Introduction

For those interested in social change and democratic politics, the fundamental insight of the COVID-19 pandemic arguably does not pertain to disease control, social epidemiology or the role of ‘the experts’ within policymaking processes. A more provocative perspective might suggest the main insight emerging from the pandemic actually relates to how it (1) exposed the existence of deep-seated embedded structural inequalities and, therefore, (2) posed a significant challenge to fundamental neoliberal ethical commitments regarding the reach and role of the state. From the United Kingdom to the United States, socioeconomic inequalities have shaped the COVID-19 burden in ways that reflect pre-pandemic social vulnerabilities. In the United Kingdom, this has been starkly reflected in the disproportionate impact that the disease has had upon black and ethnic minority communities. In this context, what is interesting about COVID-19 is how long-standing assumptions about the respective roles of the state and the individual shifted dramatically in ways that pose distinctive questions for post-pandemic models of democratic governance. The crisis opened a ‘window of opportunity’ that facilitated the discussion and implementation of policies that would otherwise have been rejected as either not credible or not ethically permissible options. A case in point is the UK Coronavirus Job Retention Scheme (i.e., ‘furlough’): a rapidly established financial safety net protecting 11.6 million jobs at a cost of around £70 billion.

Simply put, this redrew the boundary between the public and private spheres in ways that undermined long-standing neoliberal commitments to upholding formal equality between citizens by sustaining neutrality between conceptions of the good. That formal commitment to neutrality meant, for example, that promoting specific outcomes in health and wellbeing by funding prevention measures and institutions through taxation is coercive since it treats individuals as means to the ends of others. That position justified the wholesale rolling back of the state from engagement in the provision and mitigation of what are known as social determinants of health and wellbeing, crime, and many other outcomes. These social determinants include wealth, housing, food, education, and employment. On the neoliberal account, engineering social determinants is impermissible because it undermines state neutrality concerning conceptions of the good and thereby undermines equality between citizens, and, insofar as it involves restricting choice and taxing citizens to fund interventions, it is also coercive.

The neoliberal commitment to neutrality has created a series of administrative issues for neoliberal governments since a large body of evidence indicates the role of social determinants in shaping behaviour and outcomes and the need to intervene in them in order to address pressing policy issues. In terms of health, there has been a significant increase in lifestyle conditions since the deregulation of society in the late 1970s. Neoliberals generally regard these as matters of conscience that affect the functioning of society as a whole and create demands for more interventionist institutional approaches. To address this, scholars such as Thaler and Sunstein (2008) advanced what they termed libertarian paternalist tools of governance, such as ‘nudging’, that draw on behavioural scientific understandings of choice to improve outcomes in health, wealth and happiness without requiring the state to roll forward in mitigating social determinants. Nudging was presented as a system for using salient behavioural cues to guide people toward making healthier and less harmful choices. This was an ethical fudge insofar as it entailed acceptance that some outcomes are better than others and that those outcomes ought to be promoted by the state while its proponents, nevertheless, professed that it permitted the state to avoid coercive intervention by pursuing approaches that required little taxation and left open the possibility of people making conscience-based decisions that harm their outcomes.

The problem that the pandemic highlighted was that this ethical fudge was also incapable of addressing significant social problems. In the initial stages of the pandemic, the UK Government sought to deploy nudging as a means of mitigating the disease’s spread. The failure of that approach stood at odds with long-standing government investment in nudging, including through the Behavioural Insights Team. Indeed, in the wake of the pandemic, the underpinning evidence used to warrant the approach has been challenged. There appears to be little evidence for the efficacy of nudging in many policy areas independent of publishing bias, and several key publications underpinning the method have recently met with allegations of research misconduct (O’Grady, 2023).

In this article, we set out the ethical challenge that the evidence on nudging and the practical example of pandemic politics raises, suggesting that the pre-pandemic ethical compromise in the UK as a liberal democratic society is no longer sustainable. We deploy two conceptual frames to analyse this shift: (1) Shove’s politics of ABC and (2) Dowding’s critique of the state. We then use the pandemic politics of emotions to highlight the ethical and practical challenges facing policy efforts to achieve evidence-based outcomes through nudging. While we focus on examples from the UK context, the lessons are relevant to all liberal democracies that have invested substantial resources and political capital in nudging.

