Us and Them: On the Motivational Force of Formal and Informal Lockdown Rules

How do social norms and legal requirements combine to shape collective behaviour? A multi-wave ten-city panel study set during the first UK lockdown finds that compliance was a powerful in-group signalling device, driven by the expressive and coordinating power of formal and informal rules. COVID-19 pandemic laws allowed the Government to operate as an expressive agent, telling people what needed to be done and why. Acting upon mutual expectations for the common good then helped people to coordinate against the virus with a sense of a shared fate and identity. Widespread collective compliance allowed the police to continue to privilege engagement, explanation and encouragement over heavy-handed enforcement tactics that damage their popular legitimacy.


Introduction
For any government seeking to control the spread of a coronavirus like COVID-19, controlling people's behaviour is the only route, at least until 70 or 80 percent of the population is vaccinated. Governments across the world are using public health, containment and closure measures to secure public compliance, with a huge cost to people's well-being and to the economy. Many of these interventions are effectively unenforceable at scale-does anybody imagine the police knocking on every door over Christmas, checking who was inside? So their success depends on willing self-regulation.
Why, then, do people comply? Answers to this question are crucial to the design and implementation of optimal interventions. There is some evidence [1] that countries with stronger general adherence to social norms tended to have the best success limiting COVID-19 cases and deaths, even after taking into account the stringency of governmental responses. "Tight" societies have strong implicit or explicit rules that govern behaviour-they have more social coordination and self-control. "Loose" societies have weaker formal and informal punishment systems-they have more social tolerance and creativity. Gelfand and colleagues argue that self-regulation and voluntary norm abidance may have helped to keep infection and mortality rates low in tight societies and that policy interventions are needed to strengthen social norms, particularly in loose societies [1].
Social norms are clearly fundamental to public health and environmental behaviour [2,3]. Interventions such as providing people with "socially comparative feedback"-information about their behaviour relative to the behaviour of others-such as by informing them about common behaviour/attitudes and statements of public commitment and "leadership" have been shown to influence people's energy consumption [3 p6]. Some of the strongest messages during the early stages of the Covid-19 crisis involved making norms more visible and salient: "Stay home; save lives; protect the NHS" activates the prescriptive norm (staying home) by emphasising the desirable ends it will produce (saving lives and protecting the NHS). People generally want to do the latter, so reminding them of this will encourage them to do the former.
But what role does the law play? Do legal requirements combine with social norms to create motivational pressure? [4] The traditional view is that the law works through deterrence: people will obey pandemic legislation because they fear the consequences of disobedience. If social norms are informal rules that govern behaviour (people will wear a face covering because they fear disapproval from others), then the law provides formal rules that people will follow because they fear sanction from authorities (a fine for not wearing a face covering).
Yet, there is a good deal of criminological evidence that deterrence is a weak predictor of compliance. The evidence instead points to the idea that people comply with the law because they think it is the right thing to do, not because they fear the consequences of non-compliance [5]. Legitimacy is critical here: people accept that the law is entitled to dictate appropriate behaviour and internalise the norms and values attached to being a law-abiding citizen [6]. Willingness to obey is linked to positive self-definition; people believe that obeying the law is the right thing to do, and they glean value and identity from being respectful of rules and authority. Being law-abiding is who they are.
In this paper we address the psychological bases of compliance during the first UK lockdown in March, April and May 2020. This was a fascinating period in the history of the country. The powers that Parliament passed on 26 March 2020 were unprecedented, taking away basic freedoms enshrined in UK law on free movement and on free assembly. Throughout the first lockdown, public discourse and much official communication centred on the need for collective action: to unite as a community, people needed to individually remain apart. But the injunction to stay home was not a request but a legal requirement, and the police were tasked with enforcing this and other requirements (e.g., to socially distance). They faced some initial concerns about their ability to do so at scale, and there was high-level discussion about the possible unintended consequences if they went about it in a heavy-handed manner.
Yet, public compliance was remarkably widespread during those first few months, with police rarely having to intervene. Mobility data showed a significant curtailment of activity, and police records indicated low levels of sanctions. The extent of compliance seemed to surprise even the government. Soon after the easing of lockdown measures on May 13, Prime Minister Boris Johnson is said to have quipped to colleagues: "I've learnt that it's much easier to take people's freedoms away than give them back." Fast forward to just before the third lockdown was announced in January 2021, and we saw Labour leader Keir Starmer calling for a return to the "spirit of March, where there was a very strong stay at home message" 1 with few people mixing in the streets.
So how did this happen? Was adherence to the guidelines about fear of catching the virus, fear of being fined and fear of disapproval from others in the community? Was adherence due to widely shared norms and values, reflected in mantras like "stay home, save lives, protect the NHS"? What about the law: did making social distancing a legal requirement make a difference, and if it did, how?
The rest of the paper proceeds as follows. In section 2 we review some of the literature on instrumental and normative motivations for compliance. In section 3 we describe the methodology, and in section 4 we summarise some of the key findings. To foreshadow the results, we find that fear of the virus and fear of punishment played minor roles. Deterrence and legitimacy were also largely irrelevant. Key, instead, were informal rules governing behaviour (social norms) underpinned by the coordinating function of formal rules. On the one hand, lockdown law allowed the Government to operate as a powerful expressive agent, telling people what needed to be done and why. The law framed the threat and its solution at the collective level, creating a focal point around which to organise. On the other hand, compliance served to operate as a powerful in-group signalling device, driven by the expressive and coordinating power of formal and informal rules. Widespread compliance by the population, coupled with restraint in the enforcement of public health restrictions (a greater emphasis on engagement, explanation and encouragement), then helped the police to avoid damaging legitimacy and exacerbating social tensions.
Section 5 finishes with some conclusions for policy. Our data support the idea that government and public health messaging should focus on normative rather than instrumental reasons to adhere to guidelines and regulations. Pandemic legislation can play a role, but at least according to our analysis, it is not about deterrence or legitimacy, but rather, it is through communicating the rights and wrongs of certain behaviours in a way that highlights a collective solution to a collective problem. When combined with widespread internationalisation of social norms, this could be a powerful way of securing compliance. Indeed, if it is true that norms hold less sway in "loose" societies (of which the UK is apparently one [1]), then perhaps the use of the law as a coordination mechanism can go some way to addressing the challenge for collective action that "looseness" presents.

