Doing the right thing? An institutional perspective on responsible restructuring in UK local government

Correspondence Mathew Johnson, Alliance Manchester Business School, The University of Manchester, Booth Street West, Manchester M15 9PB, UK. Email: mathew.johnson@manchester.ac.uk Abstract Drawing on institutional theory, this paper advances understanding of how restructuring practices are shaped by the organisational and institutional context and the concerns for legitimacy. Longitudinal case study findings from UK local government suggest a pattern of incremental deviance where repeated cutbacks are legitimised through a narrative of continuous improvement. This results in increased labour flexibility, work intensity and managerialism within services. We argue that the lack of resistance to change can be explained by the steady desensitisation of workers to repeated cutbacks as opposed to the inherent fairness of management actions. Through context‐specific research the theoretical and empirical distinctions between responsible and pure restructuring, and soft and hard Human Resource Management policies, are challenged.


| INTRODUCTION
Two global recessions in just over a decade have placed downsizing and restructuring at the centre of debates about the future of Human Resource Management (HRM) (e.g., Cook et al., 2016;Goyer et al., 2016;Harney et al., 2018;ILO, 2020). One line of literature suggests the increasing focus on short-term organisational survival undermines mutual gains HRM predicated on long-term economic growth (e.g., Dobbins & Dundon, 2017;Thompson, 2013). These recessions have also further exposed the tendency within HRM literature to downplay the role of various contingencies such as macroeconomic policy, institutional rules and power relations in shaping changes in work organisation and labour relations (e.g., Cooke, 2018;Goyer et al., 2016).
Other studies note the remarkable stability of employment levels in many countries, and limited conflict with workers even as managers pursue cost savings through a range of mechanisms (e.g., Conway et al., 2014;Johnstone et al., 2019;Van Wanrooy et al., 2013). This continuity may be interpreted as indicating organisations having implemented socially responsible restructuring, where they successfully balance the need for cost savings against the need to maintain worker morale and cooperation (e.g., Ahlstrand, 2010;Cascio, 2002;Forde et al., 2009;Tsai & Shih, 2013). Studies show how organisations may avoid negative outcomes of restructuring by using alternatives to mass layoffs, such as reduced pay and conditions and internal redeployment, measures that are underpinned by consultation with unions and workers (Harney et al., 2018;Sahdev, 2003;Teague & Roche, 2014;Tsai & Shih, 2013).
However, the prescriptive approach of many HRM studies of restructuring often overlooks how the institutional and organisational context shapes the legitimacy of specific practices, and how the presence of multiple stakeholders creates the scope for conflict and contestation during processes of organisational change (Chaudhry & Rubery, 2019;Lewis et al., 2019). In this paper, we focus on the under-researched context of UK local government which has a historical (but not unequivocal) model employer tradition characterised by soft HRM, collective bargaining and partnership working with trade unions (Beszter et al., 2015). The challenge is that deep workforce cutbacks resulting from enforced budget cuts may disrupt and delegitimise embedded models of HRM (Bach & Stroleny, 2017;Kessler & Purcell, 1996).
To explore the contested nature of restructuring within UK local government, we apply an institutional perspective in the form of Scott's (2008) three pillars of legitimacy-regulative (rule setting), normative (legitimate means) and cultural-cognitive (common beliefs and cultural frames)-to longitudinal case study data gathered from three large local (municipal) authorities. Through a detailed and context sensitive analysis of the restructuring process over a period of 5 years, we reveal that individual local authorities have avoided significant conflict with workers and trade unions by making apparently 'soft' labour cost savings such as voluntary severance, internal redeployment and changes to the skill mix. These workforce cuts formed part of locally designed 'transformation programmes' and were generally agreed through processes of individual and collective consultation.
However, we question whether the lack of resistance and conflict is the result of the 'legitimacy and perceived fairness of management actions' (Teague & Roche, 2014, p. 179), or rather the gradual desensitisation of workers to market discipline in the form of redundancies, internal restructures and increasing work intensity which are increasingly normalised in liberal market economies (Thompson, 2013). Our findings reveals how a 'persistent state of instability' (Chaudhry & Rubery, 2019, p. 552) legitimises a mixture of negotiated and imposed changes to workplace HRM. Furthermore, by drawing on institutional theory, we show how managers attempt to legitimise insidious organisational changes that have adverse outcomes for workers through cultural-cognitive frames of 'continuous improvement'. Such symbolic reforms, whether or not they succeed, serve to legitimise the actions of public sector managers in the eyes of service users and taxpayers (Czarniawska-Joerges, 1989).
In the next section, we review literature on restructuring models before presenting our analytical frame that, based on institutional theory, emphasises the issue of legitimacy in the process of restructuring. We then discuss the research context of UK local government and its relevance for studying responsible restructuring, before 2presenting the research methods. The findings are then reported in two parts, aligned to our two primary research questions. The final section offers a discussion and implications for theory and practice.