Pre-pandemic politics

Elizabeth Shove’s work on public policy and social change across a variety of areas, including climate change and the obesity crisis, focuses on the need to recognise and expand the interpretive lenses through which we seek to understand and address major societal challenges (Shove, 2010; Blue et al., 2021). The notion of ‘re-framing’ is a central element of Shove’s work. Drawing upon this notion can help us consider ways in which the COVID-19 crisis facilitated re-framed dominant assumptions about the role of the modern state. Shove’s work focuses on reconceptualizing theories of social change through a critique of what she interprets as the dominant ‘ABC model’: social change is framed as a matter of individual Attitudes that drive Behaviours that people Choose to adopt. This is a rational-choice-derived model of social action in which the notion of choice, under conditions of state neutrality, is central. It emphasises personal responsibility and, therefore, dovetails with neoliberal conceptions of the state in which the role of politicians and public servants is to ‘persuade, price, and advise’ individuals on the basis that when given better information and more appropriate incentives individuals (i) will change their attitudes, (ii) alter their behaviour and/or (iii) make choices that are better aligned with addressing social challenges (i.e., ‘ABC’). It valorises the individual, freedom of choice, and a model of almost transactional, technocratic or consumer-based politics.

Shove’s account is best illustrated by policymakers’ pre-pandemic commitment to nudging, which focuses on shifting and re-framing the ‘choice architecture’ presented to individuals in the hope that re-biasing individuals’ intuitions toward some options that promote certain outcomes while disincentivizing ‘bad’ choices. Examples in public health practice include reducing alcohol serving sizes (Bryant, 2020), displaying pictures of disease-ridden body parts on cigarette packets, and presenting fresh fruit at head height. Such policies attempt to recalibrate cognitive biases while preserving individual choice, an approach that has been criticised for undermining autonomy (Rizzo and Whitman, 2019) and which Thaler and Sunstein (2008) have emphasised cannot address some of the critical challenges we face as a species, such as climate change.

This pre-pandemic ‘libertarian paternalism’ seeks to uphold formal neutrality between conceptions of the good and, thereby, equality between citizens by eschewing coercive means of promoting choices. In so doing, it places responsibility almost exclusively on individuals to make good choices since the value of those choices has, supposedly, already been made apparent to them. For Keith Dowding (2020), this position is the result of five decades of ideological domination by the ‘the cult of personal responsibility, propagated not only by politicians but also by moral and political philosophers’ (Dowding, 2020: xii). The problem is that the pre-pandemic approach has failed in its own terms. He writes that ‘while there is individual responsibility with regard to our behaviour, the major responsibility for social failures is that of the government’ (Dowding, 2020: xi). What Dowding charts with great effect through a focus on gun crime, obesity, homelessness, gambling, and drugs policy is either the gradual withdrawal of state-based control mechanisms (i.e., liberalisation, regulation, etc.) and/or a refusal to countenance the implementation of such measures. Whereas the government once took on a degree of direct responsibility for the health and wealth of all their citizens, Dowding argues that ‘a process of privatised blame-shifting’ has occurred, whereby responsibility for dealing with major social challenges has been, through ‘the politics (and logic) of ABC’, placed on individuals. Obesity, homelessness, gun crime, gambling and drug addiction are redefined not as governmental failure but as the consequences of poor individual choices. Hence, the sub-title of Dowding’s book, It’s the Government Stupid: How Governments Blame Citizens for Their Own Policies. This critique not only politicises personal choice but also focuses attention on the existence of choice vis-à-vis the selection of policy tools and instruments. As Table 1 illustrates, taking this forward, it is possible to differentiate between three broad governmental responses:

Table 1 Nudging, shoving, and pushing.
Full size table

There is a strong sense in which ‘the politics of ABC’ is imbued with a highly technocratic set of cost-benefit assumptions that risk almost depoliticising public policy, with political issues coming to be regarded as little more than ‘lifestyle’ choices. Shove and Dowding also emphasise the fundamental weakness of over-emphasising personal responsibility and under-emphasising the existence of other more sophisticated framings and the latent capacity of the state to protect vulnerable individuals from the impact of systemic economic, technical or scientific tides of change. The pandemic, arguably, bore out that weakness more starkly than any crisis in recent memory, not least because it saw the government abandon its commitment to one of the crown jewels of the politics of ABC, namely, nudging.

