Democratic Oversight and Political Direction of Chief Police Officers in England and Wales: Implications for Police Legitimacy

This article uses data from recent interviews with chief police officers in England and Wales to assess the connections between how chief police officers are overseen and given political direction and police legitimacy. The research found that governance changes started in 2011, particularly the election of police and crime commissioners, led to chief officers feeling more anxious, and the reforms reduced their operational independence. This may result in chief officers being less able to resist demands to encroach on civil liberties, prioritize the needs of the powerful over the marginalized, or to challenge policies that are likely to be ineffective or which neglect many peoples’ priorities. This endangers effective police leadership and legitimacy. Suggestions are made for consideration by policy makers, practitioners, and researchers to enhance police governance, leadership, and legitimacy. It is contended that the findings have implications for governance, leadership, and legitimacy in other countries and sectors.


Introduction
This article uses empirical research (Shannon, 2018) to consider the implications for police legitimacy of how chief police officers are held to account and given political direction (their governance) and the effect this has on them. It argues that the mechanism and manner of police governance is important because it influences what chief officers do, and how they do it. In turn, this article contends, chief officers affect what more junior officers do and how they do it, and what police officers do and how they do it largely determines police effectiveness and legitimacy (Loader, 2020). Although, how much influence chief officers have over their staff is unclear, as junior officers have considerable discretion about how they act (Cockcroft, 2013, p. 20).
This article contributes to debates about police legitimacy by using chief police officers' accounts of the effect of structures and processes of governance on them to examine the implications for legitimacy arising from recent reforms to police governance. Specifically, attention is paid to the impacts of reformed police governance on their attitudes and experiences.
These accounts suggest that governance changes introduced by the 'Police Reform and Social Responsibility Act 2011' (PRSRA), particularly the advent of Police and Crime Commissioners (PCCs), increased chief officers' anxieties, due to the threats they perceived the reforms posed to their job security, career progression, and operational autonomy. These anxieties may undermine their confidence and make them hesitant to challenge priorities set for them that encroach on civil liberties, or that fail to take account of widely held expectations of police, or that result in less effective policing. These concerns may also cause chief officers to feel fettered in exercising power to protect marginalized people. Altogether, it is argued, this endangers police legitimacy.
The article starts by discussing legitimacy, its construction, and chief officers' contribution to it, before assessing contemporary police governance. The challenges governance poses for chief officers and legitimacy are outlined, focusing on the power that PCCs hold over chief constables. Then the scrutiny of chief officers applied by other institutions in the policing landscape is considered. Next, the methods used for the research are addressed and chief officers' narratives about how they are governed are explored and their concerns identified. Finally, the implications of governance reforms for legitimacy are discussed and suggestions made concerning governance and legitimacy for consideration by policy makers, practitioners, and researchers. Although this study was conducted in England and Wales, it is contended that the findings have implications for governance and legitimacy in other countries.

Legitimacy
Policing is continually constructed and constrained by a quest for legitimacy (Bowling et al., 2019, p. 65-98) and this research is framed by a selective consideration of legitimacy literature. Beetham (1991) proposed that legitimacy is a right to exercise power that is broadly accepted and underpinned by rules rooted in widely shared values and is claimed by all governments and their institutions. These rules provide parameters for the use of power and safeguards to minimize its abuse (Beetham, 1991;Weber, 1922-23, p. 294-296). Procedural justice theorists argue that trust, founded on fair and respectful processes, builds legitimacy and compliance (Hough and Maffei, 2013). Legitimacy also requires distributive justice and policing that supports marginalized people, otherwise policing reproduces inequalities (Reiner, 2007, p. 153). Police effectiveness also enhances legitimacy (Goldsmith, 2005), as maintaining order and protecting people justifies the exercise of power (Hobbes, 1968(Hobbes, [1651). However, legitimacy is contested and viewed by some as a concept constructed to sustain the status quo (Habermas, 1976).
Leaders' conduct is crucial in constructing legitimacy (Dyzenhaus, 2001) and chief police officers have multiple audiences to persuade when doing so (Peck and Dickinson, 2010). These include their various publics, overseers, and junior police; the latter need to be convinced to act in ways that foster legitimacy (Hoggett et al., 2019). Overall chief officers can contribute to legitimacy by acting and leading lawfully, fairly, and effectively and by protecting people, especially excluded groups in society (Loader, 2020, p. 6). Not doing so, particularly when serving partisan interests, undermines the normative foundations of legitimacy, as it does not reflect widely shared values about fairness, liberty, and protection (Beetham, 1991). But chief officers' contribution to legitimacy is contingent on their governance.

