The contemporary relevance of Marshall to coworking space communities Open Access

Abstract

Coworking spaces (CWSs) are a relatively new form of industrial organisation which have grown exponentially in the 21st century. Early authors on CWSs have likened them to micro-clusters. Yet this simple analogy may underplay the intricacies of CWS ‘communities’, which differ by type, goals and location with the burgeoning of this organisational form. We seek to show how Marshall’s work on industrial districts (1890, 1919) and that of his Italianate followers, can inform the communitarian aspects that create distinctions across a range of CWS types that exist today. We apply Marshallian/Italianate perspectives to a qualitative study of CWSs in three English provincial regions and find it offers a useful focussing device through which to explain the subtle differences between CWSs types, giving Marshall renewed importance in the current era.

1. Introduction

This paper aims to show how Marshall’s (1890, 1919) work on industrial districts (IDs) and that of his Italianate followers¹ helps us to understand the communitarian aspects of coworking spaces (CWSs)—a relatively new form of industrial organisation, which have emerged within the social and knowledge economy of the 21st century (Micek et al., 2024). CWSs cater for the rise in geographically distributed work enabled by digitalisation and flexible, remote and hybrid working patterns (Heinzel et al., 2021) and have experienced accelerated growth since 2021 (Statista, 2023). By offering shared workspaces and amenities at relatively low cost and on flexible terms, CWSs bring together a diverse range of co-workers (e.g. freelancers, digital nomads, individual creatives, private and public sector workers) in one space, who choose to work within them, together (Jamal, 2018). CWSs may facilitate social and business networking, collaboration, ideas exchange and knowledge-sharing opportunities among co-workers (Moriset, 2017). Critically, they may also foster a sense of community (Capdevila (2013, 2015)—indeed, building a community of co-workers is a core value emphasised within coworking circles (Coworking.org, accessed 6 June 2024).

The early literature on CWSs drew on observations of their presence in large cities and likened their internal dynamics to those of firms operating in much larger business clusters, with some authors describing them as ‘micro-clusters’ (e.g. Capdevila,2013, 2015; Fiorentino, 2019A, 2019B). This literature acknowledged the collaborative activities, social networking and the building of internal communities within CWSs, and since such activities can foster local innovation and entrepreneurship, it is argued CWSs can play a key role in wider placemaking initiatives (Fai et al., 2024). Yet these early depictions of CWSs sit uneasily with comparisons to clusters from either economic geography (e.g. Krugman, 1991; Gordon and McCann, 2000) or business strategy perspectives (Porter, 1990, 1998), both of which tend to articulate the importance of colocation and agglomeration, while largely ignoring any ‘communitarian’ aspects.

Moreover, and largely in response to the increased demand for, and provision of, hybrid and remote modes of work that have ensued since the COVID-19 pandemic, the nature of CWSs, their location and business models have been changing. For instance, several corporate CWS office rental providers (mainly in large cities) have repurposed their space for ‘coworking’, while there has also been growth in independent/small chain CWSs², especially in non-urban settings such as rural and coastal areas (Yates et al., 2024). CWSs differ by type and outlook which may vary across spatial settings, and this has implications for the nature of coworking communities and the local economy in which they reside.

It is for these reasons this paper revisits Marshall’s notion of IDs, and the scholarly work he inspired on Italianate districts in the 1980s–1990s to delineate a scalar of CWS types beyond ‘micro-clusters’ found in urban settings. In doing so, our empirical sample includes CWSs in peripheral, rural and coastal areas, which have largely been neglected in previous studies (Vogl and Ahkavan, 2022). For Marshall, ‘community aspects’ were a focal feature of IDs and the evolution of places. Later, Italian scholars such as Brusco (1982) and Beccattini (1990, 1991) emphasised the ‘softer’ aspects of local production systems such as social capital, reciprocity and trust—intangible benefits embedded within a local community (Konzelmann and Wilkinson, 2017) which are collectively captured within ‘the Italianate model’. By considering the communitarian aspects of Marshall’s work within the CWS setting, we bring greater granularity and distinctiveness to the types of CWSs based on their internal communities and the experiences they offer their users. Importantly, for Marshallian scholars, we demonstrate the continued relevance of Marshall’s thinking on the importance of economic communities in the 21st century through its application to modern modes of work and their facilitation in new organisational forms, and for scholars of organisations, we move beyond simple CWS analogies with the mainstream ‘cluster’ literature and use Marshallian concepts to add further nuance to our understanding of the value and operations of CWSs.

The paper is set out as follows. Section 2 provides an ecology of CWSs, outlining how their users, locations and conceptualisation have evolved from a one-size-fits-all, to a dichotomous categorisation of types. In Section 3, we begin with recent analogies of CWSs as ‘micro-clusters’, before suggesting a more appropriate lens to explore the communitarian aspects of CWSs is to revisit the Marshallian/Italianate ID model. We highlight four key characteristics associated with this Marshallian/Italianate literature, which have conceptual parallels with CWSs communities. Section 4 delineates the boundary of our own empirical study and outlines our primary data and qualitative methodology. In Section 5, we examine our interview findings with CWS managers and co-workers through a Marshallian/Italianate lens. Finally, we offer some further thoughts on how CWSs may also stimulate further discussion of Marshallian concepts in the 21st century, before concluding.

2. Coworking spaces (CWSs)

2.1 Overview of CWSs

There is no universally accepted academic definition of a CWS (Leclercq-Vandelannoitt and Isaac, 2016), although Spinuzzi (2012) describes a CWS as an open-plan office environment, where unaffiliated professionals choose to work alongside one another, for a fee (paid to the CWS provider). CWSs were established (in 2005 by Brad Neuburg in San Francisco, USA) with the purpose of creating communities among economically active freelancers and digital nomads who were working remotely and often alone, professionally (Howell, 2022). The importance of building a work community from the ‘bottom up’ (i.e. by resident users) is supported by the flexible terms (associated with the use of the space) offered by CWSs and the lack of a formal structure, which helps to facilitate social interaction among users. Community is a core value of a CWS and ultimately distinguishes it from more structured organisations such as business incubators and accelerators where residence is limited to set terms and programmes (Howell, 2022).