Nudging

In the two decades leading up to the pandemic, nudging became established as a popular policymaking device among liberal policymakers within the United Kingdom. The progenitors of nudge, Thaler and Sunstein (2008: p. 6), define it as ‘any aspect of the choice architecture that alters people’s behaviour in a predictable way, without forbidding any options or significantly changing their economic incentives’.

Nudging is underpinned by several empirically informed assumptions. The first is that inherent features of human cognition mean that people are bad at acting in ways that realise their own interests, which, according to Thaler and Sunstein, must at least include at least health, wealth, and happiness. The second is that choice architecture that exploits those features of human cognition intentionally or unintentionally is ubiquitous. The third is that the reason many people act against their assumed interests—for example, by smoking, drinking too much, or failing to save for their retirement—is because the extant choice architecture induces them to do so. This led Thaler and Sunstein to conclude policymakers cannot be politically neutral in the sense understood by Hayek, Friedman and other key figures in the liberal tradition (Thaler and Sunstein, 2008: pp. 4, 246; see also Yeung, 2012: p. 147); if bad outcomes are to be avoided, policymakers must become choice architects, that is, they must intervene. However, according to Thaler and Sunstein, nudging allows policymakers to do so while remaining neutral about what a good life is.

With a major point of disagreement between the left and the right putatively resolved, policymakers of all political stripes have been able to unite under the banner of nudge. In the UK, the David Cameron-led governments from 2010 invested heavily in nudging as a key policy instrument. The Behavioural Insights Team was established in the UK Government’s Cabinet Office in 2010 specifically to apply methods like nudging to UK public policy (Institute for Government 2023). Known informally as the ‘Nudge Unit’, it has now formally adopted the moniker within its branding as a part of the innovation charity Nesta (Behavioural Insights Team 2023).

The contemporary debate on nudging in political and philosophical scholarship has focused on whether it is an ethically permissible means of government intervention, largely because of its implicit nature (Schmidt and Engelen, 2020). However, a more fundamental question is whether nudging is even an effective means of intervention. Both supporters and detractors of nudging tend to assume that it is. However, recent empirical findings suggest otherwise. Notably, in the context of the COVID-19 crisis, evidence for the effectiveness of nudging appears to be mixed at best (Kantorowicz-Reznichenko et al., 2022). Several studies across different national contexts have found that standard nudging methods failed to increase adherence to COVID-19 guidelines (e.g., Blayac et al., 2022; Hume et al., 2021) or decrease vaccine hesitancy (e.g., Moehring et al., 2023; Sinclair and Agerström, 2023). It is not just in the exceptional circumstances surrounding the pandemic that nudging has failed to live up to the hype. A 2017 review observed that the effectiveness of key nudging methods has been thrown into doubt in relation to numerous different problems, including physical inactivity, harmful drinking, and tobacco use (Lin et al., 2017), problems which are supposedly ideal targets for nudging (Sunstein, 2016: Ch. 2).

Of course, these claims about the ineffectiveness of nudging are based on a handful of studies from a large and growing body of research. Perhaps overall findings about the effectiveness of nudging are more promising? A recent meta-analysis of choice architecture research seemed to show that the answer was yes. Mertens at al (2022) claimed to have found robust evidence for the effectiveness of nudging and its versatility across different contexts, ranging from health and food to finance and the environment. However, they also noted ‘a moderate publication bias toward positive results’. That appears to have been a considerable and distorting understatement.

Several subsequent re-examinations of the data used in the meta-analysis showed that publication bias toward positive results in the nudging literature is actually severe (Bakdash and Marusich, 2022; Szaszi et al., 2022; Maier et al., 2022). Maier and colleagues (2022; 2023) used their robust Bayesian meta-analysis method to correct for this bias and found an absence of evidence for an overall effect. Further, with regard to the three intervention categories they assessed, they found evidence against an effect in ‘information’ and ‘assistance’. For ‘structure’, they describe the evidence as undecided. When using only the most precise estimates, for methodological reasons, to assess the effect on specific policy domains, they found evidence against an effect in ‘health’, ‘environment’, ‘finance’ and indecisive evidence for ‘other’, ‘food’ and ‘prosocial’. This use of precise estimates found ‘weak evidence’ for the overall effect.

