Spanish Workers’ Judgement of Telecommuting during the COVID-19 Pandemic: A Mixed-Method Evaluation

Abstract

This study explores the drivers of employees’ attitudes towards home teleworking with Tobit regression and fuzzy-set qualitative comparative analysis (fsQCA). Drawing from technology acceptance models, it derives hypotheses regarding variable relationships and telecommuting perceptions. Data were obtained from a survey with 3104 responses conducted by the Spanish Agency “Centro de Investigaciones Sociológicas” in Spring 2021. The results emphasize the pivotal role of the family-life impact in shaping telecommuting perceptions, alongside factors like location, ICT satisfaction, employer support, and job adaptability. The results from fsQCA reveal an asymmetric influence of input factors on the positive and negative evaluations. Positive perceptions are associated with family-life positivity, firm support, strong ICT, and non-provincial residence, while negative attitudes relate to family-life negativity, lack of employer support, and poor connectivity. The main innovation of this paper lies in the combined use of correlational and configurational methods, enriching insights into employee telecommuting perceptions beyond traditional regression analysis.

1. Introduction

The advantages of teleworking (TW) are embedded in employees, firms, and society, as the literature has widely argued [1,2]. The energy crisis in the early 1970s and the alleged advantages of TW led several academicians to predict its common adoption when the 20th Century ended [3]. However, this prediction was not accomplished at the beginning of the 21st Century [4] or the 2010s [5,6] in many territories. Differences in work regulations and practices, labor market idiosyncrasies, and the degree of development of the ICT infrastructure can explain the differences in their degree of development across countries and territories until 2020 [6]. For example, in European countries, although telecommuting experienced a remarkable development in Anglo-Saxon and Nordic countries, in Mediterranean states, such as Spain and Italy, the expansion was clearly scarcer [7].

As is well known, this panorama changed abruptly at the beginning of March 2020. Most states adopted rigorous measures to limit COVID-19 transmission; therefore, practically all firms and public agencies were obliged to develop their activities remotely [8].

In short, the COVID-19 pandemic expanded telecommuting and accelerated its implementation in firms’ work organizations [9], including enterprises whose employees had no experience in TW [8]. Thus, in 2020, the majority of workers experienced TW, especially if the tasks were well-suited. However, even in less adaptable jobs, procedures with firms’ administrative departments were conducted with the help of remote communications techs. Moreover, the majority of the population joined non-presential services and procedures, such as medical assistance, shopping, or tax payments, in a telecommuted manner [10].

Thus, the widespread adoption of the majority of world labor markets during the SARS-CoV-2 crisis represents an extremely valuable scenario for testing the limits and real applicability of teleworking practices [11,12]. Moreover, a massive shift toward teleworking is expected and will likely continue long after the pandemic [13]. It is commonly agreed that the COVID-19 pandemic has accelerated the process of digitalization, which has produced relevant changes in labor relations and, consequently, in companies’ organizations [14].

The reflections outlined in the above paragraphs motivate this research, which evaluates the impact of sociodemographic, environmental, and organizational/job factors on judgments about the influence of telecommuting. Although a great part of our analysis can be extended to all types of TW, it is constrained to home teleworking, also referred to by the literature as working from home (WFH), which was the main form of telecommuting practice experienced during the COVID-19 pandemic. This work was built with a survey carried out by the Spanish Government Agency Research Centre of Sociology in March 2021. Thus, it was developed one year after working from home became mandatory in the Spanish labor market because of the measures to limit the SARS-CoV-2 spread. Specifically, the research questions (RQ) analyzed in this paper are:

RQ1: What is the average influence of assessed sociodemographic, environmental, and organizational and job factors on the judgment of the influence of home teleworking on employees?

RQ2: How do sociodemographic, environmental, and organizational and job factors combine in the configurations linked with a positive and negative perception of home teleworking on employees?

Quantitative analysis by all reviewed literature is grounded in conventional correlational methods, such as regression analysis. The novelty of this study lies in the utilization of fuzzy-set qualitative comparative analysis (fsQCA), developed by Ragin [15,16], to complement the results obtained through regression methods. Our study demonstrates that fsQCA and regression are complementary techniques, as they allow for the analysis of data from two non-excluding perspectives.

In RQ1, we investigated the average influence of each variable on workers’ overall perception of working from home. Regression analysis is an appropriate method for this purpose because it is a variable-oriented technique that enables the evaluation of the average effects of input factors on output variables [17]. On the other hand, RQ2 aims to identify the ways in which explanatory factors combine to produce both positive and negative perceptions of the effects of teleworking. For this analysis, fsQCA is an appropriate method [18] owing to its case-oriented nature. This technique does not quantify the influence of each explanatory variable on the output value through a coefficient but rather displays sufficient conditions (which are not unique) for generating a given output [17].

In a regression analysis, an input variable can only be associated with an output variable with a single sign (positive or negative), and its significance is measured by the p-value. However, fsQCA allows for the presence or absence of a factor as conditions that jointly produce different configurations of output variables. It enables the capture of all nuances in the influence of the input variables on the studied outcome. As an example, we can outline the existence of studies reporting that younger workers have higher self-efficacy in the use of information and communication technologies (ICT) [19,20,21]. Therefore, it is logical to expect greater adherence to teleworking among younger employees. However, other authors indicate that adaptation to WFH during the COVID-19 pandemic was greater in older employees [22], which aligns with the literature suggesting that mental health issues related to remote working are less likely in older age groups [12]. Therefore, the use of fsQCA will reveal that, in certain profiles, being young, in conjunction with other circumstances, is associated with the perception that teleworking is favorable for workers, whereas in other profiles, being young may be a condition for a negative perception [23].

