Workplace experiences as push and pull factors – How are perceived office conditions associated with teleworking preferences? Open Access

Telework refers to work that could be performed on an employer’s premises but is carried out elsewhere on a regular basis (2002 EU Social Partners’ Framework Agreement on Telework). Since the COVID-19 pandemic, work time spent away from the workplace has increased (Eurofound, 2023; Gensler, 2023). This shift has led organizations to navigate a new balance between working remotely and working at the workplace, and to develop office design to support such hybrid working (Vartiainen and Vanharanta, 2023; Eurofound, 2023). Simultaneously, low occupancy of offices makes higher space efficiency and optimization an attractive option for organizations to better reflect their sustainability goals and savings targets (CBRE, 2023). For real estate owners, these trends mean increasingly empty premises and tougher competition for tenants. Consequently, new needs have emerged to research the changing role of the office (Gensler, 2023; Castellum, 2023; Appel-Meulenbroek et al., 2022).

Before the COVID-19 pandemic, user experiences of offices and teleworking were mostly studied in separate lines of research. Flexible work and the development of office design are, however, closely related in practice, as has been recognized in research on multilocational working (e.g. Bosch‐Sijtsema et al., 2010; Vartiainen et al., 2007) and activity-based offices (ABOs) (e.g. Brunia et al., 2016; Appel-Meulenbroek et al., 2020). Yet, even studies on ABOs have tended to only report telework descriptively or in parallel to workplace experiences (e.g. Brunia et al., 2016; Appel-Meulenbroek et al., 2020) and have lacked analysis of whether and how the amount of teleworking contributes to workplace experiences and vice versa (for exceptions, see e.g. Bosch‐Sijtsema et al., 2010; Haapakangas et al., 2023). Teleworking has, however, been recognized as compensating for a distractive office environment (e.g. Appel-Meulenbroek et al., 2021; Kaarlela-Tuomaala et al., 2009) and potentially moderating the effects of office design on, for example, sickness absence (Borge et al., 2023).

Likewise, research on telework (also, “remote work” or “working from home”) has paid little attention to conditions at the office, even when telework preferences (Nguyen, 2021) and the success factors of telework have been the focus of the studies (Gohoungodji et al., 2023). As shown by several literature reviews, the work environment of telework has been approached in terms of the psychosocial environment or physical arrangements at home (De Macêdo et al., 2020; Gajendran and Harrison, 2007; Gajendran et al., 2024; Lunde et al., 2022; Vleeshouwers et al., 2022; Wütschert et al., 2022). Furthermore, post-pandemic teleworking generally requires new research, as it may involve new phenomena in comparison to pre-pandemic telework, which was less common, and telework during the COVID-19 pandemic, which, in turn, occurred under exceptional circumstances (e.g. mandatory telework, social distancing and health concerns).

The increased use of the terms “hybrid work” and “hybrid workplace” signifies a shift towards recognizing on-site and off-site work as more interrelated (Vartiainen and Vanharanta, 2023). This systemic understanding is reflected in organizations’ efforts to establish optimal hybrid work arrangements (Eurofound, 2023). Workspace design can be viewed as an integral component of the broader socio-technical systems of organizations, contributing to organizational effectiveness in dynamic interplay with other system components (Davis, 2019). To better understand the systemic role of the workplace in hybrid work, investigating the dynamics between perceived office conditions and teleworking preferences is important.

Both physical and psychosocial workplace characteristics are relevant to employee well-being and productivity (Lamb and Kwok, 2015; Lunde et al., 2022), and offices and telework locations may each have different advantages and drawbacks (e.g. Gajendran et al., 2024; Engelen et al., 2019). The time spent working remotely versus on-site affects the amount of exposure to any location-related risk factors and resources, thus potentially moderating their effects on employees. Teleworking may function as a way of coping with an unsatisfactory office environment, such as noise (Appel-Meulenbroek et al., 2021; Kaarlela-Tuomaala et al., 2009; Lamb and Kwok, 2015) or other workplace stressors (Fonner and Roloff, 2010; Shao et al., 2021). According to a recent meta-analysis, a higher amount of telework seems generally beneficial in terms of, for example, greater job satisfaction and performance, whereas its downside of perceived isolation appears to be minor (Gajendran et al., 2024).