Maier and colleagues acknowledge that, given the heterogeneity of the interventions in all domains apart from finance, the absence of a mean effect does not mean that all nudge interventions are ineffective. For example, Last et al.’s (2021) systematic review of clinician-directed nudges found evidence for the efficacy of changing default options or enabling choice, albeit with the caveat that, once again, there was a substantial risk of bias, with just 13 studies (33%) considered low risk across all criteria. Given the paramount and enduring importance of health, the environment, and finance historically, but also specifically in the modern post-global financial crisis and post-pandemic era, Maier et al.’s (2022) findings should give pause for thought. A method that has been found to have little efficacy in these areas should not be relied upon to fundamentally underpin the design and implementation of almost all public policy, even if there may be some limited applications, such as making some choices defaults, such as organ donation.

Notwithstanding its apparent inefficacy, the rise of nudge appears to have had a considerable distorting effect on policymaking, likely fuelling the trends that Shove and Dowding highlight. More specifically, it has obscured the fact that humans are social beings situated within social systems. Nudging frames policy problems and solutions in individual terms, focusing on weaknesses in human cognition and how to exploit them to produce better individual behaviours (Chater and Loewenstein, 2022; Mols et al., 2015). To paraphrase Chater and Loewenstein (2022), nudging effectively seeks to adapt individuals to a hostile world rather than to make the world less hostile. Amid the pandemic, however, this approach clearly became untenable, as we seek to illustrate in the next section by examining the role of fear in that crisis.

Pandemic politics

In pre-pandemic politics, there was a presumption that public preferences for policies that differed from those promoted by libertarian paternalists were driven by irrational emotion. This can be seen, for example, in the two kinds of rhetoric related to emotions deployed during the Brexit debate. First, there was emotional rhetoric encouraging those who felt ‘left behind’ by politicians and society to vote to leave the European Union. (Similar narratives have been used in relation to ‘the forgotten people’ in the United States or ‘the peripherique’ in France.) Second, there was meta-emotional rhetoric stressing that people’s votes would be decided by how they felt about themselves (Degerman, 2022: pp. 81–2).

This has two apparent implications. Firstly, politicians cannot be blamed for pitching the wrong policies since people do not care about policies anyway. Secondly, politics is about manipulating cognitive biases and related emotions, such as fear (Sunstein, 2002), as in images of illness on cigarette packets, rather than addressing the structural issues that have given rise to those emotions in certain populations (see Degerman, 2022: p. 122). In specific contexts, emotions can be a means to divert attention away from their interpretation and acceptance as ‘public issues’—i.e., collective challenges that can only be addressed through a collective societal response and are therefore within the legitimate purview of the state—to ‘private troubles’ where the emphasis and responsibility are placed on the individual. The pandemic changed this.

Early in the pandemic, it became clear that nudging was insufficient to address the spread of disease. Merely encouraging people to practise better hygiene or to socially distance proved ineffective. Indeed, people’s emotional responses to the pandemic appeared quite rational—there was a genuine risk of illness with a possibility of death and fear served to motivate mitigation of that risk (see e.g., Harper et al., 2021). Most emotion theorists agree that fear tends to involve the identification of something as a threat, an unpleasant feeling, increased heart rate and activity in the amygdala, widened eyes, and a tendency to escape the threat (see Scarantino and de Sousa, 2018). Because of these components, fear has several potential epistemic and motivational benefits (Brady, 2018: pp. 78–80). Fear tends to consume our attention, focusing our thoughts on understanding the nature of the threat and how we can escape it, and, crucially, it motivates us to act on our conclusions. It is also communicable and socially symbolic in ways that undermine atomistic accounts of individual agency. Fear can mobilise individuals and communities to unite in the face of a shared fear, thereby helping to foster trust, solidarity, and a collective identity among people who may otherwise have little in common (Brady, 2019, pp. 122–126).

