The rhetorics of ‘agile’ and the practices of ‘agile working’: Consequences for the worker experience and uncertain implications for HR practice

Abstract The various rhetorics of ‘agile’, ‘agility’, and ‘agile working’ (AW) set an agenda for new ways of working and have recently gained traction in popular management discourse, particularly in the wake of the COVID-19 pandemic. Yet conceptually, these rhetorical varieties of ‘agile’ are underdeveloped in the academic literature. In this article we examine the stream of AW as being a particularly influential rhetoric. AW is critically evaluated by first identifying separate streams and rhetorics of ‘agile’ in the literature, and AW is then situated within this typology. To understand the particular version of reality being mainstreamed by the AW rhetoric, we then examine AWs conceptualisation as ‘a new way of working’, as promoted by dominant actors within the UK work context. We then consider existing studies of worker experiences under different employment arrangements that can be subsumed under the heading of ‘AW practices’. Our analysis highlights voids between what may be considered as mainstream HR practice when applied to standard employees compared to a spectrum of ‘non-standard’ workers. The implications for the role of HR in the implementation of AW and in managing the worker experience are discussed and future avenues for this under-researched area are offered.


Introduction
' Agility' is claimed to be a strategic means of harnessing a necessary dynamic capability for organisations operating in a highly unpredictable environment (Teece et al., 1997;Nijssen & Paauwe, 2012). In recent years, the notion of 'agile' has emerged as a rhetoric behind a broader agenda of new ways of working, displacing that of 'flexibility' which had come to be seen by some as being 'too employee-oriented' , with its association with work-life balance. While the broader notion of 'agile' would appear to have emerged from common origins, there has been divergence in what 'agile' means within multiple streams of management literature, with each stream being seemingly unaware of each other. This article focuses on one such underdeveloped stream of the agile literature, 'agile working' (AW). It is this specific stream that has been mainstreamed and promoted by an influential combination of management gurus, consultancy practitioners, and professional bodies, with the potential to influence organisational practices and, consequently, to impact on people's working lives. This is particularly important in the wake of the COVID-19 global crisis which abruptly altered the ways in which paid work was being done. Practice often precedes theory (McMackin & Heffernan, 2020) and it is for this reason that academic research needs to keep up in this area and offer analysis with implications for HR practice.
AW, as we explain in more detail below, refers to 'a set of practices that allow businesses to establish an optimal workforce and provide the benefits of a greater match between the resources and the demand for services, increased productivity, and improved talent attraction and retention' (CIPD & Agile Future Forum, 2014, p. 3). This set of ' AW practices' span four dimensions, which we explore below. These are (i) working time, (ii) workplace location, (iii) job role and (iv) composition of the workforce (CIPD & Agile Future Forum, 2014;Agile Future Forum, 2021;Holbech, 2015).
We argue that AW, as a particular stream of 'agile' , requires scrutiny, because of the interests that are promoting it. AW's apparent appeal is through its claimed solution to organisations' need to respond and adapt to change, but also because of a claimed workforce expectation driven by a preference to hold multiple jobs over the course of a career and the diminishing centrality of work in individuals' lives (CIPD & Agile Future Forum, 2014). We, therefore, need to examine the notion of 'agile' conceptually as a prescription for new ways of working. The article is guided by the following research questions. First, what are the different streams of agile and where does AW fit into this? Second, how do different groups of workers experience AW practices? In answering these questions, we can then consider what is HR's role in the administration of AW and what are the implications of AW for HR's domain of authority/control? The article proceeds as follows. First, a typology of 'agile' is offered, starting by locating and mapping the different scholarly-based streams and rhetorics of 'agile' and then placing the under-researched AW stream within the context of the agile literature. Second, a consideration of the origins and rhetoric of AW and its place in its wider conceptual context is offered. Third, a consideration of the worker experience of different working practices that may be subsumed under the heading of ' AW practices' as advanced by the influential dominant actors is made. As there is no existing research specifically on the worker experience of ' AW practices' per se, our analysis is based on a narrative literature review on workers' experiences that map onto the four different dimensions of AW. We then consider and discuss the implications for HR's domain of control/authority in relation to HR's role in managing the worker experience and in the implementation of AW. The article concludes by presenting the theoretical and practical implications of our work and by offering future avenues for this under-researched area.

