Here, there and everywhere: hybrid multilocational work and academics’ outcomes

In the post-pandemic era, the ways in which higher education professionals work have undergone significant changes. A key concern that has emerged is how hybrid and multilocational work (i.e. a combination of office and offsite work arrangements, Hislop and Axtell, 2009) affect important professional outcomes. The COVID-19 pandemic precipitated the shift toward hybrid and flexible work arrangements in academia. The traditional boundaries of academic spaces have blurred, marking the beginning of a new era characterized by hybrid work environments (Halford, 2005; Richardson and McKenna, 2014). In this evolving realm where hybrid working has become extremely diffused, “space” remains the key distinction between working from one location or another (Reissner et al., 2021). The academic community’s pursuit of knowledge is inherently tied to their workspaces on campus (Toker and Gray, 2008), raising questions about the impact of transitioning work beyond campus boundaries. Recent research in management and economics has explored the productivity gains of work from home (WFH) and hybrid work for knowledge workers (Bloom et al., 2022; Choudhury et al., 2024), while literature in architecture and built environment studies has emphasized the role of physical workspaces in facilitating work outcomes (e.g. Haynes et al., 2017; Weber et al., 2023). Studies that specifically focus on academics are limited. Some studies have highlighted the benefits of remote work in enhancing academics’ flexibility while expressing concerns about the erosion of organizational culture and the challenges of maintaining work−life boundaries (Watermeyer et al., 2022; Watermeyer and Rowe, 2022). However, little is known about academics doing research in different spatial contexts and how this influences various academic outcomes. This paper investigates whether academics’ hybrid, multilocational work (i.e. working from different workspaces) influences individual self-reported outcomes within an academic workforce, and asks: What are the effects of hybrid work (i.e. working from the university, the home, and other spaces) on self-reported work outcomes? Are some spatial characteristics of these work locations relevant to academic outcomes?

To answer these questions, this research examines not only the dichotomy between remote and on-campus work (i.e. home working vs office working) but also acknowledges “other spaces” (Tagliaro et al., 2022), which encompass all those spaces that academics occupy to fulfill their professional responsibilities (e.g. companies’ sites, cafés and public spaces). In so doing, this paper conceptualizes academics as hybrid, multilocational workers (Kojo and Nenonen, 2015). In addition, this study conceptualizes offices, homes and “other spaces” only as ‘containers’ of academic work but also as locations with specific ‘spatial’ characteristics that are crucial for either supporting or hindering academic activities. Furthermore, this study benefits from a unique data set that includes a nationwide sample of Italian academics. This broad scope enhances the robustness of the results compared to more limited studies based on academic survey data, such as those focusing on a single organization or city (e.g. Parkin et al., 2011).

Ultimately, this research seeks to inform institutional practices and policies. Universities are making huge investments in their facilities to attract their staff and students, but research confirmed the persistence of low occupancy rates in university buildings, which was a fact even before the COVID-19 period (Lansdale et al., 2011). This paper aims to enable higher education institutions to optimize work environments for the betterment of their faculty and the communities they serve.

Academics, akin to other knowledge workers, find themselves immersed in nonroutine work dynamics (Gornall and Salisbury, 2012), a consequence of their profound autonomy, minimal formalization and unconventional organizational structures (Salancik and Pfeffer, 1974; Crang, 2007). Academic work is characterized by diverse activities such as teaching, research, student supervision and administration. It requires both isolation for thinking reflection and collaboration for idea development with colleagues (Indergård and Hansen, 2024). In this context, academics navigate a complex web of flexible and fixed temporal and spatial boundaries (Gornall and Salisbury, 2012), which typically define a hybrid, multilocational workstyle. Temporally, academic work strikes a balance between the scheduled time, encompassing project deadlines, lectures and administrative obligations, and the more nebulous realm of timeless time of reading, writing and deep thinking (Ylijoki and Mantyla, 2003). Spatially, academic work necessitates a dynamic blend of workspaces (i.e. the university offices, their homes and other spaces) to optimize productivity while maintaining the requisite levels of proximity and privacy (Indergård and Hansen, 2024).

