1 Introduction

Coworking spaces (hereinafter, CSs), defined as “shared workplaces utilised by different sorts of knowledge professionals, mostly freelancers, working in various degrees of specialisation in the vast domain of the knowledge industry ([21]: 194), have become a global phenomenon since the first CSs opened in San Francisco in 2005. In particular, between 2015 and 2020 the number of CSs increased by 200% and CSs users increased by almost 400% [12], with a total of 26,300 CSs and 2,600,000 CSs users worldwide at the end of 2020.Footnote 1 However, this growth has been limited due to the COVID-19 pandemic and the consequent waves of lockdowns which led some CSs to limit or cease their activities and services. Despite that, the flexibility of CSs still seems to be attractive to some users [25].

As regards the location patterns of CSs, although CSs mainly remain an urban phenomenon concentrated in large cities worldwide, the demand for CSs in rural areas is increasing [10], and this increasing demand in rural areas has become a new concern for regional policies. Despite the increasing importance of CSs, evidence on their location patterns and drivers is limited. We can find some exceptions at the urban level in studies by Moriset [26], Mariotti et al. [29], Coll-Martinez and Méndez-Ortega [7] or [27], and at the rural level in studies by Capdevila [5], Mariotti et al. [24] or Tomaz et al. [36].

This chapter adds new evidence to a growing literature on the CSs phenomenon [21] by using Geographical Information Systems (GIS) and distance-based measures to analyze the agglomeration of CSs without considering the administrative borders at the country level. Particularly, the Kd function provides information on the CSs localization, that is, the tendency for CSs to cluster relative to overall economic activity at a given distance. Concretely, the Kd function compares the distribution of distances between pairs of establishments in a given economy to the distribution of distances in hypothetical industries with the same number of establishments randomly distributed across the area under consideration [14, 15].

The focus is on Spain, which ranks as the fourth top country in terms of number and capacity of CSs, only behind the United States, India, and the United Kingdom [17]. Spanish figures are mainly explained by the data recorded mainly in the cities of Barcelona and Madrid. The Catalan capital is one of Europe's most important creative hubs in terms of knowledge-based, creative, digital, and sharing economy [2, 22]. Moreover, the Catalan capital has been recently been highlighted for its potential as the European city with the greatest growth margin in terms of CSs’ creation in the coming years [10]. Regarding Madrid, it is considered one the European cities with the highest potential for coworking with reference to four different factors, i.e., scale, business, environment, people and catalyst. In 2019 Madrid ranked 10th place with a European Coworking Hotspot Index of 100 [9]. Despite the importance of the two larger cities in Spain, there is little evidence about the general location patterns of CSs in urban, periurban, and rural areas in Spain [5]. Against this background, our work provides notable implications. Analyzing the agglomeration of CSs from a continuous space point of view allows us to identify whether this phenomenon, attracting the most innovative start-ups and creative freelancers, tends to cluster beyond urban areas and how intense its agglomeration is in the Spanish geography and illustrate the potential challenges of this phenomenon.

Our main results confirm that CSs are highly concentrated in the most populated areas of Spain, since these areas offer greater chances to meet customers and suppliers, proximity to urban amenities, and with their specific image and reputation add to those of the individual CSs. Another reason is that they offer freelancers the opportunity to operate in the most vibrant city areas while paying competitive fees. Thus, the results show that (i) CSs are highly concentrated in the most central areas of Spain: Barcelona and Madrid; and (ii) CSs are significantly agglomerated at short distances (70–90 km), and this agglomeration rapidly disappears as distance increases. This confirms that the CSs’ location is still a urban phenomenon in the country.

The remainder of the chapter is structured as follows. In section 2, we review the main factors behind the location of CSs. Then, in section 3, we present our methodological approach and data. In section 4, we analyze the location and agglomeration patterns of CSs in Spain. Finally, in section 5, we discuss our main conclusions.

2 Literature Review

CSs emerged in a context the proliferation of three interlinked movements: the creative economyFootnote 2 [18], the spread of information and communication technologies [30], and the sharing economyFootnote 3 [3] as ‘serendipity accelerators’, designed to host creative people and entrepreneurs who endeavour to break isolation and to find a convivial environment that favours meetings and collaboration” [29].