| SOCIALLY RESPONSIBLE RESTRUCTURING: A NEW SOLUTION TO AN OLD PROBLEM?
Downsizing and restructuring have come to be accepted as legitimate business strategies to reduce costs and improve corporate performance, despite weak evidence that either outcome is likely (Cook et al., 2016;ILO, 2020;Johnstone et al., 2019;McLachlan et al., 2020;Vaughan-Whitehead, 2013). Negative results from restructuring can be explained by the loss of human capital arising from heavy workforce reductions, and deteriorating morale and commitment described as 'survivor syndrome' (Cascio, 2002;Tsai & Shih, 2013). The risk is that when faced with short-term cost pressures, many organisations focus on organisational survival at the expense of worker wellbeing and cooperative workplace relations.
These negative consequences have led to the search for a strategic approach to restructuring that takes a longer term view of organisational needs, and systematically includes stakeholders such as trade unions and workers in the process. One specific strategy which has attracted attention is 'responsible restructuring' that differs from 'pure restructuring' in various ways (Table 1). Whereas pure restructuring aims for quick and indiscriminate headcount reductions, described by Cameron (1994) as 'throwing a grenade in a room', a responsible approach views staff as a long-term investment who deserve care and support to adjust to new organisational structures and work routines (Tsai & Shih, 2013). As a result, organisations following a responsible approach, try to find loyal employees alternative internal job opportunities and support them throughout processes of change (e.g., Ahlstrand, 2010;Forde et al., 2009).
Responsible restructuring 'bundles' consist of technical and behavioural practices. Technical practices are oriented towards labour cost savings that avoid compulsory redundancies, for example pay cuts, voluntary severance and internal job moves. Behavioural practices are designed to protect worker motivation and morale during the process of change. They include maintaining strong communication with workers and unions, and offering training and counselling for both victims and survivors (Cascio, 2002;Forde et al., 2009;McLachlan et al., 2020;Teague & Roche, 2014;Tsai & Shih, 2013). Studies suggest that firms that rely on skilled workers and the cooperation of trade unions may be more likely to adopt a responsible restructuring approach (Butler, Glover, & Tregaskis, 2011;Forde et al., 2009;Tsai & Shih, 2013). However, survey data from across firms suggest that restructuring bundles often include a mix of both pure and responsible practices (Teague & Roche, 2014;Van Wanrooy et al., 2013).