CWSs encapsulate two dimensions: economic and social. In the economic dimension, CWSs offer cost-effective alternatives to traditional office setups. They are shared workspaces within buildings, have a brand name of their own and normally a skeletal staff—owner, space manager, technical person, possibly an in-house marketer, although in some small, young CWSs all these roles may be undertaken by a single individual. Private CWS providers seek to profit by renovating under-utilised premises and kitting them out with a range of traditional office amenities, including reliable, fast broadband and ‘hot’/dedicated desks, kitchenettes, meeting rooms and rooms/pod/cells for private conversations and offering these shared facilities to users at relatively low prices and on flexible lease terms (Waters-Lynch and Potts, 2017). CWSs appeal to a variety of users including freelancers, entrepreneurs, digital nomads and individual creatives who work independently on their own projects, and employees from different organisations (and sectors) undertaking tasks that can be completed remotely, but all choose to share a space working alongside each other—‘coworking’ (Jamal, 2018). Some CWSs have even begun to cater for corporate clients, who use these spaces as flexible ‘satellite’ offices for their employees (Yates et al., 2024). A common practice within coworking communities (and the scholarly literature) is to address all users of the space as ‘co-workers’ irrespective of their personal business and/or employment situation³.

On a social–societal dimension, CWSs can help to overcome some of the disadvantages of working from a standard corporate office. These may include long commutes into major cities, poor work-life balance and sustainability issues associated with travel (see Fiorentino, 2019A; Burgin et al., 2021; Hölzel and de Vries, 2021; Mariotti et al., 2021). A notional egalitarian status among co-workers can help foster a sense of community and belonging among users. Moreover, coworking can help to mitigate the adverse impacts of social isolation and consequential poor mental health, when working-from-home (Burgin et al., 2021; Merrell et al., 2021; Schifano et al., 2021; Merkel, 2023). The opportunity to network, the potential for knowledge/idea exchange, learning and social capital formation is often a key, (yet usually) implicit part of the advertised CWS offer (Waters-Lynch and Potts, 2017). Indeed, the social value of being part of a coworking community can, in turn, generate economic value, as some co-workers may start to form value chains, operate as suppliers to each other, generate ‘side hacks’ which become new ventures, or collaborate on joint funding bids (Fai et al., 2024).

2.2 The growth and geography of CWSs

Since their emergence in 2005, there have been ‘waves’ of global growth in the number and type of CWSs, across a range of spatial settings (Avdikos and Papageorgiou, 2021). Initially, this growth emerged in urban areas, as CWS providers made use of vacant offices, retail space and disused industrial buildings to provide flexible space for freelancers, and even start-ups (Capdevila, 2015). They became especially relevant in declining US manufacturing industrial clusters such as Detroit, Philadelphia and Pittsburgh, where US policy has sought to support such spaces to revive local economies (Fiorentino, 2019A). Enabled by advances in computing, mobile and digital technologies, which facilitate the freedom to ‘work-from-anywhere’ (WFA), the popularity of CWSs has subsequently risen and they can now be found in every major city in the world⁴. More recently, CWSs have emerged outside the metropoles and in peripheral regions –suburbia, towns, rural and coastal areas—stimulating growth in the work-near-home, as opposed to the work-from-home phenomenon (Foertsch, 2021). This follows the normalisation of hybrid work practices (Fiorentino et al., 2022) following the pandemic where, according to surveys, 68% of people prefer hybrid work (World Economic Forum, 2022), and 74% of employers now offer hybrid work arrangements (Forbes, 28/12/2022). In addition, the ongoing cost-of-living crisis in the UK has seen a rise in knowledge workers moving to suburban areas, towns and villages (Brouwer and Mariotti, 2023; Florida et al., 2023) - albeit the choice of non-urban location is restricted to areas with reliable, stable and fast, broadband connectivity (Burgin et al., 2021; Hölzel and de Vries, 2021).

By 2020, it was estimated there were 19,345 CWSs globally, with a forecast of this trend more than doubling to 41,975 by the end of 2024 (Statista, 2023). The COVID-19 pandemic temporarily halted this trend, as CWSs struggled with much reduced permitted occupancy levels. Yet, fewer closed than expected—many were helped by government subsidy schemes, landlord payment holidays and/or cross-subsidisation by the CWS owner-entrepreneur’s other business interests or used the time to refurbish their premises. Indeed, by the summer of 2020 (after the first wave of pandemic restrictions were loosened), a Deskmag (2020) survey found that 46% of CWSs reported demand for hot desks had increased, 20% reported no change, with 31% reported a reduction. Similarly, 50% or more of CWSs reported demand for individual offices, dedicated desks and team offices had either increased or stayed the same after the first wave of COVID-19 (Fiorentino et al., 2022; Yates et al., 2024). By 2030, hybrid working (which includes the utilisation of CWSs), is forecast to apply to 20–25% of the workforce in advanced countries and 10% in emerging markets (McKinsey & Company, 2021). Post-COVID-19, the global CWS market has been and continues to be strong (see also Coworking Space Business Trends, 2023).

2.3. Types of CWSs and their associated communities

As outlined in Section 2.1, a core value of coworking as an activity, and the physical space in which it takes place, is the creation of a ‘community’ (additional CWSs values include ‘collaboration’, ‘sustainability’, ‘openness’ and ‘accessibility’; see Coworking.org, [accessed 6/6/24]; Gandini and Cossu, 2021). The term ‘community’ is ubiquitously applied to CWSs by marketers and positively linked with narratives of sharing and ‘collaborative work environments’ (e.g. Füzi, 2015, p. 465). However, in practice, the presence of community within any specific CWS and the extent to which they represent coworking values (if at all) will depend on the individual CWS’s type, goals and location (Spinuzzi, 2012; Blagoev et al., 2019; Mariotti and Akhavan, 2020).

In this regard, Spinuzzi et al. (2019) provide a simple dichotomous conceptual distinction between ‘market-orientated’ and ‘collaborative’ CWS communities. The most prevalent are ‘market-orientated’ CWS communities. In these, the ‘sense of community’ may appear somewhat superficial, since the space offers the opportunity for an individual to belong to a group of people who work independently together (Jamal, 2018) (alongside low cost, flexible rental terms and access to professional equipment and services), they do not take up this opportunity and choose to remain highly individualistic, with empathies that align with neoliberal employability agendas and individual survival (see McRobbie, 2016; Avdikos and Iliopoulou, 2019). Any cooperation among co-workers tends to arise through individualistic incentives and actions, notably in seeking to achieve specific (usually commercial) outcomes.

