Defunding the police in the UK: Critical questions and practical suggestions

Correspondence Email: j.fleetwood@gold.ac.uk Abstract Calls to defund the police emerged from Black Lives Matter (BLM) protests of 2020, inspiring USA cities to shift funding from policing to social welfare. Here we consider how defunding might translate to the UK, raising critical questions about our distinct funding arrangements, and social welfare traditions. Next, we consider how the spirit of defunding could be adapted in the UK drawing on the left realist proposition of minimal policing, radically restricting police powers and autonomy. In contrast to many abolitionists, we foresee the state continuing to play an important role in ensuring justice through the development of specialist non-police led agencies to respond to serious crimes and residual conflicts.

easy translation of the slogan to a UK context in which police have been 'defunded' by government for at least a decade with no reduction in police violence against racialised minorities. Keir Starmer, Labour Party leader and former Director of Public Prosecutions dismissed the slogan as 'just nonsense' (BBC Breakfast, 29 June 2020). While Starmer's comments were quickly contested (Elliot-Cooper, 2020;Onapa, 2020), there are profound difficulties in importing US policy vocabularies (even radical ones) into UK politics.
The purpose of this article is therefore to re-articulate core elements of the defunding debate for a UK context, acknowledging our distinct institutional structures and history, drawing on academic work proposing radical reductions in policing, in particular the left realist proposal of minimal policing (Kinsey, Lea & Young, 1986). After outlining key aspects of the US debate we argue that we cannot adopt defunding wholesale. Rather than literal defunding (reduced financing), we focus on radically restricting police powers and autonomy but retaining -in contrast to some abolitionists -the key role of the state in delivering justice (Kinsey, Lea & Young, 1986, p.205), particularly for crimes and harms by powerful offenders and institutions immune to community control. Finally, we begin to sketch out how specialist agencies, rather than the police, could be involved in responding to social conflicts. In this way we hope to apply the spirit of defunding to the particular circumstances of the UK.

ORIGINS OF 'DEFUND THE POLICE'
The call to defund police is an international, activist response to excessive, long-standing racialised violence by police against Black people, especially in the USA. On the face of it, the idea is to take funding from police, and invest it in community initiatives that build safety. For some, defunding is an end in itself. Defunding makes particular sense in the USA, where it confronts escalating budgets and militarisation of the police (Bouie, 2014;Vitale, 2017, p.28), even as crime rates overall keep dropping (Morgan & Truman, 2019). For others, defunding is an immediate step towards police abolitionism. Police abolitionism mostly emerges from the USA (Chazkel, Kim & Paik, 2020;For a World Without Police, nd;McDowell & Fernandez, 2018;Vitale, 2017;Williams, 2004), and to a lesser extent the UK (Abolitionist Futures, 2020;Duff, 2021). It emphasises state racism, capitalism and the prison industrial complex, distinguishing it from liberal, European abolitionism (McDowell & Fernandez, 2018). Following prison abolitionism (i.e., Alexander, 2010;Davis, 2011;Gilmore, 2007), abolitionists aim to make the police, eventually, obsolete. That is, to challenge the 'assumed inextricable relationship between the modern state and the police: the belief that violence is necessary to the maintenance of public order' (Chazkel, Kim & Paik, 2020, p.5). Here, we focus primarily on defunding as an immediate, necessary and achievable intervention, setting aside longer-term questions of abolitionism.
Calls to defund the police come from diverse academic and activist contexts and are, of course, varied. Nonetheless, we identify three central aspects of the defunding argument. First the USA police, historically tainted by slavery and labour repression, exhibit a history of armed violence towards the Black community and are considered essentially unreformable (Chazkel, Kim & Paik, 2020;Duff, 2021;Vitale, 2017;Williams, 2004). It is argued that police and prisons are rooted in, and so integral to, maintaining racial and class inequality that no racial justice is possible without their abolition (inter alia, Chazkel, Kim & Paik, 2020;Davis, 2011;Gilmore, 2007). Decades of reform have failed to bring about change, instead reforms shore up a failing institution.
In the USA, defunding advocates argue compellingly that the police are essentially nonreformable. Vitale observed that prior to the murder of George Floyd, Minneapolis Police Department undertook massive reforms involving community dialogue, racial bias training, promotion of Black police leadership, use of body cameras and other innovations (Vitale, 2020; see also Minneapolis Police Department, 2018). For Vitale this unreformability derives from the historical origins of police in the protection of slavery and the repression of early labour organisation.
Neither do the police effectively deter or detect crime (Vitale, 2017). In the UK, as in the USA, the clear-up rate for serious crimes, like rape and sexual assault, are persistently low, despite decades of reforms. Racialised minorities are especially under-protected. The Metropolitan Police's failure to effectively investigate the murder of Stephen Lawrence in 1993 (Macpherson, 1999) has clear parallels with the more recent failure to investigate the deaths of Shukri Abdi in 2019 (MEND, 2020) and Blessing Olusegun in 2021 (BBC, 2021).