The multiple meanings of agile: multiple origins, aims and conceptual grounding
Before examining the rhetorical model of AW that is the focus of this article, we first consider some conceptual grounding for the notion of 'agile' more broadly. The concept of 'agile' operates at multiple levels: '…from philosophy, culture and mindset, through to processes and methodologies, and affecting roles and behaviours of leaders, teams and individual employees' (McMackin & Heffernan, 2020, p. 3). It claims relevance to all sectors and all functions (Rigby et al., 2016). Yet, closer examination reveals that there is more than one interpretation of agile. We now bring together the various streams and rhetorics of agile in the literature, starting with five different 'agile' streams of literature before introducing a sixth stream of AW.
A first stream is 'agile manufacturing' . As a term, 'agile' emerged in a variety of management (and sometimes non-management) disciplines, in publications both academic and practitioner, with its first mention from as early as 1985 (Walter, 2021). It is generally acknowledged that the concept of 'agility' was first introduced in mainstream business literature by the Iacocca Institute in 1991 and in the specific context of 'agile manufacturing' (Yusuf et al., 1999). In this first articulation and stream, agile is the ability to produce a broad range of low-cost, high-quality products with short lead times in varying lot sizes and built to individual customer specifications (Vokurka & Fliedner, 1997). This manufacturing-based stream of agile has developed its own body of literature, including Burgess (1994), Gunasekaran (2001), Meade and Sarkis (1999), Narasimhan et al. (2006) and Vázquez-Bustelo et al. (2007). A consistent pattern in the early articulations of this stream is its explicit referral back to 'lean production' and just-in-time (JIT) manufacturing (Womack et al., 1990;Oliver & Wilkinson, 1989), as well as business process re-engineering and total quality management (Hill & Wilkinson, 1995). In their own time, these 1980s innovations were being claimed as signifiers of a major paradigm-shift, away from Fordist mass production (Hirst & Zeitlin, 1991;Jessop, 1988;Piore & Sabel, 1984).
A second stream of agile is 'agile as project management' . It is oriented to team-based project management, termed as 'scrum' methodology (Schwaber & Beedle, 2002) in software development (Birkinshaw, 2018;Pope-Ruark, 2015). In this stream, multi-disciplinary project teams are formed and then dissolved according to relevance of expertise and project timescales. This stream is bolstered with the formation of the Agile Alliance in 2001 and the publication of The Agile Manifesto (Fowler & Highsmith, 2001) to promote this approach among 'creative' and/or 'knowledge workers' (Thursfield, 2015). The origins of this stream of agile, like the manufacturing variety, also traces its origins to the just-intime manufacturing paradigm of the 1980s (Hodgson & Briand, 2013).
A third stream of agile, 'agile as workplace ergonomics' , focusses on workplace ergonomics and has a particular sectoral bias towards workplaces in public services contexts, including healthcare (Joroff et al., 2003), social work (Jeyasingham, 2019), andprobation (McDermott, 2016). This stream, while making some reference to 'lean production' , is more explicitly tied to spatial and temporal dimensions of agile, considered in more detail as part of the AW stream which we present later in this section.
A fourth agile stream is 'organisational agility' (OA) and is by far the most developed in terms of recent academic conceptualisation. OA, also seemingly emerging from the manufacturing-based stream of agile, is geared more specifically in the strategic capabilities of organisations on the ability to adapt to external change. Walter's (2021) systematic review of OA identifies some ambiguities in whether OA is a capability, a process, or a 'system of practices' (Narasimhan et al., 2006, p. 441). Some place OA as a 'philosophy' (Bernardes & Hanna, 2009;Sharp et al., 1999), claiming OA as encapsulating a firm's philosophy, values and culture. This high level of abstraction, with identifiable characteristics disappearing, risks a conflation of 'what' and 'how' issues (Narasimhan et al., 2006). More concrete definitions identify OA as a 'strategic capability' (e.g. Chakravarty et al., 2013;Paixão & Marlow, 2003) or a 'dynamic capability' (e.g. Bessant et al., 2001). Felipe et al. (2016) define OA as an organisation's capability to sense environmental changes and to respond efficiently and effectively to them, while Teece et al. (2016, p. 17) define it as 'the capacity of an organization to efficiently and effectively redeploy/redirect its resources to value creating and value protecting (and capturing) higher-yield activities as internal and external circumstances warrant' . Nijssen and Paauwe (2012) observe that OA includes the ability to create organisational knowledge fast, the ability of having a scalable workforce, and having a highly adaptable organisational infrastructure as a prerequisite for workforce scalability.
A fifth stream of agile is an apparent offshoot from the OA stream, which identifies itself as 'workforce agility' (Al-Kasasbeh et al., 2016;Alavi, 2016;Braun et al., 2017;Sherehiy & Karwowski, 2014). However, despite its title, the common attribute of this manifestation of 'agile' is not of the workforce per se, but more specifically of the idealised attitudinal attributes of workers within the workforce. For an organisation to have the capacity for agility, 'workforce agility' requires workers willing and able to be 'resilient' , 'adaptable' and 'proactive' in the face of change. This stream of agile seems firmly embedded within the organisational psychology literature.
Through our consideration of the different streams of agile, we depict the consistent origin of these five different agile streams. Figure 1 below is our typology of agile streams and illustrates common roots of all versions in the just-in-time/lean production regimes of the 1980-1990s and where each stream deviated from other streams. It also visually presents the rhetorical origins and links of a sixth stream-the ' AW' stream-within the broader agile typology, which we examine in detail below.
The AW stream is oriented around a managerial agenda for making specific interventions around the reorganisation of work. It shares with all other agile streams an implicit reference back to the managerial 'transformation' rhetorics of the 1980s and 1990s, though in this case the obvious reference is to the rhetoric of 'flexibility' . While AW has a lesser academic grounding than other agile streams, it can claim powerful rhetorical influence as a heuristic for forcing change, which we now explore.