Contemporary police governance
This article addresses how chief police officers are overseen and directed now, but context is required to explain why governance was changed by the PRSRA. After the 'Police Act 1964' an accrual of legislation, government guidance and performance regimes eroded the influence of police authorities. Notable among these changes was the Conservative Government's 'Police and Magistrates Court Act, 1994'. This Act strengthened the Home Secretary's position in the tripartite governance arrangements by setting national objectives, weakened local government influence by introducing lay members to police authorities (McLaughlin, 2007, p. 183-184) and increased the autonomy of chief constables (Loveday, 2018, p. 29).
The operational independence of chief constables, intended to promote legitimacy by protecting police from 'improper political interference' (Policing Protocol Order, Home Office, 2011, p. 12), was critiqued as preventing proper scrutiny (Jefferson and Grimshaw, 1984;Simey, 1988;Patten, 1999;Savage et al., 2000). Concerns were raised that police authorities failed to hold chief officers to account during the Miners' Strike of 1984-85, and that the Government was increasingly controlling the police (Loader and Mulcahy, 2001, p. 258-259). The failure of the South Yorkshire Police Authority to effectively scrutinize the 'exercise in misinformation and disinformation which was initiated and overseen by the chief constable', following the Hillsborough disaster in 1989 increased demands to reform police governance (Hillsborough Independent Panel, 2012;Loveday, 2018, p. 32). The centralizing tendency of national government continued under New Labour (Golding and Savage, 2008) reducing the influence of local governance and the autonomy of chief officers, through mechanisms such as the Police Performance Assessment Framework (Caless, 2011, p. 148).
Concerns about the ineffectiveness of police authorities in controlling chief constables, coupled with the Conservative Party's commitment to localism, led to the Coalition Government introducing the PRSRA, which shifted power towards local governance (Lister, 2014, p. 23). Yet, the Home Secretary retains influence, particularly by steering institutions of national governance (Jones and Lister, 2019) and by intervening in forces perceived to be under-performing. 1 Foremost among the PRSRA reforms was the election of PCCs in 2012. PCCs hold chief constables to account, set their budgets, renew, or end chiefs' contracts and set their priorities through local police and crime plans. Crucially, Section 38 gives PCCs the power to recruit and remove chief constables. These reforms deliberately weakened the autonomy of chief constables and were envisaged, by the Government, as democratizing policing (Cooper, 2020). However, by diminishing the operational independence of chief officers, these arrangements risk partisan influence being exerted over them, which could damage legitimacy.
Limited checks on PCCs' powers are provided by Police and Crime Panels (PCPs), consisting of local councillors and some independent members. Lister (2014), Bailey (2015Bailey ( , 2017, and Loveday (2018) argued that the scrutiny of PCCs provided by PCPs is ineffective, as they lack adequate powers and resources and, ironically, this creates an accountability deficit for PCCs. PCPs' powers include seeking advice from Her Majesty's Inspectorate of Constabulary and Fire and Rescue Services (HMICFRS 2 ) if a PCC decides to remove a chief constable. In the one case where such advice was given Her Majesty's Chief Inspector of Constabulary (HMCIC), Sir Tom Winsor, described the removal of David Crompton, the Chief Constable (CC) of South Yorkshire, as 'so unreasonable that I cannot understand how the PCC has reached this view' (BBC, 2016). This did not stop the removal, illustrating the limited capacity of HMICFRS to moderate PCCs' power over chief constables. Although the consequent judgement may increase the influence of HMCIC's advice to PCCs in relation to the removal of chief constables in the future (R v Police and Crime Commissioner for South Yorkshire, 2017, para 33 3 ).
HMICFRS is legally independent of Government and reports on the efficiency and effectiveness of police in England and Wales, but its budget is determined by the Government and its inspection programme approved by the Home Secretary (HMICFRS, 2020a). The influence of the Inspectorate over chief officer appointments diminished following the PRSRA. Her Majesty's Inspectors (HMIs) no longer advise the Home Secretary about shortlists for chief officer posts, or guide PCCs about the merits of potential chief constables, or mentor chief officers (HMICFRS, 2019a). Additionally, HMIs are, with one exception, no longer former chief constables. HMICFRS has no power to inspect PCCs, arguably an anomaly, as they set police priorities. The need for oversight of PCCs may be greater than for police authorities (which were inspected), as power is vested in an individual, not a group who can challenge each other. And the quality and experience of some PCCs has been called into question (Bailey, 2015(Bailey, , 2017Loveday 2018, p. 30).