Similarly, the mechanisms that lead to acceptance and resistance to new technologies are not symmetrical, and this holds true in the context of telecommuting [24]. In this regard, fsQCA allows for the capture of asymmetric explanations of the presence or absence of a specific outcome [17,23]. For instance, the need to commute long distances from home to the workplace is associated with a favorable perception of working from home [19,25]. This may explain the greater acceptance of working from home among residents of suburban areas compared to those residing in city centers [26], where administrative work centers are typically located. However, this does not necessarily imply that residing in the city center is a condition in configurations that explains resistance to telecommuting.

2. Theoretical Ground

2.1. Initial Considerations

In this study, we examine the factors influencing employees’ adherence to telecommuting, considering the suitability of remote working [27] and employees’ adaptation to (WFH) during the SARS-CoV-2 crisis [22]. We categorized these factors as sociodemographic, environmental, and organizational and job-related variables. The sociodemographic factors included sex, age, and educational level. The environmental variables encompass place of residence, perception of connectivity services received during the COVID-19 lockdown, and the perceived impact of WFH on family life. Organizational and job variables embed a firm’s support for teleworking, experience in home teleworking before the COVID-19 crisis, and job adaptability to remote work.

Considering that WFH implementation relies on the use of ICTs, workers’ motivations can ultimately be analyzed within theoretical frameworks, such as the technology acceptance model (TAM) proposed by Davis [28] or the more comprehensive unified theory of acceptance and use of technology (UTAUT) model developed in [29], which focuses on the evaluation of technology acceptance in the workplace. TAM suggests that the main influencing variables are performance expectancy and ease of use [28,30], whereas UTAUT incorporates social influence and facilitating conditions into these variables [29,31].

Following TAM, a worker’s perception of WFH is likely to be more positive if they perceive it as providing greater utility (e.g., reducing commuting costs) and if the technologies and systems used are straightforward (e.g., they are accustomed to using ICTs in their daily life). TAM has been used to explain attitudes toward teleworking from the firms’ perspective [32,33] as well as from workers’ points of view [34,35,36]. The inclusion of the social influence and facilitating conditions constructs from UTAUT offers additional insights into an employee’s judgment of WFH. Therefore, the perceived receptivity of the organization and colleagues toward WFH is a crucial factor that influences worker adherence to this work arrangement [13]. Additionally, organizational support, such as providing learning resources, enhances facilitating conditions [37] and thus enables greater acceptance of remote working practices [35].

The conceptual ground of our empirical analysis is displayed in Figure 1.

2.2. Sociodemographic Variables

The statistical relevance of sociodemographic variables to explain employees’ adherence to telecommuting has been extensively shown by a great deal of papers [38,39,40,41,42,43,44,45].

There is a stream of findings outlining a correlation between practitioners of teleworking and being male [46]. On the other hand, there are several reports pointing out that women often tend to have a more favorable attitude toward the WFH measure [19,21]. A commonly suggested explanation is that there has traditionally been a societal perception that home care plays a feminine role, and working from home allows for a balance between family and work obligations, thereby reducing work absences related to family matters [2]. Paradoxically, this social norm also explains why various reports indicate that women experience more negative effects from WFH. They have a more challenging role than men in managing the balance of household and work responsibilities [40] and typically perceive greater conflicts between working from home and family life [23,44,47] in a WFH setting. Thus, empirical experience is not conclusive regarding the impact of gender on employees’ judgments of home teleworking [48]. Therefore, from a statistical perspective, it is reasonable to assume that gender has no statistical effect on WFH acceptance and the following statistical hypothesis is proposed:

Hypothesis 1 (H1):

Gender has no significant impact on a positive perception of the effect of home telework on employees.

Younger workers have higher self-efficacy in utilizing ICTs [19,20,21]. Therefore, it seems logical to assume that telecommuting effort expectancy increases with age. However, it is worth noting that senior employees typically perform remote work [6], which explains why remote work-related mental health issues are less likely among older individuals [12,39,49]. This also aligns with findings such as [22], which indicate that adaptation to working from home during the COVID-19 pandemic was greater among older employees.

The perception that members of a specific generation hold regarding the role of ICTs in their daily lives is also relevant. For instance, Baby Boomers generally value face-to-face interactions and tend to perceive mediated activities as useless, whereas X-generation members tend to embrace autonomy and flexibility in work relationships [21]. Furthermore, Tripati and Bajpai [50] emphasized the significance of being technologically native to experience well-being in the use of ICTs, with the millennial generation being the first to be considered as such [51]. Thus, there are reasons explaining both the higher inclination toward telecommuting among younger workers and the similar inclination among older workers, possibly resulting in a null combined effect. Consequently, we propose the following hypothesis:

Hypothesis 2 (H2):

Age has no significant impact on employees’ judgment of the effects of home teleworking.

Teleworkers often develop jobs linked with knowledge-oriented tasks. The principal product of these tasks is information; thus, they have the relevant content of management and information transmission [52]. Therefore, highly skilled and high-status employees who often possess high education levels are more likely to be employed in telework-suitable jobs [7,53]. Thus, it is logical to suppose that they perceive the potential advantages of home teleworking more intensely than employees with lower educational degrees and tasks that involve less information load [35]. Likewise, in general, workers with a higher educational degree had, at the onset of the COVID-19 pandemic, more experience with this type of work arrangement and a better infrastructure to make HWT. Therefore, they must perceive lower effort expectancy and greater facilitating conditions to develop WFH. This explains why, before the SARS-CoV-2 crisis, in many companies, WFH was perceived as a workplace advantage associated with employees possessing greater skills and stronger negotiation abilities [20]. The mainstream literature often reports the positive influence of an educational degree on teleworking acceptance [25,35,46]. Therefore, the following hypothesis is proposed for statistical testing:

Hypothesis 3 (H3):

Having a higher educational level is linked to a positive perception of the effect of home teleworking on employees.