Although the capacity of fear (or ‘fearfulness’) to act as a form of social glue or a tie that binds is rarely considered in the broader literature, there is also a related dynamic that deserves identification and elaboration—fear as a tool of governance. There is, of course, an immediate need to lay down some normative and analytical markers around this notion of fear as a tool of governance. As a vast body of historical and comparative scholarship has explored in great depth, fear as a tool of control is a common dimension of authoritarian and totalitarian regimes (see e.g., Arendt, 1958). The simple and far softer interpretation of ‘fear as a tool of governance’ that we seek to highlight in this paper is, although not wholly unrelated, quite different and can be explained in two simple steps:

Step One: Fear is a critical emotion within social psychology due to the manner in which it focuses public attention and support for defensive strategies around a specific topic.

Step Two: In designing and delivering their strategies to mitigate the sources of specific societal threats and fears, politicians and public servants may have to maintain and manage ‘the fear factor’ in order to ensure public compliance.

What we fear, how much we fear it, and what we take to be appropriate responses to our fear in the context of a crisis is contingent on policy and other factors (see Harbin, 2023). Hence, whether fear functions effectively as a tool of governance is also contingent. When the threat is misconceived, the intensity of our fear exceeds the threat (or prevents us from responding to it), or the means of addressing it are noxious, the epistemic, motivational, communicative, and social-cohesive functions of fear may become destructive too. Harnessing the functions of fear to fight the pandemic was, thus, a challenge for governments.

That challenge can, at least in the UK context, be schematically divided into two temporal stages in which different sets of problems come into focus: early in the pandemic and late in the pandemic. Early in the pandemic, the UK Government needed to engender and align public fears in order to minimise the most serious risks of health services being overwhelmed and law and order breaking down (Degerman et al., 2020). This is a critical point. The task of engendering fear pertains to the virus and related vectors and risk factors (see Alaszewski, 2023: Ch. 4). After living through nearly two years in what might be termed the ‘COVID context’, it was easy to forget that we do not naturally fear social interaction in the way we might naturally fear snakes or sharks. Humans are social animals and, as such, physical contact with other people (touching and hugging), being in crowded spaces (night clubs or football stadiums) and seeing someone’s face (free from masks or personal protection equipment) are generally sources of pleasure which trigger positive emotions (excitement, joy, hope, love, etc.). The emotional trauma of COVID-19 reflected the way basic human interaction became something to fear.

The critical point we are trying to make, however, is that to some extent, the UK government and governments around the world had to reinforce and legitimate these fears in an attempt to encourage public compliance with mitigation measures (wearing a facemask, wash your hands, stay at home, etc.). At this point, a key example of nudging concern attempts to stop people from touching their faces (e.g., Parkinson, 2020; Mageit, 2020). Statistics were one notable tool in these efforts. Ubiquitous charts representing rates of infections, hospitalisations, and deaths effectively became a kind of prescriptive barometer of fear (Alaszewski, 2023: p. 159). While infections and fear were low, the government stuck to nudging, but that clearly was not enough. So, when infections spiked, it had to move to shoving and eventually to pushing (Table 2).

Table 2 Nudging, shoving, and pushing during the COVID Crisis.
Full size table

However, the problem that most starkly illustrates the shortcomings not just of nudge but the broader ABC frame that provides scope for nudge as a policy measure is the task of aligning fears. Engendering fear of a particular object or set of objects is insufficient to harness the functions of fear. In the context of the pandemic, many people had other fears that they perceived as equal to or outweighing the threat of infection, and, as Ned Lebow (2021) points out, ‘people and politicians are reluctant to make difficult trade-offs between highly valued ends’, which can lead them to act harmfully or too slowly. Fears, therefore, also need to be aligned. Government messaging that there was no trade-off between addressing COVID-19 and protecting other ends, like wealth, could be interpreted as one means of achieving this. Claims like these were, of course, unlikely to convince individuals facing the threat of being unable to provide for themselves and their families because of COVID-19 restrictions. The introduction of the Coronavirus Job Retention Scheme was, therefore, probably the most effective means of aligning other fears (i.e., job loss and poverty) behind the fear of COVID-19 since it removed the need to choose between the end of trying to contain the virus and the end of staying financially afloat. In this instance, the need for individual choice was, hence, taken out of the equation because the government moved beyond ‘nudge’, ‘shove’, and ‘push’ by instead beginning to ‘catch’ people. ‘Catch’ in this sense relates to the creation of an economic social safety net in the form of the furlough scheme.