| THE INSTITUTIONAL LEGITIMACY OF RESTRUCTURING
From a management perspective, responsible restructuring shows how organisations can theoretically avoid the damaging effects associated with blunt headcount reductions. However, given the short-term financial pressures on many organisations during recession and the growing normalisation of redundancies, there remains the question why would organisations do more than is legally required? To understand why and how organisations adopt responsible restructuring practices, it is important to consider the institutional and organisational context within which they operate (Cooke, 2018;Lewis et al., 2019). This is important as restructuring practices that are considered responsible in one context may not be considered so in another (e.g., Ahlstrand, 2010;McLachlan et al., 2020).
Institutional theory suggests that the adoption of particular practices, and the avoidance of others, may be motivated by concerns for legitimacy (e.g., Powell & DiMaggio, 1991). Legitimacy is defined as 'a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially JOHNSON AND WATT -3 constructed system of norms, values, beliefs and definitions' (Suchman, 1995, p. 574). While some literature suggests that restructuring has come to be regarded as a legitimate management strategy in itself, organisations may also be keen to avoid reputational damage and maintain the cooperation of workers by following certain procedures. Scott (2008) refers to three pillars of legitimacy: Regulative, cultural-cognitive and normative which, in conjunction with one another, provide stability and meaning to social behaviour and help cement institutional durability over time. The regulative pillar concerns the need for actors to avoid state sanctions by following established procedures. For example, under EU law, employers are obliged to consult collectively where more than 20 workers are at risk of redundancy, but in countries such as France, Germany and Spain, national law also requires organisations to develop a 'social plan' with trade unions that outlines the measures to mitigate the negative consequences for affected employees (Gaudu, 2011). In the UK, the 1992 Trade Union and Labour Relations (Consolidation) Act 1 sets out minimum consultation periods where there is a risk of redundancy (known as Section 188), the offer of alternative internal employment, and the entitlement to redundancy pay (roughly 1 week per year of service). However, since 2012 workers in the UK with less than 24 months continuous service are not entitled to statutory redundancy pay, and the low level of trade union membership density outside the public sector means that collective consultations typically relate to the timing rather than substantive content of restructuring proposals (Cook et al., 2020;McLachlan et al., 2020). However, following legal rules such as these may be the least that workers and trade unions expect. A responsible restructuring approach therefore can mean honouring additional collectively agreed standards around severance pay and retraining (Ahlstrand, 2010).
The cultural-cognitive pillar reflects what actors consider contextually appropriate behaviour given their shared identity and understanding of how particular issues, such as restructures, should be dealt with. This is important as enhancing workers' perceptions of fairness in procedures and communication can enhance perceptions of fairness in outcomes (e.g., Tsai & Shih, 2013). For example, there may be implicit rules that govern the 'fairest' selection of workers for redundancies such as the principle of 'last-in first-out' (Turnbull, 1988, p. 208) or conversely inviting older workers to take early retirement to shield younger workers from job losses (McLachlan et al., 2020). A socially responsible approach also involves forewarning workers about potential cutbacks and changes to working practices, and delegating the tactful delivery of redundancy notices to appropriate personnel such as line managers (Forde et al., 2009;Tsai & Shih, 2013). Empirical studies also suggest that the negative reaction of survivors to changes in work routines can be mitigated by managers setting out a clear vision for how they will contribute to the long-term success of the organisation, over and above reactive changes to short-term cost pressures (Sahdev, 2003).  Tsai and Shih (2013).
The normative pillar reflects the obligation to uphold the shared values, norms and ethics that give purpose and social legitimacy to the organisation. This involves a degree of symbolic management whereby organisations demonstrate how their actions are congruent with existing norms and values, and satisfy the expectations of key stakeholders (Fernández-Alles & Valle-Cabrera, 2006). For example, organisations may publicly announce restructuring plans and signal their commitment to helping laid-off workers find alternative employment so as to enhance perceptions of transparency among investors, politicians and the general public (Ahlstrand, 2010;Forde et al., 2009). At the same time, one might (cynically) argue that these 'ethical' gestures are a way to enhance the Corporate Social Responsibility credentials of organisations, and to pre-empt closer scrutiny by regulators (Ahlstrand, 2010). Responsible restructuring practices such as providing training and counselling may also be intended to secure worker acquiescence rather than being motivated by a genuine concern for worker welfare (McLachlan et al., 2020). Scott's (2008) institutional pillars help explain the durability of organisational practices in the face of external shocks, and also patterns of convergence across organisations operating within similar institutional contexts, driven by concerns for legitimacy. Institutional theory suggests that organisations typically change in incremental path dependent ways as managers are cognisant of embedded norms, rules and cultures (Powell & DiMaggio, 1991). At the same time, adopting an institutional perspective also provides a framework for understanding change where embedded practices lose their legitimacy because of waning organisational consensus and changes in the wider context (e.g., Chaudhry & Rubery, 2019;Czarniawska-Joerges, 1989;Lewis et al., 2019;Sahdev, 2003).