These seemingly contradictory demands have contributed to the emergence of two contrasting archetypal images of academic workers which closely relate to different workspaces. On one hand, there is the enduring image of academics toiling in solitude in their study rooms at home. On the other hand, there exists the romanticized notion of the “wandering scholar” (Kim, 2009), reminiscent of medieval monks who traveled across Europe during the Dark Ages. According to this view, academics are globetrotters, traversing several spaces for conferences and research projects to foster networking and professional growth (Morley et al., 2021). Contemporary academic work embodies a harmonious blend of both paradigms. Academics navigate a multitude of work locations, including the university as their primary fixed workspace, their homes as the primary alternative workspace (Dowling and Mantai, 2017) and various other locations as needed (Baldry and Barnes, 2012; Gornall and Salisbury, 2012; Kuntz, 2012). Even if these choices are closely related to academics’ disciplinary background (Huhtelin and Nenonen, 2019), all academics can be considered multilocational workers (Kojo and Nenonen, 2015; Tagliaro et al., 2022). This conceptualization aligns with the understanding in higher education studies of a “liquid” academic era – à la Bauman (2000) – where academics navigate multiple “learning spaces” simultaneously as part of their daily lives: workspaces, nonwork spaces, family, leisure, social and professional networks and various virtual spaces. This multiplicity allows them to find respite from the pressures of academic life and to engage with knowledge, ideas and new experiences (Savin-Baden, 2007). However, in spatial terms, not every location is the same. In contrast with the “anytime, anyplace” rhetoric of work (Brown and O’Hara, 2003; Petani and Mengis, 2023), space plays a crucial role in shaping and constraining what tasks workers can carry out where (Brown and O’Hara, 2003; Hill et al., 2003). In this study, I analyze office spaces on campus, home spaces and “other” spaces.

First, workspaces on-campus “are assigned at the time of the scholar recruitment” (Stephan, 2012, p. 105); these are often tied to their roles and play a pivotal role in shaping their productivity and academic identity (Temple, 2009; Dowling and Mantai, 2017; Indergård and Hansen, 2024). Working on campus positively ensures physical proximity, fostering serendipitous interactions and collaborative endeavors (Toker and Gray, 2008). However, the outcomes of working at the university are not always positive. Research has shown that academic offices designed without considering individual preferences (e.g. open-plan offices, activity-based offices) can lead to negative outcomes such as decreased productivity, lower job satisfaction, increased stress and a diminished sense of well-being among staff and faculty (Jemine et al., 2022; Nooij et al., 2023; Indergård and Hansen, 2024). In addition, research has recognized several spatial aspects of the office (i.e. psychosocial aspects, layout, individual space, indoor climate, aesthetics, accessibility and, services and facilities) as direct and indirect determinants of work outcomes (Appel-Meulenbroek et al., 2018; Indergård and Hansen, 2024).

Second, academic pursuits that require concentrated focus, such as writing scientific papers, often lead academics to WFH (Gornall and Salisbury, 2012; Dowling and Mantai, 2017). For other workers categories, WFH was generally associated with increased productivity (Bloom et al., 2015), although the overall contribution of remote working to productivity gain is highly variable and differentiated by sector, level of education and the extent of their household duties and personal privilege (Brynjolfsson et al., 2020). Before the pandemic, research by Bloom et al. (2015) and Choudhury et al. (2020) highlighted productivity improvements when employees shifted to working from home or adopting a work-from-anywhere model. Particularly, Choudhury et al. (2024) found that workers who spent around three days working from home (and two days in the office) each week on average self-reported greater work−life balance, job satisfaction and lower isolation from colleagues compared to workers who spent more or fewer days working from home. Full-time WFH may heighten feelings of isolation and have mixed effects on wellbeing (Reuschke, 2019). In addition, while it offers the allure of flexibility, it concurrently presents the challenge of blurred boundaries between work and personal life, potentially encumbering the pursuit of an equilibrium between the two (Bailey and Kurland, 2002). Spatially, the literature generally acknowledges that home workers attempt to replicate familiar aspects of their office in the home environment (Ng, 2010) and aspects of the physical home environment such as air quality, lighting, access to daylight, noise, views and ergonomics have direct effects on employees (Sander et al., 2021).