In contrast, Spinuzzi et al’.s (2019) ‘collaborative communities’ exhibit a greater sense of cooperative interdependence and collectivism (among co-workers) with ‘grassroots’ initiatives and joint activities serving to meet a wider set of shared goals. These communities play an important social role, often through the provision of ‘mutual support’ infrastructures (Merkel, 2019A; Crovara, 2023), such as supporting those in precarious employment (Avdikos and Kalogeresis, 2017), and/or facilitating not only (business) production, but also social reproduction activities (Avdikos and Pettas, 2021) including activities like care (e.g., babysitting), in-kind exchanges, mutual aid and volunteer work (Merkel, 2019B; Merkel et al., 2024). Garrett et al. (2017) notes co-workers in these communities engage with each other out of genuine interest rather than obligation, fulfilling a need for social, authentic connections while preserving their autonomy and independence. The communities are based upon reciprocity and shared resources (see Waters-Lynch and Duff, 2021), and foster a sense of belonging (Merkel, 2019B).

Spinuzzi et al’.s (2019) categorisation of CWS communities broadly aligns with the different types of CWS observed across the sector. In this regard, Yates et al. (2024) provide a useful trifurcate typology, distinguishing between neo-corporate/commercial real estate CWSs, independent CWSs and those funded by the public sector/private benefactors (which they consider to be largely insulated from competitive pressures within the broader CWS sector). Neo-corporate CWSs have been described as highly commercialised coworking agglomerations (Gandini and Cossu, 2021), whose primary focus is on maximising real estate returns (see also Fiorentino, 2019B). Examples here include the now-bankrupt WeWork⁵ and/or office space providers such as International Working Group or IWG (Regus and other brands) who, challenged by the post-2008 financial crisis and post-COVID environments, have changed their business models from long-term commercial leasing to flexible space offerings (Echeverri et al., 2021). Their clientele is relatively homogeneous consisting of largely corporate tenants, and/or independent professionals (whose tenancies are paid by their employer) who conduct their daily routines alongside professional peers, largely working in the same sector (Gandini, 2015) working-together-alone or in parallel, as ‘good neighbours’ (Spinuzzi, 2012). Consequently, the extent to which community is created among users of these types of CWSs is relatively limited, more symbolic rather than real (Bandinelli, 2020). Such coworking relationships and dyads are generally transactional and ‘market-orientated’ (see Spinuzzi et al., 2019). The pricing and contractual terms are largely exclusive and do not necessarily cater for smaller individual users, while little collaboration arises and where it does, it tends to be between co-workers in the same or related corporate sector (Spinuzzi et al., 2019). Moreover, when collaboration does occur, it may be inhibited (or not realised) due to frequent changes in spatial layouts by neo-corporate CWS managers, whose strategic priorities to raise occupancy, scalability (of the CWS) and profitability are often detrimental to layouts favouring informal social interaction (see Yacoub and Haefliger, 2024). Additionally, ‘neo-corporate’ CWSs are located, but not embedded, within their local geographic community (Gandini and Cossu, 2021).

In contrast, privately owned independent/boutique-style CWSs (and those funded by the public sector) tend to be more diverse in their outlook extending from profit-seeking alone to filling gaps in the local market and/or addressing wider socio-economic challenges. Some niche providers may have a specialist focus for their space such as promoting social inclusion (Fiorentino, 2019B) and/or supporting female entrepreneurs, and/or catering for community social enterprises (Surman, 2013). These CWSs align more closely with Spinuzzi et al’.s (2019) notion of ‘collaborative communities’. For many independent CWSs, creating a collaborative coworking community among their users is often critical to their appeal, and is important to their own resilience and survival in the competitive coworking market (Waters-Lynch and Potts, 2017). Hence, in these CWSs, the creation of community and generating social impact, both within the space and beyond it to the wider locale, are often at the heart of their raison-d’etre (Gandini and Cossu, 2021).

The types of CWSs (and their communities) also broadly correlate with physical location. Neo-corporate CWSs tend to be prevalent in major cities where they seek corporate clients, while smaller independent CWSs, are more visible in peripheral and rural/coastal regions. The smaller, independent CWSs may also be present in ‘urban villages’ and districts within cities, but it is harder for them to compete (see Yates et al., 2024). Publicly funded CWSs may be found in a variety of locations that are unattractive to private funders, but where the local government authority sees the potential to bring new life to under-utilised municipal buildings and support local co-workers and entrepreneurs (Fai et al., 2023). In a study of three UK city-regions, Yates et al. (2024) worryingly found cost pressures were leading independent CWSs based in both major and even in some lower-tier cities to adopt valorisation business models to attract more (high-paying) enterprise clients. This movement towards a neo-corporate type of CWS may lead to a more concentrated (and less diverse) CWS market within urban areas and could lessen the impetus to build genuine coworking communities with implications for local geographies (Yates et al., 2024; see also Coll-Martínez and Méndez-Ortega, 2023).

Outside the major cities, it is likely the competitive pressures on independent CWSs are weaker since such locations are less profitable for neo-corporate providers, due to a lack of suitable premises and potential corporate clients to attain economies of scale. These locations offer opportunities for independent CWSs to build fruitful communities based on the five original coworking values mentioned above (Gandini and Cossu, 2021). In peripheral regions, CWSs can become embedded within locales and provide a physical and service offering that brings together workers (with skills and knowledge) that might otherwise not meet. Those CWSs that best build specialised communities serving local user needs and contribute to their wider locality are the most likely to survive and thrive (Mariotti and Akhavan, 2020). These may include CWSs established by local entrepreneurs who wish to contribute to the local area in which they reside and provide flexible space for freelancers and start-ups (Capdevila, 2015) or those that support social enterprises with motives to raise local skills (Fiorentino, 2019B), or local chambers of commerce who create CWSs to enhance their offerings of support to members, and university-based CWSs that support (student) entrepreneurial endeavours (Fai et al., 2023).