Instrumental and normative compliance
Criminologists, psychologists and others concerned with why people comply with legal rules and regulations, and what social institutions can do to motivate such compliance, typically offer two complementary, but also rather contrasting, accounts. First, compliance can be instrumental. In a general sense, people may comply because they fear the consequences of non-compliance, whether these arise from the behaviour of agents of formal social control (e.g., most obviously, but by no means solely, the police) or from informal social control actors and processes (e.g., action by parents in the case of their children). In the context of Covid-19, fear of catching the virus clearly offers a further prudential reason for complying with regulations aimed at preventing its transmission.
Instrumental reasoning tends to dominate "common sense" accounts of compliance, which often revolve around the need for stricter enforcement. Much of the reasoning within the criminal justice system and the politics of crime control is that the best way to encourage compliance with Covid-19 restrictions would be to generate fear of non-compliance. This is the "tough man" approach. At the time of writing, for example, the government is suggesting ten-year prison sentences for people found to have concealed the fact they have travelled into the UK from a "red listed" high infection country. Yet, evidence for the importance of instrumental reasoning in decisions to comply with laws is rather thin and often contradictory. While it is clear that some people, in some contexts, are "deterrable" in a general and specific sense, the evidence suggests that, overall, instrumental reasoning plays only a relatively minor role. Other explanations are needed if we are to understand the "bulk" of compliance behaviour.
The second account of compliance, and the policies and practices best placed to promote it, can be broadly defined as relational. Theories of legitimacy and procedural justice stress that people wish to retain positive relationships with authorities they perceive as legitimate and representative of their values [5,6]. This provides a motivation to comply with rules enacted and enforced by those authorities separate and distinct from any credible risk of sanction they signal to citizens. Compliance becomes a value in and of itself-something that instantiates and demonstrates the social bonds between individuals and governing institutions. Authorities can be powerful representatives of superordinate social groups, and willingness to obey is linked to individuals positively self-defining their position within the group. People believe that following the instructions of legitimate authority and obeying the laws they enforce is the right thing to do, and they glean value and identity from being respectful of rules and authority.
However, broader social norms are almost by definition likely to be important for compliance. People are strongly motivated to adhere to widely accepted norms of behaviour, at least in part because they wish to maintain positive relationships within the communities in which they live, rather than just because they wish to retain the approval of authority figures. Sustaining such relations means maintaining a positive sense of self in relation to continued membership of salient social categories. It also means avoiding the various forms of sanction disapproving community members can inflict if norms are transgressed, suggesting there may be significant over-lap between instrumental and relational reasons for compliance.
Unlike the "purely" instrumental accounts outlined above, evidence for relation-based compliance seems much firmer. Work from a wide range of literatures-procedural justice theory, "nudge" theory and so forth-converges on the central finding that the desire of individuals to maintain social relations with and within various institutional and social contexts exerts a strong pressure towards compliance. To comply is to act as an expressive agent, saying to authorities and others in the group that one respects widely-shared norms and priorities.
Given all this, it is striking that in official statements and guidance to officers, UK police have relied explicitly on ideals of public consent and procedural justice in their response to the crisis. The College of Policing guidance states "people are more likely to comply after a police encounter if they feel they have been treated fairly, have received an explanation, and have been given the opportunity to state their view", and the "four E's" approach-Engage, Explain, Encourage and Enforce-equally stresses that enforcement activity is a last resort, only to be used if engagement, explanation and encouragement fail. To put it another way, if all else fails turn to enforcement, but consensual compliance, built by maintaining positive relations between police and public, is likely to be more immediately effective and enduring.
In the context of policing the Covid-19 pandemic-by which we mean not just the activity of the uniformed police but the wider effort, involving many other state and non-state actors, to motivate compliance-all this means that a range of levers are available and, in all likelihood, needed. Yet, during the first lockdown relational factors seemed most prevalent. This was well expressed in the 4E's approach, with its focus on encouraging compliance by invoking values, building legitimacy and, only as a last resort, turning to enforcement. Public health messaging also clearly sought to build normative compliance by invoking the shared fight against the virus, with the NHS, and NHS workers, the physical embodiment of this effort.