Chief constables can resist PCCs' power through judicial review. R v Police and Crime Commissioner for South Yorkshire, 2017, quashed the decisions made by the PCC in relation to CC Crompton's removal. Muir (2017) and Loveday (2018) identified that this case is important. The judgement illustrates the tensions in balancing democratic oversight of chief constables and their operational autonomy. It may diminish the operational independence of chief constables and extend the powers of PCCs over them, as it obliged PCCs to hold chief constables to account for all aspects of their performance (para 76); this may be the most significant feature of this case for police governance (Loveday, 2018). Yet, a PCC should also work with the CC 'to maintain that independence' (para 90). The judgement narrows the grounds for removing chief constables to establishing whether their actions are unreasonable (para 94); although this does not offer the protection provided before the PRSRA by Section 29 of the Police Act 1964, which only permitted removal in the interests of police efficiency (Cooper, 2020). The judgement emphasizes that trust and goodwill between chief constables and PCCs supports good governance (para 73).
Judgements on PCCs' use of power might be provided by elections, but these have seen low turnouts (Lister and Rowe, 2015), possibly casting doubt on PCCs' legitimacy, due to inadequate 'evidence of consent' (Beetham, 1991, p. 16).
Governance introduced by the PRSRA added to anxieties amongst chief officers, shortened tenures, and reduced applications for chief officer posts (National Police Chiefs' Council [NPCC] 2018; HMICFRS, 2019a 4 ). The power of PCCs, and elected mayors performing the role in some cities (hereafter PCC indicates both), has contentiously ended the careers of some high-profile chief officers 5 (Reiner, 2016;Cooper, 2020). The power placed in one person may also induce instability in police leadership, leading to less effective policing (Cooper, 2020). The controls on PCCs' power arguably do not satisfy Beetham's (1991) and Weber's (1922-23, p. 294-296) requirements for 3 R (David Crompton) v Police and Crime Commissioner for South Yorkshire and Ors, (2017) EWHC 1349 (Admin). 4 The author declares an interest as leader of the team that conducted this inspection. 5 Examples include the retirement of Commissioner Sir Ian (now Lord) Blair in 2008, following pressure from Boris Johnson, then Mayor of London (Caless, 2011, p. 237), and the lack of confidence shown in Chief Constable Sir Norman Bettison by PCC candidates following the Hillsborough Inquiry, leading to his effectively enforced retirement in 2012 (Bettison, 2016;Conn, 2018). effective safeguards to prevent abuse of power and protect legitimacy.
Other governance changes have occurred. Chief officers have received greater scrutiny from HMICFRS ( , 2019a. Sir Tom Winsor is the first Chief Inspector of Constabulary not to have been a CC and he was previously a prominent critic of police pay and conditions of service (Winsor, 2011). The judgements of some inspections have created tensions between HMICFRS, chief constables and PCCs; contentiously in Bedfordshire, where the Chief and PCC complained about the unfairness of the assessment (Hutber, 2017;HMIC, 2016).
The Independent Office for Police Conduct (IOPC, before 2018 the Independent Police Complaints Commission [IPCC] 6 ) has also increasingly sought to hold chief officers to account (Caless, 2011, p. 237). The IOPC is independent of police and Government, although its budget is determined by the Home Secretary; it oversees the police complaints system, investigates serious complaints, and has doubled in size since 2013 (IOPC, 2020). The quality and timeliness of IOPC investigations has been criticized by PCCs, senior police officers and HMICFRS, causing the Home Affairs Select Committee (2020) to investigate its role.
The College of Policing-which aims to prevent crime, protect the public and secure public trust by developing knowledge about what works in policing, educating police and setting standards (College of Policing, 2020)-commissioned research into chief officers' conduct (Hales et al., 2015) and took over some functions from chief officers, including setting national operational guidelines. Mike Cunningham, Chief Executive of the College, has identified that changes need to be made to make it more relevant to police at all levels .
This increased scrutiny by governing institutions was apparently prompted by concerns, held by some in governance, about the conduct of chief officers. 7 Chief officers' accounts of this scrutiny are now considered.