Finally, a variety of “other spaces” characterize academic work. Academic knowledge production frequently necessitates collaborations with external institutions, prompting academics to use workspaces and laboratories affiliated with private or public organizations (Baldry and Barnes, 2012). Recent trends have witnessed the emergence of public libraries (Schopfel et al., 2015) and coworking spaces as alternative venues for individual and collaborative academic work (Bouncken, 2018). These “third spaces” (Oldenburg, 1989) are neither office nor home but can facilitate the relational aspects of academic work. In this paper, I use the term “other spaces” as an umbrella term to refer to all these environments which are not the office nor the home. However, research on these spaces, particularly within academia, is limited, and the outcome implications of working from such spaces remain unclear. In addition, given the variety of these venues, it is difficult to grasp their spatial characteristics. Therefore, in this paper, they are referred to more as alternatives to home and office rather than as specific spaces with measurable spatial characteristics.

This study is exploratory and uses data aggregated from a variety of sources to form a distinctive database on Italian academics. The main data source was a survey of all academics employed in Italian public universities. In line with the requirements for pre-testing survey instruments outlined by Collins (2003), I conducted a preliminary pre-test on a representative sample of the population under investigation. After testing the survey, I distributed it via email to academics at Italian public universities (N = 52,630). This target population was identified through the publicly accessible lists provided by the Italian Education and Research Ministry (MUR) ^[1], which also supplied secondary data regarding the academics’ research fields, tenure, gender and university affiliations. These secondary data were later matched to survey responses.

The survey collected information on (1) academics’ hybrid and multilocational workstyles (i.e. their work locations and the frequency with which they access these locations ^[2]), (2) the spatial characteristics of the spaces where they work (i.e. layout of their work environments on-campus and at home, as well as perceptions toward them and satisfaction with specific spatial features) and (3) their self-reported work outcomes (i.e. individual productivity, job satisfaction, work−life balance and occasions to socialize). In addition, the survey gathered data on the frequency of collaborative work, the use of tools, leadership roles (such as being the department head) and household composition, recognizing their potential impacts on self-reported work outcomes. The survey was distributed between July and September 2020, post the initial COVID-19 pandemic wave in Italy, when strict mobility restrictions were eased in the whole country and before the onset of the second pandemic wave in mid-October 2020 (Tagliaro and Migliore, 2022). In this period, many academics returned to their campuses and working from home and working from the office have become equally appealing work environments for academics, more so than before the pandemic. The survey yielded 11,634 responses, translating to a 22.11% response rate, consistent with expectations for surveys without financial incentives. Of these, 7,861 responses were fully completed and therefore deemed valid and coherent for quantitative analysis.

The analysis included four different dependent variables: Self-reported Individual Productivity, Job Satisfaction, Work-life Balance, Socialization. Self-reported Individual Productivity is a self-reported rating of productivity (mean = 2.92; SD = 1.02; min = 1; max = 5). Respondents were asked to rate their individual productivity in the observed period using a six-point Likert scale ranging from 0 = very negative to 5 = very positive. Job Satisfaction was computed through the five-items scale from Brayfield and Rothe (1951) ^[3] (mean = −0.12; SD = 1.60; min = −5.22; max = 2.76). Worklife Balance is a self-reported rating of work−life balance, measured through a Likert type scale from 0 = very negative to 5 = very positive (mean = 1.97; SD = 0.85; min = 1; max = 5). Socialization is a self-reported rating of occasions of socialization at work, measured through a Likert type scale from 0 = very negative to 5 = very positive (mean = 2.65; SD = 1.14; min = 1; max = 5). Figure A1 in the  Appendix reports the distribution of all the dependent variables through histograms.