In summary, CWSs have proliferated and evolved since their emergence in 2005. A simple ‘one-size fits all’ type of CWS does not exist—CWSs vary greatly, but subtly. Their internal communities shape and are shaped by, the type and set-up of the CWS they are in and the broader geographic location in which they reside. As we discuss below, some authors have likened CWS’s to ‘micro-clusters’ (Capdevila, 2013, 2015). However, given the importance of communities, and the role of location in forming those communities for the characterisation of CWSs, this analogy may not be appropriate. Instead, more granularity and detail might be better encapsulated through considering the communitarian aspects of the Marshallian/Italianate model. This may allow for a more accurate conceptualisation of CWSs as existing on a scalar between ‘market-oriented’ neo-corporate providers and (privately/ publicly provided) independent CWSs embodying greater degrees of ‘collaborative’ community, both among its users, and with the wider community in which the CWS is embedded. We consider these issues below.

3. Marshall, economic geography and CWSs

3.1 CWSs as ‘micro-clusters?’

Some authors have described CWSs as ‘micro-clusters’. This analogy is based upon CWSs sharing similar economic traits as observed between colocated firms in industrial clusters (see Capdevila, 2013, 2015; Fiorentino, 2019A, 2019B). For instance, co-workers (each with their own specialisations) share a physical space and facilities and may engage in knowledge exchange and networking within the CWS. In a similar vein, the cluster literature—from both economic geography (Martin and Sunley, 2003) and business strategy perspectives (Porter, 1990, 1998) - emphasises how cluster firms may access an ‘industrial commons’ of local public goods such as R&D support (Harrison, 1992; Pisano and Shih, 2009, 2012), can benefit from industrial specialisation, agglomeration economies and the positive knowledge externalities arising from their colocation (Krugman, 1991; Gordon and McCann, 2000). Capdevila (2013) is the most explicit in the comparison, writing that CWSs in Barcelona ‘present similar knowledge dynamics as the ones identified in industrial clusters, but at a lower scale’ (ibid., p.1).

The origins of the cluster literature lie in Alfred Marshall’s (1890, 1919) notion of IDs, where the proximity of firms in the same industry benefitted from the ‘localisation’ economies of specialised labour markets, local suppliers and an ‘industrial atmosphere’ of local knowledge spillovers (Marshall, 1923, p. 284). Yet, while contemporary cluster studies embody these core Marshallian concepts, critically this literature has downplayed and even ignored Marshall’s recognition of IDs as not just being an economic mode of production, but as a socially organised one of local communities of people. Marshall’s heightened cognition that IDs were an interpenetration of productive forces and communities bounded by place, wherein ‘social forces here co-operate with economic’ (Marshall, 1920, Book IV, Ch. X, p. 226) is critical to why our application of Marshall to describe types of CWSs, offers value and understanding over and above their analogy to clusters as depicted so far (e.g. by Capdevila, 2013, 2015) and Fiorentino, 2019A, 2019b). Indeed, the ‘industrial atmosphere’ was observed not only in production but, for Marshall, also reflected the sociability of a district’s wider community and evolved with different organisational and social contexts (see Whitaker, 1975; Becattini, 2002). As Konzelmann (2023, p. 139) notes Marshall observed IDs ‘occupy both geographical and social spaces, with their own economic and social histories, structures and prospects’.

In contrast, while some of the cluster literature has begun to recognise the value of weak ties emerging via inter-organisational social trust and reputation (Maskell and Lorenzen, 2004) which are a key feature of Marshallian IDs, for the most part ‘models of [cluster] agglomeration assume no form of co-operation between actors beyond what is in their individual interests’ (Gordon and McCann, 2000, p. 157) or where inter-firm ‘complementarities’ arise, they are purely through business activity (Porter, 1998). In business clusters, there is limited interaction between cluster firms and local citizens (Capdevila, 2013). Meanwhile, networks and dyads are typically formal and organised by corporate firms, such as in ‘hub and spoke’ clusters where large vertical integrated corporates are supported by suppliers, and/or in satellite business clusters dominated by externally owned and headquartered firms (Markusen, 1996; Porter, 1998). Capdevila (2013) observed similar inter-organisational structures in his Barcelona study, viewing CWSs as smaller organisations subservient to larger corporates in the area, either on the basis of relative costs or as hubs of innovation—CWSs are ‘a way of externalising the innovation processes from the large organisations thus reducing the high costs that an equivalent exploration effort would represent if they would take place within large firms or research centres. In the same vein, the risks associated with high explorative practices are also absorbed by the micro-clusters’ (ibid., p.12). The lack of ‘community’ in the cluster model is evident and may account for why early studies of CWSs considered them to be akin to ‘micro-clusters’.

3.2 Marshallian/Italianate districts and CWS communities

Given the central importance of community building to their original purpose and concept, arguably CWSs are better explored through the lens of the communitarian aspects of Marshall’s (1890, 1919) IDs model. In this regard, it is useful to draw on the scholarly narratives about the Italian districts—located in central and north-east Italy (the so-called ‘Third Italy’) - during the 1980s/1990s, which revived interest in, and held explicit reverence for Marshall’s earlier work (Becattini, 1990). As well as considering the impacts of colocation, the Italianate scholars emphasised the ‘softer’ aspects of local production systems—social capital, reciprocity and trust—intangible benefits embedded within a local community (see Schmitz, 1999; Konzelmann and Wilkinson, 2017). Marshallian communitarian traits are prevalent throughout these accounts. For Brusco (1990, p. 16) ‘production in the industrial district is carried on as a social process by virtue of a social structure that encourages interaction’. Similarly, Becattini (1990) considered the Italian districts as a ‘merged’ community of both people and firms bound together with socio-economic development moving together in an integrated process. Bianchi (1993) echoed similar sentiments, noting the lesson from the Italian districts is that local, endogenous growth is very much a community-driven process (see also Bellandi, 2003).