2.3. Environmental Factors

We understand environmental factors as those that will influence the conditions under which home teleworking takes place, distinguishing them from the organizational factors that pertain to specific aspects of the worker’s home environment rather than the characteristics of the company and its human resources policies. Thus, essential factors that characterize the employee’s home and impact their work include the residential address, the available information technology infrastructure, and the worker’s particular circumstances regarding the influence that remote work may have on their family, such as marital status, the presence of dependents, etc. Among the environmental factors that affect workers, this study focuses on the distance to work, the adequacy of ICT to perform WFH, and the perception of how working from home impacts family life.

Working from home eliminates time spent commuting to the workplace. The perceived usefulness of telework can be summarized by its ability to save time, as it reduces or eliminates commuting time that cannot be utilized for work, family, or leisure activities [2]. Additionally, teleworking can also lead to cost savings [48]. It is widely recognized that the time taken to travel to the workplace negatively impacts job-related well-being [47], primarily due to factors such as inconvenience, lack of control and predictability in commuting situations [54], increased probability of infection in crowded public transportation [55], and the risk of traffic accidents [56].

It has been reported in several papers that the time spent commuting to the workplace is often a relevant variable to explain adherence to telecommuting practices [19,25,48]. Thus, this explains why WFH has greater acceptance among people who do not live in administrative centers [26]. Therefore, the following hypothesis was formulated:

Hypothesis 4 (H4):

Living in and administrative capital is negatively related to employees’ judgments of home telework.

The adequacy of household resources for WFH is usually a key aspect of attitudes toward telecommuting [44,57] and embeds good connectivity and ICT equipment [10]. In fact, this is true not only in a telecommuting setting but also in deciding the place of residence [58]. WFH relies on the use of ICTs to perform tasks such as communicating with colleagues and supervisors [12] and accessing firms’ documents and data [20]. For an employee to be able to communicate or collaborate with colleagues adequately, the various technologies used at home, such as Zoom or Skype, should fit the purpose of teleworking [59].

Therefore, having a reliable ICT infrastructure is essential to facilitate satisfactory mediated interactions and ensure good access to data [2,20]. A common telecommuter’s concern is the drawbacks of obtaining information to carry out tasks from home [48] and, consequently, losing control of work [60] as well as isolation [12]. Poor ICT infrastructure creates discomfort, thereby acting as a potential stressor [7,50,61]. Therefore, perceiving ICT resources as adequate to develop jobs induces a perception of greater job relevance and results in demonstrability, which increases perceived usefulness [62], diminishes stress, increases self-efficacy and ease of use [28], and makes workers’ personal conditions compatible with tasks to carry on in-home, which is a condition for facilitating conditions [13]. Therefore, we propose the following hypothesis:

Hypothesis 5. (H5):

Perceiving that ITCs allowed for carrying on activities of sufficient quality during the COVID-19 crisis positively impacts the judgment about the impact of home teleworking on employees.

That WFH improves the balance between a person’s job and personal life has been outlined repeatedly [1]; thus, this balance has been proven to be an explanatory factor of job satisfaction and well-being at work [2,63]. Therefore, working from home reduces stressors linked to personal relations in the workplace, increases workers’ flexibility and autonomy, and allows a better compromise between familiar duties and jobs [2,64]. The COVID-19 pandemic has also led to the adoption of WFH by companies to protect employees [65] because isolation not only drastically diminishes the probability of being infected by a respiratory virus but also the stress linked with this fact [26]. Therefore, theoretically, the acceptance of WFH should come through both the perception of the utility of the work arrangement and the potential positive influence that working from home can have from the opinions of people in their own family unit.

On the other hand, blurred limits between work and home in regard to how time and resources are distributed between these two fields often induce conflict between the job and personal life [47,64]. The forced and rushed implementation of working from home during the SARS-CoV-2 pandemic has led to various studies confirming that WFH does not necessarily improve work–life balance [59]. The emergence of misunderstandings, tensions, and conflicts with family and friends, as well as an increase in stress in the private sphere, is frequently observed, both from the point of view of teleworkers and their loved ones [64]. An important reason for this may be that working from home can lead to an intensification of work and longer working hours, triggering health issues [66].

Having children to care for increases home duties, and therefore, attaining a compromise between home and work becomes more difficult. This fact may increase the usefulness of working from home, but paradoxically, it poses a challenge because of the limited resources in homes, such as spaces and ICT devices, which must be shared with other members of the household [5]. Therefore, there is a great deal of literature outlining that children at home amplify their conflict perceptions [38,42,44,57,67,68].

The physical space of the home dedicated to implementing telecommuting is relevant in explaining well-being at work [46]. Its limitations can facilitate work interruptions and, consequently, conflicts in households [69]. WFH environments need to use private places and material resources, such as laptops, which can lead to tense situations in families [39,68,69] and endanger job–family balance [37]. These drawbacks of home teleworking on work–family balance may negatively influence the perceived effort expectancy and facilitating conditions [13] of this work arrangement. Therefore, we propose the following hypothesis:

Hypothesis 6 (H6):

The feeling of the positive impact of home teleworking on family life positively influences employees’ judgments of this work arrangement.

2.4. Organizational and Job Factors

If the organization demonstrates a strong commitment to teleworking, it is expected to have a beneficial effect on how employees perceive their job performance when using information and communication technologies (ICTs) for teleworking. This, in turn, will influence the relevance and demonstrability of their work outcomes, as well as their perception of the usefulness of these tools [30]. Additionally, if the organization supports WFH, it will have a positive impact on the usability of ICTs for job execution, thereby diminishing the effort expectancy of the evaluated practices [62].

Employers’ behavior toward home teleworking can enable or act as a barrier to working from home since it influences social norms [65]. Supervisors’ trust and support are important resources for teleworkers and are related to perceived career opportunities and satisfaction with telework [48]. The attitudes of coworkers and managers moderate the image or reputation perceived by an individual when they make use of telework and the importance that it would have for completing work with this modality [33]. It has been widely documented that widespread employers’ and managers’ reluctance to remote working practices is due to the difficulty associated with monitoring workers who are not working from the office [2,24].