'Agile working' within the rhetorics of agile
The rhetorical promotion of the idea of AW by a variety of management gurus, consultancy practitioners (e.g. Cantelo & Clarke, 2017;Holbech, 2015;McKinsey, 2021) and professional bodies (e.g. CIPD & Agile Future Forum, 2014) are somewhat disconnected from an academically-grounded conceptual model of it. This phenomenon of disconnect is not new. Management 'fads' have previously been identified as being important to study for the rhetoric and interests that lay behind the concepts as much as the concepts themselves (Newell et al., 2001). AW as a rhetorical discourse, therefore, affects organisational practices and thus, has real effects on people's working lives. This is particularly important in the wake of the COVID-19 global crisis, which abruptly altered the ways in which paid work was being done. Across the world, the multiple national lockdowns over 2020-2021 necessitated many workers to work from home at different times of the extended working day and business commentators speculated about a 'new normal' way of working in a post-COVID-19 world (European Commission, 2020). During this period, the usage of the term AW was propelled from managerial to popular discourse and often used interchangeably with the term 'flexible working' (ILO, 2020). These crisscrossing definitions point to a 'conceptual slippage' (Ackers & Payne, 1998) in the rhetorical usage of 'agile' . As demonstrated in our typology and in Figure 1, 'agile' can be seen as a contested concept through its various deployments in separate streams of literature.
While conceptual clarity and empirical reality are important, there is also a need to examine the rhetoric of agile in its own right, because rhetoric holds constitutive power for stakeholders who employ the rhetoric to 'forge a certain version of reality' (Ercek, 2006, p. 650). As Delbridge and Keenoy (2010, p. 804) emphasise, 'language is never neutral-it always implicates and privileges particular social values if not also specific socio-economic interests' . Again, this has precedent. As Wood (1991, pp. 582-583) points out, in the conceptualisation of JIT production systems at the point in which they were permeating into workplaces outside their origins in Japan, 'real' manifestations of JIT were more important to understand than the 'pure' versions existing in prescriptive management accounts. It is important to note, then, that the word 'agile' is itself not a neutral or mere technical word: it is an honorific term, making any claim to oppose any initiative under its name to appear unreasonable and reactionary. Dominant actors, including management elites, 'gurus' , and consultants, stand to gain from upholding specific claims about 'new' managerial discourses and concepts to provide certain advantages for themselves (Ercek, 2006).
In the UK work and employment context, two influential national-level stakeholders stand out in this regard. The first is a consortium known as the Agile Future Forum (AFF), an employer-led group with 23 founder member organisations, representing leading businesses across a range of sectors and launching itself via an open letter to the prominent UK establishment newspaper, the Telegraph (2013). The second UK-based national-level stakeholder is the professional body for HR practitioners, the Chartered Institute of Personnel and Development (CIPD). Together, these influential stakeholders have driven AW as an agenda. As actors in a position to promote particular versions of agile, the CIPD and AFF are both influential. For the AFF, this influence is through their ability to implement policy affecting the half-million workers directly employed in their own organisations (Agile Future Forum, 2021) and through their supply chains. In institutionalist terms, this influence can be recognised as coercive and mimetic isomorphism (DiMaggio & Powell, 1983). It is coercive isomorphism in the sense that the promotion of particular agile practices by those large organisations signed-up to the AFF creates an imperative both in the organisations themselves-through direct imposition-and also in smaller suppliers in the value chain obliged to adapt to the changed circumstances AW would impose. Isomorphism could be considered mimetic in the sense that the promotion of a particular stream of agile, by such 'leading' organisations, offers a demonstration to others to mimic 'proven' best practice. The CIPD's influence is through its defining of 'best practice' i.e. via normative isomorphism (Roper & Higgins, 2020), using its membership base of over 150,000 (CIPD, 2021a). This makes it vital for researchers to scrutinise the concept and rhetoric of AW and to ask questions around the implications of this concept and rhetoric for how work is to be reformed.

The dimensions of 'agile working' and the legacy of the 'flexible firm'
In this section, we argue that AW as a stream of agile is both an amalgam of a selection of practices taken variously from the more technically grounded streams of agile described above and derived from another antecedent of many of these: the 'flexible firm' (FF) model. Proposed by Atkinson (1984), the FF model informed much of the backdrop to changing working practices associated with neoliberalism from the late 1980s onward in the UK-despite receiving biting critique as to its desirability and its efficacy (Hunter et al., 1993;Pollert, 1988Pollert, , 1991. We outline each dimension of AW and consider how it compares to the FF model, as both propose four dimensions of working practices. The first AW dimension of working time, which relates to 'when people work' , is comparable to the FF dimension of temporal flexibility (sometimes internal numerical flexibility). The second AW dimension of job role maps onto the FF dimension of functional flexibility. This relates to the extent to which workers can take on different tasks or responsibilities within the organisation. The third AW dimension of labour source, which refers to 'who is employed' , parallels the FF type of external numerical flexibility, which creates segmentation into 'core' and 'peripheral' groups of workers. The fourth AW dimension is workplace location, where the notion of the fixed workplace is opened up to include remote working and homeworking as part of the spatial mix where work is conducted. It does not map directly on the FF model. However, we argue that while this dimension may be 'new' in relation to the FF model, it does relate closely to the FF type of (temporal) internal numerical flexibility and the AW dimension of working time. In practice, homeworking breaks both location and temporal boundaries concerning the work-life interface. There is one dimension of FF model, financial/ wage flexibility, that has not been translated into the AW four dimensions of working practices. One can speculate as to why AW is needed when the existing FF model is readily available. One answer is that AW offers novelty and an incentive to begin 'change management' with a new label. A second reason may be that, despite the influence of the FF model over the years, advocates of AW have asserted that 'flexibility' has come to be associated with being 'an employee benefit' through its association with work-life balance. AW shifts this back toward an employer agenda. Finally, while there are overlaps, there are also some differences. Both FF and AW propose working practices that segment the workforce within the organisation into core ('permanent') and periphery ('non-standard') workers through the mechanism of FF's external numerical flexibility and AW's labour source dimension. Non-standard contingent workers include those who have a limited temporal attachment to organisations (e.g. fixed-term/temporary workers, zero-hours workers, and freelance workers), those with limited physical attachment to the organisation (e.g. teleworkers and remote/homeworkers), and those with limited administrative attachment to the organisation (e.g. through labour intermediaries, such as agency workers, or independent contractors, such as freelance workers) (Pfeffer & Baron, 1988). Whereas 'permanent' employees engage with organisations in a standard employment relationship with close connection to the employing organisation, non-standard workers are involved in a non-standard employment relationship with a fairly loose connection to the employing organisation (Bernhard-Oettel et al., 2017). Bolton et al. (2012, p. 122) argue that while non-standard work comes in many forms, a key feature is 'its strategic use as an employment relations mechanism. This may involve the use of short-term or casual contracts to give numerical flexibility or plug-in short-term gaps, or it may be through longer-term structural means such as outsourcing' . In this article, it is this AW dimension of labour source that we focus on in our analysis of different workers' experiences of ' AW practices' and its implications for HR's implementation of AW in relation to managing the worker experience, including HR's domain of authority/control