Second, defunding is not only about reducing funding for police, but diverting resources to alternatives that would increase safety and security. USA-based abolitionists Critical Resistance (2020) give an indicative list: 'quality, affordable, and accessible housing, universal quality health care, including community-based mental health services, income support to stay safe during the pandemic, safe living wage employment, education, and youth programming' (p.3). Rather than relying on police, crime can be effectively prevented through investment in welfare, housing, education and job creation, arguably more effective than police in crime reduction. Traditional liberal social policy sees welfare and policing as complementary mechanisms for social stability: the defund movement replaces this 'welfare plus policing' with 'welfare instead of policing'. While defunding might appear radical, its advocates include the police. In the UK the retiring Chief Constable of Merseyside, Andy Cook, told the Guardian that if he were given £5 billion to cut crime, he would invest only £1 billion in policing and the rest in measures to tackle poverty (Dodd, 2021a).
Third, as we have seen, the main assumption of the movement to defund the police is that police intervention, and the consequent construction of situations as 'crime' and requiring criminal justice solutions, could be better dealt with as social problems by a range of other agencies such as social and community work, probation, education and local authority agencies. Some of these at first sight seem obvious: 'Rough-sleeping, mental health crises, drug and alcohol-related problems, domestic violence, issues relating to Covid-19 . . . could be far more effectively and humanely addressed by well-funded, community-based emergency support teams, crisis intervention and violence de-escalation workers' (Duff & Kemp, 2020).
The fact that in some cases police officers may deal sensitively with such cases simply serves to highlight the essential redundancy of the policing perspective over so much of what the police actually do. It is well known that the vast majority of calls by the public to police -both in the USA and the UK -are not crime-related but stem from the fact that the police are available and capable of rapid response, and that other agencies have been progressively defunded. Where police are involved there is always a risk of criminalisation and so rather than relying on the police to thoughtfully refer cases to expertise in these agencies, it is these agencies -suitably funded and equipped -which we propose would request police assistance in situations where the use of legitimate force is deemed necessary as a last resort. The issue of how the agency of first response could be decided will be discussed below.
Abolitionist defunders contend, further, that conflicts which do arise can be dealt with by communities themselves. Instead of needing police, communities will be 'empowered . . . to solve their own problems' (Vitale, 2017, p.30). In the USA the Black Panther Party's 'community defense' offers an important historical precedent (Williams, 2004, p.226). For example, a contributor to 'For a World Without Police' proposes that communities create 'No Cop Zones' to disempower police. In an American spirit of self-reliance, communities can study and share conflict mediation strategies and develop 'support networks, crisis centers and self-defense groups' (For a World Without Police, 2016). Likewise, a special edition of Radical History Review (Chazkel, Kim & Paik, 2020) highlights an array of community responses to crime, outside of formal state justice institutions. In the UK, the edited collection Abolishing the police (Duff, 2021) highlights alternative responses to crime, including fostering cultures of care and accountability (Lamble, 2021) and exploring alternatives to police coercion (Aitchison, 2021). However, abolitionists have scant discussion of terrorism, serious organised or corporate crime whose harms likely 'outstrip in extent and seriousness either victim-survey or police-recorded statistics' (Reiner, 2016, p.114). The question of which agencies might attend to such crimes, and some of the problems with community conflict resolutions are discussed in more depth below.

DEFUNDING THE POLICE AS 'POLICY TRANSFER'?
In the UK, calls to defund the police are comparatively novel but quickly garnered support from academics and activists (i.e., Begum, 2020;Duff, 2021;Duff & Kemp, 2020;Elliott-Cooper, 2020;Onapa, 2020). While defunding is a highly effective campaigning slogan, we urge caution in straightforwardly adopting USA ideas, even progressive ones, wholesale. Policies -whether they originate from the state or communities -bring with them ways of imagining crime, responses to it and the state (Newburn, 2002;Wacquant, 1999). For example, in the 1990s and early 2000s concepts like 'zero tolerance' or 'three strikes' became common-sense solutions to crime problems in Europe, reflecting neoliberal ideologies about crime and punishment originating in the USA political climate (Newburn, 2002). Likewise, much writing on police abolitionism reflects peculiarly USA forms of common sense, suggesting appropriate targets of campaigning, and the kinds of alternatives that can be imagined.