Chief police officers' accounts of governance
The empirical data used in this article were gathered in semi-structured qualitative interviews conducted in 2016 (Shannon, 2018) with nine chief constables, three deputy chief constables, and four assistant chief constables, including the equivalent ranks in London (the rank designations applied outside London are used for all). The officers served in metropolitan, rural, and mixed forces. In two forces, two officers of different ranks were interviewed; in London, three officers of different ranks participated. Four interviewees were females. Chief constables are overrepresented but otherwise the sample broadly reflects the demographic profile of chief officers across England and Wales. These interviews are used to examine key institutions in the policing landscape.
The researcher's approach unavoidably influenced responses, and analysis of institutions involved in police governance is inevitably selective, as networks of governance are complex (Crawford, 2006), and systems for formulating 6 The Independent Police Complaints Commission became the Independent Office for Police Conduct in 2018, with a slightly extended remit and more resources. For simplicity IOPC is used in this article, other than for quotations or citation of reports. 7 Examples of worrying conduct that may have provoked these concerns include the resignations in London of two assistant commissioners, in 2007 and 2011, linked to their relationships with the press and poor judgement and leadership (Colbran, 2017) (Holdaway, 2017;Jones and Lister, 2019). Analysis involved immersion in the data, including recordings, transcripts, and handwritten notes, followed by coding using QSR NVivo (a qualitative data analysis tool) and subsequent refinement of coding, analysis, and interpretation.
PCCs and others in governance positions are likely to have different perspectives but it was judged that they were not in scope for this study, given the volume of data to be gathered and the need to interview a range of chief officers. Hence, the research is limited by its reliance on their narratives. However, other researchers (Reiner, 1991;Wall, 1998) have demonstrated the value of critically considering police governance through chief officers' accounts.
The researcher is a former chief police officer and inevitably knew many of the interviewees. The risk of subjectivity was controlled by supportive but challenging relationships with academic colleagues and reflexivity. Overall, the benefits of being a former insider outweighed the disadvantages, notably in gaining access and building rapport.
The research found that chief officers held a pattern of apprehensions about their governance by institutions within the policing landscape. However, they stressed the importance of governance in providing accountability and parameters for the use of power. Examination of their apprehensions starts with PCCs and the Home Office, before turning to HMICFRS, the IOPC, and College of Policing.
The impact of PCCs on participants varied. A few claimed to have good working relationships with PCCs, involving considered challenge and understandings of their respective roles. Others said PCCs were unreasonably intrusive and tried to direct how policing priorities should be pursued. All contended that PCCs' influence on them was greater than the Home Secretary's, largely because of PCCs' power to recruit and remove chief constables. A CC commented: All suggested that an unhelpful aspect of the relationship between CC and PCC was that it was not mediated by different perspectives and because power rested in one person: There was actually better governance and scrutiny under the Police Authority than what there is under the PCC . . . you had a balanced challenge because you had a balanced group of people representing different areas, different parties. (CC) PCCs were seen by some participants as helping justify the exercise of power, through a democratic mandate and as an intermediary between police and their publics, and contended that this democratic directive constrained and guided them: It would be totally wrong if the [PCC's] Police and Crime Plan then had absolutely no impact whatever on operational activity . . . that would be . . . the chief constable obstructing the local democratic will (CC).
But four raised concerns about the mandate, due to low turnout at PCC elections, and most suggested that engagement by PCCs with their publics, particularly the marginalized, was inadequate. These reservations echo Loader's (2002, p. 137) warning that democratization of policing threatens civil liberties and minorities, who may be more policed than protected: [The PCC] is there as a proxy for the public but only about [XX] per cent bothered to turn out and vote (CC). I don't think [the PCC's] consultation tends to get near that [engagement with marginalized people] . . . we tend to consult with the great and the good (CC).
Increased localism was an intended outcome of the PRSRA, and PCCs have displaced the Home Office as the institution of governance that chief officers pay most attention to and are causing them more concern than police authorities had. Reduced Home Office influence was reflected in this comment: The big player is your Police and Crime Commissioner, but interesting isn't it, because . . . Home Office Circulars [government guidance] were like the Bible (CC).
However, the Home Office continued to exert influence over participants, though pressure was often distanced and applied through HMICFRS. This raised their anxieties about the potential for the Home Office to exercise power over them, while evading responsibility and blaming others, by claiming that accountability for policing failures lay with PCCs and chief constables, echoing a point made by Jones and Lister (2019) Office] ought to commit to that position and not just say it is down to chief constables or PCCs -who it will be easy to blame if things don't go well (CC).