| RESTRUCTURING IN UK LOCAL GOVERNMENT
The UK local government sector is an interesting context in which to study the implementation of restructuring bundles, and to explore how concerns for legitimacy shape the process of change. Viewed through the prism of Scott's (2008) three pillars, local authorities might be expected to pursue a responsible, rather than pure, approach to restructuring. In regulative terms, a two-tier framework of collective bargaining in local government allows employers and trade unions to negotiate nationally over core issues such as pay and conditions, supplemented by local negotiations over staffing levels and working practices (Grimshaw et al., 2017). In normative terms, local politicians and managers have significant 'sunk investments' in a model employer approach underpinned by job security, internal career progression and paternalistic HRM that are hard to unpick (Beszter et al., 2015). In cultural-cognitive terms, managers historically relied on the skill and public service ethos of workers rather than close control of the labour process, and may therefore be inclined to proceed cautiously with changes to working practices and cultures (Cunningham et al., 2020).
On the other hand, the limited restrictions on hiring and firing in the UK and the weak tradition of social plans mean that simply following legal procedures on collective redundancies may be seen as a legitimate attempt to act responsibly during restructuring. Furthermore, the two-tier framework of collective bargaining in local government provides significant lateral room for individual local authorities to set their own policies around redundancies and severance pay in line with political agendas, funding pressures and local trade union strength (Connolly, 2020;Grimshaw et al., 2017;Kessler & Purcell, 1996). More broadly, a prolonged period of austerity may call into question the legitimacy of 'soft' approaches to public sector HRM when others industries have suffered job cuts and downward pressure on pay and conditions (Bach & Stroleny, 2017).
An in-depth qualitative analysis of the process and outcomes of restructuring is required to shed more light on the complex institutional legitimacy of change at the organisational level. Our empirical research is guided by two main research questions.
1. To what extent do local authorities follow a responsible approach to restructuring as set out in Table 1 and what is the interplay between technical and behavioural practices? JOHNSON AND WATT -5 2. How do different actors (managers, HR professionals, trade unions) view the institutional legitimacy of restructuring practices?