However, teleworking is more likely to be beneficial when it is perceived as voluntary (Kaluza and van Dick, 2023; Kaltiainen and Hakanen, 2023). Unsatisfactory work conditions at the office might weaken the perception of the voluntary nature of teleworking, reducing its benefits (Borge et al., 2023). It also seems possible that any negative effects of teleworking, such as isolation (Gajendran et al., 2024) and ergonomic risks (Wütschert et al., 2022) might be aggravated if negative workplace experiences lead employees to excessively favour telework. Consequently, needs have emerged for new research on how office design could attract people to the workplace (Gensler, 2023; Castellum, 2023; Appel-Meulenbroek et al., 2022).

Consulting and real estate firms have been the first to survey user experiences of hybrid work (e.g. CBRE, 2023; Castellum, 2023; Gensler, 2023; Steelcase, 2023; Leesman, 2023), trying to identify work activities that the future office should support and the central pull factors for the office. Some reports raise social and collaborative activities as well as learning as the most important factors (Castellum, 2023), while others emphasize supporting a range of activities (Gensler, 2023; Leesman, 2023) or considering individual needs at the workplace (Steelcase, 2023). All the reports emphasize expectations of high-quality work environments.

However, the scientific contributions to these questions are still scarce and focus mainly on task-related work location preferences and perceptions (Shao et al., 2021; Yang et al., 2023; Rücker et al., 2024), paying limited attention to the characteristics of these locations (for exceptions, see Appel-Meulenbroek et al., 2022; Bergefurt et al., 2024). These studies show that the office is generally preferred over teleworking for interactive needs and activities (Shao et al., 2021; Yang et al., 2023; Appel-Meulenbroek et al., 2022; Rücker et al., 2024), although Bergefurt et al. (2024) have found that support for activities that require concentration at the office also predicted higher productivity. Distractions generally appear to be an important determinant of work location choices (Appel-Meulenbroek et al., 2022; Asmussen et al., 2023). However, most of these studies used data collected during the COVID-19 pandemic, and the extent to which they can be generalized to post-pandemic (i.e. post-lockdown) hybrid work is uncertain.

To address the research gap concerning the role of office design in hybrid work, this study examines workplace experiences as push and pull factors for on-site working. We approached this by identifying the perceptions of the physical workplace that are related to willingness to increase or decrease teleworking. The first research question was:

The environmental features of the office affect workspace choices within the office (Kim et al., 2016; van den Berg et al., 2020), suggesting that they may also be relevant to remote versus on-site preferences. Therefore, we investigated factors that are commonly included in user satisfaction surveys, such as privacy, ergonomic comfort and interaction (Brunia et al., 2016; Kim et al., 2016), as well as the attributes relevant to ABO (e.g. person–environment fit and workspace-switching, Haapakangas et al., 2023). In line with the salutogenic approach to office design (Ruohomäki et al., 2015), we also included being able to detach oneself from work during breaks in workplace experiences.

To deepen our investigation, we also compared workplace experiences with other potential determinants of telework preferences. Employee well-being and psychosocial factors have been linked to office design (Herbig et al., 2016; Pejtersen et al., 2006) and teleworking (Lunde et al., 2022; Vleeshouwers et al., 2022), suggesting intertwined relations among these factors. Work ability has also been related to workplace experiences (Haapakangas et al., 2023), and is shaped by the interplay between many factors, such as the individual’s resources (e.g. health, competence) and work, including environmental conditions (Ilmarinen et al., 2008). Teleworking could serve as a coping mechanism for employees who experience a negative work climate at the office (Fonner and Roloff, 2010) or who wish to avoid additional sources of stress or to adapt their work conditions due to diminished work ability or well-being (Montreuil and Lippel, 2003). For example, clinical burnout may reduce a person’s ability to cope with workplace distractions (Krabbe et al., 2017). Furthermore, the current amount of teleworking most likely predicts preferences for teleworking. Thus, we also posed the following research questions:

This cross-sectional study involved four Finnish organizations located in Helsinki. The survey data were collected electronically between September and December 2022. The national teleworking recommendations had been lifted at the end of February 2022, and since then, the workplaces have encouraged normal working. We also systematically inspected each office site and gathered documents on workspaces (e.g. layout drawings and workplace guidelines) to evaluate the quality and design of workspaces independently of the survey. The study was approved by the ethics review board of the Finnish Institute of Occupational Health, Helsinki, Finland. All the respondents provided their written informed consent to participate in the study.