Participants reported that HMICFRS influenced them, half claimed it was used by the Home Secretary to exert pressure on them and 12 resented significant aspects of Inspectorate oversight. The connecting theme was that HMICFRS made them anxious and added little to the quality of policing. Concerns about its independence and resources were evident: I don't see them (HMIC) as independent; I see them as being driven very much by the Home Secretary (ACC).
[The Home Secretary exerts influence] partly through the Inspectorate . . . [they] have got more resources now than they ever had so do more inspections, spend more time with us . . . it is not uncommon for-for there to be a missive arrive from the Home Secretary following an inspection (CC).
[HMIC] are meant to be the fierce advocates of the public. I don't really see any connection between what they do and the public (CC).
The relationship between HMICFRS and chief officers appeared to have deteriorated since 2011 and this added to their anxieties about governance. The relationship with the IOPC seemed worse, participants' common view being that the IOPC was ineffective, and its motivations were questioned: The quality of their investigations are poor, the quality of their investigators are poor, I-I don't get it at all (CC).
[The IOPC are] making announcements . . . which are more about establishing their credentials than actually about, is this the best way to Democratic oversight and political direction of chief police officers Article Policing be dealing with the circumstance? (CC) The IOPC increased chief officers' worries about the fairness and competence of their oversight. They reported that the College of Policing caused less anxiety, but its role seemed peripheral to 14 of the 16 participants, and its relevance was questioned: [The College] do their Authorized Professional Practice . . . they're endorsing the national guidance that used to be endorsed by Chief Constables' Council. They're not central to my thinking at all (ACC).
The College of Policing seemed to have little influence over chief officers, but its impact on their anxieties was limited. The College was a product of the 2011 reforms, which also modified the application of the concept of operational independence.

Operational independence after the PRSRA
Operational independence and the benefits it may provide in building legitimacy by protecting police from improper political interference (Royal Commission on the Police, 1962) remains, as discussed earlier, contested (Jefferson and Grimshaw, 1984;Simey, 1988;Patten, 1999;Savage et al., 2000). The 'Policing Protocol Order' (Home Office, 2011) required by the PRSRA, guides relationships between chief constables, PCCs, PCPs, and the Home Office. It envisages operational independence as PCCs setting the 'what' and chief constables deciding 'how' (Lister, 2013), and describes operational independence as context driven. However, it is not clearly defined. This was evident in participants' accounts of operational independence as indistinct, and diminished since 2011: PCCs' power to recruit or remove chief constables caused participants' greatest anxieties and appeared to impinge on operational independence and have a chilling effect on the will of chief constables, and aspiring deputy and assistant chiefs, to challenge PCCs, echoing concerns voiced by Cooper (2020) and the Independent Police Commission (2013): it's really hard saying to the person with the headline 'I'm going to hire and fire you': 'Back off, I'm not going to do it'. . . . you get all of these election manifestos-'I'm going to put 100 extra police officers on the neighbourhood streets.' . . . that is an operational decision. But quite a brave chief to push back (CC).
Chief officers were anxious about the diminution of their operational autonomy, arising from the roles and responsibilities of PCCs but also from interventions by other governing institutions in the policing landscape and the threat this posed to their career prospects. This risks them lacking the confidence to challenge those in governance, who might steer them towards directing policing that neglects civil liberties; a risk posed by the democratization of policing (Loader, 2002). It may also discourage them from resisting policies and operational mechanisms that are not evidencebased, which could undermine effectiveness and legitimacy (Goldsmith, 2005).
Illiberal direction and disregard of evidencebased policy by PCCs and Government is, arguably, rare. However, some mavericks have become PCCs (Bailey, 2015(Bailey, , 2017. The first PCC for Surrey used Facebook to say of a criminal 'I wanted to break his legs' (BBC, 2016a) and his emphasis on intolerant enforcement (Hurley, 2012) does not seem to be evidence-based. The current PCC called for unauthorized traveller encampments to be criminalized (Office of the Police and Crime Commissioner in Surrey, 2020), a policy that would probably be detrimental to minorities and distributive justice. The Home Secretary's demands for 'a clampdown on Black Lives Matter protests' (Hamilton, 2020) also, debatably, illustrates the risk of delegitimating political interventions in policing. These examples arguably illustrate the need for operational independence, balanced by democratic oversight, to protect liberties, and effective and legitimate policing.