Even though there is no official definition for such an innovative workplace, several definitions of CSs have been proposed.Footnote 4 Herein, according to the aim of this chapter, we understand CSs as in Mariotti et al. [26], (pp. 6): “Coworking spaces are innovative workplaces where independent knowledge-based, creative, and digital workers—mainly freelancers or self-employed professionals—share their work spaces […]”.

The CSs phenomenon is the subject of public debate. On the one hand, the new working practice comes with some risks related to a potential coworking “bubble” and real estate speculation [26, 29], leading to increasing rental prices, gentrification or increasing inequalities in the core neighborhoods of big cities. This is especially relevant in a context that lacks clear rules and regulations for the housing and labor markets to effectively control all potential interactions, uses, and conflicts arising from the use of private buildings as CSs. On the other hand, CSs are seen as a strategic tool to facilitate the development of creative cities as well as peripheral areas by reinforcing the concentration of high-skilled creative workers. CSs are seen as drivers of revitalization, community building, and improvement of surrounding public spaces [18, 19, 26, 29].

Even if there is a clear preference for CSs to be located in urban areas, given the advantages of agglomeration economies, there are some examples of successful semi-private initiatives that developed CSs and coliving spaces in less populated areas.Footnote 5 However, the increasing attraction to urban areas should make us think about the fact that in the next years, more than 70% of the population and jobs will be concentrated in urban areas. This is the case in Spain, where most of the population lives and works around the two major cities, Madrid and Barcelona, and to a lesser extent around the other regional capitals, which highlights the serious problem of depopulation in rural areas. Thus, rural population drift is currently a great challenge for policy-makers in most countries, due to the lack of territorial cohesion arising from this phenomenon.

However, the spread of new technologies that facilitate teleworking and the increasing phenomenon of the creative economy may facilitate the gradual success of using CSs and cohousing spaces in rural areas. Moreover, the availability of more affordable housing in smaller cities and villages, even with entrepreneurial projects linked to housing rehabilitation, and the reduction of commuting are also advantages in terms of living and home-based business conditions [4, 32]. In this regard, new generations, characterized by college-educated professionals in their mid-twenties and their late-thirties who primarily work within the creative industries, such as web development, graphic design, and programming, or new media [20], and look for alternative lifestyles, may be willing to locate in rural areas because of the inspiring and slow-way of life that is typical of these places.

Most of the studies in Spain focus on the two largest cities, Barcelona and Madrid. Capdevila [5] focused on the Catalan case through the analysis of the network of coworking spaces in the region and showed that the diffusion of the practice of coworking from urban to rural areas is not a replication but is an adaptation to a new context. In Madrid, Alonso-Almeida et al. [1] considered how the presence of coworking operators and spaces is changing the way people go to the office in Madrid and argued that coworking spaces help to work safely and overcome the problems related to working at home.

Because of all the above, this phenomenon becomes a new concern for regional policies, since CSs may have several implications for the daily life of inhabitants and bring new regulatory challenges.

3 Data and Methods

3.1 Data

The data used in this chapter about CSs in Spain comes from different sources. Firstly, we used data at Eurostat NUTS3 level (province level in Spain). The NUTS3 socioeconomic information (i.e., population) is based on INE data (the Spanish statistical service).

Data on Coworking spaces (599 spaces) was gathered as part of the COST Action CA18214 “The geography of NewWorking Spaces and the impact on the periphery” based on the data available from coworker.com website. All information refer to 2021.

3.2 Methods

This chapter uses a combination of Geographical Information Systems (GIS) and distance-based methods (concretely the Kd function) to analyze the location and agglomeration of CSs in Spain.

First, in the context of the location of economic activity, GIS can be used to analyze and visualize data on businesses, industries, and economic activity in a particular area. In this case, we used it to display the location of CSs across the country.

Second, the Kd function, also known as the “density function,” is used in urban economics to measure the relationship between land use and land value [13, 14]. This function can take different forms depending on the specific assumptions about the relations between density and land value. However. it is typically used to estimate the optimal density level for a given parcel of land. In this chapter, the function is used to measure the agglomeration of CSs for different distances.