Furthermore, home teleworking can negatively impact the social support of coworkers and supervisors in the workplace due to issues such as weak personal linkages with the quality and dependability of ICTs to develop personal relationships [69]. This explains the relevance of supporting employees by providing them with learning sources [2,37], promoting mediated communication, and providing effective platforms to implement these practices. These measures may help to avoid undesired issues, such as feelings of isolation [12,70], and to increase job well-being and adherence to remote working practices [11,35]. Providing learning resources and organizational support increases the self-efficacy and usability of the system and reduces anxiety; thus, perceived ease-of-use benefits [62]. This support for teleworking embeds material and technical support in the home office [71], which is a direct enabler of facilitating conditions [13]. In many countries, there is no clear legal basis for how home offices must be equipped. Therefore, the equipment provided by employers is often not sufficient [72]; thus, employees have to obtain resources for telecommuting by themselves, such as greater connectivity or electronic devices. Therefore, the following hypothesis is proposed:

Hypothesis 7 (H7):

Organizational support has a positive impact on judgments about the influence of home telecommuting on workers.

Being experienced in home telework prior to the SARS-CoV-2 pandemic positively influenced employers’ judgments, as they were already accustomed to this job arrangement [34]. Workers who had previously practiced working from home before the COVID-19 lockdown measures allegedly had access to sufficiently large technical and hardware resources to effectively work from home [57]. Therefore, their facilitating conditions must be better than those of new telecommuters, who had to adopt telecommuting unexpectedly and under mandatory circumstances. Having more familiarity with telecommuted practices can enhance self-efficacy, leading to higher levels of well-being [59], reduced mental issues [12,49] and, consequently, lower effort expectancy.

Not all individuals are suited to working from home [7], as it requires specific qualities such as self-discipline, competence, and autonomy [59] to perform effectively in such an arrangement. Considering the limited prevalence of telecommuting in Spain until Spring 2020, it is reasonable to assume that experienced employees in this domain were well-suited for remote work. Based on this premise, the following hypothesis is proposed:

Hypothesis 8 (H8):

Having prior experience with remote working is linked to a positive perception of the effect of home telework on employees.

A job must attain several conditions to be fully adaptable to telecommuting, such as being cerebral rather than manual or having a high degree of autonomy [27]. Thus, employees in knowledge-intensive industries, advanced services, and the education sector are more likely to be eligible for teleworking. This also clarifies why individuals who typically perform intellectual tasks [20], several types of managers and professionals [14], and public sector employees [35] tend to show a greater acceptance of telecommuting.

However, jobs requiring highly skilled employees are not necessarily uniquely suitable for telecommuting. Those that do not require much communication are highly standardized and have clear goals and outputs, such as many clerical jobs, and are often regarded as suitable for working from home [46]. Likewise, teleworking has traditionally suited tasks in which employees control their work pace and have little need for face-to-face interaction and others’ input; therefore, it requires concentration and adaptation to individual work rhythms [7].

The perceived usefulness of teleworking is related to expectations of productivity and efficiency, which are greater if the work is adaptable to the use of ICTs. This productivity and efficiency are indicated by the expectations of work autonomy and improvement in work productivity and performance [33]. Thus, we propose the following hypothesis:

Hypothesis 9 (H9):

The degree of job adaptability to telecommuting positively affects the judgment of the effects of teleworking on employees.

3. Materials and Methods

3.1. Materials

This paper analyzed the survey of the Spanish Agency “Centro de Investigaciones Sociológicas”, commonly known by its Spanish acronym, CIS, in March 2021 and had 3014 responses. To obtain a relatively homogeneous sample, we were interested in the perceptions of employees, but not by employers and self-employers, who actually worked when the survey was responded to, in such a way that we analyzed a subsample of the survey that was selected by following the schema in Figure 2. We omitted entrepreneurs (employers and self-employers) since they have traditionally been reluctant to embrace telecommuting arrangements. A commonly cited reason is that home teleworking (HTW) is often linked to reduced organizational commitment, as teleworkers might develop a stronger commitment to working from home rather than to their organization, leading to a more transactional perspective on their relationship with their employer [2]. Furthermore, the implementation of telework arrangements requires an initial effort from the company, which can be substantial both in terms of financial investment and organizational adjustments [24].

Within the group of workers, to prevent potential biases arising from their differing status within the active population, we focused on those who were actively engaged in their work at the time of the survey’s administration. Therefore, responses from individuals within the active population who were not employed due to reasons such as being unemployed, having a disability due to illness, or being temporarily inactive due to a temporary employment regulation scheme were not considered. Thus, the sample used in this study had 1543 observations (51.19% of the overall survey).

Table 1. Profile of the sample and questions and issues used to state input variables.