The main explanatory variables are: Work from university, Work from home, Work from other spaces, which are continuous variables capturing the percentage time of the work week devoted to work from university, home and other spaces, respectively. Respondents were asked to report how often they accessed six work venues, measured in days ^[4]. For the analysis, all venues other than the office or home were combined into a single category referred to as “other” spaces. To focus on the proportional allocation of work-week time across the three different venues, I converted the data into percentages, ensuring that their total represents 100% of the workweek. This aligns with the quantitative nature of the data and allows for clear comparisons across workspaces.

Other explanatory variables included in the models point to the physical characteristics of these work locations. Dedicated office is a dummy variable valuing 1 whether the academic has the availability of a dedicated workspace on-campus, 0 otherwise. Dedicated home office is a dummy variable valuing 1 whether the academic has the availability of a dedicated office at home, 0 otherwise. University vs Home spatial quality captures the spatial quality of the on-campus workspaces compared to that of the home. This latter variable is built through a factor analysis on 14 spatial items. These items are inspired by Appel-Meulenbroek et al. (2018) and measured through a five-points Likert scale (from −2 better at home to +2 better on-campus):

• internet connection quality;

• availability of space to take a break;

• availability of team working spaces (e.g. meeting/calls, etc.);

• ability to organize the space (e.g. personalization);

• lack of distractions;

• privacy;

• availability of individual space;

• availability of storage for own items/work items;

• inspiration given by the environment (e.g. atmosphere, colors);

• functionality of the workspace (layout);

• ergonomics of the workstation (e.g. desk);

• indoor environmental quality (e.g. temperature, air quality, light, etc.);

• aesthetics; and

• outside view.

The survey did not gather equivalent information on the “other spaces” (beyond home and campus) where academics work, given the diverse array of spaces that fall under this category, making it challenging to capture comprehensively in a survey. The variable University vs Home spatial quality provides an overall comparison between the office and home environments. If the variable scores higher than 0, it indicates that the academic rated the office more positively. Conversely, if the score is lower than 0, it suggests that the academic rated the home more positively ^[5].

The econometric estimations contain several control variables. Namely, I included the demographic variables Gender and Age as well as geographical dummies (i.e. North of Italy, South of Italy, Center of Italy). In addition, I included work-related variables such as those pointing to seniority (i.e. Full Professor, Associate Professor, Researcher), leadership roles within the faculty (i.e. Leadership Role), discipline (i.e. Physical Sciences, Life Sciences, Civil Engineering and Architecture, Industrial Engineering, Social Sciences and Humanities). Together with these variables, I considered: Digital which captures the extent of adoption of digital tools such as tools for online meetings before the COVID-19 pandemic; Collaboration which captures the extent of collaborative work; Lab which captures the need for a laboratory for conducting research. Finally, I also considered WFH before covid which captures the frequency time of the work week spent at home before the pandemic, Residence close to campus which is a dummy variable valuing 1 if the academic lives in the same province of their campuses and it is used as a proxy of commuting time, and Children at home (y.o. 0–14), which values 1 if academic live with children aged between 0 and 14 years old. Table 1 reports variables descriptive statistics.

To evaluate the influence of different work locations where academics work on their work outcomes, I first resorted to descriptive statistics to explore data patterns, while then I resorted to OLS regression to understand the influence of working from multiple workspaces and work outcomes. Specifically, I first checked whether Work from university, Work from home, Work from other spaces differ across geographical areas where academics work, their disciplines, roles, gender and household composition (Table 2). These descriptive statistics served to provide a basis for interpreting correlational results from the OLS regressions. Then, I resorted to OLS regressions. I have run four OLS regressions (each for every DV, Self-reported Productivity, Job Satisfaction, Self-reported Worklife Balance and Self-reported Socializaion) for each work venue (Work from university, Work from home, Work from other spaces), totaling 12 econometric analyses. In addition, I also explored the interaction effects between (1) Work from university and Private Office, (2) Work from university and University vs Home Space Quality, (3) Work from home and Home Office and (4) Work from home and University vs Home Space Quality to explore whether spatial characteristics moderate the effect of working from a work venue and outcomes.