Distinct funding for police
The injunction to 'defund the police' reflects the funding arrangements in the USA. Without getting into the complexities, local authorities manage the budgets for police as well as housing, social welfare, education, and so forth. While some cities like New York spend just 8% of their budget on policing, in Minneapolis it is 35% (VERA Institute of Justice, 2021). Numerous USA cities have cut police funding in response to campaigns (McEvoy, 2020), including Minneapolis (Levin, 2020). People's Budget Los Angeles (2021), in coalition with BLM, successfully campaigned to permanently ring fence at least 10% of city budget for direct investment in communities and alternatives to incarceration. Over $300 million dollars will be invested annually in rent assistance, supporting small businesses, offering training and youth services as well as funding restorative justice community measures. A different approach is needed in the UK, where most police funding comes directly from central government (Brogden & Ellison, 2012). While Scottish Police are completely funded by government, around 30% of funding for police in England and Wales is raised locally through the council tax precept (Home Office, 2020). Local councils do not have control over the precept, or how it is spent -in fact Metro Mayors and Police and Crime Commissioners (PCC) do but they are not able to redirect the precept towards housing or health budgets. Furthermore, it seems difficult to imagine a PCC or Metro Mayors campaigning on the basis of defunding police without being able to offer an alternative, given that the main political parties aim to increase police numbers. Thus, the notion of 'defunding' does not reflect institutional arrangements or funding for police in the UK.

Police and social services have been defunded
In the USA, calls to defund the police are meaningful against a backdrop of expanding budgets. Curbing police resources, and thereby police power, follows logically as a way to reduce police violence. This logic does not fit the reality of the UK, where police budgets are shrinking. In England and Wales police have been steadily defunded since the economic crisis of 2009, resulting in about 19,500 fewer police officers (about 14% of 2009 levels) (Home Office, 2019b, p.4). Likewise, financial austerity has reduced funding for policing in Scotland (Millie, 2014) and Northern Ireland (Brogden & Ellison, 2012). Moreover, in the UK the balance between police and social services has moved against the latter. Despite financial restraint, police are increasingly responsible for a wide range of prevention and emergency responsibilities previously undertaken by social services. Millie (2013Millie ( , 2014 calls this 'policification' of social policy. Thus, due to massive financial cuts to social services over the last decade, the police are often the only state agency available in a crisis. Funding cuts have also seriously impacted on police capacity to engage with communities, or to respond adequately to victims of crime, especially Black/working-class communities and other racialised minorities. Austerity has also hollowed out other state agencies (community centres, youth work, drug and alcohol services) that might have previously been able to prevent and manage crime and conflicts in the community (Millie, 2013). For example, despite the recognition of 'hate crime' and mobilisation of police, hate crimes continue to rise, and the needs of victims go largely unmet (Chakraborti, 2018). Thus, defunding campaigns need to recognise our distinct starting point: that both social services and police have been defunded, and the police currently stands as an agency of last resort for all sorts of social problems.
The experience of the past decade makes crystal clear that 'defunding from above' has in no way reduced police violence towards Black people or other racialised minorities in the UK. Since monitoring was introduced in England and Wales in 2017, police use of force has increased by 36%, from 313,000 incidents in 2017/18 to 428,000 in 2018/19 (Home Office, 2018, 2019a). Further, use of force against Black people constitutes 16% of incidents of use of force (Home Office, 2019b), despite representing only 3.3% of the population of England and Wales (Office for National Statistics, 2011). In London, use of Tasers is racially disproportionate (Noor, 2018). Black people are also still six times more likely to be searched for drugs than are white people (Shiner et al., 2018) and are more likely to die in police custody (INQUEST, 2020) for which, since 1969, no officers have been held accountable.
Much USA defunding debate seeks to replace police repression with community-based lay violence control. Community alternatives are not without problems, however. Looking at community courts in 1970s Northern Ireland and street courts in apartheid South Africa, Williams (2004) notes that there is a tendency for community justice to 'uphold traditional hierarchies of gender and age' (p.230) Furthermore, in the UK a strong welfare state tradition remains, despite the impact of austerity. Nonetheless, as Jacobs et al. (2021) warn, social work is not straightforwardly a good alternative to police. Social workers -like the police -have historically been involved in the oppressive management of Black, migrant and poor populations (Jacobs et al., 2021), an orientation heavily supplemented in recent years by coercive risk-management strategies in both probation and social work generally (Fitzgibbon, 2011).
Defunding the police does not translate well to the UK. First, funding arrangements rule out an easy reassignment of funds from police to alternatives. Second, a decade of cuts to both police and social services mean that we are at a position of profound deficit. Third, the notion of community conflict resolutions do not sit well with the UK's relatively strong welfare state. We can now inspect more closely how a 'defunding' orientation might tackle the entrenched problems of policingincluding police violence -in the UK.

POLICE AUTONOMY AND THE COMMUNITY
In our assessment, the key issue is not simply police funding but organisational autonomy. The debate should be not just about police funding but also about police organisation. This is where 'defund the police' is a partial orientation. Two examples illustrate the centrality of the problem of police organisational autonomy.