Rule content and expression
A necessary counterpart of the above is the content of the regulations and restrictions themselves-what people were being asked to comply with. From some perspectives, the content of rules can seem less important than the mere fact of their existence. The concept of legitimacy, for example, revolves around the idea that legitimate authorities are able to command obedience regardless of the content of the rules they enact and enforce-the ability to do so, in a sense, is legitimacy. Many non-formal norms (such as wearing a tie to dinner) stress the maintenance of "appropriate" behaviour over any consideration of whether the specific behaviour defined is justifiable, beneficial or even rational.
Such accounts can appear overly reductive, however. People clearly do attend to the content of rules, and this may particularly be the case when, as during the early phases of the Covid-19 crisis, they were confronted with a raft of new regulations that severely curtailed their freedom. Aside from legitimacy ("here's the law I should comply with because it's the law") and prudential concerns ("here's what I must do to avoid sanction"), consideration of the content of rules can provide two important pieces of information. First, that the behaviours involved are morally appropriate (or inappropriate) and thus should be adhered to (or avoided). Assuming the process through which the rules are decided maintains some baseline level of legitimacy, rule content becomes "what we do", playing a vital signalling role and helping establish new social norms.
Second, the content of rules allows people to coordinate behaviour. As Richard McAdams [7] notes, specific rules, such as stopping at red lights, allow us to manage interactions with unknown others. Because both parties at an intersection know that the person at the red light will stop, at least in part because they will assume that the person at the green light will go and hit them if they do not, their behaviour is coordinated effectively without further interaction. This process does not actually require a legitimate rule-maker, and it may be very useful in understanding, for example, mask-wearing during the pandemic. Because the primary benefit from masks accrues not to the wearer but to others in their environment, putting a mask on only becomes worth it, from a rational choice perspective, when they have confidence that others will do the same. The existence of the rule may provide such reassurance (which is not to say that other reasons, such as a desire to help others, may not be as or more important).
"Us and we", "me vs. them" All this invokes a key distinction that we sought to explore in our panel study. On the one hand, people may comply with lockdown restrictions for "us" or "we" related reasons: because they buy into the idea of collective action against the virus and because they wish to maintain positive relations with legitimate authority and/or others in their communities. A key claim in both cases is that they have internalised the idea that compliance is the right thing to do, and it becomes an important part of their identity and self-concept.
On the other hand, people may comply with what we might characterise as "me vs. them" reasons. Compliance may be entirely prudential, premised in fear of the virus or of sanctions for non-compliance, and/or it may be based in fear of informal social sanction-of what others might think-that does not reflect internalised commitment but, rather, the desire to avoid disapproval and/or stigma.
We sought, in other words, to test two potentially contrasting accounts: 1. That compliance is rooted in fear of catching COVID-19 and fear of formal and informal punishment (police deterrence and social norms based on disapproval, "them"). 2. Compliance is rooted in collective solidarity and coordination (expressive function of the law and a broader conception of social norms based not only on prescriptive and proscriptive norms but also ingroup moral norms, "us"; we are all doing it for each other).