Workers' experiences of 'agile working' and the 'new' core-periphery
We identify no existing research specifically on the workers' experience of AW as a package of working practices, nor on any of the five other agile streams described in the typology above. However, we can draw on extant literature that has considered workers' experiences of the various types of 'non-standard' working practices separately. These categories of workers form part of the 'new periphery' under the AW dimension of labour source. We now consider the definitions, characteristics, and experiences of each category of worker, presented hierarchically by the degree of precarity (Kalleberg, 2018).
The discussion draws on published academic journal articles we identified using a narrative literature review approach (Green et al., 2006), which is an approach leading to a comprehensive narrative synthesis of previously published information. This approach is used for integrating or comparing findings from studies, looking for themes or constructs that lie in or across individual studies and synthesize them through a qualitative narrative approach. Post et al. (2020) note that such an approach often takes the form of qualitative discussions of ideas produced by authors who are experienced in their fields and develop arguments drawing upon prior knowledge (Baumeister & Leary, 1997;Jones & Gatrell, 2014). For the purpose of a thorough search, we looked for published academic journal articles in databases of online libraries such as EBSCO, Emerald, JSTOR, SAGE, Scopus, Springer, Taylor and Francis, Web of Science and Wiley. We used keywords to look for articles published on each of the category of workers we were interested in: fixed-term/ temporary, agency, zero-hours, freelancers/e-Lancers and homeworkers. The criteria to choose articles for review were: (a) peer-reviewed scholarly articles that were either empirical, review or conceptual in nature and (b) and that were written in the English language.

Fixed-term and temporary work
As a 'non-standard' category, employment on a fixed-term basis is reasonably categorised as the least precarious. In the UK, workers who are employed on a fixed-term basis are legally classified as employees and enjoy comparable employment rights as permanent employees up and till the point when the contract ends. That said, there is a hierarchy of precarity within this group defined by the length of the contract. Fixed-term employment can vary between project-based work, which could be measured in years, through to seasonal work and casual work (ILO, 2016). As a broad category, this type of employment is long established. There is overwhelming evidence of the negative experiences of fixed-term and temporary workers with work. Our review finds numerous studies that have measured the adverse differences in the employment situation of such workers in a range of areas, such as working conditions including pay (Feldman et al., 1994), access to voice opportunities (e.g. Oyetunde et al., 2022;Hoque & Kirkpatrick, 2003;Piasna et al., 2013;Qian et al., 2020), job satisfaction (Wilkin, 2013), turnover intentions (Mauno et al., 2015;Nuhn et al., 2018), career development (Kompier et al., 2009), job insecurity (Cappelli & Keller, 2013;De Cuyper et al., 2018;Håkansson et al., 2020) and well-being (Baluch, 2017;Imhof & Andresen, 2018;Sirviö et al., 2012;Virtanen et al., 2005).

Agency work
Further down the precarity hierarchy are agency workers. Agency workers are part of a triangular employment relationship consisting of the work agency, the workers, and the client organisation (Kalleberg, 2000;Svensson et al., 2015). In the UK, workers who access employment via a third-party agency and have a 'contract for services' with the agency, not with the client organisation, are currently legally classified as workers (ACAS, 2020). To date, they have been covered by some minimal protection via the EU Agency Workers Directive. Agency workers are workers who have been outsourced to provide labour and services, where they are location-and time-bound. In terms of workers' experience, our review finds that agency workers are usually characterised as of relatively low value and easily replaceable (Lepak & Snell, 1999). For this reason, such workers are usually neglected from collective agreements (Benassi & Vlandas, 2016). Working conditions are poor, characterised by a risk of unemployment, lack of on-the-job training (Hoque & Kirkpatrick, 2003), lack of promotion prospects and higher risk of hazardous work conditions (Jahn & Pozzoli, 2013;Nienhüser & Matiaske, 2006) that may damage the workers' health and well-being (Kompier et al., 2009;Virtanen et al., 2003).