Stop and search
Decades of critical debate dating from at least the Scarman Report of 1981 (Scarman, 1981) has failed to dislodge stop and search as a police practice inflicted disproportionately on poor and Black communities despite its acknowledged ineffectiveness as a form of crime control (Bradford, 2016;Lea, 2000;Tiratelli, Quinton & Bradford, 2018) and despite numerous attempts to reform its operation (see Dodd, 2021b). Prevent/anti-terrorism strategy extends over-policing to Muslim communities (Philips & Bowling, 2017). A recent official admission of the futility of stop and search as effective crime control by HM inspectorate (HM Inspectorate of Constabulary, Fire & Rescue Services, 2021) states: 'Over 35 years on from the introduction of stop and search legislation, no [police] force fully understands the impact of the use of these powers. Disproportionality [in stopping racialised minorities] persists and no force can satisfactorily explain why' (p.5). The report notes that stop and search relates overwhelmingly to drugs but deals with possession rather than supply and is therefore ineffective as a crime control measure (p.2) and that 'the potential damage to public trust and police legitimacy is substantial' (p.6). Most searches are based on officer suspicion rather than hard intelligence (p.34), and usually find nothing! (p.32).
That officer suspicion is precisely what leads to disproportionality has been known for decades (see Lea, 1986;Lea & Young, 1984). Yet the report wheels out the same tired recommendations, many of which go back to Scarman and Macpherson, such as better recording of grounds for stop and search or analysis of body camera data.
Intelligence-led stop and search operations requested by communities or welfare agencies, focused on real crime problems -such as evidence of a significant volume of weapons entering a locality -might be acceptable, unlike stop and search for small amounts of drugs. Yet whatever the area crime profile, drug possession (and in particular the dubious claim that the officer could 'smell cannabis') is almost always the rationale for stop and search (see Delsol & Shiner, 2015).
Why do police continue with a practice both inefficient as crime control and which, by alienating support for the police in poor and Black communities, reduces the flow of information about crime thereby further undermining police effectiveness? Senior Black police officers have criticised the practice in precisely these terms (e.g., Gayle, 2018;Logan, 2018;Taylor, 2020).
First, because there is no direct authority which can stop police commanders initiating such measures. Second, stop and search is rooted in the historical origins of police, particularly in disciplining of the poorest sections of the working class. The father of British policing, Sir Robert Peel, founded two distinct forces. In 1829, his London Metropolitan Police were allegedly 'citizens in uniform' and the model for uniquely British peaceful policing. But earlier, in 1814, he founded the Royal Irish Constabulary (RIC), as a paramilitary gendarmerie devoted less to crime control than the pacification of a rebellious Irish peasantry under British rule. The RIC became the model for British colonial policing and was unfavourably -and misleadingly -contrasted with the 'consensus policing' of the Metropolitan Police. There was, in fact, some direct overlap. As the 'new police' spread beyond London: 'during the 1840s, for example, a few English counties believed that the Irish Constabulary constituted a better model than the police of London for their own circumstances' (Emsley, 2014, p.10). Also, the Metropolitan Police in London avoided building upon existing crime-control innovations by magistrates such as the Fielding brothers and concentrated instead on 'the moralisation of the poor and the continual harassment of those identified as the least moral sections of the poor' (Rawlings, 1999, p.77).
The role of police in the harassment and repression of the working class continued (Brogden, 1982;Storch, 1976) until the increasingly organised working class forced a 'grudging acceptance' between police and community (Brogden, 1982, p.184; see also Cohen, 1979) in which crime control came to predominate over general repression. During the post-Second World War welfare state and strong organised labour movement of the 1950s and 1960s the police gained legitimacy (Reiner, 2010) as efficient crime fighters and relatively impartial guardians of multi-class access to public space.
The capacity to force the police into anything resembling 'grudging acceptance' of its rights was always denied to the colonial masses. From the 1960s the latter provided Britain with a politically marginalised immigrant proletariat (Lea, 1980;Rex & Tomlinson, 1979), which then became part of the growing 'precariat' (Standing, 2011) but was constructed as pathological, criminal and in need of police control (Bhattacharyya et al., 2021;Centre for Contemporary Cultural Studies, 1982;Hall et al., 1978).
The British state responded to colonial immigration with colonial policing merged with domestic traditions of general repression of the working class. The result was the current institutional racism of stop and search. The political marginalisation of Black communities results in a lack of local political power -despite strong local activism and the determined support of some local authorities -sufficient to force a still overwhelmingly white police force acting as the 'state on the street' (Brogden & Ellison, 2012) into any sort of compromise (Lea, 2000). Black communities, especially the young, experience police as an 'occupying force'. The issue is police autonomy and power rather than police funding as such.