IQ1: Sex	IQ2: Age	IQ3: Academic Level
Female (49.45%) Male (50.55%)	≤35 years (18.47%) >35 and ≤45 years (25.08%) >45 and ≤55 years (25.02%) >55 years (31.43%) Mean = 49.69 years. SD = 16.76 years.	Graduate and upper graduate (57.16%) Secondary/vocational training (9.46%) Primary school (20.16%) Others: (13.22%)
IQ4: Place of residence	IQ5: Connectivity
Capital of region (65.39%) Cap. of prov. (10.69%) Others (23.91%)	The failures in internet service have not been an added problem during the lockdown (79.68%). Having access to a wide range of audiovisual and internet services from home has made the lockdown more bearable (63.77%). Video conferences through the internet or mobile phones have allowed us to stay in touch with family and friends over the past year (71.22%). Working from home has enabled many professional activities to be carried out with high quality (56.90%). We have been able to enjoy high-quality movies without having to leave the house (52.62%). CONNECTIVITY: mean = 2.96, std. dev. = 1.94.
IQ6: Consequences of working from home on family
Positive issues Helps with family balance (51.98%). It allows parents to spend more time together (38.30%). It enables parents to take better care of their children (41.41%). It allows for better organization of work and nonwork times (42.45%). It saves time and money on commuting (1.10%). It is beneficial for people’s health and helps prevent contagion (0.32%). Other reasons (1.30%). Negative issues It subjects couples to constant cohabitation (9.79%). It prevents meeting, being with, and talking to other people (10.17%). It hinders the care of children (in their case) due to the dedication and attention to work (10.76%). Homes are not prepared to separate work from family life (12.38%). It is difficult to disconnect (1.36%). It entails extra expenses (0.13%). Other reasons (0.58%). FAMILY: mean = 1.32; std. dev. = 2.41
IQ7: Organization support		IQ8: Experience
Positive issues Your employer had already provided you with a laptop before the pandemic (16.59%). Your employer gave you a laptop that you did not have before the pandemic (10.24%). Your employer asked you to use your own computer until they provided you with a laptop (9.01%). You had to upgrade your communication capacity, and your employer reimbursed the cost (0.84%). Your employer organized technical support to assist employees in effectively implementing remote work (26.18%). Negative issues Your employer asked you to use your own computer during the lockdown (13.55%). You paid for the cost of internet connectivity (35.58%). SUPPORT: mean = 0.14, std. dev. = 0.77.		Teleworking before March 2020 was: Habitual work arrangement (2.66%). Frequent work arrangement (3.18%). Occasional work arrangement (10.50%). Was new for me/others (86.32%).
IQ9: Adaptability
Answering customer, user, patient, or student inquiries over the phone, email, or website (34.15%). Handling administrative tasks, troubleshooting technical issues, conducting studies, or other routine tasks by accessing digital databases or documents, both internal and external (37.27%). Engaging with employers, suppliers, users, customers, etc., through digital platforms (33.05%). Participating in work meetings through digital platforms (37.20%). Teaching classes or training activities (24.50%). Managing organization, coordination, and leadership tasks (30.59%). ADAPTABILITY: mean = 1.97, std. dev. = 2.52.

displays the composition by gender (49.45% females and 50.55% men). From Figure 2, it can be deduced that 87.10% of the answers came from employees of private firms, and 12.90% were provided by workers in the public sector. Table 1 displays that the respondents’ average age was 49.69 years, with a standard deviation of 15.76 years.

We believe that the sample is representative of the Spanish working population, primarily due to the stratification conducted [73]. Additionally, the sample’s size, surpassing 1000, permits us to accept a margin of error below 3%, considering that the analyzed population is large [74] since it encompasses over 20 million individuals [75].

The complete questionnaire presented by CIS [73] used in this paper inquired about a wide range of topics related to the digital society. Table 1 and

Table 2. Judgment of the impact of home teleworking on employees.

Output Question: Judgment of the Impact of WFH on Employees

Positive arguments Employees/workers have ownership of their time to organize their schedules (33.38%). Eliminates commuting (39.92%). Promotes work–life balance (40.38%). Increases productivity and professional development (more efficiency, higher performance, better concentration) (1.17%). Saves costs (0.84%). Because of convenience (in communications, less stress) (1.23%). Offers more flexibility and freedom (0.71%). Because of health safety and avoiding contagion (0.58%). Serves as a complement and a different form of work (0.58%). It prevents job loss (0.52%). Other reasons (0.45%). Negative arguments Promotes isolation (12.31%). It increases stress (9.40%). Makes it difficult to disconnect from work during rest times (10.50%). Allows working many more hours than the established ones (11.21%). It poses health problems (excessive screen time, discomfort, sedentary lifestyle) (0.58%). Decreases productivity (0.65%). Job losses and salary reductions (0.78%). Entails additional expenses (0.65%). Hinders work–life balance (0.39%). Other reasons (0.65%). JUDGEMENT: mean = 0.73; std. dev. = 2.05

solely display the questions and items used in this paper, which are confined to the model proposed in the second section, along with the descriptive statistics associated with the employee sample.

To have an appropriate perspective on the relevance of telecommuting in the Spanish labor market, according to data from the Ministry of Labor and Social Economy [75], 75.4% is engaged in the services sector, while the remaining 24.6% is distributed across agriculture, industry, and construction. If we discount the fact that in the sector, around 30% is dedicated to activities where a significant portion is difficult to implement through telecommuting, such as the logistics and transportation sector and the hospitality industry [76], more than 50% of the jobs in the Spanish economy are susceptible to being carried out, at least partially, through the use of information and communication technologies (ICTs).

3.2. Definition of the Variables and Membership Functions Used in the Study

The variables have been defined from the questions displayed in Table 1 and Table 2 in such a way that they are modeled to allow running Tobit regression to test the hypothesis developed in Section 2 and fsQCA. Thus, these variables are first quantified in an intuitive way and subsequently are calibrated to let them vary between [0, 1]. With regard to sociodemographic variables, we define the following:

FEMALE = It comes directly from IQ1 in Table 1 and takes 0 for males and 1 for women. Therefore, its normalized value $m_{FEMALE}$ is simply the value of that variable.

AGE = It differentiates the Baby Boom generation from younger generations. In Spain, the members of the Baby Boom generation are considered born in a period with an upper fuzzy boundary in the middle 1970s. This variable is defined from IQ2, which is simply the age at the date on which the survey was answered. Its normalized value (or membership function) is:

$$m_{AGE}\left(x\right)=\{\begin{array}{cc}1& x\ge 55\\ \frac{x-45}{10}& 45\le x<55\\ 0& x<45\end{array}$$

where x is the age of the respondent at the moment of being surveyed.