Zero hours contracts
Further down our hierarchy of precarity lies zero-hour contracts (ZHCs). In the UK, workers who work on ZHCs are workers who are on contracts that may specify the nature of the work to be done but do not guarantee a minimum number of hours (ONS, 2018)-and therefore no minimum associated pay. In 2018, 901,000 people in the UK were on ZHCs (ONS, 2018). The term ZHCs does not have a specific meaning in law (CIPD, 2019), so they can be legally classified as either employees or workers, with the latter being entitled to fewer employment rights protection.
The Labour Force Survey (LFS) define ZHCs as 'where a person is not contracted to work a set number of hours and is only paid for the number of hours that they actually work' (ONS, 2018, p. 3). This definition captures an explicit focus on employer-induced workforce flexibility (Wood & Burchell, 2014). Another definition offered by the UK government (BIS, 2013, p. 7) goes further to imply workers can also exercise 'choice' , defining ZHC as an 'employment contract in which the employer does not guarantee the individual any work, and the individual is not obliged to accept any work offered' .
This 'choice' interpretation of ZHCs is also supported by the CIPD (2013, p. 5), which claims that 'ZHCs, managed properly, can work for both employers and individuals' , based on a survey of employers and ZHC workers in UK. The same report found 'very little difference' in overall job satisfaction between ZHC workers and non-ZHC workers. Given that ZHC workers are hourly paid, this precarity caused by high levels of unpredictable variability in ZHC workers' schedules has led to a particular form of job insecurity, which Wood and Burchell (2014) have termed schedule insecurity. Much, if not all, of the 'choice' argument for ZHCs is challenged in the Taylor Review (Taylor et al., 2017), where enhanced protections in employment law were advocated, but yet to be implemented.
Our review reveals that ZHCs are associated with indicators of inferior job quality such as low pay and underemployment (Koumenta & Williams, 2019). The consequences of ZHCs for workers are job insecurity and unpredictable income, lack of access to work-related benefits and entitlements, lack of opportunity for career development, poor work-life balance, and stress (Farina et al., 2020). Work intensity is also higher, while line management and peer support weaker, and the threat of dismissal and job loss negatively impacts on the well-being of these workers (Felstead et al., 2020;Ravalier et al., 2019Ravalier et al., , 2017.

Freelance work
In the UK, individuals who do freelance work are legally classified as self-employed independent contractors working for their own business and are therefore neither employees nor workers of the hiring organisation. Freelance work can be done by independent professionals, knowledge-or project-based workers, and freelancers. The work done by this broad group of people tend be characterised as high-skilled and they can be either location-bound or fulfilled virtually and is therefore done remotely.
In recent years, there is growing research focus on a particular group of workers who engage in a form of freelance employment accessed via digital labour platforms, known as e-Lancers (Abubakar & Shneikat, 2017). Workers become 'sellers' by registering on websites that act as digital 'marketplaces' for their services. Digitally-mediated matches can then be made between e-Lancers and 'buyers' , who are the hiring organisations or individuals (Aguinis & Lawal, 2013;Jabagi et al., 2019). New research estimates that 163 million workers globally have registered with digital labour platforms (Kässi et al., 2021). Such platforms provide transportation and delivery services (e.g. Uber, Deliveroo), household and personal services (e.g. TaskRabbit), specialised services (e.g. TakeLessons), clerical services (e.g. Fancyhand, Microwork, Amazon MTurk), or creative/technical work (e.g. Fiverr, Upwork, Freelancer.com) (Jabagi et al., 2019).
Our review reveals mixed experiences of workers who do freelance work. In a diary study of portfolio freelance workers, Wood and Michaelides (2016, p. 131) find that participants exercised choice to escape from 'the politics of the organisation' by becoming independent contractors. While they saw themselves unlikely to return to standard work, they also experienced comparable pressures in terms of work demands and their fluctuations. Leighton (2016) reported that independent professionals often find themselves as outsiders of the organisations. E-Lancers in Southeast Asia and Sub-Saharan Africa reported positive experiences, including high level of flexibility, autonomy, task variety, and complexity (Wood et al., 2019). Similar positive experiences were reported in a study of e-Lancing workers in Turkey (Abubakar & Shneikat, 2017). But they also report negative experience of low pay, social isolation, unsociable and irregular working hours, overwork, sleep deprivation, and exhaustion (Wood et al., 2019). Kuhn and Maleki (2017) found that high levels of surveillance used by digital labour platforms have a negative impact on workers' experience, including their sense of autonomy and motivation. In the case of Uber, while drivers can exercise some autonomy in terms of when and where they turn off the app, they are penalised by Uber for low acceptance rates and cancellation rates (Kuhn & Maleki, 2017). Wood et al. (2019) emphasised the role of algorithmic control and management in producing commonalities in the mixed experiences of e-Lancers. The lack of power to negotiate contracts and rates contribute to the negative experiences of e-Lancers (Kuhn & Galloway, 2019). The legal employment status of e-Lancers is more contentious, as demonstrated in the landmark UK Supreme Court ruling of Uber BV v Aslam [2020] on 19 February 2021 to recognise the status of Uber drivers as 'workers' . In a review of e-Lancing as a working practice, Aguinis and Lawal (2013) argued that e-Lancing work differs from independent contracting work, because the employment relationship is not only with the worker and the employer, but also with the marketplace i.e. the digitally mediated labour platform.