Multi-agency collaboration
The role of the police in multi-agency working further illustrates the need to restrict their autonomy. The issue is not that an element of police presence may not be necessary. In many situations which can be diverted away from policing and criminal justice towards mental health, conflict resolution or removal of vulnerable people to safe spaces there is, nevertheless, frequently a high risk of violence as an element of the situation. Rather, the problem is precisely how police work with welfare, community work and other agencies as part of multi-agency collaboration and what structures of accountability exist. There are two aspects to the problem. First, welfare and community work are continually undermined by budget cuts allowing police, despite their own defunding, to encroach on the proper territory of welfare agencies, and of probation and education (Millie, 2013). Harm problems then come to be defined as crime problems purely because the police deal with them. But, second, problems of police invasiveness are compounded by doctrines of constabulary operational independence. Police independence from other agencies -even those with which they are legally required to work closely and share information, as for example with multi-agency public protection arrangements (MAPPA) under the Criminal Justice Act 2003, for the supervision of offenders in the community -undermines attempts to shift emphasis from police to welfare. Many police forces have well-established and good collaboration with local authorities and welfare agencies but police autonomy, especially in the initiation of stop and search operations, continually undermines welfare approaches to social problems. A current example is that of combatting knife crime.
In Glasgow, in 2005, a public health orientation to knife crime saw the establishment of a Violence Reduction Unit (VRU) to which young people are referred as an alternative to movement into the criminal justice system. In fact, Strathclyde police were instrumental in setting up the initiative whereby young people involved in knife crime are seen as at risk rather than as criminal offenders and are offered a range of services by welfare, health and voluntary agencies. But at the same time, police in Scotland have one of the highest rates of stop and search anywhere in the UK and the 'debatable role of stop and search as a general police tactic is certainly one aspect that complicates the picture of impact' of the public health approach of the VRU (Grimshaw & Ford, 2018, p.22).
Meanwhile, in London, attempts to emulate the Glasgow approach contrasted with simultaneous policing strategies involving an intensification of stop and search. The Metropolitan Police had been defending the measure as an important strategy against knife crime (Weaver, 2017) notwithstanding its detrimental effects on their relations with young Black men (Keeling, 2017), a sentiment echoed by several senior Black police officers (see Gayle, 2018;Logan, 2018). The police, with the support of central government, began trialling new 'knife crime prevention orders' which a police spokesperson welcomed as a 'valuable tactic in preventing violence on the streets of London'. However, Sarah Jones, chair of the All-Party Parliamentary Group on knife crime and violence reduction retorted that such orders 'risk criminalising a generation of young people who . . . need care and support rather than punitive action' (BBC, 2020). Nevertheless, it will be our argument that 'defunding' -a shift of resources from police to public health and welfare -can be combined with organisational changes forcing police to function as a subordinate part of a welfare approach to social harm. How such a reorganisation might function we discuss below.

Community alternatives
Could community-based alternatives be established as an alternative to police intervention? As we have seen, USA defunding discussions tend to assume a community resilience sufficient to replace police as 'occupying force' by varieties of community-based lay-violence control. In the aftermath of the 2011 riots, the UK Home Office showed some official interest in 'violence interrupters'originating in the USA -in which ex-gang members attempt to prevent retaliatory action in cases of inter-gang violence (see Home Office, 2011, pp.44-5). But inter-gang mediation hardly constitutes a general crime control system sufficient to avoid the need for police authority as a last resort particularly for the growing amount of crime (see below) that does not originate in local streets or communities even though it may impact them. But even for crime which does originate in local communities -for example, theft and violence, domestic violence and sexual assault -the notion of lay community intervention as a sufficient response presupposes a resilient community with a strong consensus around values, problems and how to solve them. Much effort by community and social work may aim at establishing this but it cannot be taken for granted. Community resilience has rather to be continually rebuilt in the face of economic decay, and in Britain the still strong cultural and political traditions of the welfare state are vital in this respect. In divided and fragmented communities, political groups, the far right for example, may capture control of 'community alternatives' while middle-class residents may -and increasingly do -choose to employ private security. In seeking to solve one problem we may create others. Last, community alternatives may be in conflict with police. In response to Scotland's tragic rise in drug deaths, Peter Krykant set up a mobile drug injection facility, offering a safe, clean place to inject drugs. Despite evidence that such facilities save lives, Krykant's facility lacks legal status. In preventing police entering his van to search service users, he was charged with obstructing justice (the charge was later dropped) (McGivern, 2021). Likewise, BLM activists report being surveilled by police, called on at home and having their social media accounts disabled (Gayle, 2021). In this context, it is hard to imagine that police would keep their distance sufficiently for communitybased alternatives to flourish.