The investigated workspaces were environments for knowledge work. The workspaces were in altogether 13 buildings and included variations in office types and design although the majority were ABOs. Based on our observations, the design of the ABOs was mostly modern and of good quality compared to typical contemporary office design in Finland. The traditional office types varied more in terms of when they had been constructed and, therefore, also in quality.

Two workplaces (facilities built or refurbished in 2018–2020) were completely designed as ABOs and had a variety of workspaces for different tasks and mainly non-assigned workstations. The plan layout in these ABOs was mostly open, with enclosed rooms dividing the space into smaller sections, including separate workstation areas for silent work that required concentration. The open areas mainly had workstations, but also some informal meeting or collaboration spots. The functions of the enclosed spaces varied from support facilities (printing, mail, etc.) to meeting and project rooms (bookable and non-bookable), and non-assigned rooms for individual work and virtual meetings. Larger meeting rooms for internal and external meetings were mostly centralized close to the main people traffic areas or on separate floors. Both facilities also featured work cafés on each office floor for informal work and breaks.

The other two organizations’ facilities (built or refurbished mainly between 2001 and 2015 with some small changes during the COVID-19 pandemic) included some ABOs corresponding to the description above, but the plan layout was more often based on closed rooms or areas and separate corridors. In addition to the ABOs, they had traditional rooms for one to four people and open-plan offices (both with primarily assigned seating), as well as some special-purpose facilities (e.g. laboratories). Those working in the more traditional office areas also had access to work cafés or break rooms, support facility spaces (printing, mail, etc.) and some meeting rooms for internal and external meetings.

At the time of the survey, the organizations had not yet reduced their office spaces in response to lower occupancy. Thus, the availability of different workspaces was better than normal and some employees in the ABOs had monopolized workstations or rooms, despite desk-sharing policies.

The telework practices were very flexible in the organizations and most employees had no compulsory minimum on-site requirements. Two organizations recommended working at the office at least one to two days per week, and the personnel who had job roles that required them to mostly work on-site in special-purpose facilities were a minority.

Of the total 1,017 survey respondents, we included those whose responses included their telework preference, age, gender and education (n = 923). The organizations’ response rates varied from 31% to 73%. The respondents were on average 47.8 years old (SD 10.2), and the majority were women (77.4%). Table 1 presents the descriptive statistics of the sample.

The questionnaire addressed several themes concerning the physical and psychosocial work environment, telework, well-being and health. The study included the following items.

Telework preference was measured by the question “Would you prefer to increase or decrease the amount of time you spend working remotely?” (Baker et al., 2007). A five-point scale (Decrease it a lot, Decrease it a little, Continue about the same, Increase it a little, Increase it a lot) was recoded into three categories by combining the two highest and lowest options.

Workplace experiences included a three-item task privacy scale (adapted from Oldham, 1988; scale: 1 strongly disagree – 7 strongly agree) that measured the distractions and being able to concentrate at the employer’s premises. The perceived availability of workspaces was rated separately for quiet spaces for work requiring concentration, spaces for spontaneous discussions or meetings and meeting spaces that can be booked, with three options (1 = Yes, sufficient, 2 = Varies, not sufficient, 3 = No, not at all; adapted from Bodin Danielsson and Theorell, 2019). The ease of workspace-switching was measured by the question “How easy is it for you to find a more suitable workspace during the working day and to go there (e.g., if you need to concentrate or have a confidential discussion or phone call)?” (Scale: 1 very easy – 5 very difficult). The remaining single statements (Ruohomäki et al., 2013; scale: 1 strongly disagree – 5 strongly agree) included person–environment fit (“The work premises are well-suited to my work tasks”), support for interaction (“The work premises support interaction between individuals”), availability of colleagues (“Colleagues are easily available in this environment”), being able to detach oneself from work in break rooms (“I am able to detach myself from work for a while in the break room”), furniture ergonomics and comfort (“The furniture at the workplace is ergonomic and comfortable”), cleanliness (“I am satisfied with the cleanliness of the work premises”), the amount of workspace (“There is sufficient space for working”) and the amount of storage space (“I have enough room to store my items”).