The colocation of users of CWSs in a physical space (a building) confers benefits to the users in a manner which parallels the colocation benefits of firms within an industrial district. The sense of community created among co-workers in the CWS similarly parallels the role of the ‘softer’ intangible benefits embedded in the ID community and just like IDs, the strength of this community will vary from location to location. Moreover, particularly in peripheral/rural/coastal areas, often CWSs may have a social mission to help not only their direct users but also to connect to and support the wider local community of their geographic place (see Section 2.3; Fiorentino, 2019B), echoing the sentiments of Brusco, Becattini and Bianchi about the mutual support within the context of Italianate ID communities.

We note, that in the late 20th century, the fabric of inter-firm relationships within Italy’s IDs changed in response to globalisation and remains today. Now regions often centre on anchor firms that emphasise global outsourcing, while the tradition of ‘familism’ (family firms and friendly ties) has declined (Sacchetti and Tomlinson, 2009; De Marchi and Grandinetti, 2014). In doing so, the Italian IDs have moved to a more business-cluster type model and may be increasingly indistinct from clusters in academic and policy literatures (McCann, 2009; Rocha and Audretsch, 2022). Nevertheless, there remains a rich literature on the experiences of the Italian districts during the 1970s through to the 1990s which is of relevance to the CWSs of today. Space precludes a full exposition (for a valuable and insightful review, see Konzelmann and Wilkinson, 2017), although of pertinence to our study the Marshallian and Italianate District literatures identify four key Marshallian features summarised below:

• i) An ‘industrial commons’ of quasi-public goods (Bellandi, 2009)⁶: a common pool of infrastructural and other services [such as R&D support] & knowhow available to all district firms and provided by local trade associations and/or by municipalities (Harrison, 1992; Schmitz, 1999; Pisano and Shih, 2012).

• ii) A socio-economic community: firms and actors rooted in their locality and communities of people which creates a sense of ‘belonging’ to the community, based upon the shared history and culture of the district, with socio-economic development based on people and firms working together in an integrated community-driven process (Becattini, 1990; Bianchi, 1993; Bellandi, 2003).

• iii) Dyads, networks and collaboration: a web of informal inter-firm subcontracting, collaborative and ‘reciprocal’ relationships among district firms, embodied by ‘joint actions’, co-learning and technology and knowledge sharing, often facilitated/supported by municipalities and local trade associations (Brusco, 1982; Schmitz, 1999).

• iv) Knowledge sharing, creation/co-creation: an ‘industrial atmosphere’ (Marshall, 1919) of tacit and codified knowledge, which generates new ideas and innovation. IDs have been described as a ‘creative milieu’, a source of creativity, learning, innovation and entrepreneurship (Becattini, 1991; Schmitz, 1999).

As much as these Marshallian characteristics underpinned the communities of the Italianate districts, we will provide evidence that they are relevant constructs for analysing the communitarian aspects of CWSs today. Over the last decade, and since the early analogies with the cluster literature, the CWS market has grown and changed in terms of CWS types, with a broader range of users, and increased geographic dispersion from urban to non-urban settings (see Section 2). This has had implications for the nurturing of CWS communities—their breadth and depth of the benefits they bring, to both internal to CWS connections, and connections to the wider locale in which they are situated, and as such, our understanding of a CWS as a type of ‘micro-cluster’ needs to evolve to better capture greater sensitivities about their roles as (socio)economic actors.

4. Methodology

We explore the characteristics of CWSs and their communities within three English Local Enterprise Partnership (LEP) bounded regions (outside London and larger metropole cities): Staffordshire and Stoke-on-Trent—SSLEP, Heart of South-West—HoSWLEP and Thames Valley Berkshire—TVBLEP. These were chosen based on their distinct socio-economic geographies and associated challenges. We deliberately chose CWSs in peripheral, rural and coastal areas; these types of areas have largely been ignored in previous CWS studies (Vogl and Ahkavan, 2022), but have recently seen growth in CWS activity (see Section 2.2). Thus, our sample generated qualitative data from CWSs located in an old North Midlands industrial region seeking to reverse a period of long-term decline, a second-tier vibrant city in the South-West, and several CWSs in rural towns (some within commuting distance of London) and coastal areas.

Between February and May 2023, we visited 8 CWSs across all three LEP regions (an additional CWS manager interview (CWS 7) was conducted via Microsoft Teams in August 2023). Cases were ‘selected in proportion to one or more characteristics in the population’ (Gorard, 2013, p. 81), with a relatively even mix of CWSs in rural, coastal and lower-tier urban locations allowing for a diversity of research participants, which is important for accurately characterising CWSs and maximising internal validity (Henderson et al., 2002; Bryman, 2012; Billing and Bryson, 2019). Our sample predominantly included owner-manager independent CWSs, some publicly funded CWSs and two partnership-type CWSs (where independent CWSs partnered with a national CWS operator for marketing/administrative support)⁷. In total, we conducted 12 interviews with CWS owners and/or managers and ran 11 focus groups with CWS users⁸. The focus groups facilitated wider deliberation and reflection (among users) on the key issues, which complemented the data insights obtained from the more in-depth manager interviews (Stokes and Bergin, 2006). Tables A1a-c in the  Appendix provide a detailed (and anonymised) summary of the CWSs visited, and study participants.

Both the interviews and focus groups were held at the CWS premises and lasted between 60 and 90 minutes. Topics covered the characteristics and communities of CWSs—specifically, the nature of CWS dyads and networking, knowledge transfer and co-learning among users, and the wider role of CWSs in reviving regional economies. For the purposes of this study, our attention is on those responses relating to CWS communities. Our sample of different CWS types in peripheral areas, towns and lower-tier cities thus allows us to present a broader perspective on how they resemble the communitarian aspects of Marshallian IDs.

The interviews and focus groups were conducted by the research team, digitally audio-recorded and professionally transcribed. The research team were aware of the danger of response bias, so to ensure reliability and validity in the data collection, we adopted appropriate questioning techniques (see Eastery-Smith et al., 2002; Yin, 2014). This included issuing each participant with an information sheet, and a consent form and providing assurances about the anonymity of interviewees.

Following transcription, we used a directed content analysis (Hsieh and Shannon, 2005; Bowen and Miller, 2022) and organised the data around the four core constructs (or characteristics) associated with Marshallian/Italianate district communities (as highlighted in Section 3.2); industrial commons, socio-economic community, networking, dyads and collaboration and knowledge exchange, creation/co-creation. As scholars of Marshall recognise, these constructs are not to be considered in isolation—indeed, they coalesce, support and re-enforce one another. Nevertheless, they each offer the opportunity to observe CWS communities through a Marshallian/Italianate lens.