Method
In April 2020 we started a multi-wave panel study to track the experiences, attitudes and behaviours of 1,200 people-300 living in London and 100 living in each of Edinburgh, Newcastle, Cardiff, Leeds, Liverpool, Manchester, Birmingham, Sheffield and Glasgow. It is important to stress that our sample was not representative of the 10 cities. The strength of the data lie in our ability to field a large number of measures quickly and our ability to track individuals over time (every 3 weeks). Key to our analysis is, therefore, not only modelling the predictors of compliance in each wave separately, but also the tracking of trajectories over time.

Sample
Our data come from the Policing the Pandemic longitudinal study. 2 The surveys were hosted on Qualtrics and fielded on the online platform Prolific Academic. Prolific is similar to other crowdsourcing platforms such as Mechanical Turk but has a larger, more diverse pool of UK participants. We analyse here data from the first, second and third waves collected on 21 April (wave 1), 11-14 May (wave 2) and 1-5 June (wave 3) 2020. Quota weights were calculated for gender and age. Although quota sampling is not probabilistic, the stratification involved with the sampling strategy meant that after the data collection, quota weights could adjust the results, making them slightly more representative of the 10 cities. The first wave was fielded one month after the first lockdown began. The second wave was fielded the day after UK Prime Minister Boris Johnson's address to the nation that announced an easing of the initial lockdown restrictions. These changes involved revised messaging from "stay at home" to "stay alert" as well as some rule changes (e.g., allowing more outdoor activities within households, and more contact between them) that came into force on 13 May. The data collection of the third wave took place 10 days after the nation learnt about the lockdown breach by Dominic Cummings, the prime minister's chief advisor, which was followed by a national outcry. The data collection for wave 3 also coincided with further easing of the restrictions, including permitting up to six people to meet outside, the reopening of some childcare facilities, and so on.

Modelling
Survey respondents were asked: "How often during the past week have you engaged in each of the following behaviours?": (1) Socialised in person with friends or relatives whom you don't live with. (2) Gone out for a walk, run, or cycle and spent more than a few minutes sitting somewhere to relax. (3) Travelled for leisure (e.g., driven somewhere to go for a walk). The response alternatives were [0-4: Never, rarely, sometimes, often, or very often]. We added up people's answers to each of these three questions, creating a "mock-count." Figure 1 shows the distribution of self-reported compliance in each of the three waves.
In each wave, we compared two models (both included city, age, gender and the interaction of age and gender as controls): 1. Instrumental and prudential factors: concern about getting COVID-19, self-assessed knowledge about the virus, police deterrence perceptions and prescriptive/proscriptive social norms based on the belief that other people follow the rules and would disapprove of others if they did not follow the rules. 3 2. Adding legitimacy of the law and the expressive function of the law and swapping prescriptive/proscriptive social norms for a broader conception of social norms based also on in-group moral norms, "us"; we are all doing it for each other. 4 We examined the predictors of compliance for each wave separately, and we fitted cross-lagged panel models to assess the predictors of change over time. We briefly summarise these results here. 5