The following psychosocial risk factors and resources were included. Job demands (five items, Fransson et al., 2012) related to high workload and time pressure. Relational justice (six items, Moorman, 1991) referred to interaction experiences with one’s supervisor, for example, whether they are kind and can suppress personal biases. The Effort-Reward Imbalance (ERI) (Kivimäki et al., 2007) measured the balance between perceived effort (one item) and perceived rewards (three items, e.g. income, personal satisfaction). An ERI score above 1 indicates that the effort is greater than the reward, and a score below 1 indicates the opposite. Team climate was measured using four dimensions of cooperation relevant to team creativity and innovation (vision, participation safety, support for innovation and task orientation) on the short version (14 items; Kivimaki and Elovainio, 1999) of the team climate inventory (Anderson and West, 1994).

Work engagement (the three-item Ultra-Short Measure for Work Engagement, UWES-3, Schaufeli et al., 2017) and burnout symptoms (Burnout Assessment Tool BAT-12, 12 items, Hadžibajramović et al., 2022) were selected to cover the different dimensions of employee well-being. Work engagement refers to positive work-related experiences of vigour, dedication and absorption (Schaufeli and Bakker, 2004). The BAT-12 measures four defining dimensions of burnout:

1. exhaustion;

2. mental distancing;

3. cognitive impairment; and

4. emotional impairment (Schaufeli et al., 2020).

We measured current work ability using a single item from the work ability index (WAI) (Tuomi et al., 1991; scale: 0 Unable to work – 10 Work ability at its all-time best). This item correlates strongly with the WAI and predicts health-related outcomes (Ahlstrom et al., 2010).

The respondents rated their normal amount of teleworking on a six-point scale (1 I do not do any telework – 6 Daily, see Table 1; adapted from Ruohomäki et al., 2023).

Age, gender (female/male) and education were included as demographic factors.

We used multinomial regression analysis to examine how the independent variables were associated with telework preferences, using “continue about the same” as the reference category. This method ignored the ordinal nature of the outcome variable and tested the push and pull nature of workplace experiences independently of each other. Firstly, we examined how each independent variable was associated with teleworking preferences separately, adjusting for gender, age and education (RQs 1 and 2). The p-values of these multiple tests were adjusted according to the Benjamini–Hochberg procedure to control for false discovery rate. Secondly, to answer RQ3, we specified a new model by including all the variables that were statistically significant, i.e. below the alpha level of 0.05, in the first step and the covariates (gender, age and education). Odds ratios (ORs) with 95% confidence intervals are reported. Significant ORs greater than one indicate an increase in the odds of preferring a decrease or an increase (depending on which is examined) as opposed to continuing the same. Similarly, ORs smaller than one indicate decreased odds of preferring a change. All analyses were performed using R (version 4.3.0), and its nnet package (version 7.3-19).

Tables 1 and 2 show the descriptive statistics and Table 3 shows the correlations. The workplace factors were mostly rated positively. Teleworking was very frequent in the sample. Most respondents preferred to continue with their current amount of teleworking (73.1%), but preferences for increasing (16.8%) were slightly more common than preferences for decreasing (10.1%). The preference groups had slightly different profiles (Table 1). Those who preferred to increase telework had, on average, the lowest education level, and those wishing to decrease had the highest. The respondents who preferred to decrease telework were, on average, slightly younger and teleworked less than those with other preferences. Gender and office type distributions were similar across the groups.

Tables 4–5 reveal the results of the multinomial regression analyses. We interpreted independent variables as push factors when they were associated with a preference to increase telework, and as pull factors when associated with a preference to decrease telework. It should be noted that lower odds (i.e. OR <1.0) of preferring to increase telework do not indicate a greater likelihood of preferring to decrease telework, but of “continuing the same”.