Chief officers' concerns about career security and progression
Participants claimed PCCs pose the greatest risk to their jobs and professional prospects, but other institutions were also said to raise anxieties. Increased HMICFRS and IOPC activity were repeatedly raised, and there was nostalgia for a Home Office that participants perceived had engaged meaningfully with them, reminiscent of Loader's (2006) account of a collaborative criminal justice elite: HMIC has doubled. IPCC has doubled. Home Office has become more distant. And then crime commissioners are now political. So, all of that has created added tensions . . .I've had more complaints about me in the last year than I've had in [the rest of the participant's career] . . . it takes its toll (CC).
A common theme was the intensity of scrutiny and the threat this posed to their job security and career progression, leading to risk aversion: One of my huge issues and concerns in policing . . . is the "heads must roll" mentality . . . It stifles people from being allowed to take proportionate risk and creates a sense of fear (CC).
Participants attributed many of these anxieties to PCCs' power to recruit and remove chief constables and to influence the career progression of all chief officers. They inferred that consequently their impartiality might be compromised. 8 The potential that chiefs become wedded to the commissioners and you get a more, almost Americanstyle system where a mayor is booted out of office and then a chief goes as well . . . that starts to politicize the police by osmosis (CC).
Any chief constable that stands up to a PCC and says: "wind your neck in, this is operational policing" is immediately vulnerable because they can be hired and fired by that PCC (ACC).
These anxieties were partly based on awareness of colleagues leaving policing prematurely. The average term for a CC has fallen by 1.2 years since the mid-1980s (NPCC, 2018. HMICFRS (2019a, p. 60) described the trend, which was accelerated by the PRSRA, as 'marked and worrying'.
Deputy and assistant chiefs do not face the direct threat of removal from office by PCCs, but they worried about PCCs' influence on their careers. They no longer have the potential protection from chief constables provided by police authorities (which appointed and removed them; this now rests with chief constables), and their career prospects can be contingent on the view of the CC and PCC. Deputies seeking to be chiefs may be particularly exposed: possibly the most vulnerable beast in policing today is the deputy chief constable aspiring to be the chief (CC). you aren't going to particularly challenge somebody when they are going to select you are you (DCC).
I've always felt that deputies and ACCs were more vulnerable [than CCs], knowing some of the chiefs I've met . . . if a chief is under threat what happens next? (CC) Chief officers are no longer supported and mentored by HMIs. Institutional and professional distance between inspector and inspected may have benefits; nonetheless, this remoteness increased many participants' sense of isolation: HMIC was, I think, very much supportive and [had an] advisory role; and to become adversarial-I'm not 100% sure has been, actually, such a good move, because quite often forces who might be underperforming have suddenly felt very, very vulnerable . . . it becomes a really, really dark place for some people (ACC).
Some participants identified the IOPC as a source of anxiety and consequent risk aversion: something like 20 per cent of all chief officers . . . are subject to some sort of enquiry by the IPCC . . . it makes some people very risk averse-can paralyse organizations or individuals (DCC).
Many participants suggested these anxieties made them less likely to use power, particularly to challenge their overseers, or in circumstances which might provoke scrutiny. They felt vulnerable due to their perceptions of threats to their careers posed by unreasonable scrutiny from the IOPC, HMICFRS, and PCCs, by the apparent withdrawal of Inspectorate support, and the risk of being blamed by the Home Office. Their foremost worry was what they saw as the danger to their careers posed by PCCs. These apprehensions could be debilitating: When it comes to power, there's something about protecting that office [of chief constable]. The less often that office wields power, the safer it is because every time you wield your power, you make yourself vulnerable to challenge (CC).
These quotations do not adequately capture the angst that emerged, in the tone and context of the interviews but they reveal some of this anguish: I had a tortuous first year . . . it certainly feels a lonely place. . . . the performance indicators don't look good and people are getting worried and the perfect storm because everyone is on your case, there is a real temptation to doubt your own personal judgement (CC).
[I have] been a chief since XXXX. It feels like 400 years . . . you're making different-different, kind of, calls on the power you have got because of some of that pressure to be seen to be accountable . . . I'd be lying if I said it doesn't affect you (CC).
Chief officers reported that governance changes introduced in 2011 led to disproportionate scrutiny, which exacerbated their anxieties. The implications of this for police legitimacy are discussed next.