THE CRISIS OF GENERIC POLICING
In recent years, conventional police thinking has been preoccupied with a 'crisis of complexity'. Crime has become increasingly differentiated such that it can no longer be effectively dealt with by a single generic police agency. If community-based lay volunteers cannot deal with international child trafficking or cybercrime as well as local community disputes then neither can a single generic police force. While radical reformers want to replace police with welfare agencies dealing with a spectrum of harms and crimes in a non-coercive way, some police thinkers are contemplating the inevitable replacement of 'the police' by a spectrum of specialist agencies, a tendency already evident with such agencies as the Serious Fraud Office (2021) and the UK National Crime Agency. In July 2020, the Police Foundation, the leading conventional UK police policy think tank, published the first of a two part Strategic review of policing in England And Wales (Police Foundation, 2020). The report details the decline, since the mid-1990s, in traditional volume crimes such as burglary and vehicle theft and the rise in a new spectrum of crimes including domestic abuse, child sexual abuse and sexual assault, hate crime, drug trafficking, modern slavery and knife crime, all of which 'pose a significant challenge to the police who have limited resources and whose traditional ways of operating have struggled to keep up with the scale and complexity of demand' (Police Foundation, 2020, p.4).
As far as the Police Foundation is concerned, the key issue is: given that other specialist agencies must make an increasing contribution, what should their relationship with the police be? This discussion is deferred to the second part of the report, expected later in 2021. Nevertheless, the report contains a broad discussion of some approaches derived from the conventional policing paradigm. The first is the existing 'multi-agency' approach in which police collaborate with other specialist agencies. The problem here, as we have argued above, is that constitutional autonomy undermines any guarantee that police intervention will be coherent with, and not undermine, the work of other agencies.
A second solution is to put the police in overall charge -at the top of the pyramid -as the coordinating agency with other specialisms reporting to them. The notion of a new type of 'generic' policing which co-ordinates an array of specialist agencies as information providers for police intervention is problematic. In particular it is vulnerable to what Ian Loader, in a paper specially written to accompany the Police Foundation report, calls 'ambient policing': 'a form of policing that becomes intrusively pervasive across much of social life' (Loader, 2020, p.12). The concept has some similarities with Millie's (2014) idea of 'policification' noted above. Thus, child mental health practitioners might be deployed in cases of online harm but, in reporting to police, would inevitably help expand the policing and criminalisation perspective into what should remain an education and mental health issue. However, this co-ordination of a range of specialist agencies does not itself have to be a police function. Below we shall discuss a possible alternative to 'ambient policing', The Police Foundation, in the research for its report, ran focus groups of ordinary citizens and concluded that the fact that the bulk of calls to the police are not crime related led group participants to sense a 'mission creep' and to voice a feeling that police intervention, in, for example, mental health issues, should be limited to what was 'crime-related' or 'just the immediate crisis' (Police Foundation, 2020, p.55). This problem of course becomes more acute as welfare agencies themselves experience government defunding which, despite decline in police funding and personnel, pushes the police into the position of first responders in areas where they might not otherwise intervene.
However, a third alternative, and the one advocated here, is to turn the issue on its head and make the specialist and welfare agencies -with suitable increases in both funding and training -potential first responders with police as backup where necessary. What needs to be ended is police insistence on not only automatic status of first responder but also the discretionary power to reframe complex issues -such as mental health, homelessness or drug use -as crime and therefore with supposed criminal justice solutions. In simple terms, we propose a radical shift in the role of police, radically restricting their scope and autonomy.
The unique police contribution to any situation is the availability of non-negotiable force (Bittner, 1990, pp.121-2). Loader (2020) succinctly states the issue: 'If such force is not at least potentially required, it is difficult to see what legitimates police involvement' (p.14). In each case the necessity of police intervention needs to be evaluated as much by welfare agencies rather than the necessity of welfare intervention being decided by police. We argue that conflicts can be resolved by placing the decision as to the nature of intervention in a particular case in the hands of a third party.
One candidate for the status of third party might be the call handlers who receive emergency calls from the public or other agencies. However, in such a model, handlers would have to make swift decisions about the nature of complex situations under conditions in which callers may be unable to clearly understand the nature of the incident or which agency would best be able to solve the problem. It is one thing to distinguish fire, police or ambulance in caller requests, it is another entirely for call handlers to acquire the skill to distinguish the need for mental health, dispute resolution and several other possible modes of social intervention. Many situations are precisely ambiguous and multifaceted. What may ultimately turn out to be a mental health or housing issue may, at the time of requesting assistance, be very much characterised by violence or the threat of violence. Even if a welfare agency took on the role of first responder, the agency would likely request accompanying police presence 'just in case', as happens in any case with many social work interventions.
None of these three possibilities is satisfactory. Further, we note that crimes (or social conflicts) may not always require an emergency response. With this in mind, we propose another way forward.

TOWARDS MINIMAL POLICING
Here we begin to sketch out how we might shift state responses to social problems from criminal justice to state-based, non-criminalising alternatives. Rethinking the role of police inevitably has implications for other aspects of criminal justice (i.e., courts, prisons, etc.). Our focus here is on the police. We can imagine an ongoing role for police in managing public order. The governance of road traffic, the guaranteeing of peaceful and non-discriminatory access to public spaces and events are all important functions for the police. In the French system there is a distinction between police judiciare, in which police work under the instructions of the public prosecutor in criminal investigation, and police administrative which concerns the latter problems of the management of public space and populations. In this latter aspect the police work under the authority of the mayor. We see no reason why this arrangement should not have more influence on policing in the UK and lead to a better sharing of responsibility and authority with other agencies concerned with public space and safety such as Fire and Rescue Services.