Overall, there were substantially more “push” than “pull” factors among the investigated variables (Table 4). That is, the preference to increase telework was associated with more negative ratings of many workplace experiences (ORs <1.0), including the availability of workspaces for quiet work and for spontaneous discussions, the ease of workspace-switching, perceived person–environment fit, being able to detach oneself from work in break rooms, sufficient work and storage space, and task privacy. Other workplace experiences were not related to telework preferences (Table 4).

Task privacy was the only pull factor among workplace perceptions, as higher privacy increased preferences to decrease teleworking.

The current amount of teleworking predicted telework preferences, that is, a lower amount of teleworking was associated with preferences to increase teleworking. Of the psychosocial factors, relational justice was most consistently related to telework preferences: lower perceived justice was a push factor and higher justice showed a tendency towards being a pull factor. A higher ERI was associated with preferring to increase teleworking. Job demands and team climate did not predict telework preferences.

Well-being and work ability were not associated with telework preferences, although lower work engagement showed a non-significant tendency towards being a push factor.

Table 5 and Figure 1 show the results of the final model, which included all the statistically significant independent variables of the first step and the covariates (age, gender and education). Task privacy and the current amount of telework were the only predictors with both push and pull qualities. Of the other workplace experiences, less ease of workspace-switching and amount of workspace showed some association, although not consistently, with preferences to increase telework.

Psychosocial factors did not predict teleworking preferences in the final model.

Bridging the research on telework and office design, this study showed that workplace experiences function as push and pull factors for on-site working in hybrid work. Although the majority of the respondents preferred to continue with their current amount of teleworking, more negative perceptions of several workplace factors predicted preferences for increasing telework (RQ1). Only task privacy was also a pull factor, meaning the absence of distractions and the ability to concentrate at the workplace (RQ1). The current amount of telework and a few psychosocial factors were associated with telework preferences when investigated separately from workplace experiences (RQ2). When all the predictors were analysed together, task privacy remained a statistically significant predictor of telework preferences, but psychosocial factors did not (RQ3). Overall, the results suggest that office design may play an important role in the dynamics between on-site and off-site working.

The results highlight that workplace experiences are primarily a push factor in hybrid work. This was related to poorer task privacy, insufficient availability of workspaces for work that requires concentration and spontaneous interaction, difficulties workspace-switching, a poorer person–environment fit, difficulty detaching oneself from work in break rooms and insufficient work and storage space (RQ1). More positive perceptions of these factors were associated with a lower likelihood of preferring to increase telework (Table 4), but aside from task privacy, were not relevant to decreasing telework. The correlations between the variables (Table 3) suggest that many aspects of workplace experiences, such as the ease of workspace-switching and the availability of different workspaces, contribute to the overall experience and to task privacy. Thus, despite being non-significant in the final model, they may be relevant to mitigating the push nature of workplace experiences. The fact that these push factors emerged in workplaces that were generally rated positively (Table 2) and mainly represented good-quality modern design is in line with previously observed heightened quality demands of the office users (Gensler, 2023; Leesman, 2023).

Another key finding was that task privacy was highlighted as the only workplace experience that was both a push and pull factor. This finding aligns with observations that distractions are a key determinant of work location choice in hybrid work (Appel-Meulenbroek et al., 2022; Asmussen et al., 2023). The push nature of insufficient privacy complies with teleworking being a coping strategy (Appel-Meulenbroek et al., 2021; Kaarlela-Tuomaala et al., 2009). The results also agree with previous findings that ABOs are perceived to support collaborative tasks better than concentration (Haapakangas et al., 2023; Engelen et al., 2019). With increased freedom over work location, our results suggest a risk that the downsides of inadequate privacy in modern offices may push employees to telework more.

The pull nature of task privacy contradicts the dichotomy that employees mainly view offices as places for interaction and collaboration, and prefer their homes for work that requires concentration (Shao et al., 2021; Appel-Meulenbroek et al., 2022). Increased teleworking may also involve increased distractions at home (Gohoungodji et al., 2023; Kaltiainen and Hakanen, 2023), and supporting work that requires concentration at the office appears to be related to higher productivity (Bergefurt et al., 2024). Planning working days for only face-to-face collaboration at the office might be challenging, and appropriate privacy is also increasingly needed for planned and virtual meetings regardless of the work location (Leesman, 2023). Thus, we agree with the views on balancing interactive and individual work in workplace design (Bergefurt et al., 2024; Gensler, 2023; Leesman, 2023).