4 Results

This section introduces the main results of the chapter. Figure 1 shows the density of CSs by province in Spain in 2021. This measure shows the presence and importance of CSs at province level (NUTS3) in relative terms.

Fig. 1
A map of Spain on which the locations of co working spaces are marked. The center has a cluster of C S and most of the C S are along the coastline. Few C Ss are located on the islands.

Source elaboration by the authors

Location of CSs in Spain.

Full size image

It can be observed that the two provinces with the highest density of CSs (i.e., the lowest number of people per CSs) are the provinces of Barcelona and Madrid, followed by Valencia, Malaga, Castellon, and Granada. The aforementioned provinces correspond to the most important cities in Spain in terms of economic activity as well as creative and technological industry [33]. There provinces with high densities, such as Orense or Valladolid; this is due to the fact that these provinces have very low population, which makes the concentration of CSs in them more relevant. Then come provinces such as Seville, Zaragoza, and Vizcaya, which contain very important cities in terms of density and economic activity.Footnote 6

Lastly, on the one hand, we can observe that there is a pattern regarding the density of CSs at a provincial level. This is seen in the provinces along the Mediterranean coastline (area known as the “Mediterranean Corridor”), which stretches from the province of Gerona to the province of Cadiz. Along this corridor, only the province of Murcia has a low density of CSs. Except for the Mediterranean Corridor, we find a high concentration in the Madrid region, corresponding to the city of Madrid, the capital of Spain. The concentration of economic activity in these regions is due to a combination of historical, geographical, and political factors. The Mediterranean Corridor has traditionally been an area of commerce, with the presence of important seaports facilitating trade. Madrid's central location in Spain makes it a strategic location for the distribution of goods and services. Furthermore, both regions have received significant investments in infrastructure, technology, and human capital, contributing to their economic development and attractiveness for businesses. Overall, the concentration of economic activity in these regions has led to a solid economic and commercial infrastructure that allows for competitiveness at the national and international levels, with a strong economy, reflected by the presence of CSs.

On the other hand, we find regions in with no CSs identified, or a very low density of CSs. These regions correspond to rural Spanish provinces (e.g., Caceres, Badajoz, Jaen, La Rioja, Teruel, or Toledo), characterized by low population densities and where the most important economic sectors are the primary and secondary ones, which have very low added value.

Observing Fig. 3, which shows the average entry price per desk in a CSs, an interesting pattern emerges. While in Fig. 2, the highest density of CSs corresponded to the entire Mediterranean axis and Madrid, when looking at the average entry price at a provincial level, we find that the lowest price does not correspond to those regions with a high density of CSs (indeed, in provinces such as Barcelona or Madrid the aver price per desk is higher than average). This may be due to the fact that in these provinces, despite the higher presence of CSs and greater competition, the price of land and office rent (necessary for CSs) is significantly higher than in the other provinces, and this causes an increase in the entry price of the CSs.

Fig. 2
A map of Spain on which the areas of population 0 to 74194, 74194 to 124596, 124596 to 202358, 202358 to 278265, 278265 to 389558, and 3889558 to 709403 per number of C Ss in different shades. Malaga, Granada, Castellon, Tarragona, Barcelona, Soria, and Orensa have the least population per C Ss.

Source elaboration by the authors

Population per CSs by province (NUTS3) for Spain. Year 2021.

Full size image
Fig. 3
A map of Spain on which provinces with average monthly rate of C Ss less than 99 euros, 99 to 150 euros, 150 to 218 euros, 218 to 346 euros, and more than 346 euros are shaded in different shades. Caceres has the highest average monthly price. The data of a few provinces are unknown.

Source Elaboration by the authors

Average monthly rate of CSs by province (NUTS3) for Spain. Year 2021.

Full size image

The provinces where we the average price is lower are 1) provinces with a lower density of CSs but located in the Mediterranean axis (i.e., Castellon and Almeria, followed by Tarragona, Malaga, or Murcia) and some clearly rural interior provinces (such as Cuenca, Badajoz, Jaen, or Ourense) where prices are lower either due to low demand for CSs or because most of CSs in these areas are publicly managed or receive public subsidies.

Finally, Fig. 4 shows the Kd function of the location of CSs in Spain, at a radius of 200 km (to visualize regional agglomeration) and at a radius of 20 km (to visualize urban agglomeration).