England and Wales as a criminal jurisdiction suffers an exaggerated form of police autonomy due to the historic absence of the office of public prosecutor. One does not even have to leave the UK to discover the uniqueness of this situation. In Scotland police are legally the agents of the Procurator Fiscal to whom they report and who, in serious cases such as murder, will take an active part in directing police investigations (Donnelly, 2014). This does not reduce the predominance of the criminalisation paradigm in responding to harms but it does establish the principle that the police report to, and may take instruction from, a higher legal authority. In our opinion, placing the control of response in the hands of an official, whom we shall provisionally call the Controller, would solve the problem of police autonomy, and of police acting as the de facto co-ordinator of the response of a diversity of specialist welfare and other agencies -Loader's 'ambient policing' problem.
The Controller would hold a position analogous to a Scottish Procurator Fiscal with two important differences. First, the responsibilities of the office would not be simply directing police activities but the decision as to whether the police are, indeed, the appropriate agency to continue the response to a situation once the initial report has been made. Second the office would be subject to democratic oversight. This is discussed below.
The police, having less power and autonomy, could be first -blue light -responders to emergency situations involving violence or its threat. But their main aim would not be to gather evidence, make arrests and define the situation as one of criminality but simply to 'make safe' the situation by eliminating the possibility of further violence. This role accords with public opinion that the police should attend to 'the immediate crisis' (Police Foundation, 2020, p.55). This could, of course, involve arrests but these would lead to a social, rather than criminal justice, response. In urgent situations, the police would immediately make a report to the Controller whose task it would be to characterise -with specialist advice -the nature of the incident and direct the appropriate specialist agencies to intervene.
In non-emergency circumstances the Controller could directly refer the situation to one or more welfare agencies oriented to a 'problem solving' approach. A neighbour might request a response to antisocial noise in their neighbourhood to which the Controller might convene a mediation session. A GP might refer over safeguarding concerns for a child, and the Controller might convene a specialist panel of social services appropriate to the situation. A community organisation might report persistent racist vandalism, to which the Controller might trigger a police investigation.
Police would remain as backup in case force is required. But in such arrangements, there would be no power for police to decide the nature of the case at issue or to act autonomously or to assume leadership. This would be a decision of the Controller whose remit would be guided by principles of conflict resolution and violence reduction. It is not ruled out that the case will be characterised as criminal and police will be the lead agency -as in burglary, other volume crimes and when the perpetrator is not yet identified. The police are uniquely able to lead criminal investigations and would retain legal powers to do so. Police would also be directed as lead agency in incidents such as homicide, terrorism or other serious violence where the safety of the public is the immediate priority. In cases of international trafficking networks other specialist police agencies would be involved at a national and, indeed, international level.
But the important point is that causality is reversed: police become involved because it is deemed by the Controller a criminal matter, rather than it becoming a criminal matter because police are involved. Other agencies meanwhile would avoid any tendency to acquire police powers but would call on police support at the discretion of the Controller. This principle is already operative in embryonic form in England and Wales in certain areas. An example is the Serious Fraud Office (SFO) in which police fraud experts work as part of a multi-agency team including corporate law, computer technology experts and forensic accountants under the leadership of a non-police directorate (Serious Fraud Office, 2021). The SFO, although having some investigatory powers, does not have powers of arrest and needs to rely on police agencies for this (see Pennington Manches Cooper, 2019). In this model non-police agencies -including welfare agencies -can avoid becoming themselves forms of police if they find themselves conducting investigations.
It might be argued that the Controller would simply become an all-powerful legal figure and produce another version of Loader's 'ambient policing'. The democratic nature of the Controller's office would be the main antidote to such tendencies. The office would be best, in our opinion, as an appointment by the mayors of major cities and with strong input from representatives of community groups including trade unions, women's groups and minority ethnic organisations, for example. This would be an alternative to the current arrangements for the involvement of mayors in the appointment of Chief Constables and the Met Commissioner. The importance of the latter officials would be displaced by that of the Controller who would be responsible for detailed direction of police and a spectrum of other agencies as part of an attempt to remove the centrality of police in the definition and control of crime and the criminalisation of social problems. The police would also be accountable to the Controller, in the first instance, for any illegalities, for example, racism, sexism, or excessive use of force.