To check the robustness of the results and given the ordinal nature of Self-reported Productivity, Socialization and Work−life Balance, I used ordinal probit models to explore the relationship between space and work outcomes.

Before performing the analysis, pairwise correlations among variables were checked. I also checked multicollinearity by inspecting the variance inflation factor (VIF) values for each regression. The mean VIF value for all the independent variables is below the threshold of 2.5, suggesting that multicollinearity of independent variables is not a concern. Since both the DVs and EVs of the regressions are collected through the same survey instrument, I resorted to several remedies to minimize the risk of common method bias in the results. I assured respondents of complete anonymity, which decreased their tendency to provide socially desirable answers (Podsakoff et al., 2003). Moreover, I fine-tuned a cover story (Podsakoff et al., 2012) to reduce the likelihood of respondents guessing the relationship between the dependent and independent variables. In addition, I conducted Harman’s one-factor test (Podsakoff and Organ, 1986), loading all items into an exploratory factor analysis. No dominant factor emerged.

Before conducting OLS regressions, I performed several descriptive tests to explore the data set and identify significant data patterns. While no significant difference was found in the time spent working from the three venues (i.e. Work from University; Work from Home and Work from Other Spaces) for performing research across geographical areas, through ANOVA tests, I found the most interesting differences by describing academics’ time spent working in the three venues among disciplinary groups (F = 497.71, p = 0.000). Figure 1 provides a visual representation of the average values of work venues’ usage across disciplines. Those who research in life sciences work more frequently from the office than their colleagues in other disciplines (mean = 0.39) and, interestingly, they also work from other spaces (mean = 0.10) more than any other scientist in other disciplines. Conversely, those in social sciences (mean = 0.87) and humanities (mean = 0.89) are the ones working mostly from home compared to other disciplinary groups. This distinction likely derives from the STEM academics’ need for specialized equipment and laboratory access, which are integral to their research and are predominantly located on university premises or in other spaces.

Tenure also plays a role. The time spent at the university office was found to be slightly higher among associate professors and researchers compared to full professors, with a statistically significant difference (p = 0.0249). This variation may indicate that senior academics have greater flexibility in their work responsibilities.

Gender analysis revealed that work location preferences are broadly consistent across male and female academics. However, the t-test showed that there is a significant deviation observed in the propensity to WFH, with women working from home more than men at the expense of limited access to other spaces (p = 0.0002). Household composition, particularly the presence of school children, also plays a role. T-tests showed that academics with children at home are more inclined to WFH (p = 0.0001), even if the difference between the two groups is small and equals 0.02.

Finally, those living close to the university (i.e. with short commutes) spent, on average, more time at the university office, whereas those living farther away (i.e. with long commutes) tend to WFH more frequently (T-test; p = 0.0000). No significant differences are observed in the time spent working from other spaces between long commuters and noncommuters (T-test; p = 0.0735).

Table 3 shows the description of the three explanatory variables of interest used to proxy the spatial characteristics of the space across groups of academics (i.e. Home Office, Single Office, University vs Home Space Quality). Academics living in the South of Italy are more likely to benefit from a home office compared to those living in the North, according to the ANOVA test (66% vs 54%, p = 0.0000). The urban fabric between the North and South of Italy is highly different, and real estate prices are lower in the South because of the relatively lower economic development compared to the North. There is a notable disciplinary divide in workspace preferences and availability according to the ANOVA test (p = 0.0000). Academics in life and physical sciences and engineering disciplines show a higher preference for university workspaces and a greater availability of single offices. In contrast, academics in the Humanities and Social Sciences report a stronger preference for home spaces, with academics in the humanities showing the lowest preference for university spaces (mean −0.36) and the highest availability of home offices (70% of the sample). Seniority makes a difference in the availability of spaces. Full professors exhibit the greatest prevalence of home and single-office availability, trailed by associate professors and researchers. This may signify the financial autonomy of senior academics, enabling them to afford larger residential spaces. Moreover, access to private offices within university contexts is linked to academic seniority, as offices are often tied to roles. Men reported a higher average availability of home offices and a stronger preference for university spaces (mean +0.09) compared to women, who showed a slightly higher preference for home spaces (mean −0.10). Having children at home influences the preference of the home vs the office spaces significantly (T-test; p = 0.0000): parents show a more favorable view of university spaces (mean 0.16) compared to those without children, who show a slight preference for the home spaces (mean −0.07).

