Weil, Hutchings, Zuriaga, Satterthwaite, Yip, Eshareturi, Billett, Hepworth, Dodhia, Schwartz, Penniston, Mordaunt, Bulmer, Barratt, Illingworth, Inskip, Bury, Jenkins, Mounier-Jack and Raine2023). However, at a national level, such engagement was largely absent. Instead of fostering dialogue and trust, government representatives voiced disdain, labelling those who remained unvaccinated as ‘selfish.’ (BBC, Reference BBC2021).

Boosting competences

The boosting approach to behavioural public policy works by harnessing and building on human strengths (Hertwig and Grüne-Yanoff, Reference Hertwig and Grüne-Yanoff2017). Boosts are interventions designed to improve people’s competences to make informed choices that align with their goals, preferences, and desires. To illustrate, Table 1 lists six societal challenges alongside the competences needed to address them and boost interventions that have proved successful in developing those competences.

Table 1. Six examples of global challenges, competences needed to address them, and tried-and-tested boosts

By fostering existing cognitive and motivational competences or instilling new ones, boosts make it easier for people to exercise their own agency. The emphasis on agency in the boost approach has both ethical and efficacy dimensions. From an ethical standpoint, individual autonomy and freedom cannot be achieved without nurturing citizens’ agency. From an efficacy perspective, interventions that fail to promote agency – as many nudging interventions do – risk leaving citizens in the dark, unable to take ownership of the process of behaviour change. This lack of agency can have several negative outcomes: non-persistent treatment effects, compensatory negative spillovers, or psychological reactance and backfiring effects (Banerjee et al., Reference Banerjee, Grüne-Yanoff, John and Moseley2024). What is more, failing to engage with and listen to people may undermine trust (Tyler and Blader, Reference Tyler and Blader2003), which, as we have seen, is critical for successful behaviour change.

Boosts can be classified according to the kinds of competences they build or enhance. Digital literacy boosts involve strategies like lateral reading, modelled after the methods used by professional fact checkers to efficiently and effectively assess the credibility of unfamiliar websites, posts or information (Kozyreva et al., Reference Kozyreva, Lorenz-Spreen, Herzog, Ecker, Lewandowsky, Hertwig, Ali, Bak-Coleman, Barzilai, Basol, Berinsky, Betsch, Cook, Fazio, Geers, Guess, Huang, Larreguy, Maertens, Panizza, Pennycook, Rand, Rathje, Reifler, Schmid, Smith, Swire-Thompson, Szewach, van der Linden and Wineburg2024). Risk literacy boosts include experienced simulations of risks that help people understand the temporal and cumulative nature of health risks (Wegwarth et al., Reference Wegwarth, Ludwig, Spies, Schulte and Hertwig2022). Financial literacy boosts might employ simple heuristics to help people understand compound interest and exponential growth (Foltice, Reference Foltice2017) or enable microentrepreneurs to clearly separate business and private accounts with the help of physical analogues (Drexler et al., Reference Drexler, Fischer and Schoar2014). Statistical reasoning boosts might contrast correct solutions with typical biased responses in one-shot (Morewedge et al., Reference Morewedge, Yoon, Scopelliti, Symborski, Korris and Kassam2015) or repeated trials (Franiatte et al., Reference Franiatte, Boissin, Delmas and De Neys2024), or train people to transform complex probability representations into simpler frequency representations (Sedlmeier and Gigerenzer, Reference Sedlmeier and Gigerenzer2001). Health literacy boosts enable people to understand drug labels (Sahm et al., Reference Sahm, Wolf, Curtis, Behan, Brennan, Gallwey and McCarthy2012), recognize and monitor symptoms, and employ simple strategies like using implementation intentions to improve self-regulation (Oettingen et al., Reference Oettingen, Hönig and Gollwitzer2000). Decision-making boosts such as decision trees support decision-making in domains such as finance, medicine, human resources, and science education (Katsikopoulos et al., Reference Katsikopoulos, Şimşek, Buckmann and Gigerenzer2021; Osborne and Pimentel, Reference Osborne and Pimentel2022). Motivation boosts can help to overcome maths anxiety and enable parents and children to engage playfully with maths tasks (Berkowitz et al., Reference Berkowitz, Schaeffer, Maloney, Peterson, Gregor, Levine and Beilock2015).

Boosting can also empower citizens to turn the choice architectures around them into strategic allies, a concept known as self-nudging. In self-nudging, the individual serves as both the nudger and the nudgee, effectively becoming a citizen choice architect. This approach addresses several major objections to nudging in addition to those discussed above, such as concerns about paternalism and the uncertainty and heterogeneity of preferences (Reijula and Hertwig, Reference Reijula and Hertwig2022). An example of self-nudging is the mobile phone app one sec (Grüning et al., Reference Grüning, Riedel and Lorenz-Spreen2023), which empowers users to substantially reduce their social media consumption. (See Herzog and Hertwig, Reference Herzog and Hertwig2025, for a comprehensive review of boosts.)

Boosting involves more than just enhancing people’s competences; it also means shaping the environment to maximize individuals’ ability to use those competences. All behaviours involve an interaction of competences, opportunities, and motivations (Michie et al., Reference Michie, van Stralen and West2011). Competences are only useful to the extent that the environment enables them to be deployed effectively. For example, understanding the importance of ventilating indoor spaces to avoid the spread of airborne viruses is useless if buildings do not allow for windows to be opened. Similarly, understanding the importance of self-isolating when infected with a dangerous pathogen will only get a society so far if citizens lack the necessary economic, practical, and social support to do so.