Homeworkers
Homeworking falls under the AW dimensions of workplace location and working time ('where' and 'when'). Aguinis and Lawal (2013) differentiate e-Lancers from teleworkers, because while both work remotely with the aid of technology, it is only teleworkers who are in traditional employment relationships with their employers, i.e., they work only for one employer. While temporal flexibility formed part of the narrative of the FF, technological advances-particularly information technology-enhanced the scope for homeworking by the 2000s. Thus ONS (2020) report that technology has enabled homeworking in the UK, though with significant sectoral variations: hospitality experiencing only a 10% take-up compared to ICT with 53%.
While AW may already be a catalyst for this, following COVID-19, it is now being widely reported that homeworking is no longer going to be the option for the privileged few within the 'core' , but will, for many categories of jobs, be the new normal (European Commission, 2020). As of April 2020, some 46% of workers in the UK experience some level of homeworking, with 86% of them directly as a consequence of COVID-19 (ONS, 2020). The CIPD (2021b) reports that a significant proportion of UK employers are actively planning to increase homeworking as a permanent measure after COVID-19.
What may be the predicted consequences of making homeworking the new normal? Our review finds some predicting significant benefits for gender equality from the increased homeworking resulting from COVID-19 (Alon et al., 2020). Increased homeworking, however, may not benefit all equally as the impact of 'family-work conflict' (De Clercq, 2020) may disadvantage those with unfavourable childcare and home environment situations. Conversely, the intrusion of work into the home may also not be felt evenly: some workers experiencing increased autonomy whilst others feeling the intrusion of 'limitless worker surveillance' (Ajunwa et al., 2017). Work-life balance, like other forms of worker experiences, is shaped by the intersectional characteristics of the worker (Beauregard et al., 2020), as different workers and their families experience 'different life worlds' (Kelliher et al., 2019). Lockdown has already disadvantaged working mothers (Andrew et al., 2020), so considering if this will have longer term impacts on equality is prescient. What is also prescient is how this AW dimension of workplace location intersects with the dimensions of labour source and working time.

The rhetoric of 'agile working' versus the empirical reality
In the first part of the article, we discussed the importance of the rhetorics of 'agile' in general, the variety of 'agile' streams available and why we focused specifically on the AW stream, despite it having perhaps the lowest degree of scholarly support. AW is important, we argue, because it has the support of powerful advocates, serving as catalysts for certain 'management fads' that have become important in the past (Newell et al., 2001). Employer organisations such as the AFF, supported by professional bodies such as CIPD, are promoting AW as a particular rhetoric with real life effects on organisational practices and on people's working lives.
Because AW is rooted in practitioner advocacy, not academic purity, our review finds no research on workers' experience of AW practices as a package. However, it does produce evidence of research that has considered workers' experiences of the various practices we have identified as the four dimensions of AW when looked for separately. We presented an overview of this research, showing how worker experiences vary. For fixed-term, temporary and agency workers, the research evidence, on balance, points to a negative experience for such workers across multiple dimensions of working conditions and work quality. ZHC workers have similar, but more pronounced, negative experiences, because of their higher job insecurity and inability to plan work. Freelancers and e-Lancers have mixed experiences depending on their agency to 'choose' and manage such work. Finally, homeworkers may experience the benefits of where and when work is done, but research also indicates inequalities in the experience of these benefits. It should be noted that evidence from homeworking comes from publications prior to the reporting of the experiences of homeworking during COVID-19 in 2020. Our analysis of the existing literature points to the need to carefully consider the rhetoric of AW, as rhetoric is translated to policy and practice, and this has implications to the increasing number of non-standard workers in the workforce.