Figure 2 shows the distribution of the 14 spatial variables used to compute the variable University vs Home Space Quality through factor analysis. Some extreme values are noticeable. Significant differences across various groups of academics emerged (see Table A1 in the  Appendix). For example, the aesthetic aspects and external views of home workspaces are generally rated higher, especially by women and academics in the humanities who strongly prefer their home environments. Conversely, on-campus team working spaces receive high ratings, particularly from those with shorter commutes, while longer commutes correlate with lower ratings of space quality, emphasizing the subjective nature of workspace quality assessments. Notably, the ratings for ergonomics and indoor environmental quality differ significantly across disciplines: physical scientists, industrial scientists and life scientists rate these aspects higher on campus, whereas those in social sciences and humanities prefer their home environments regarding these aspects. Similarly, the ability to organize and personalize space varies, with academics in the humanities favoring home spaces and physical scientists preferring on-campus workspaces. Distraction and privacy preferences vary. Academics with children find their offices quieter, while those without children find their home environments quieter. These results highlight the diverse and subjective nature of workspace quality perceptions among different academic groups. In the following paragraph, I report the results of OLS regressions which analyze the relationship between the time spent working in the three different work locations and work outcomes, taking into account some spatial characteristics of the work locations (i.e. Private office, Home office and University vs Home Space Quality).

Tables 4–6 display the results of the OLS regression models. In the  Appendix, Tables 2, 3 and 4 report the same results, including the coefficients for the control variables, which are not discussed below for brevity.

All else equal, I found that Work from the university has no significant direct effect on Self-reported Individual Productivity (β = −0.035; p = 0.510), while it has a positive and significant effect on Job Satisfaction (β = 0.406; p= 0.000), and Socialization (β = 0.384; p = 0.000), and a close to significant positive effect on Work-life Balance (β = 0.101; p = 0.057). Work from home has no significant direct effect on Self-reported Individual Productivity (β = −0.017; p = 0.718) and Work-life Balance (β = −0.059; p = 0.205), while it has a significant and negative effect on Socialization (β: −0.445; p = 0.000) and Job Satisfaction (β = −0.402; p = 0.000). Finally, working from other spaces has a slight positive and significant effect only on Socialization (β = 0.004; p = 0.000), while it remains insignificant for the other outcome variables.

In models I-VIII B, I added interaction terms. Namely, I considered whether working from university (and working from home) have magnified (or lower) effects when academics rate higher (or lower) their workspace on campus compared to the one at home. I found that the interaction term Work-from-university x University vs Home Space Quality has a positive effect on Self-reported Individual Productivity (β = 0.311; p = 0.000), Job satisfaction (β = 0.467; p = 0.000), Socialization (β = 0.249; p = 0.000) and Work-life Balance (β = 0.316; p = 0.000). Therefore, spatial quality magnifies the effect of working from the university spaces on all outcome variables.

The interaction term Work-from-home x University vs Home Space Quality magnifies the negative direct effects of Work-from-home on the outcome variables. Namely, Work-from-home x University vs Home Space Quality show a negative and significant effect on Self-reported Individual Productivity (β = −0.287; p = 0.000), Job satisfaction (β = −0.507; p = 0.000), Socialization (β = −0.210; p = 0.000) and Work-life Balance (β = −0.313; p = 0.000). Therefore, when spatial quality is rated better on-campus than at home, the effect of Work-from-home on all outcome variables is adverse, especially for job satisfaction which show the highest regression coefficient.