5. Insights from interviews and focus groups

Our interpretative observations are summarised and presented in Table 1 which also provides an overview of the core features of both neo-corporate and those (private and publicly owned) independent CWS which embody ‘collaborative communities’ (see the discussion in Section 2.3). The neo-corporate features are largely derived from prior research on CWSs, both academic (e.g. Yates et al., 2024) and ‘grey’ (e.g. coworking industry reports such as Coworking.com) literatures. The features of independent CWSs are derived from participant observation from our fieldwork⁹. For conceptual clarity, Table 1 presents the extremes of a scalar of characteristics, although in the real world, both neo-corporate and independent CWS may exhibit different mixes and degrees of each characteristic (and activities) associated with the other (see also Spinuzzi et al., 2019; Yates et al., 2024).

5.1 Industrial commons

The equivalent of the Marshallian ‘industrial commons’ (Harrison, 1992; Pisano and Shih, 2012) within CWSs are the ‘communal’ facilities available to all CWS users (Howell, 2022). Typically, they include a kitchenette and wi-fi, and (depending on the CWS) access to computer monitors, printers, mailbox services and private meeting spaces although the provision and quality of such facilities across each CWS varies. Some CWSs (known as Fabrication Laboratories - FabLabs¹⁰) also included equipment such as lathes, computer-aided design software, 3D printing and augmented/virtual reality (AR/VR) equipment; items which are too expensive/experimental for individual users/micro-firms to purchase, but which they can utilise within the space and familiarise themselves with before making an investment purchase.

As with Marshall’s IDs, the utility of such facilities depends upon the availability and quality of additional support services (and activities) which formulate the commons. These additional services are not generally a feature of neo-corporate CWSs, which seek to offer only core office functionalities within a ‘corporate setting’ (Yates et al., 2024). However, across our sample of independent CWSs, we observed a range of support services from technical guidance on equipment and digital platforms to advice on business operations (such as finance, marketing and exports). Some CWSs also offered a range of training sessions (e.g. in general support of business growth through to specific skills on specific digital platforms) to their users on an ad-hoc to regular basis (depending on demand). Some hosted public events with opportunities for CWS users to ‘showcase’ their businesses to the wider community.

These additional services and activities were clearly part of the wider offer of independent CWSs with collaborative communities and were tailored to local demands to support CWS users and their businesses—in a similar vein to some of the quasi-public goods available in IDs (Bellandi, 2009). Such supports were not universal across our sample, and just as with IDs, the fertility of the ‘commons’ varied across different CWSs. In some cases, support services came as part of a package associated with membership in a public/business organisation, such as a Chamber of Commerce (offering its own CWS). Or they were integral to specific types of CWSs, such as a FabLab or part of the CWS offer from more upmarket and/or niche providers.

Pivotal to the successful functioning of the ‘commons’ was the presence of a CWS manager playing a role similar to that of a trade association/industry body within an ID, by acting as the gatekeeper, facilitator and organiser of support services and CWS activities. This was especially evident in those CWSs where a manager was in situ daily¹¹. Typically, CWS managers will ‘look after, all the maintenance, health and safety issues, fire alarms and everything, we look after them [CWS users], everybody in the building and any requests with additional IT support’ (CWS4/Mgr/SS). In addition, a proactive CWS manager will often ‘ask the community [of CWS users] what is holding you back in your business. What do you not know? What can we help you with?’ and look to host appropriate workshops and training events to address particular challenges (CWS6/Mgr4/HoSW).

5.2 Socio-economic community

The vitality of the ‘commons’ reflects the sense of community within CWSs. As in districts, community building involves nurturing trust and other facets of social capital (such as reciprocity, and shared values) among CWS users through ongoing social interactions and engagement with the space—the communitarian factor (De Marchi and Grandinetti, 2014; Mariotti and Akhavan, 2020). Facilities are important in this ambition, and across our sample, there were examples of CWSs providing high quality, often ‘funky’ social spaces, such as common rooms and/or ‘playrooms’, to enable social interaction and ferment new professional and communitarian relationships.

However, community vibes differ widely across CWSs. In neo-corporate CWSs, more market-orientated communities are common. Relationships in these CWSs may be more impersonal, transactional and transient (see Spinuzzi et al., 2019): ‘I’ve worked at some [neo-corporate CWSs] recently and just walked in thinking, this is just desks with no soul’ (CWS6/Mgr3/HoSW). However, within our sample of independent CWSs, there was a strong sense some CWSs were seeking to develop communities with shared values and a ‘collective identity’, which co-workers engaged with wholeheartedly and sought to shape (see also Garret et al., 2017). Expressions of ‘collective identity’ and shared values, customs and norms resonate with the core traits of Marshallian districts (see Sammarra and Biggiero, 2001).

Again, CWS owners (and managers) played a key role in community building and cultivating a sociable environment (see also Merkel, 2019B) by acting as conduits in connecting users, through direct (one-to-one) introductions and/or organising social and work-related events to ‘create those sorts of “spark moments” for people that don’t normally encounter each other’ (CWS7/Mgr/HoSW). These included regular ‘lunch and learn’ and ‘biscuit club’ activities, where CWS users were encouraged to share (and learn from) their (often different) work experiences within an informal setting.

‘The thing I was most proud of was the community that was building […]. I could create this [space], but I couldn’t dictate who came… it was almost like this space had a kind of energy that attracted a certain kind of people and there is this real lovely community feel... a real generosity of spirit amongst the people, and that’s something I couldn’t have [created] with the best business plans in the world’ (CWS8/Mgr/TVB).

‘Yes, community is core. You get a sense of what they are when you walk in [to a CWS][…] it is a sense the actual community has been established by people […] we’ve always described trying to build a culture more than trying to build set-desks in a space, and it’s the cultural impact of what we’re doing is the bit that matters[…] that’s the sort of the backbone of the ecosystem building’ (CWS6/Mgr1/HoSW).