The introduction of the furlough scheme matters because it reflected a fundamental rejection of the dominant pre-pandemic conception of state-society relationships. Pre-pandemic, as the previous section outlined, a highly individualised and market-based set of assumptions about the role of the state and the individual had existed. Following Shove, this conceptualisation can be summarised as ‘the ABC approach’ to ‘doing politics’. This section has explored how the pandemic challenged and changed this conceptualisation. Looking back and using the UK as an empirical case study, it is possible to see an initial attempt to reframe the ‘ABC approach’ in the face of the pandemic through an emphasis on the provision of information and data that was intended to influence changes in individual behaviour (i.e., ‘hands, face, space’). Tracing the evolution of the crisis, what is particularly noteworthy is how the scale of the pandemic almost outgrew the response capacity of conventional, individually focused modes of response and that ministers were gradually obliged to expand the Overton window to encompass a broader range of policy tools. It was broader, critically, in the sense of involving a far more direct and collective role of the state. The policy spectrum, therefore, changed to include elements of ‘push’ and ‘shove’ in which the state explicitly constrained individual freedom; consequently, it became impossible (qua. Dowding) for the government to blame the public.

The practical and intellectual malleability of the dominant pre-pandemic conceptualisation was tested to breaking point, and, in the end, the government was forced to fundamentally recalibrate the boundaries between the public and the private sphere. The scale of the state in terms of both financial spending and social intervention increased to a level only ever previously experienced in times of war. Having previously rejected such intervention on the grounds of it promoting the interests of specific individuals within society (the unemployed, etc.) and thereby contravening neutrality as equality, the pandemic’s indiscriminate nature meant that intervention to address social determinants became the only practical means of upholding the interests of citizens.

Post-pandemic politics

There is a large body of evidence to suggest that the key outcomes that UK policymakers identify as being of national importance—such as health—are not only correlated with one another but also shaped by social determinants. The social determinants of health were most clearly articulated by the Black Report (1980). It demonstrated that such goods as wealth, housing, diet and various other phenomena shaped health outcomes that could only be partially mitigated by health services. More recently, Wilkinson and Pickett’s (2010) The Spirit Level brought together a large body of evidence from various fields to suggest that poverty and inequality are associated with poor health, educational, employment and criminal justice outcomes.

The timing of the pandemic was particularly unfortunate in terms of the status of those social determinants: decades of de-industrialisation had stripped communities of wealth that might provide sources of local resilience; the Global Financial Crisis had destroyed many small and medium-sized businesses and subsequent austerity politics had critically reduced safety nets developed in the wake of Europe’s last great crisis, World War II. Public services have been stretched further by a general trend, identified by Standing (2021), toward deskilling. Domain-specific entities have been stripped of resources (see Pettit, 2021b) and asked to support previously distinct domains, with medics taking on the role of benefits agents (Johnson, et al., 2019) and police officers the role of social workers (Simpson, 2019), leaving both completely overwhelmed (see Dewhurst, 2020).

Had the virus not struck after public health had been affected by decades of welfare cuts and underfunding of state services, it may have been subject to more robust obstacles. There may have been entrenched safety nets to catch people falling into destitution from the financial fallout of the pandemic. Instead, it struck at a time in which those with the worst immunity and greatest susceptibility to the virus (see Mena et al., 2021; Sá, 2020) were also those most likely to need to be in public spaces in order to work or to care for others (see Hussein et al., 2012). If they did not become seriously ill themselves, they were liable to spread the infection to those who would become so: their parents or grandparents or those to whom they were providing services.

The notion that the pandemic was an aberrant policy frame and that policymakers could return to pre-pandemic forms has been challenged by a condition of perma-crisis. The environmental emergency, the cost-of-living crisis and geopolitical conflict are all compounding crises.