Similarly, I considered whether working from university and working from home have magnified effects when academics have access to a single office on-campus and at home. When considering the interaction term Work-from-university x Private Office, I found that when working from the university and academics have a private office on-campus, self-reported productivity increases significantly (β = 0.184; p = 0.045), while job satisfaction, socialization and work-life balance show no effect. When considering the interaction term Work-from-home x Home Office, I found that when academics have a home office during WFH, they only slightly significantly increase self-reported individual productivity (β = 0.012; p = 0.040), while the interaction term does not affect job satisfaction, socialization and work−life balance significantly.

Models IX-XII A present the results pertaining to working from other spaces. This variable does not show a significant relationship with any of self-reported outcome variables considered in this paper, except for Socialization. Working from other spaces is significantly and positively related to socialization outcomes (β = 0.004; p = 0.000). However, the low coefficient indicates a minimal influence.

In an era where academic workspaces transcend traditional boundaries to become increasingly multilocational and hybrid (Savin-Baden, 2007), it is of utmost importance to understand whether hybrid, multilocational work influences individual self-reported outcomes and whether the spatial characteristics of these venues moderate the relationship. Drawing on data from over 7,000 academics in Italy, the findings of this study reveal that working from the university fosters positive perceptions of job satisfaction and socialization, while it has a positive effect on self-reported individual productivity and work−life balance, only contingent on the favorability of self-reported spatial characteristics. This finding represents a significant advancement in the literature (e.g. Appel-Meulenbroek et al., 2018), highlighting that improved productivity and work−life balance when working from the office depend on the perceived spatial quality of campus spaces. Contrarily, working from home shows a stark divergence in impact on job satisfaction and socialization. Working from home offers no direct benefits to self-reported productivity and work−life balance under standard conditions. Similar to what happens at the office, the benefits of working from home on self-reported productivity and work−life balance are contingent on the perceived spatial quality of the home spaces compared to their offices.

These results highlight the greater importance of perceived spatial quality over the layout configuration, such as having a single office or a home office setup. While most literature has recognized academics’ preference for single offices, attributing dissatisfaction to the lack of such spaces in new university projects (e.g. Jemine et al., 2022; Nooij et al., 2023; Haynes et al., 2017), the findings of this paper suggest that the presence of single offices on-campus and home offices only limitedly improves self-reported productivity and has no significant impact on other work outcomes. Access to single offices and home offices is mainly associated with academics’ tenure and gender, suggesting that single offices serve more as a status symbol rather than providing an actual enhancement to – at least self-reported – work outcomes.

Finally, the positive correlation between nontraditional workspaces (i.e. other spaces) and socialization is limited in size. This indicates that the significance of these alternative workspaces may not be as essential for academic professionals. In addition, descriptive results show that the time spent working from “other” spaces does not vary significantly across different groups of academics, except for life scientists, who exhibit a higher usage rate of these spaces. Life scientists are characterized by a higher rate of collaboration for research purposes (Huhtelin and Nenonen, 2019), making them a suitable group for future analysis on the impact of working from other spaces, particularly “third spaces,” which are known for their socialization benefits (Oldenburg, 1989).

These findings add to previous studies on the outcomes of work from several locations, finding that academia has unique specificities that are not found in other contexts (Baldry and Barnes, 2012; Gornall and Salisbury, 2012; Kuntz, 2012). While WFH is generally associated with productivity gains in pre-pandemic studies (e.g. Bloom et al., 2015; Choudhury et al., 2020), this study shows that it hurts productivity and job satisfaction, especially when spatial conditions are not favorable. This adds to the emerging literature which is pointing to the relevance of home space design to allow productive WFH (Sander et al., 2021). In contrast, working from the office is not a panacea, as already pointed out in other post-Covid studies (e.g. Watermeyer et al., 2022), since its benefits hold only when some spatial characteristics – especially related to psychosocial and perceived aspects – exist (Appel-Meulenbroek et al., 2018). The results acknowledge the differential time spent across spaces among respondents. Academics in STEM disciplines demonstrate a higher use of office spaces on average, reflecting the equipment-intensive nature of their work (Baldry and Barnes, 2012), while social sciences and humanities tend to favor remote work settings. These findings corroborate Huhtelin and Nenonen’ (2019) observations regarding discipline-specific workspace needs. Furthermore, disparities in workspace use based on gender and family circumstances underscore broader issues of work−life balance and equity in academia, aligning with discussions on the quality of work environments (Weber et al., 2023). This unequal use of multiple workspaces may reinforce existing privilege dynamics, which were intensified in the aftermath of the pandemic (Brynjolfsson et al., 2020).