More widely, nurturing a community vibe was generally more purposeful when CWS users lived and worked in the same locale, where they could share similar experiences, interests and/or backgrounds, including a sense of local pride (see also Mariotti and Akhavan, 2020), where CWSs were truly embedded within, not just located in, a place. Indeed, some CWSs chose to be selective over tenants who were unlikely to contribute which in some cases meant the exclusion of digital nomads (for whom the CWS notion was originally conceived):

‘I’m trying to build a community, so I want people who are going come in all (or most of) the time. Not ephemeral people who just need a desk for today, and then they’re off to the next place. I want the people who I can get to know, I can recommend them, connect them to other people…I don’t just want (or take) the desk bookings. I want (to select) the right people from my local area’ (CWS6/Mgr3/HoSW).

These examples highlight an interpenetration between business and social domains within a physically circumscribed place (the CWS), again drawing comparisons to the embeddedness of Marshallian/Italian IDs, where the activities of businesses, workers and residents in daily life overlap (Becattini 1990; see also Tomlinson and Branston, 2018). A related issue is that while large city-based neo-corporate CWSs tend not to engage with their wider locale (see Fiorentino, 2019B; Yates et al., 2024), across our sample, several CWSs actively sought to do so through a range of outreach activities. These included promoting local charities, hosting wider community events, offering work placements to locally based students and supporting the local business community.

‘[We host] community events so that we can contribute back into this area and give people an experience of things that they might not necessarily be able to access’ (CWS8/User1/TVB).

‘We’re partnering with a couple of charities. The revenues that we’re bringing in is then feeding into that local charity’ (CWS3/Mgr/SS).

In addition, the presence of a vibrant CWS can bring much-needed custom to local high streets (and/or in rural villages), while CWS managers are often keen to prioritise local suppliers (for example, catering or stationary supplies, or premises maintenance). Some CWSs also participate in Business Improvement District (BID) schemes¹² and/or other local stakeholder initiatives to enhance their neighbourhood. Where CWSs undertake these types of external activities, they become more embedded within their wider locality (see also Gandini and Cossu, 2021).

5.3. Networking, dyads and collaboration

Inter-firm, intra-district collaborative ties and dyads are a critical component of the ID model; they underpin the socio-economic community and district dynamism. In CWSs, opportunities for serendipitous networking, dyads and collaborative ties also exist through the informality of the premises’ social and workspaces, and other in-house activities (such as organised workshops/events). Networking opportunities arise in both neo-corporate and independent CWSs, although there are subtle differences. For instance, within independent CWSs, managers (as noted in Section 5.2) will draw on their background knowledge of their tenants (and their skills) and take a more proactive role in arranging introductions. In addition, compared to neo-corporates, independent CWSs outside core cities are likely to host a greater diversity of co-workers (in terms of skills/expertise) since they do not have the density to attract tenants from any single industry, as they might in large cities. This will impact the nature of collaboration, support and advice sharing arising within the CWS.

‘why coworking? Well, there’s expertise in other areas. So, if you need help… there’s always someone good at social media or digital marketing… HMRC…legal etc’ (CWS6/User4/HoSW).

‘we run a program with a bank [to empower women]. One of our members [on the program] has just established a coworking group (slightly more networking than coworking) …. they meet once a month, and basically, they get together and have a nice lunch…they talk about what’s working and what’s not working (CWS1/Mgr/SS).

Occasionally, more formal collaborative networks arose between CWS co-workers, and—like coopetition within Marshallian districts—sometimes this comprised co-workers who work for, or own, rival organisations. This was typically the case where co-workers in the same or related sector (e.g. defence or cyber-security industries) sought to collaborate on joint funding bids for government or research funding contracts. Here, the proximity of co-workers and the informality of the CWS and its community were key to the collaboration.

Most co-workers also have their own external business and social networks, which can be extended to other co-workers through interpersonal networking within the CWS. This has echoes of district firms (and actors) engaging in external networks (such as supply chains) and establishing links to the district (Bathelt et al., 2004). However, in the CWS (relative to IDs), such connections are typically much looser, ad-hoc and informal. Nevertheless, they sometimes open wider opportunities for CWS users; for instance, in terms of access to training courses, conferences and additional expertise.

‘So we will put our people [CWS users] in touch with other work hubs in Devon and also business support centres… we work quite closely with XXX down in Exeter and we get them to help with business support, but we also work quite closely with another [nearby] coworking space up because they do start up programmes which they offer free training and for which they’ve had funding’ (CWS5/Mgr/HoSW).

In some cases, external links can enable more formal collaborative actions and support:

‘we had a case where a company was falling apart, and it was turning a bit nasty… the company was spread across three different cities, but a lot of [it’s operations] were in our CWS… and people’s livelihoods and shares, investments were on the line. So, we sort of leaned into our networks and we pulled in some quite experienced CEOs to come in and help them negotiate their way through the contract and that was all done voluntarily. They came in and got experts to come and fight [and successfully resolve the situation]’ (CWS6/Mgr2/HoSW).

5.4 Knowledge exchange, creation/co-creation

Networking and new dyads feed into knowledge exchange and creation/co-creation of new business opportunities, epitomising successful IDs. The ad-hoc nature of new dyads and network formation within CWSs is also manifested in knowledge exchange. Where knowledge exchange arises, it tends to relate to informal advice and technical guidance between CWS tenants to assist on specific tasks (e.g. advice on software). This may facilitate learning among CWS tenants that can enhance their own productivity or help establish a baseline of knowledge that provides sufficient absorptive capability for the individual to understand what they need to be aware of when contracting a supplier to provide a product/service. Indeed, as in IDs, CWS users can benefit from informal conversations, gossip and knowledge flows from ‘just being there’ (Gertler, 1995). CWS managers again play a key role in facilitating communications between users; ongoing discourse between users can reduce the cognitive distance between them to enable fruitful knowledge exchange (Boschma, 2005; Mariotti and Akhavan,2020).

Marshallian-type knowledge flows are more likely observed in neo-corporate CWSs, where the clientele is largely drawn from the same and/or related corporate and professional sectors (Neffe et al., 2011). Conversely, in independent CWSs, the greater diversity of co-workers from a range of sectors, and who may well reside in the locality (but work remotely from the CWS for companies elsewhere) means that any knowledge exchange will largely be cross-sectoral and of the Jacobs (1969) type¹³. Across our sample, there were several examples of CWS acting as platforms to obtain informal feedback (and fresh thinking) on new ideas.