Standing (2021) and others have referred to such individuals who could not afford not to be in public as members of a precariat whose work is increasingly elided with labour, whose labour has been stripped of skill, and whose participation in the economy piecemeal and insufficient, not just to advance their interests, but to sustain them independently of state support. For such people, the pandemic served to foreground existential fear of destitution, which serves as an extrinsic mortality cue for death and contributed to a rise in mental health problems, such as anxiety, depression and trauma (see Johnson et al., 2020). Given the breadth of cues, there is every reason (Koob et al., 2020) to suggest that these factors will increase what Case and Deayton (2020) describe as ‘deaths of despair’. While we have previously argued that fear of communicable disease can be perfectly rational (Degerman et al., 2020), fear of destitution in a modern, affluent society ought to be seen as a consequence of failure in policy: people recognise that the state will no longer catch them and that the precarity of their condition has real, existential meaning.

This structural failure was only compounded by laissez-faire concern for personal responsibility. At the beginning of the pandemic, there was a widespread sense that dealing with transmission ought to be seen in the same sense as sustaining an income (see Christensen, 2009): that individuals ought to bear the consequences of their decisions. As such, individuals made choices based on relative risk of mortality and security of income. For young people in the gig economy, not getting a COVID-19 test or not abiding by isolation periods because of the absence of a safety net was perfectly rational (see Tapper, 2021; cf. discussions in Faulkner, 2021; Pettit, 2021a). Those rational decisions led to macro-level disease transmission and imposed unprecedented financial burdens on societies. Neoliberal governments had few answers (see Pettit, 2021b). As the pandemic wore on, it became clear that even people in societies with long traditions of privatising responsibility for health and privatised healthcare were unwilling to be subject to such thoroughgoing levels of personal responsibility. Having initially sought to avoid restricting liberties, as fear of COVID-19 increased, public demand for intervention was satisfied in all but the most extreme cases (see Kalil et al., 2021; Honeycombe-Foster, 2020). That intervention, through lockdowns and self-isolation mandates, often required substantive constraints on negative liberty (see Berlin 1969). Increasingly, it became apparent that decades-old settled neoliberal principles had actively contributed to the crisis and been rendered obsolete by conditions on the ground (see Johnson et al., 2021). There was no market solution to the need for intensive care beds, creating medicines, funding and training researchers, or restricting transmission of new variants via international travel and commerce. Only governments could close borders and limit transmission.

Accordingly, public opinion about what is politically possible shifted. Having experienced a decade of austerity politics, public support for government intervention in social security began to rise (see Britain Thinks, 2020, p. 35). Politicians across the spectrum disavowed a return to the old normal (Schomberg and Sandle, 2020). However, what ‘new normal’ policies would look like was unclear. The Chancellor, Rishi Sunak, was heavily influenced by trade unions in adopting furlough schemes and mortgage payment holidays that provided unprecedented levels of support to individuals (e.g., Unite, 2021). He also presented a range of gimmick-like stimulus packages, such as the ‘Eat Out To Help Out’ programme (see Hutton, 2020), which was an effective enormous subsidy to the leisure industry and contributed to the second lockdown (Fetzer, 2022), and stamp duty holidays, which contributed to house price inflation (Scalon et al., 2021). Each of these policies was clearly crisis-dependent, but their elements hinted at a possible new normal.

While there has been increased talk among politicians of a return to normal politics and the importance of fiscal responsibility (see Committee of Public Accounts, 2021), it is simply impossible for the government to do so. The challenges are just too great. Automation is likely to produce unprecedented job losses (see Standing, 2016), reducing the capacity of private sector employment to catch people and raise them out of poverty (see Barry, 2012), while the ability of individuals to consume independently of state funding (see discussion of contradictions in Jessop, 2013) is already changing and will continue apace. Moreover, much economic activity will become unsustainable by virtue of its contribution to climate change (see discussion in Clark and York, 2005)—the contradiction of infinite growth on a finite planet (see Park, 2015) is simply unsustainable. Mass unemployment and human extinction via climate change are two direct and existential threats worthy of fear.