| RESEARCH CONTEXT AND METHODS
In order to address our two core research aims, we adopted a longitudinal qualitative case study design (Yin, 2003).
We sampled three English local authorities (LAs) in the north of England that undertook significant programmes of restructuring after 2010. While this might limit the generalisability of our findings, as large Labour (left wing)controlled authorities in urban areas with high levels of trade union membership we viewed them as 'most likely' cases for taking a responsible restructuring approach (Bach & Stroleny, 2017). To protect anonymity, we label these councils 'Local Council', 'Metropolitan City' and 'North Borough'.
We conducted 45 interviews across the three LAs (see Table 2). Most interviews were with HR and departmental managers at varying levels, from front-line managers to heads of service. Twelve interviews were conducted with trade union representatives, largely from non-managerial backgrounds, making the data especially rich. Interviewing was done in three stages across a 5-year period (2014-2019). First, key actors at each LA were emailed, including the head of HR and local and regional trade union representatives. Second, using snowball sampling from these initial interviews, further interviews were conducted with departmental managers at different levels. Although snowball sampling risks subsequent interviewees reinforcing similar viewpoints, the range of interviewees across different levels of seniority and service areas helped maintain diversity. Third, follow-up interviews with HR managers and trade union representatives were conducted to ask how things had changed since the initial interviews. This final stage helped us develop a longitudinal perspective on the implementation of restructuring bundles, and the responses of trade unions and workers over time.
All interviews were semi-structured, and explored the nature of restructuring at a local level, and the perceived importance of institutional rules and standards (in regulative, cognitive-cultural and normative terms) in shaping and legitimising the restructuring process. Interviews lasted between 45 and 120 min and (with permission) were audio recorded. In terms of analysis, no coding software was used; interviews were coded manually across two stages. First, transcripts were searched at 'surface-level' to identify the practical steps that were taken at each council to restructure, and the extent to which they were combined into distinct responsible bundles. To corroborate the interview data at this stage, key documents such as strategic service plans, annual accounts and workforce reports were also analysed. The second stage of analysis sought to understand how different actors viewed and legitimised specific actions. This required more interpretation on our part, and we acknowledge the risk of researcher bias in the presentation of these results. However, the fundamentally grounded nature of our analysis, having no pre-determined expectations of what we would find limited this potential bias. For this second stage, we applied thematic analysis; mapping issues that emerged in the transcripts onto Scott's (2008) institutional pillars. This allowed us to reveal the salience of different formal and informal rules in shaping views of legitimacy or deviance in relation to specific restructuring practices, and how different actors attempted to make sense of the restructuring process both during and after the event.

| FINDINGS
The findings are presented in two parts aligned to our overarching research questions. Attention is paid to variation within and between the three case studies but, reflecting the common approaches to retrenchment and restructuring that emerged from the data, the findings from across all three councils are synthesised to develop a general picture of change. Table 3 summarises the specific technical and behavioural restructuring practices adopted within our sample and identifies common and idiosyncratic practices adopted by individual councils.

JOHNSON AND WATT
The data in Table 3 highlight three important findings. The first is that none of the three LAs appears to have adopted a strongly responsible approach to restructuring as defined by the practices in Table 1. While LAs exceeded their legal responsibilities in respect of some technical practices, such as offering enhanced severance packages and pay protection for redeployed workers, these were actually the most expedient ways of making cutbacks as it avoided the need to draw up detailed selection criteria for layoffs. Furthermore, even where councils did adopt symbolically 'softer' restructuring strategies such as voluntary redundancy (VR) and early retirement, these were often followed by deep cuts and outsourcing that may have heightened the sense of injustice among workers about the erosion of standards. For example, there were significant one-off costs attached to VR schemes (£25m at Local Council and North Borough and £75m at Metropolitan City) which were funded from cash reserves that could have been used to offset subsequent cutbacks. It was even suggested by an HR manager at Metropolitan City that the offer of enhanced severance packages and 3 years pay protection meant outsourcing became more likely in subsequent years, a point echoed by HR managers at North Borough.
The second related finding is the seeming disconnection between technical and behavioural practices. Contrary to expectations, the case studies reveal that two key behavioural practices that underpin a responsible approach, consultation and systems of worker support, became increasingly individualised and concerned with following legal procedures rather than meaningful engagement and compromise with workers and trade unions over substantive issues. In addition, cutbacks to HR functions within each LA meant that individual workers were expected to take greater responsibility for their own training, career development and personal wellbeing.
The third finding is the ad hoc nature of cutbacks which challenges the idea that LAs could be considered to have adopted coherent bundles of practices. For example, Metropolitan City and North Borough committed to no compulsory redundancies, but specifically targeted senior roles for staffing reductions and cuts to terms and conditions, which the unions argued was largely symbolic and opportunistic as it delivered only marginal cost savings. In contrast Local Council imposed 3 days of unpaid annual leave on all staff (equating to a 1.15% pay cut) in order to try and reduce the need for job losses, but at the same time made 25% of staffing reductions through compulsory means. Similarly, all councils raised pay for the lowest grades up to the living wage (10% higher than the statutory minimum wage), but this was partly offset by reductions to car mileage allowances and unsocial hours premiums which adversely affected low paid and part-time workers (e.g., care workers). Given the various tensions and trade-offs emerging between technical and behavioural practices, we now turn to an analysis of the legitimacy and legitimation of specific restructuring practices from different perspectives.