HR and its domain of control/authority -where do non-standard workers fit?
Given the diversity of non-standard workers, what are the implications of AW for HRM? There appear to be some considerations of this question emerging. McMackin and Heffernan (2020) provide one such approach to redesigning the HR function, though in this case, the approach is aligned to the OA stream of agility. As our analysis has shown, the implications of AW as a managerial agenda for new ways of working pose different dilemmas for HR, which we now discuss.
HR's very existence is arguably a product of the paradox defined in 'the Coase theorem' (Coase, 1937(Coase, , 1960, where the open-ended contract of employment has been the norm. HRM deals with the central problem of control and the employment relationship as 'structured antagonism' (Edwards, 1986). While Coase did not explicitly make reference to HRM as a management function, the Coase theorem implicitly explains the existence of a function to deal with dilemmas faced in using people in organisations not based on the spot market. Yet, even before AW, developments in how labour is deployed has shifted away from an exclusive focus on work organised through the open-ended employment contract and with this shift, the role of HRM is less clear for those 'non-employed' workers. From the high Fordist notion of the all-encompassing vertically integrated manufacturing company employing labour at every point in the value chain, to the disaggregation of such monolithic enterprises inspired by 'transaction cost economics' school (Williamson, 1981), the Coase theorem was not, it seemed, absolute and outsourcing grew in influence.
The most prominent and explicit call for a redefinition of purpose of HRM to meet contemporary challenges has been that associated with Ulrich (1997) and the call to move away from transactional HR to become a 'strategic business partner' . While this model has undoubtedly become the most prominent 'best practice' model of HRM, it has been called into question: as a utilitarian instrumentalist project, it conflates shareholder-value with the interests of wider society and in doing so, puts HR is the position of becoming merely the perfect agent of senior management, thereby facilitating its own demise (Kochan, 2007). Dundon and Rafferty (2018) and Marchington (2015) similarly identify that the activities HR has prioritised have been at the expense of those other roles oriented toward the welfare and wellbeing of workers. This creates voids in HR outcomes for non-traditional workers.
Therefore, in the case of AW, HR's pursuit of the role of a strategic business partner, itself criticised as being overly instrumentalist, may undermine Ulrich's assertion that one pillar of HR's role should be in managing workers' welfare and wellbeing in the organisation. For those AW workers not directly employed, who looks after their welfare, career development and even conduct?
HR's experience of dealing with a more heterogeneous workforce, as asserted in AW is not completely new. Changes in the labour and employment markets have been recognised and along with it the role of HRM in addressing the issues relating to such non-standard workers (Burgess et al., 2013). Some have argued the need for a differentiated approach to HRM, allowing for the difference in content and quality of the employment relationship for different types of workers (Koene & Van Riemsdijk, 2005). Lepak and Snell (1999) framework follows this reasoning by distinguishing four different 'employment modes' linked to four types of 'human capital' with implications for the HR architecture for managing each category. However, this framework treats temporary employees and workers as of limited strategic value to organisations as their employment relationship is transactional. They argue that because of this transactional employment relationship, HR activities need only focus on securing compliance with the terms and conditions of the contract, upholding specific provisions regarding work protocols, and ensuring conformance to pre-set standards. Similarly, Lepak et al. (2007) find that organisations deploy high investment HR systems only for core employees, as non-core ones are perceived as neither strategically important nor unique.
Others, however, have identified a less utilitarian view on the role of HR for non-standard workers. Organisations have a moral, if not also legal, duty of care for those working for them within and outside the boundaries of organisation (Cross & Swart, 2022). However, under AW, if there are growing numbers of people working for and in the organisation, but not formally a member of the organisation, who deals with the less straightforward and less transactional issues of wellbeing, personal development, and conflict? On a more utilitarian level, who deals also with the issues of conduct, discipline, the interactions with customers, and of vicarious liability?
The rise in the rhetoric of the FF occurred simultaneously with the range of structural changes associated with neoliberalism from the late 1980s. AW seems aligned more nebulously with the post-2008 financial crisis but possibly given a whole new impetus post-COVID-19. Yet there is also continuity. Despite its rhetorical appeal in the 1980s, the FF model ceased to be commented on as an ongoing model for firm behaviour from the 2000s. AW seems to be a repackaging of this forgotten-yet-taken-for-granted FF model. AW continues many of the FF features directly and amplifies them (i.e. external numerical flexibility) and takes others for granted as being so embedded they hardly seem worth explaining (i.e. functional flexibility). Returning to Pollert's (1988) critique of flexibility and the FF, it is remarkable how much has changed empirically (the examples given skewing toward manufacturing sectoral agreements) but also how prescient the critique of the conceptual underpinning of the entire FF rhetoric remains and continues for AW. We focus here on HR outcomes. That is, in contrast to HR processes, where the focus is on what HR practitioners do, we are concerned with the gaps that emerge as a result of AW removing key areas of the remit for the control of work, from the domain of HR's authority. Bolton et al. (2012) identify the current mismatch in HR's domain, covering only employees within the boundary of the firm, as inadequate and that HRM in its wider sense has a wider duty through the lens of moral economy. While Thompson (2011) questions whether the very premise of HRM is capable of delivering its objectives, it is no longer possible for organisations to shield themselves for culpability for the consequences of their labour practices within global value chains (Taylor et al., 2015). If such accountability may find its way into HRM discourse through the rhetoric of, if not the practice, of sustainable HRM, then the gaps in AW that emerge for workers closer to home seem harder to justify.
HR retains its domain for employees, whether permanent or fixed-term, albeit in its reformed post-Ulrich model of strategic business partner advisory role. Its role is less clear for those on non-employment arrangements (Cross & Swart, 2022). While some statutory protections must still be adhered to for those non-employees still deemed 'workers' (minimum wage, holidays), in many national jurisdictions, still others, those deemed independent self-employed freelance contractors, are outside that domain. The model of AW further exacerbates this. This is not only an existential issue for HR, as others are now highlighting both directly (Dundon & Rafferty, 2018) and indirectly (Duggan et al., 2020), but also for organisations themselves. Leaving aside those jobs that may be entirely offshored as part of AW, there remain some categories within AW that HR's role remains ambiguous. Those categories of non-standard workers under the source dimension of AW, who work in physical proximity to their co-workers who are core 'standard' employees, have needs and responsibilities roughly equal to employees that could be categorised as HR outcomes. Yet, for this category of workers, HR's authority is not clear in relation to HR processes such as recruitment and selection, performance management, remuneration and termination. On the issue of HR outcomes, wellbeing, welfare, voice (and grievance procedures), equality and any notion of career planning would seem to be absent as an HR responsibility for these workers. Related to these HR outcomes, organisations would presumably also be potentially liable for actions of such workers under the notion of vicarious liability. So, who is responsible for HR processes? Figure 2 maps all AW work contractual categories against the domain of authority/control for HR activities, highlighting the areas in which HR can expand its scope to take into consideration the diversity of non-standard workers. We explain the contribution of this map in the section that follows.