ACADEMIC = It is linked with the attained academic degree by the respondent. It is defined from IQ3 in Table 1, and its membership function, $m_{ACADEMIC}$, takes the following values:

$$m_{ACADEMIC}=\left\{\begin{array}{c}\begin{array}{cc}1& \mathrm{u}\mathrm{n}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{u}\mathrm{a}\mathrm{t}\mathrm{e}/\mathrm{u}\mathrm{p}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{u}\mathrm{a}\mathrm{t}\mathrm{e}\end{array}\hfill \\ \begin{array}{cc}0.8& \mathrm{s}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{o}\mathrm{o}\mathrm{l}/\mathrm{v}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{g}\end{array}\hfill \\ \begin{array}{cc}0.5& \mathrm{p}\mathrm{r}\mathrm{i}\mathrm{m}\mathrm{a}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{o}\mathrm{o}\mathrm{l}\end{array}\hfill \\ \begin{array}{cc}0& \mathrm{p}\mathrm{r}\mathrm{i}\mathrm{m}\mathrm{a}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{o}\mathrm{o}\mathrm{l}\mathrm{n}\mathrm{o}\mathrm{t}\mathrm{f}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{s}\mathrm{h}\mathrm{e}\mathrm{d}\end{array}\hfill \end{array}\right.$$

With regard to the three environmental variables outlined in Section 2, we define the following variables:

CAP_ADM = This is a proxy variable for proximity to the physical workplace. Usually, jobs adaptable to home teleworking are linked with administrative tasks, and therefore, the workplaces of such jobs are located in administrative cities. This variable indicates if the observation comes from people who reside in an administrative capital and, therefore, are allegedly close to the workplace, and is built up from the questions answered in IQ4. Therefore, a negative relation of this variable with the judgment about WFH is expected. Its normalized value within the interval [0, 1], $m_{CAP\_ADM}$, is defined as:

$$m_{CAP\_ADM}=\{\begin{array}{cc}1& \mathrm{if} \mathrm{the} \mathrm{response} \mathrm{comes} \mathrm{from} \mathrm{the} \mathrm{capital} \mathrm{of} \mathrm{the} \mathrm{region}\\ 0.8& \mathrm{if} \mathrm{the} \mathrm{response} \mathrm{comes} \mathrm{from} \mathrm{the} \mathrm{capital} \mathrm{of} \mathrm{the} \mathrm{province}\\ 0& \mathrm{otherwise}\end{array}$$

CONNECT = This variable is built up from the items indicated with regard to IQ5. Therefore, this variable can vary in {0, 1, …, 4, 5}, with a mean of 2.96 and a standard deviation of 1.94. The normalized value of that sum ($m_{CONNECT}$) is obtained from the parameters outlined in

Table 3. The 5th percentile, 95th percentile, and central values to build up membership function (1) for CONNECT, FAMILY, SUPPORT, ADAPTABILITY, and JUDGEMENT.

	CONNECT	FAMILY	SUPPORT	ADAPTABILITY	JUDGEMENT
$X^{\left(5th\right)}$	0	−4	−1	0	−4
$X^{\left(c\right)}$	3	0	0	2	0
$X^{\left(95th\right)}$	5	4	2	6	3

and by using Equation (1).

FAMILY = This variable is built up from issues declared in IQ6. Its first quantification comes from the sum of the perceived positive effects and the subtraction of the negative effects in such a way that the potential values of that variable are {−6, −5, …, 0, …, 8}. Its mean is 1.32, and its standard deviation is 2.41. The normalized value or membership value of that sum, $m_{FAMILY}$, is obtained by substituting the parameters in Table 3 in Equation (1).

Variables linked with job and organizational idiosyncrasies are defined as follows:

SUPPORT = quantifies the support of the firm to work from work during the COVID-19 lockdown. It is obtained by summing the declared positive items and subtracting the negative ones of IQ7 in such a way that the possible punctuations can vary within {−2, −1, 0, 1, …, 5}. The average value is 0.14, and the standard deviation is 0.77. Its normalized value within [0, 1], $m_{SUPPORT}$, is obtained by taking into account the parameters in Table 3 and Equation (1).

EXPERIENCE = quantifies the experience of the worker before the SARS-CoV-2 crisis. Its membership value $m_{EXPERIENCE}$ is built up from IQ8 as:

$$m_{SUPPORT}=\left\{\begin{array}{c}\begin{array}{cc}1& \mathrm{I}\mathrm{t} \mathrm{w}\mathrm{a}\mathrm{s} \mathrm{m}\mathrm{y} \mathrm{h}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{t}\mathrm{u}\mathrm{a}\mathrm{l} \mathrm{w}\mathrm{o}\mathrm{r}\mathrm{k} \mathrm{a}\mathrm{r}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t} \end{array}\hfill \\ \begin{array}{cc}0.8& \mathrm{I} \mathrm{f}\mathrm{r}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{l}\mathrm{y} \mathrm{w}\mathrm{o}\mathrm{r}\mathrm{k}\mathrm{e}\mathrm{d} \mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m} \mathrm{h}\mathrm{o}\mathrm{m}\mathrm{e}\end{array}\hfill \\ \begin{array}{cc}0.2& \mathrm{I}\mathrm{t} \mathrm{w}\mathrm{a}\mathrm{s} \mathrm{a}\mathrm{n} \mathrm{s}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{i}\mathrm{c} \mathrm{w}\mathrm{o}\mathrm{r}\mathrm{k} \mathrm{a}\mathrm{r}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{t}\end{array}\hfill \\ \begin{array}{cc}0& \mathrm{I}\mathrm{b}\mathrm{e}\mathrm{g}\mathrm{a}\mathrm{n}\end{array}\mathrm{w}\mathrm{o}\mathrm{r}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m} \mathrm{h}\mathrm{o}\mathrm{m}\mathrm{e} \mathrm{d}\mathrm{u}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{p}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{m}\mathrm{i}\mathrm{c}\hfill \end{array}\right.$$

ADAPTABILITY = Quantifies the capability of the work to be adapted to working from home. Its initial quantification is performed by summing the items of IQ9 that can take values in {0, 1, 2, …, 6}, and that sum has a mean of 1.97 and a standard deviation of 2.52. Subsequently, its membership function $m_{ADAPTABILITY}$ is obtained by applying the parameters in Table 3 in Equation (1).