The reform project outlined here could take its inspiration from the concept of 'minimal policing' articulated by left realist criminologists during the 1980s as: a style of policing that is not only precluded from using, but has no generalised need for, the extended use of surveillance, criminal intelligence, special units, powers of stop and search or the extended detention of witnesses and evidence; nor need it assimilate the work of other agencies to the police task . . . The strategy rests on a series of interrelated propositions: namely, maximum public initiation of police action and maximum access to the police, and minimal police-initiated action and minimum use of coercion. (Kinsey, Lea & Young, 1986, pp.188-9) The main aim of minimal policing was to refocus police as a service devoted to the detection of crime and away from general surveillance and questionable measures such as stop and search. However, there is one theme that needs an increased stress in response to the 'defund the police' perspective. It is important in this context to avoid turning social and community workers or other crisis intervention workers into police officers by, for example, giving them detention powers. To the extent that some form of non-negotiable force will be necessary as a last resort -hopefully on a very much reduced scale -it is important to restrict it to a clearly identified institution which can be held accountable for its use and for which use key guidelines can be democratically established. Indeed, it would seriously impede the effective functioning of welfare agencies and their relations with communities if they acquired such powers. This is the main argument against complete abolition of the police. What requires abolition is a particular type of policing of communities.
Once this has been established the 'crisis of generic policing' can yield both shifting resources and response to problems of harm away from the police to welfare agencies and for those problems that do require a police response, to a range of specialist agencies -differentiated basically in terms of detective and information gathering skills, both police and non-police -responding to the growing complexity and differentiation of crime. Serial homicide, terrorism, trafficking and organised crime, and corporate fraud all require mixed groups of police and non-police expertise.
The consequence is a reconstitution of the police in accordance with the principle of 'secondary response' involving the surrendering by police of the power to initiate actions such as suspicionbased stop and search and the assumption of a variety of support roles to welfare and other specialist agencies, and in certain cases as lead agency as dictated by the Controller.

CONCLUSION
In this article we have sought to consider 'defunding' for a UK contact. We have offered some critique of its over simplification of the problems of dealing with those crimes and harms not amenable to a straightforward response either by welfare agencies or communities themselves. Abolition is a long-term project, but defunding draws our attention to the kinds of changes that can happen now. Our contribution is to reimagine the role of police in the medium term. We argue for restricting police autonomy. First, this means the cessation of suspicion-led stop and search, an archaic hangover from colonial policing and the general control of the working class, now focused disproportionately on Black people and accompanied by other forms of racism. It also means resisting the continued expansion of policing into schools, and the expansion of legal powers currently proposed in the Police, Crime and Sentencing Bill 2021. Second, it means surrendering the power to decide which issues are to be criminalised and which handed over to other agencies to deal with in accordance with principles other than those of criminal justice.
But we have stopped short of police abolition, believing that this would simply force other agencies to become substitute police forces as they found themselves undertaking de facto criminal investigations. Another unwelcome side effect would be the expansion of private security and policing. Although many social harms and problems can be dealt with through social and community measures, there will remain a relatively small number of residual incidents of coercion and violence which require a police response even if as backup to, or as elements of, non-police engagement. We note that the public have come to rely on the police as agency of first and last resort and are understandably unnerved at the thought of abolishing them completely. Our proposal radically restricts police autonomy while acknowledging their social value as an emergency service.
Finally, we have argued that such issues of coercion and violence and the manner of police involvement in them and police relations with other welfare or specialist agencies should not be decided by the police but by a higher authority, analogous to a public prosecutor but with a brief to embrace a wide spectrum of possible interventions in social problems and to take the decisions in a particular case as to what role police and other agencies will perform, underpinned by a pragmatic aim of violence reduction and conflict resolution.
We are sympathetic to the abolitionist notion of 'non-reformist reforms', that is to say changes that 'reduce the power of an oppressive system while illuminating the system's inability to solve the crises it creates' (Berger, Kaba & Stern, 2017). We are concerned not with reforming the police as presently constituted, but with reinventing their role and scope, from the default responder with the power to 'define the situation' to one kind of response among many. As we have argued above, merely defunding does not tackle core issues of police autonomy.
The importance of such reforms is that, despite critical noises from conventional police thinking discussed above, the coming period is likely to see intensified pressures towards firming up the older policing paradigm of the repression of the 'dangerous classes', now in the form of a potentially rebellious precariat -much of it non-white -in a post-Covid world of poverty, inequality and unemployment. We can expect proliferation of surveillant technologies reducing the importance of local communities as sources of information in favour of smart CCTV and similar systems. Police will be enabled to define and detect a variety of risks and crimes and even activate automatic penalties while minimising interaction with the public (see O'Malley, 2014). Meanwhile the consolidation of a paramilitary element, oriented to public order and reinvigorating histories of colonial and labour repression is likely. As former Labour prime minister, Gordon Brown, said in November 2020: 'I see a country very fragmented, very polarised and very divided -more divided than I have ever seen it before' (Elliot, 2020). Among the measures to prevent this is required a very clear commitment to exactly what the role of police should be.