April 2020: One month into lockdown
Wave 1 data were collected just one month after the beginning of lockdown on 23 March. Compliance was high but there was some variatio. Neither concern about catching COVID-19 nor knowledge about the virus was a predictor of compliance in any of the models. Deterrence perceptions, legal legitimacy and social norms based on perceptions of what others in one's community thought and disapproved of were similarly unimportant. So what explained the generally high levels of compliance? We asked people whether they thought it was right or wrong to make social distancing a legal requirement (just under 90% of our research participants said it was "right" or "completely right"); whether by making it a legal requirement the government sent the message that social distancing is important to fight the pandemic; and whether making social distancing a legal requirement helped to clarify what they should and should not be doing.
We found that peoples' endorsement of these sentiments predicted their lockdown compliance, adjusting for fear of the virus, police and law (and all the other factors). The law, through what it expresses, can influence our beliefs and coordinate our behaviour. Two complementary processes may be at work here. First, the content of law expresses to citizens what authorities and other citizens deem to be right or wrong and reveals widespread attitudes towards the behaviour the law regulates. Second, by shaping expectations about how others will behave (e.g., that they too will comply with restrictions), the content of law helps to coordinate behaviour. The law helps to create concordant expectations. As mentioned earlier, we approached social norms in two ways. The first, narrower, conception focused on people's beliefs about what "most people in their local community" thinks about rights or wrongs about social distancing and whether others in one's community would disapprove if the guidelines were not being followed. The second, broader identity/group conception added to those sentiments one's own belief about whether "everybody" should follow social distancing to help prevent the spread of COVID-19 and whether it is important that everybody does so to help protect the National Health Service (note that in this formulation, respondents were free to infer who "everybody" is).
We found that peoples' support for and acceptance of the first definition of social norms was not a significant predictor, but that the second definition was. By enabling people to express shared values, meaning and identities, norms help to make people accountable to each other: …we project a certain image and craft a sense of self: of people who care about the dignity of human beings, for whom death is a sombre and serious business, for whom the group is more important than the individual, or whatever. We are, in effect, expressing to one another, through our recognition of one another's right to demand this or that, what matters to us, who we take ourselves to be, and how we see ourselves and others [8 p37].
Overall, what explained variation in lockdown compliance at this point in time were moral in-group norms and the expressive function of the law. It seems that by following informal and formal rules, compliance enabled people to express shared values, meanings and identities. Most important to self-reported lockdown compliance was the belief that "we are all in it together and we all need to come out of it together"-a sense of a common fate, a shared identity, and acting for the common or the social good, centred around national sentiment towards the NHS.
In other words, making social distancing a legal requirement may have strengthened public compliance not through deterrence or legitimacy but through signalling that the nation needed to take social distancing seriously, underlining what people should and should not be doing, and why. By turning social distancing into legal requirement, the legal system seems to have acted as an expressive agent, sending the message to the nation that the threat is to the national group rather than the individual and that we collectively need to respond to the virus' threat, clarifying how citizens should act in order to fulfil this.

May 2020: On the cusp of the initial easing of lockdown
What about three weeks later? There were slightly lower levels of compliance, and consequently slightly more variation to be explained, although overall-with one exception-findings were similar to those from wave 1. Social norms were again important, as was the expressive function of the law. In wave 2, 94% of our respondents somewhat or strongly agreed with the statement "by making it a legal requirement, the government sent the message that social distancing is important to fight the pandemic", and 92% somewhat or strongly agreed with the statement "introducing the social distancing rules helped communicate to the public the need to do what we can to stop the pandemic from spreading". Stronger agreement with these statements was associated with greater compliance.
We also asked about citizens acting as expressive agents in response to the legal requirement, with their compliance broadcasting their support for the community. We found that 87% of respondents agreed with the statement "observing the social distancing laws shows other people in my community that I care for their safety" and 82% agreed that "following the social distancing rules helps me feel that I am part of the collective fight against the pandemic". Acting in unison seem to bind people together, especially when there was a legal requirement to coordinate at the group level against a common threat.
Compared to wave 1 analysis, however, concern about catching COVID-19 was a positive predictor of compliance: people who were most concerned were a bit more likely to comply most fully with lockdown restrictions. Levels of concern about catching the virus were a little lower, and this may have partly explained why compliance went down marginally. In addition, our panel analysis modelling change over the three waves found that levels of concern at wave 1 predicted changes in compliance between waves 1 and 2. This change was apparent in that while people who were "very concerned" in April tended to continue to be concerned in May, those who were only "concerned" in April were associated with slightly lower levels of compliance between April and May.

June 2020: Restrictions being eased
By wave 3 the behaviours that comprised our measures of lockdown compliance were no longer illegal. Yet, a fair number of people continued to curtail their behaviour, electing not to travel for leisure, to meet up with people, or to stop for a rest in public while exercising. The correlates of these behaviours remained similar, particularly social norms and the expressive function of the law. Concern about catching COVID-19 was again a predictor; indeed our panel analysis showed that people who were very concerned in wave 1 tended to continue refraining from these behaviours.