| The regulative pillar
In Scott's (2008) framework, the regulative pillar reflects the need to follow established rules and procedures. All three councils closely followed Section 188 legal procedures for consultation around collective redundancies, and attempted to move workers internally into alternative jobs rather than make mass layoffs. However, the voluntary redundancy scheme had to be re-opened several times, and the formal legal notice of possible collective redundancies became so frequent that it was seen largely as a symbolic exercise rather than an attempt to engage workers in meaningful consultation:

Forewarning of redundancies
Comments like this suggested that it was not consultation per se that helped reduce conflict, but rather how normalised the process of announcing collective redundancies became. Despite this, there was broad consensus between managers and trade unionists about the symbolic importance of formal consultation mechanisms that helped maintain the perception of legitimacy in the process of change. For example, Joint Consultative Committees (JCCs) were an expedient way for managers to discuss significant proposals with trade union officials, which in turn reduced collective resistance to change. Trade union officials also saw JCCs as an important mechanism by which to prevent changes being rushed through, and to formally hold senior managers personally accountable for decisions: …we'll make sure that the council are held accountable to do things right… [TU official, North Borough] While this process of oppositional engagement was seen as important to protect members' interests (Cook et al., 2020), there was a clear tension for the unions in that adopting a strongly confrontational approach might compromise the informal day-to-day relationships with managers, but failing to challenge managers risked alienating their own members. Indeed some union members felt JCCs only served to 'rubber stamp' decisions that had already been made by managers behind closed doors rather than offering a genuine opportunity for negotiation around alternative proposals:

How many times I've heard those words 'it's a done deal [TU official, Local Council]
The trade unions were also conscious that in view of falling union membership across the sector, direct communications were increasingly being used by HR managers to issue messages to workers either via open meetings or by email and the staff intranet, something that would have been considered highly deviant in the past.
Nevertheless, the trade unions supplemented these meetings with further discussions and briefing sessions with members where they attempted to interpret the proposals before formulating a collective response.
A clear example of increasingly shallow and individualised consultation was seen at Local Council where managers used Section 188 collective redundancy procedures to dismiss and re-engage all workers on new employment contracts that included 3 days of unpaid leave (equating to an annual pay cut for a full-time worker of 1.15%). Unilaterally imposing changes is typically considered a last resort where collective negotiations have failed, and although the trade unions offered to continue negotiating, management refused to back down. Despite the recommendation of the union that members reject the new contracts, the vast majority of staff (many of whom were not union members) agreed to sign. This suggests that formal mechanisms of consultation are increasingly being used to unilaterally impose management changes, with or without trade union endorsement.

| The cultural-cognitive pillar
The cultural-cognitive pillar reflects embedded working patterns, organisational cultures and management styles which in our study were clearly in a state of flux. For example, on one hand, the use of voluntary, rather than compulsory, redundancies reflected the long held shared understanding about how workforce reductions should be handled. On the other hand, the increasing use of individualised legal procedures to communicate and enforce change was a significant departure from historical approaches. Perhaps more important though were the insidious changes in work culture within services observed over the course of the research. From the perspective of responsible restructuring, devolving the announcement and implementation of cutbacks is meant to promote the sensitive handling of bad news, but in our case studies, it appeared that departmental managers also saw an opportunity to leverage conditions of austerity to impose a more business-like culture within their teams.
JOHNSON AND WATT -9 The first significant change in work culture was that, as a result of staffing reductions and internal labour mobility, greater individual flexibility was required. Generic job descriptions and changes to the skill mix increased the chances of individual workers being able to find alternative opportunities within the internal labour market, but it also meant increased labour substitutability as occupational boundaries and specialisms were weakened.
The second significant shift in work culture was increased workloads. Some managers appeared to embrace higher levels of work intensity as a symbol of their individual resilience, which in turn set an example to staff about increased performance expectations. Departmental managers often argued that financial pressures had forced them to stop the non-value-added work they had previously done, but staff reductions in and of themselves were also a measure of success: there's been an awful lot of re-structuring and shrinkage in the last 3 years we've lost over 2,000 staff and by It appears that managers saw these changes in work cultures and management styles that served the shortterm instrumental needs of the organisation as increasingly legitimate, irrespective of the potential negative effects on substantive workplace employment relations.