Conclusion
This article contributes to existing HRM literature on new ways of working and the implications for different sub-groups of the workforce in three ways. First, we contribute to the framing of 'agile' and its variants, by identifying the variety of conceptual and rhetorical streams that exist, seemingly independently of each other. We identify six agile streams that apply to work and organisations. We have situated the AW rhetoric in relation to these divergent streams (Figure 1) to show how AW warrants scrutiny given its weaker conceptual grounding and its potential to become influential among practitioners due to the status of the interests promoting it. We have identified AW as being a rhetoric mainstreamed by dominant actors, particularly within the UK. A second contribution is in identifying the implications of AW for the worker experience. Here, we have examined AW by drawing on a narrative literature overview of existing studies of different worker experiences that can be subsumed under the heading of AW practices. Our contribution is a reasoned conceptualisation of an amalgam term (AW) that seeks to combine a variety of forms of 'new ways of working' . By reviewing published work on workers' experiences of the various types of AW practices, specifically fixed-term/temporary, agency, ZHCs, freelancers/e-Lancers and homeworkers, our article has offered a picture of what the combination of the different dimensions of AW could look like when put together as a package in practice. Given that most of the evidence pointed to largely negative worker experiences, we relate the role of HRM within the AW to adopt a moral economy perspective (Bolton et al., 2012).
A third contribution of this article relates to the implications of widespread adoption of AW for HR practice. We identify a series of voids between what may be considered as 'normal' HR practices (applied to standard employees) and HR practices applied to a whole range of non-standard workers. We add to the recently published work by Cross and Swart (2022) which challenges current assumption that HR is purely defined, and bounded, by the employment relationship, therefore being only responsible for 'employees' found internally to the organisation. They argue that HR should incorporate all work and labour done in and around organisations, pointing to a need to fundamentally re-examine the role and function of HR. Their work focuses on independent work (self-employed, freelancers, independent contractors). Our work further extends their argument towards additional categories of non-standard workers (fixed-term/temporary, agency, freelancers/e-Lancers, homeworkers), which we study through the lens of the worker experience with such AW contractual arrangements. Our map of such workers and the authority/control of HR activities (Figure 2) highlights the voids that exist in HR activities to support a variety of workers that do not fall within the defined boundaries of 'employment relationships' (Cross & Swart, 2022). The figure captures the variety of non-standard workers and highlights the areas in which the dominant viewpoint on the authority/control of HR activities falls short, because it does not apply to the wide range of workers that our work has highlighted. This is a problem for conventional HR theory, for those workers who fall within this void, and is an existential problem for HR that now needs to re-consider its domain of authority/control to address the increasing variety of 'workers' in contemporary organisations.
We recognise some limitations to our study. The study aims to map worker experiences stemming from a number of sub-categories of a specific and underdeveloped stream of agile. As we have commented, earlier, however, there is conceptual slippage in both the definitions of AW and its sub-categories, which creates a challenge in identifying the empirical studies that help us map the worker experience of AW. This presents avenues for future research on the concept and rhetoric of AW and its implications for HRM, which we outline below.
First, the conceptual slippage in both the definitions of 'agile' more generally and each of the various streams therein, including AW and including the sub-categories within each stream, warrants further empirical investigation. That is, which of the six streams of 'agile' , if any, do organisations identify with and what is the level of understanding of whichever variety most closely pursued?
Second, there is a need for research seeking to empirically examine the worker experience with the dimensions of the AW variety, focusing on exploring the reality versus the rhetoric with primary data. Such research can seek to capture and compare the experience of non-standard workers according to demographic characteristics (e.g. nationality, workplace characteristics, and all equality-related characteristics). Crucially, research should focus not only on the disparate array of working practices, but to consider them in combination-as difficult as this may be.
Third, research should focus on the domain of authority/control for HR activities across the range of non-standard workers that we identify in this article. Specifically, to what extent can the HR function and profession claim the aspiration to promote 'better work and better working lives' (CIPD, 2019) if many of those workers, under the practices promoted by AW, fall outside the remit of what HR has jurisdiction over?
Finally, of all AW dimensions, the dimension of workplace location via homeworking has become a prominent feature of contemporary working. Homeworking has previously been associated as a way of working for the 'privileged core' i.e. standard employees. However, in light of COVID-19, homeworking has already become a major shift for HR and is likely to continue post-COVID-19. Therefore, future research can explore if homeworking will indeed become normalised as a 'new' standard way of work, whether there will be different categories of homeworkers as a result of its intersection with other non-standard ways of working (such as in the case of e-Lancers), and what role HRM would play in mitigating the negative outcomes for these workers.

Disclosure statement
No potential competing interest was reported by the authors.

Data availability statement
Data sharing is not applicable to this article as no new data were created or analysed in this study.