Fig. 4
2 graphs of K d of r versus r plot a line K d cap o b s of r. Graph A, K d function C W 20 kilometers plots a line that increases initially and falls. Graph B, K d function C W 200 kilometers plots a decreasing line.

Source Elaboration by the authors based on CSs data. Note: Radius (r) in meters

Agglomeration of CSs. Year 2021.

Full size image

On the one hand, if we look at the regional radius (200 km), it can be observed that the density of agglomeration of CSs spaces is very high at the start and rapidly decreases; this shows that CSs spaces are mostly located in major cities and purely urban areas. Also, it is noted that between 70 and 90 km, the density of CSs slightly increases and then falls again. This is due to the average distance between provincial capitals in Spain being around 90–100 km (approximately, all provincial capitals are 100 km away from the nearest provincial capital), which causes the agglomeration function to grow in that range and shows that CSs are mostly located in provincial capitals in Spain.

On the other hand, regarding the urban range (20 km), it is observed that the function increases between 0 and 3000 m, showing that in the urban setting, in Spain, CSs agglomeration generally occurs at a urban level and not by districts or clusters as in other industries (that need to be located close together to benefit from localization economies). Generally, in Spanish provincial capitals, with some exceptions in the larger cities (e.g., the 22@ district in Barcelona or technological clusters in the main Spanish cities), CSs provide a solution for remote workers and freelancers to work close to home. Therefore, they are generally distributed equidistantly in the city (hence the function reaches its maximum density of agglomeration around 3000–4000 m, which coincides with the average diameter of Spanish provincial capitals).

5 Discussion

The main aim of this chapter was to identify and explain the location patterns of CSs in Spain. Therefore, we contribute to the literature on CSs by providing a micro-analysis of the agglomeration of CSs in urban and rural areas. Furthermore, we dealt with previous methodological limitations by making use of geographical information systems and the Kd-function of agglomeration [14, 15] in the analysis of location patterns of CSs. Specifically, our results showed that (i) CSs are highly concentrated in the most central areas of Spain: Barcelona and Madrid, followed by the Mediterranean Corridor; (ii) at the regional level, CSs are significantly agglomerated at short distances, and this agglomeration rapidly disappears as distance increases, showing that it is an urban phenomenon; and (iii) at the urban level, in capital provinces CSs are distributed throughout the city so as to provide amenities and workplaces to freelance and teleworkers, except in the case of Madrid and Barcelona, where CSs cluster in some specific economic areas as 22@ in poblenou area [7].

These results confirm our preliminary expectations and complement previous contributions. Specifically, they endorse the theoretical discourse that CSs find clear advantages in agglomerating in urban cores [7, 26, 28, 29]. Moreover, they complement the findings of [5, 24], and Tomaz et al. (2022), who analyzed the location patterns of CSs in urban and rural areas in countries such as Italy and France. We also found that CSs are highly clustered around the metropolitan areas of Barcelona and Madrid, but some focal locations are also found in periurban and rural areas. However, by taking advantage of GIS and the Kd-function to test the statistical significance of the results at each distance, we could also ascertain the factors that may influence their location decision.

From the literature and our findings, it is possible to raise some questions for further discussion. First, given that CSs are highly agglomerated in urban areas, one may concern the actual impact that the agglomeration of CSs may have in increasing office rent prices in the city center and the limitations to build strong horizontal networks that facilitate connections across professionals, residents, and public sector actors [4, 6, 35]. Second, the increasing attraction to urban areas should make us think about the fact that in the next years, more than 70% of the population and jobs will be concentrated in urban areas. Thus, the location of CSs in rural areas may be a window of opportunity for policymakers to face rural depopulation. Nevertheless, acting against rural depopulation through the creation of CSs and coliving spaces in rural areas, requires developing a critical mass and appropriate regional policies and legislation design.

Despite all these facts, this chapter has some limitations. In this regard, future research can expand this analysis in two main ways. First, the period of analysis should be expanded to check for time dynamics on the location patterns of CSs. Second, it would be worthwhile to include spatial-time dynamics, particularly to examine the shock of the Covid-19 pandemic.