‘it’s also about sharing some ideas that are maybe outside of your normal thinking or even like what you’re looking at and exploring how other people’s reactions towards what you’re thinking is.... you can just organically sort of try ideas out with people’ (CWS1/User2/SS).

‘it’s nice to interact with other companies … you learn a lot from just hearing stuff in the background and having different conversations, but it’s really nice to do that outside of your immediate industry as well. I think there’s a lot of value to get from that which you wouldn’t necessarily get in an office’ (CWS8/User3/TVB).

In some cases, this can lead to new product development and/or ‘side-hack’ business opportunities. For instance, in one case (CWS6/User3/HoSW), a CWS user (working in crypto currencies) was co-developing a new NFT project (non-fungible tokens) with a fellow CWS user who was a freelance illustrator and designer. The project involved combining the former’s knowledge of crypto markets with the illustrator’s expertise to design the tokens. In our sample, such cases arose on an ad-hoc basis, although it does illustrate the potential for CWSs to foster entrepreneurial activity—in a similar vein to how in IDs, some employees (often in SMEs) would develop new ideas/products and spin-off their own firms. Occasionally, knowledge sharing arose where CWS users collaborated (with each other) on joint projects that have non-proprietary objectives such as joint funding bids; the heterogeneity of users’ knowledge bases/expertise helped to strengthen such collaborations¹⁴.

While knowledge-sharing and learning arise within CWSs, it is not all-pervasive; not all CWSs will have a knowledge-sharing culture or community. CWSs will reflect knowledge communities embedded within their own settings. As with the unique ‘industrial atmosphere’ of districts, the vibe/buzz across CWSs differs. CWS users will also tend to guard proprietary knowledge when in the vicinity of other users from rival firms and utilise private phone booths and meeting rooms to do so or choose to come in on days when the rival does not.

‘I’m quite cautious because some of my work will overlap with other people…. they work with tech firms, and I work with tech firms, so I would be very cautious about talking or being on a call with one of my clients because you could be heard’ (CWS8/User6/TVB).

Most CWSs are also relatively young (i.e. in our sample, between 2 and 10 years old); as Marshall noted, the development of an ‘industrial atmosphere’ can take some time to emerge (see also Belussi and Caldari, 2009).

6. Conclusions

It is over 130 years since Alfred Marshall (1890, 1919) first introduced the concept of IDs not only as an economic unit of analysis but also as a socially organised mode of production. Since then, scholars from across social science have studied and observed their evolution and brought new insights into their historical, socio-economic and spatial significance. However, in the late 20th century, most studies of the spatial agglomeration of firms—now referred to (and popularised by writers such as Porter, 1998) as ‘business clusters’—have largely ignored the community aspects of Marshall’s districts, focussing instead on their economic outcomes. While the experiences of Italian districts demonstrated the socio-economic vitality of places is underpinned by communitarian characteristics (Piore and Sabel, 1984; Becattini, 1990) these too have lost their familial characterisation under the pressures of globalisation, and for a time in the late 20th to early 21st century, the relevance of Marshallian communities was much diminished. One contribution of our work is the detailed application of Marshallian ideas on the importance of community to the increasingly popular CWS as a unit of analysis in the 21^st century, reinvigorating perspectives that have died since their earlier revival in the study of Italian IDs.

Since their emergence, CWSs have undergone significant growth in numbers, and have evolved in terms of their users, aims and missions and proliferated into new locations. Consequently, the conceptual identity of CWSs has become increasingly blurred and their early comparison to micro-clusters (Capdevila, 2013) becomes less useful, by only capturing the more economic-based characteristics of CWSs which maybe more strongly reflected by CWSs in urban settings, and potentially those with a corporate nature, and of little resonance for smaller CWSs outside of large cities and city-regions. Perceptions of CWSs as ‘micro-clusters’ have also been re-enforced by empirical studies that have mainly considered neo-corporate CWSs in large urban areas. Instead, our sample largely consisted of independent CWSs located in provincial cities and rural and coastal areas. Hence, our second contribution is to evidence the variance with which these CWS providers and CWSs reflect a combination of core Marshallian/Italianate ID characteristics—in terms of an industrial commons, cultivating a socio-economic community, interpersonal networking/multiple dyadic relationships, and knowledge exchange and collaboration—and to place them at different points along a scalar of CWSs (see Table 1), as opposed to sitting within a dichotomy of CWS communities (Spinuzzi et al., 2019), or within a trifurcate categorisation (Yates et al., 2024).

Returning to Brad Neuberg’s (the founder of the first CWS) original purpose of building a community for and among its users, then utilising a Marshallian lens as a focussing device enriches our understanding of different CWS communities, and the nature of their users’ social and business networks that contribute to the vitality of the space, and of the area in which they are located. This, of course, has implications for CWS managers, users and regional development policy. It is beyond the scope of this paper to offer a full exposition, suffice it to note that for CWS managers, a deeper understanding of CWS communities may provide clarity on the extent to which they could/should differentiate their CWS, based on their own preferences to develop depths and types of user experience and community building (within and beyond the space). In addition, the range of experiences provided by different CWSs enhances a potential user’s ability to find a CWS with the right ‘fit’ for their requirements. On policy, CWSs may open possibilities for revitalising regional socio-economic place-based growth. This is especially the case if CWSs foster communities of innovation, entrepreneurship and start-ups in a bottom-up dynamic, which, when accompanied by a willingness to experiment, ease of access to digital and other technologies available in CWSs (which may otherwise be beyond the means of individuals or micro-to-small firms) can lead to the generation of start-ups, and local growth (see Fai et al., 2024).

Finally, in drawing these conclusions, we offer the caveat that our study is methodologically limited by its focus on a relatively small study of CWSs from across three English regions (all outside London). While our findings cover a broad snapshot of perspectives, we are aware they are confined by a short time window and the chosen geography. Future work might look to increase the number of cases, broaden the geographical spread to possibly include international comparators or consider a longer time window.

Conflict of interest statement:

None

Bibliography

Footnotes

Appendix