In order to deal with the fallout of the pandemic and coming crises, a new normal policy framework requires a radically recalibrated understanding of what the state can achieve with regard to safety nets and the management of personal pleasures. There needs to be recognition that if we wish to address health, educational and criminal justice outcomes, we would be best placed to invest in mitigating social determinants through investment—that is, working ‘upstream’—rather than deploying ineffective nudging or mitigating outcomes ‘downstream’ once they are already being formed. Policymakers need to be guided by concern for catching people via ‘upstream’ economic interventions (NCCDH, 2021) that affect social determinants of health (Black Report 1980) and material outcomes and, as the Scottish Government has recently recognised, to roll back the state in dealing with ‘downstream’, self-regarding activities, such as their use of drugs. The domain-general entity—that is, government—needs to ensure that domain-specific entities are properly resourced and their authority limited to their specific remits: medics need to be allowed to focus on dealing with morbidity and preventing deaths rather than acting as gatekeepers to the benefits system (Johnson et al., 2019), while Police Officers and members of Criminal Justice System more generally need to focus on preventing other-regarding harm and punishing offenders, rather than acting as social workers.

The UK government is, reluctantly, beginning to understand this imperative. Its Prevention Agenda is intended to shift the understanding of the National Health Service away from the ‘National Hospital Service’ (see Department of Health and Social Care and Hancock M, 2018). It is explicitly aimed at avoiding morbidity and mortality but has yet to advance substantive policy measures beyond The Soft Drinks Industry Levy (SDIL) (2013), which constituted a limited and indirect intervention via taxation on business. This may be because it appears either to fail to understand the material basis of health or to reject ideologically the ‘pushing’ on personal economic matters (see Neville and Pickard, 2017) or the resulting radical socioeconomic implications on policy.

There is a need for policies that do not just catch people when they fall but hold people back from falling. We could call these ‘holding policies’. This term aptly evokes Donald Winnicott’s (1960) concept of holding, which roughly describes circumstances that foster and sustain a sense of security and capacities for dealing with adversity. In this regard, there is, and there needs to be, a marked shift toward increasing recognition, not only that nudging is an ineffective means of governance, but that human beings’ behaviour is affected by social determinants and addressing those social determinants is the only effective means of governance in our era of permacrisis.

Ironically, this is consistent with the ethics of those commonly cited as having justified pre-pandemic liberal reform in the first place. For Hayek (1944, pp. 124–125),

There is no reason why in a society that has reached the general level of wealth which ours has attained, the first kind of security should not be guaranteed to all without endangering general freedom…. [T. there can be no doubt that some minimum of food, shelter, and clothing, sufficient to preserve health and the capacity to work, can be assured to everybody.

Likewise, in advancing a negative income tax, Friedman states that such a system is warranted precisely to uphold neutrality as equality: ‘There is every reason to help the poor man who happens to be a farmer, not because he is a farmer but because he is poor’ (Friedman, 1962, 191). Such defence of the provision of resources necessarily appears stark in light of otherwise fundamental opposition to state intervention. However, it is presented as being consistent with neutrality and equality in its provision to all individuals because they all share specific human needs. The evidence suggests that meeting those needs securely and adequately upstream is more likely to produce the sorts of outcomes libertarian paternalists endorse but without the ineffective and ethically problematic downstream interference of nudging. Nudging may have been an attempt to address a problem that the key proponents of the neoliberal settlement did not themselves foresee.

Conclusion

The pandemic has fundamentally unsettled several assumptions about policymaking. The recognition of vulnerability and the limitations of the private sector has increased the legitimacy of the state as an economic actor in the eyes of the public. The emotions fostered by pandemic vulnerability have no natural end; climate crises and unemployment pose genuine existential threats. It is easy to see how these crises compound one another and leave society ever more vulnerable to another pandemic, made more likely by the expansion of humans into animal habitats. The only viable responses to these crises are those that have the potential to expend political capital, namely, the abrogation of any individualising nudge, push or shove policies and the arrogation of tax-raising powers to fund more substantive welfare schemes. This pragmatic response involves accepting that a portion of the population will free-ride, that crime will persist and that many will engage in hedonistic pursuits. It is guided by evidence that holding people in a state of functioning by providing a materially secure context holds the vast majority of individuals and that catching people when they engage in acts of hedonistic vice drags them down through the criminal justice system to the detriment of society. There is simply no alternative capable of addressing post-pandemic crises successfully. The question is whether political parties can overcome the short-term electoral foci of liberal democratic government to do what is in the best interests of the populace. Presently, the unwillingness to intervene to mitigate social determinants is producing precisely the outcome that neoliberals decry: the promotion of a small set of interests without providing a substantive basis from which individuals can pursue their conceptions of the good.