Our findings on workplace experiences are supported by analyses of other predictors of telework preferences. The importance of workplace experiences over psychosocial environment, in particular, highlights the need to recognize workspace design as a component of the organizational system in hybrid work. Perceptions of the physical and psychosocial environment are seldom analysed together, despite their documented interrelations (e.g. Herbig et al., 2016; Pejtersen et al., 2006). Future studies should deepen the systemic perspective and investigate what long-term consequences the interplay between workplace experiences and telework may have for individuals, teams and organizations.

Well-being at work and work ability were not related to telework preferences (RQ2). The relationship between teleworking and employee well-being in general is complex (Lunde et al., 2022). The tendencies concerning work engagement and some psychosocial factors (Table 4) imply that the relation between well-being and telework preferences may involve moderators that were not considered here. As the attention being paid to salutogenic workplace design is increasing (e.g. Ruohomäki et al., 2015), more research is needed on the relations between office design, user well-being and teleworking.

We operationalized push and pull factors as preferences to increase and decrease teleworking, but other approaches could provide additional insights. Our analysis does not reveal how workplace experiences (or other predictors) had already shaped the amount of telework, which was high in our sample. Longitudinal studies could provide more comprehensive evidence.

As preferences to decrease teleworking were relatively uncommon, the statistical power of the study may have been too weak to enable identification of the pull factors. However, our sample consisted of nearly 1,000 respondents and was characterized by high teleworking. The lack of a clear pull towards the workplace may indicate that it is a weak phenomenon, given the popularity of teleworking.

Low occupancy rates during the survey may have limited the identification of push and pull factors related to social interaction at the workplace. Thus, the results do not exclude the possibility that the social aspects of workspace design could be a pull factor under other conditions. Furthermore, we did not investigate whether more innovative design elements, including services and technology offered at the workplace, could function as pull factors at the hybrid workplace (e.g. Castellum, 2023; Steelcase, 2023).

The generalizability of the findings may be limited by differences in organizational cultures. The data were gathered in Finland, where teleworking was comparatively high even before the pandemic. Office design can also be different in different countries.

Our results suggest that organizations should integrate workplace design into their hybrid work models. The workplace experience may be particularly relevant in organizations aiming to provide highly flexible teleworking while still emphasizing the importance of on-site work.

Based on this study, satisfactory task privacy and adequate workspaces for work that requires concentration seem particularly important for making the office more attractive. This can be facilitated by transforming existing open areas in the office into spaces for quiet work. Also, adding one- to two-person enclosed workspaces can offer privacy for various tasks, including individual work, attending virtual meetings and having ad hoc discussions without disturbing others.

To accommodate different needs and effortless switches between workspaces, attention should be paid to the ease of use and uniformity of technology as well as the easy adjustability of furniture, equipment and environmental factors such as lighting. Visible instructions can also facilitate a good user experience for those coming to the office less frequently or who lack routine in working in an ABO.

To support detaching oneself from work, the break rooms should be properly separated from work areas, also acoustically. Design features that facilitate recovery, such as daylight and nature views, should be incorporated (Ruohomäki et al., 2015).

Although research on post-pandemic office design is in its early stages, previous research points to various factors that promote the successful development of work environments and user well-being (Ruohomäki et al., 2015; Kämpf-Dern and Konkol, 2017). A shared understanding between the employer and employees of the role of the office in hybrid work, a participatory design process, and collecting feedback from users after workplace changes can facilitate considering different needs and requirements in workplace development and improving user satisfaction (Kämpf-Dern and Konkol, 2017).

The workplace experience contributes to telework preferences and is particularly related to whether or not employees want to increase teleworking. However, better privacy at the workplace is also a pull factor. Privacy can be supported by, for example, quiet workspaces and easy workspace-switching. The workplace experience appears to be more important for telework preferences than psychosocial factors, and should thus be considered an integral component of hybrid work arrangements.

This work was supported by the Finnish Work Environment Fund [grant 210092], the Finnish Institute of Occupational Health and the participating organizations. The authors thank Ms. Maria Hirvonen (MSc, Statistics) for her help in the collection and management of the survey data.