With regard to the output variable, JUDGEMENT measures the degree to which employees feel that WFH is a good arrangement for them. It is initially measured by summing the declared positive arguments and subtracting the negative perceived items of the output question in Table 2 in such a way that the judgment may vary within {−10, −9, …, −1, 0, 1, …, 11}. Its mean and standard deviation are 0.73 and 2.05, respectively. Its normalized value, $m_{JUDGEMENT}$, is obtained by considering the parameters in Table 3 and Equation (1).

Table 3 shows the parameters to calibrate the membership function of CONNECT, FAMILY, SUPPORT, ADAPTABILITY, and JUDGEMENT. By naming $X$ to any variable of this set, $X^{\left(95th\right)}$ and $X^{\left(5th\right)}$ to its 95th and 5th percentile in the sample, and $X^{\left(\mathrm{c}\right)}$, a value that is considered central (for example, the median) and states the crossover point, the membership function of these variables ($X$) has been defined as:

$$m_X\left(x\right)=\left\{\begin{array}{c}1 x\ge X^{\left(95th\right)}\hfill \\ 1-\frac{X^{\left(95th\right)}-x}{2\left(X^{\left(95th\right)}-X^{\left(\mathrm{c}\right)}\right)} \mathrm{i}\mathrm{f}{X}^{\left(\mathrm{c}\right)}\le x\le X^{\left(95th\right)}\hfill \\ 0.5-\frac{x-X^{\left(5th\right)}}{2\left(X^{\left(\mathrm{c}\right)}-X^{\left(5th\right)}\right)} \mathrm{i}\mathrm{f}{X}^{\left(5th\right)}\le x\le X^{\left(\mathrm{c}\right)}\hfill \\ 0 x\le X\left(5th\right)\hfill \end{array}\right.$$

(1)

3.3. Data Analysis

The framework depicted in Figure 1 and developed in Section 2 grounds the data analysis developed in this section. To answer RQ1, we used a Tobit regression approach because JUDGEMENT varies in [0, 1]. This analysis is conducted by fitting a set of hierarchical regressions. To explain JUDGEMENT, we first introduce sociodemographic factors. Subsequently, we consider environmental factors and, finally, organizational and job inputs. To decide if a new block of variables enhances adjustment accuracy, we used the conventional information criteria by Akaike (AIC), Schwartz (BIC), and Hannan and Quinn (HQIC). The introduction of the new set of factors improves the explanatory capability of the new model if the values of the information criteria decrease. Tobit regression will allow us to state the net effect of each explanatory variable on the outcome and test the statistical hypothesis displayed in Section 2.

To answer RQ2, we performed fsQCA with the help of fsQCA 3.1 software [77] for the positive perception of home teleworking, JUDGEMENT, and for the negative perception ~JUDGEMENT, where “~” denotes negation or absence. The use of fsQCA allows us to visualize how different combinations in the presence or absence of certain factors lead to the same judgment, whether positive or negative, regarding telecommuting from home.

We first performed a necessity analysis of the conditions linked to these two outputs. A variable attains a status of necessary condition if its necessity measure is >0.9. Subsequently, we carried out an analysis of sufficiency by stating so-called prime implicates (recipes or configurations). To state configurations, we fitted two Boolean functions by means of the Quine–McCluskey Boolean minimization algorithm:

JUDGEMENT = f(FEMALE, AGE, ACADEMIC, CAP_ADM, CONNECT, FAMILY, SUPPORT,
EXPERIENCE, ADAPTABILITY)

~JUDGEMENT = f(FEMALE, AGE, ACADEMIC, CAP_ADM, CONNECT, FAMILY, SUPPORT,
EXPERIENCE, ADAPTABILITY)

Following Pappas and Woodside [17], to visualize how the factors impact output, we used a so-called intermediate solution. It is a set of configurations that embed the core conditions (those present in both intermediate and parsimonious solutions) and peripheral variables (that are only present in the configuration displayed in the intermediate condition).

The so-called consistency (cons) and coverage (cov) of prime implicates thus quantify their explanatory capability. Consistency measures the membership of a prime implicate in the outcome set [78]. For a prime implicate to be a sufficient condition, it must have a cons > 0.75. The coverage of a solution quantifies the proportion of the output set explained by that solution and can be interpreted analogously to a coefficient of determination R² [78].

4. Results

4.1. Regression Analysis

Table 4. Results of hierarchical Tobit regressions for the judgment of the impact of WFH on employees.

	Model 1		Model 2		Model 3
	Marg. Effect	p-Value	Marg. Effect	p-Value	Marg. Effect	p-Value
Sociodemographic variables
SEX	−0.058 *	0.0349	−0.001	0.9788	0.005	0.8140
AGE	−0.081 *	0.0205	−0.002	0.9498	−0.002	0.9571
ACADEMIC	0.125 **	0.0033	0.010	0.7986	−0.008	0.8522
Environmental factors
CAP_ADM	---	---	−0.062 *	0.0122	−0.066 *	0.0116
CONNECT	---	---	0.123 **	0.0010	0.100 **	0.0095
FAMILY	---	---	0.882 ***	<0.0001	0.870 ***	<0.0001
Job/organizational factors
SUPPORT	---	---	---	---	0.123 **	0.0038
EXPERIENCE	---	---	---	---	0.059	0.1167
ADAPTABILITY	---	---	---	---	0.082 *	0.0425
	Measure		Measure	Δ	Measure	Δ
AIC	2624.02		2017.28	−606.74	2007.30	−9.98
BIC	2650.73		2060.01	−590.72	2066.05	6.04
HQIC	2633.95		2033.17	−600.78	2029.15	−4.02
Likelihood ratio	29.427 ***		658.73 ***		675.868 ***

Notes: (1) “*”, “**” and “***” stand for significance at 5%, 1%, and 0.01%. (2) AIC = Akaike’s information criteria; BIC = Schwartz information criteria; HQIC = Hannan–Quinn criteria.

displays the results of the Tobit hierarchical regressions on a judgment about the influence of working from home on workers. All adjusted models are statistically significant because the likelihood ratio test rejects that the model does not improve an equation with only a constant. The p-value (p) is <0.01. The introduction of environmental factors in Model 1 to adjust Model 2 provides a clear improvement of the adjustment in all the measures. Similarly, the incorporation of job and organizational factors into the construction of Model 3 also enhances Model 2. However, this improvement is not as evident, as it is relatively small in terms of the AIC and HQIC criteria and even worsens according to the Schwartz criterion.