Discussion
Chief officers' power does not stop them feeling anxious and they can be exposed to pressures that those who are less visible do not experience. 9 Vulnerability is particularly disempowering if the 'individual does not self-identify as vulnerable' (Dehaghani, 2019, p. 40) and chief officers may be reluctant to do so. Their anxieties seemed to increase considerably following the 2011 reforms, due to concerns about their governance, operational autonomy, career prospects, and job security. Worryingly their goodwill towards and trust of PCCs, that should support good governance (R v Police and Crime Commissioner for South Yorkshire, 2017, para. 73) appeared fragile. Chief officers' anxieties can diminish their confidence to use power to protect people and to question priorities that may not reflect wider societal values. This may impede them in providing procedurally fair (Hough and Maffei, 2013) and effective policing (Goldsmith, 2005) and distributive justice that protects excluded people (Reiner, 2007). This potentially leads to withdrawals of consent to police use of power, particularly by marginalized people, and to a legitimacy crisis (Habermas, 1976;Beetham, 1991).
These threats to legitimacy may seem remote but the potential for authoritarian and partisan agendas being imposed on police was a concern of the Royal Commission on the Police (1962) and informed the governance and operational independence that followed, until reformed in 2011. Chief officers debatably have a normative duty to resist such agendas. Authoritarian threats to legitimacy may be rare in England and Wales but they remain, and are arguably illuminated by sometimes overzealous, inconsistent, and unlawful use of new police powers during the Coronavirus 19 pandemic (Smith, 2020a;The Times, 2020;Karim, 2020). This may exacerbate inequalities and threaten liberties (Parris, 2020), and lead to a disjunction between the values of the powerful and their subordinates, which endangers legitimacy (Beetham, 1991).
Chief officers' anxieties may also make them reluctant to challenge PCCs when they set priorities which fail to consider the expectations of many people or are likely to be ineffective. Examples include prioritizing protecting the vulnerable, which seem laudable, but which, without adequate funding, displaces activities such as dealing with acquisitive crime and local policing (Pratt, 2017;Higgins, 2018). HMICFRS (2020) has identified that many users of police services are experiencing increasingly ineffective policing as these activities decline, and ineffective policing diminishes legitimacy (Goldsmith, 2005).
A balance needs to be struck between democratic oversight of chief police officers and their operational independence. And individuals and institutions responsible for directing and overseeing chief officers also need to be held to account. This article has not explored the accountability of national Government, but oversight of, and checks on, PCCs' powers over chief officers appear deficient.

Conclusion
This article has shown how recently introduced governance arrangements heightened chief police officers' anxieties about their job security, career prospects, and operational autonomy. This may sap their confidence, leading to them failing to resist political directions that undermine civil liberties or prioritize the needs of the powerful over those of the marginalized. They may also fail to challenge policies that lead to ineffective policing or neglect the priorities of many people. This endangers legitimacy. Nonetheless, checks on chief officers' power and performance remain important in ensuring that police are effective and legitimate (Jefferson and Grimshaw, 1984). Therefore, the following proposals are made to improve police governance, increase chief officers' confidence and effectiveness, and consequently enhance police legitimacy.
Firstly, the relationship between PCCs and chief constables seems unduly weighted towards PCCs (Cooper, 2020). This might be ameliorated by making PCCs' more accountable, by strengthening the roles and responsibilities of HMICFRS and PCPs in relation to PCCs' governance of chief officers. Arguably the roles of PCPs should be clarified, and their powers and resources extended (Lister, 2014;Bailey, 2015Bailey, , 2017Loveday, 2018), and the remit of HMICFRS expanded to include inspecting PCCs.
Secondly, deputy and assistant chiefs' concerns about their oversight and direction by institutions in the policing landscape include the impact that chief constables and PCCs have on their career trajectories. This issue needs further consideration by researchers, practitioners, and policy makers.
Finally, decisions about policing priorities are inevitably political (Loader, 2020, p. 8), but chief officers need protection against improper partisan influence. Consequently, the indistinct concept of operational independence and its application should be revisited by researchers, practitioners, and policy makers.
The preceding discussions have explored the implications of oversight and political direction of chief police officers for police legitimacy. Consideration of the findings would inform deliberations about policies and practices intended to improve police governance and legitimacy. The recently announced review of PCCs (Home Office, 2020) may provide an opportunity for such consideration.