In this study, I have explored the relationship between hybrid, multilocational work (i.e. working from different workspaces) and individual self-reported outcomes in academia.

The results offer several implications for real estate and facility managers of higher education organizations.

The findings suggest that higher education institutions should prioritize improving the quality of their campus environments to ensure positive outcomes for their faculty. This quality encompasses psychosocial factors such as personal control, the absence of distractions and privacy, as well as indoor climate, ergonomics, aesthetics and architectural design, as highlighted by Appel-Meulenbroek et al. (2018). Leveraging these factors is crucial for facility managers and workplace managers who want to improve academic experience on-campus and related work outcomes. The layout of office space (namely the availability of single offices) is related to academic outcomes, but it has a lower impact on enhancing these outcomes compared to perceived psychosocial aspects. While managers are attempting to encourage academics to transition from individual offices to shared space models, there is notable resistance to this change (Van Marrewijk and Van den Ende, 2018; Jemine et al., 2022). These findings corroborate evidence indicating that effective academic workspaces facilitate various work styles and collaborative interactions rather than solely providing individual offices (Adenipekun et al., 2021). Although the extant literature has identified academics’ preference for single offices and attributed dissatisfaction to the absence of such spaces in new university projects (e.g. Jemine et al., 2022; Nooij et al., 2023; Haynes et al., 2017), our findings suggest that while having a single office on campus does impact work-related outcomes − such as productivity, socialization and job satisfaction − as well as life-related outcomes like work−life balance, its effect is not stronger than that of the perceived spatial quality of the office. These spatial qualities show a strong effect also when considering WFH effects. This advances prior understanding of the space of WFH (Ng, 2010; Sander et al., 2021) as the perceived quality of the home space – and not the availability of a home office – has the capacity to turn positive the adverse effects of WFH on the outcomes considered in this study. Facility managers can leverage these perceptions through several strategies, including enhancing virtual collaboration tools, providing ergonomic home office equipment and offering guidelines for creating effective homework environments.

Moreover, it is crucial to acknowledge that perceptions of space quality vary across demographic and disciplinary groups, as well as personal experiences. Facility management can address these differences by conducting regular surveys to understand the specific needs of diverse academic groups and designing adaptable workspaces that cater to a wide range of preferences and requirements.

This study is not without its limitations, which open avenues for future research. The unique context of the COVID-19 pandemic, which has significantly altered organizational dynamics and worker isolation, limited the generalizability of the results and suggests that the observed effects may require re-evaluation in a post-pandemic, hybrid work landscape. Future research should pivot toward examining the evolving landscape of working spaces and their categorization by use and academics’ characteristics, to test whether the findings remain relevant in a more stabilized, post-pandemic environment.

In addition, this study primarily addressed academics as multilocational workers but remains limited in its examination of the influence of alternative locations (i.e. neither the office nor home) on work outcomes. Future research should investigate how third spaces, such as coworking spaces and cafes, impact academic productivity and socialization.

Finally, this study had not deepen in the econometric regressions the influence of the covariates age, gender, seniority and disciplines. Future research should disentangle the interaction effects between these variables, space quality and work outcomes to provide a more nuanced understanding of how different factors collectively influence academic work outcomes for different groups.

Funding: This study received support from PRIN PNRR Grant P20224FPTT “The relation between workspaces and gender in academia: an interdisciplinary approach,” Grant No. P20224FPTT.

Figure A1 

Table A1 

Table A2 

Table A3 

Table A4