Policing the pandemic
Finally, we turn to the role of police during the lockdown. In wave 1 only 27 out of 1,200 people had contact with police (15 of whom had reported not completely respecting the lockdown rules) and all but one or two agreed that the police were polite, treated them with respect and explained why they had been approached. In wave 3 we asked about the previous 6 weeks, with similar results.
Police activity in enforcing the lockdown seemed to largely be taking place at the margins, at least from the perspective of our respondents. This was necessarily the case-policing on its own would never be enough to maintain the restrictions. However, police do sometimes need to take action, moving people on, issuing fixed penalty notices and even making arrests.
Our admittedly limited data on police contact during the first lockdown suggests that the 4Es policing strategy was successful among our research participants. We found that perceptions of police legitimacy and procedural fairness remained high across the three waves. To measure legitimacy, we asked our research participants whether they thought it was their moral duty to obey the police, to support the decisions of police officers even if they disagree with them, whether they believe that the police usually act in right and proper ways, and finally if the police stand up for values that are important for people like them [9]. The distribution remained relatively consistent across the three waves, and a quantile-quantile plot also showed little change [10].
To measure procedural justice, we asked people whether they thought police officers in their community make fair, impartial decisions in the cases they deal with, explain their decisions to the people they deal with and treat people with respect. As with legitimacy, the distribution remained pretty similar across the three waves, and a quantile-quantile plot showed little change.

Conclusions
We have provided above a range of possible "drivers" of lockdown compliance. Understanding which seemed to shape people's compliance behaviours during the first lock down of 2020 may provide important information for policymakers and others, particularly as, writing in March 2021, some form of restrictions seem likely to be in place for a few more months to come. We are still in lockdown 3.0, and just like before, many elements of the restrictions and guidelines, like legally mandated social distancing and quarantine measures, are effectively unenforceable. Police cannot present a credible threat of sanction to most/all people in most/all situations.
Some of our findings are consistent with prior evidence: social norms were important (albeit in our work, only when they include group-based moral norms), as was fair and respectful policing (so as not to threaten trust and legitimacy), and deterrence rarely plays much of a role. But three things stand out.
First, when defined solely as proscriptive ("they" do it, "they" will disapprove), social norms were not important. It does not seem that social stigma or informal sanctions played much of a role in people's thinking. What made social norms important in our analysis was broadening the definition to include the moral and in-group norms regarding what people should be doing to help protect the NHS ("we" not "them").
Second, law was important, but not because of legitimacy, nor deterrence, but rather because of the expressive function of the law. Using force, or fear of sanction, to ensure compliance seems not to be the most effective way of procuring it. People seem to be more responsive to normative motivations to comply (I'll do it because it's the right thing to do) than instrumental motivations to comply (I'll do it because I'm afraid of formal punishment). Enshrining restrictions in law seems to have had the dual effect of signalling their importance to people-indicating that it was now inappropriate to behave in particular ways-and providing a set of guidelines around which they could organise. To put it another way, the law indicated to people how "we" should behave and afforded moral and practical under-pinning to the collective effort against the virus. While it is conceivable such effects could have been obtained by issuing non-enforceable guidelines, we doubt this would have had the same effect. The law provides a moral force to regulation that is lacking in other edicts and communications.
Third, concern about catching COVID only came in as a predictor of compliance when lockdown was starting to be eased. It is intriguing that prudential compliance based on a desire to avoid infection only "kicked in" once the easing of lockdown restrictions indicated that the threat of the virus had receded (which felt like a reasonable assumption in May 2020 even if it does not now). Perhaps the easing of restrictions indicated that the collective effort had been successful, meaning "me" related factors emerged as more important (and recall that the social norms became significantly less important once restrictions were eased).
Overall, it seems that by abiding by the social distancing guidelines, people express to each other a sense of in-group identity in a way that is additive, not reducible, to the role that adherence to social norms plays. Informal rules governing behaviour bind people together, but so too do formal rules-at least in the current public health context.
Finally, the Covid-19 crisis has thrown into sharp relief the fact that, absent sweeping authoritarian powers and their widespread use, the only way democracies can be policed is via consent. And because most compliance has sources outside formal policing arrangements, when police do intervene they need to do so in ways that maintain, rather than undermine, the bonds between individuals, institutions and wider society. This spells out the danger of any switch to more assertive or aggressive policing styles. Not only is any effect on compliance likely to be marginal at best, but in as much as such styles are experienced as unfair and illegitimate by those on the receiving end, their commitment to the collective effort is likely to be (further) weakened. The apparent success of police in walking this difficult line during the first lockdown is something to be applauded, and moves towards more enforcement-led approaches now should be treated with some caution.

Competing Interests
The authors have no competing interests to declare.