| The normative pillar
Despite significant job losses, internal redeployment and rising workloads, our results show remarkably little explicit conflict with workers or trade unions. HR managers argued this was largely the result of the legitimacy and fairness of 'soft' cost cutting measures such as VR and internal redeployment, and the ongoing commitment to worker consultation. Strategic transformation plans were also important in establishing the narrative that cutbacks were legitimate and not ad hoc or random (Czarniawska-Joerges, 1989), while also persuading workers to 'buy into' the council's corporate vision. One HR manager explained: Echoing these concerns, the trade unions at all three councils noted that despite the espoused commitment to wellbeing, sickness rates had increased sharply since 2010, with stress and mental health issues by far the most common cause of staff absence because of rising work intensity. All three councils experienced increased turnover among critical front-line roles such as social workers and had to use agency staff to compensate; a finding echoed by Cunningham et al. (2021).

| DISCUSSION AND CONCLUSION
This article has revealed the complex realities of downsizing and restructuring and shown how and why organisations adopt specific restructuring practices shaped by a range of contextual and institutional factors. We were motivated to look beyond the short-term resilience of employment levels and HRM practices after the great recession to understand how restructuring is implemented and legitimised within organisations, and the response of different stakeholders to specific recessionary bundles.
Through a detailed and context sensitive analysis of the restructuring process at three large local authorities in the UK, putative model employers, this article makes three substantive contributions. The first contribution is to shed light on changing management models in the public sector; a context often overlooked despite the clear threat posed by austerity to the historical commitments to paternalistic HRM and collective employment relations. The longitudinal research design allowed us to capture the cumulative effects of repeated incremental change over 5 years that eluded studies undertaken shortly after the 2008-2009 recession (e.g., Beszter et al., 2015;Conway et al., 2014;Van Wanrooy et al., 2013). Our case studies reveal that local authorities have made significant labour cost savings largely through voluntary redundancy, internal rather than external labour flexibility and changes to the skill mix. However, the trade-offs for initially avoiding mass layoffs were a somewhat 'take it or leave it' offer of alternative job opportunities, rising work intensity and managerialism within services. A lack of observed conflict is not due to the inherent fairness or responsibility of cost cutting measures, or the investment in behavioural practices such as training and consultation (Teague & Roche, 2014;Tsai & Shih, 2013). This lack of conflict seems to reflect the weakened bargaining power of workers, and their steady desensitisation to redundancies, internal restructures and the threat of outsourcing (e.g., Turnbull, 1988). The dissonance between insidious workforce cutbacks and the espoused commitments to fairness and ethics is yet more evidence of the subordination of worker wellbeing to the logic of market discipline, even in sectors not directly exposed to processes of financialisation (e.g., Thompson, 2013).
Second, by drawing on institutional theory, we are able to move beyond the simple dichotomy often found in the literature between either radical change or resilience in models of HRM when faced with external threats such as recessions (Harney et al., 2018;Johnstone et al., 2019;Vaughan-Whitehead, 2013). Institutional theory allowed us to develop a more nuanced account of the dynamics of workforce adjustment within and across organisations, and provided a framework for exploring how actors frame specific courses of action based on the perceived congruence with embedded organisational values (e.g., Ahlstrand, 2010;Lewis et al., 2019). Drawing specifically on Scott's (2008) model, we see that the regulative and normative features of public sector HRM retain their legitimacy in part as a result of the reference to basic legal minimums and the lower standards offered in the external labour market, rather than substantive notions of good or model employment. Furthermore, our multi-actor perspective allowed us to explore the subtle differences between legitimacy (e.g., what practices are considered acceptable or deviant a priori within a particular institutional context) and legitimation, which is the process of justifying potentially deviant practices within that context (Fernández-Alles & Valle-Cabrera, 2006). Our case studies reveal how managers seek to legitimise the erosion of key standards such as job security, internal promotion and worker autonomy through the narrative of 'continuous improvement'. This narrative helps justify the short-term focus on doing 'more with less', but crucially it is vague and open-ended, and demands ever increasing levels of individual flexibility, resilience and work effort. Indeed, it is the 'persistent state of instability' created by 12internal reorganisation and reform that provides cover for the erosion of HRM standards and practices (Chaudhry & Rubery, 2019, p. 552). Here, we build on earlier institutional analyses of public sector change (Czarniawska-Joerges, 1989) to show how ongoing reforms confer symbolic legitimacy on managers, whether or not reforms actually achieve any measurable improvements.
The third, and perhaps most provocative, contribution is to question the theoretical and practical distinction between pure and responsible restructuring, and by extension the sharp contrasts sometimes made between 'hard' and 'soft' models of HRM that supposedly underpin distinct bundles of practices (e.g., Ahlstrand, 2010;Teague & Roche, 2014;Tsai & Shih, 2013). Despite the strong tradition of progressive HRM in the public sector, we found an ambiguous relationship between behavioural and technical practices, and an uneven engagement of HR professionals in substantive processes of cultural-cognitive change within services. As a result, we argue the concept of responsible restructuring not only lacks a context-specific framework for determining the likely effectiveness of specific HR policies (e.g., Cooke, 2018), but also assumes it is feasible for the HR function to reconcile the fundamentally competing logics of market discipline and worker wellbeing (Dobbins & Dundon, 2017;Guest, 2017).
The steady erosion of employment conditions within the public sector raises questions about the legitimacy of HR professionals as employee champions in the face of top-down pressures for labour cost savings and the opportunism of operational managers who leverage conditions of economic uncertainty to take greater control of working practices (Cook et al., 2016). In this context, the trade unions may see value in adopting more confrontational strategies that sacrifice local partnership deals (Connolly, 2020) in order to put the responsibility back onto managers to 'do HRM properly' (Cook et al., 2020(Cook et al., , p. 1277. Taken together, the findings of this study suggest a pattern of incremental deviance characterised by steady, rather than radical, cost cutting, the hollowing out of consultation mechanisms and the ratcheting up of work intensity as workers leave and are not replaced. As Cameron (1994) argues, repeated minor cutbacks, however carefully they are announced and implemented, can be equally as damaging as one-off restructures. While the public sector may appear institutionally resilient, a decade of austerity has levelled down standards, increased work intensity and reduced career development opportunities for public sector workers, all of which create significant risks for recruitment, retention and burnout (e.g., Cunningham et al., 2021). While care must be taken not to overly romanticise the historical 'model employer' approach in UK local government, the core institutional reference points of decent wages, job security and career development are stronger foundations on which to reform public services in the 21st Century than those of short-term cost cutting and market discipline.

| IMPLICATIONS, LIMITATIONS AND FUTURE DIRECTIONS
It is worth noting that our findings, although rich and contextualised, are limited to the UK public sector where the scope for perceived deviance is greater than in the private sector given the historical commitment to progressive HRM. There are clear policy lessons for national governments, not only in terms of the short-term underinvestment in public services, but in the risk to long term recruitment and retention of skilled workers arising from deteriorating job quality and the normalisation of fiscal cutbacks (Cunningham et al., 2021). Our findings also complement studies of large private sector organisations that have experienced the contestation of embedded HRM standards