We can observe that in the simplest model (Model 1) that only includes the sociodemographic variables, all three considered variables are significant. Being male is associated with a higher acceptance of working from home, with a marginal effect (me) of −0.058 and a p = 0.0349. Additionally, younger employees (me = −0.081, p = 0.0205) and those with higher academic degrees (me = 0.125, p = 0.033) seem to be related to a better assessment of the impact of WFH on workers.

In Model 2, with the introduction of environmental variables, the sociodemographic variables lose their significance. Instead, the three new variables become significant. While residing in an administrative capital is inversely related to a favorable perception of WFH (me = −0.0621, p = 0.0122), being satisfied with the existing connectivity at home and the suitability of ICT (me = 0.122, p = 0.001) and perceiving a positive influence of telecommuting on family life (me = 0.882, p < 0.0001) are directly related.

The inclusion of organizational and job-related variables maintains the lack of significance of sociodemographic factors. Similarly, the environmental factors maintain their significant influence and, of course, the direction of that influence. Thus, for CAP_ADM, we observe me = −0.063, p = 0.0116; in the case of CONNECT, me = 0.1 (p = 0.0095); and for the perception of the positive impact of WFH on family, me = 0.870 (p < 0.0001). Among the three job and organizational factors, perceiving that the organization supports WFH, SUPPORT (me = 0.123, p = 0.0038) and ADAPTABILITY (me = 0.082, p = 0.0425) are variables with a significant positive influence on the judgment that employees make regarding working from home for themselves.

Table 5. Acceptance and rejection of hypotheses in Section 2 (H1-H9) after adjusting hierarchical regressions.

	Hypothesized Sign	Model 1	Model 2	Model 3
FEMALE	Null effect	Not supported	Supported	Supported
AGE	Null effect	Not supported	Supported	Supported
ACADEMIC	Positive	Supported	Not supported	Not supported
CAP_ADM	Negative		Supported	Supported
CONNECT	Positive	---	Supported	Supported
FAMILY	Positive	---	Supported	Supported
SUPPORT	Positive	---	---	Supported
EXPERIENCE	Positive	---	---	Not supported
ADAPTABILITY	Positive	---	---	Supported

summarizes the results of testing H1-H9 that allow answering RQ1.

4.2. Fuzzy-Set Qualitative Comparative Analysis

As a first step in this analysis, we performed a necessity analysis to check how much the presence or absence of a specific condition could be considered a cause for a positive or negative judgment.

Table 6. Necessity analysis of the simple conditions on positive and negative judgements.

	JUDGEMENT (Y)		~JUDGEMENT (~Y)
Condition (X)	Consistency of X⇒Y	Consistency of ~X⇒Y	Consistency of X⇒~Y	Consistency of ~X⇒~Y
FEMALE	0.63	0.67	0.37	0.37
AGE	0.66	0.72	0.42	0.37
ACADEMIC	0.70	0.71	0.37	0.51
CAP_ADM	0.63	0.66	0.39	0.35
CONNECT	0.75	0.66	0.36	0.49
FAMILY	0.85	0.63	0.34	0.73
SUPPORT	0.81	0.78	0.46	0.55
EXPERIENCE	0.64	0.69	0.37	0.34
ADAPTABILITY	0.75	0.63	0.33	0.40

Notes: (1) “X” stands for any input factor, and “Y” stands for JUDGEMENT. (2) With “~”, we denote absence (or negation).

shows that there are no necessary conditions. In regard to positive judgments, only FAMILY (cons = 0.85) and, to a lesser extent, SUPPORT (cons = 0.8) come close to being necessary conditions, which is usually at cons = 0.9. For negative judgments, there are no conditions that come close to being necessary.

Figure 3 shows the results of the sufficiency analysis solutions for the perception of positive judgment (JUDGEMENT). The solution is fairly consistent (above 0.86), and the coverage is 0.437. FAMILY, CONNECT, SUPPORT, and EXPERIENCE should be present in configurations where they are core conditions. Additionally, CAP_ADM must always be absent in the configurations of JUDGEMENT that it is a part of. The two variables that appear most as core conditions are FAMILY (present as a key condition in 8 out of 14 conditions) and connectivity (present in 7 out of 14). Regarding being female and age, we found that these variables can be both present or absent in configurations where they appear as conditions. Both academic level and job adaptability are conditions that should be present in the solutions they are part of, but always as peripheral factors.

Figure 4 displays the prime implicates linked to a bad perception of the consequences of telecommuting for workers. The set of recipes explaining ~JUDGEMENT has an adequate consistency (cons = 0.80) and a clearly smaller coverage than for JUDGEMENT (cov = 0.29). We can check that the influence of FAMILY and CONNECT have a univocal sign in the configurations where they take part. We can check that the main condition in configurations is the negative perception of WFH on family (seven times, core condition) and the absence of adequate connectivity (three times, core condition). The factors EXPERIENCE and SUPPORT (CAP_ADM) tend to be absent (present) in configurations where they are conditions. However, there are exceptions. Having experience is a core condition in the fourth configuration, the presence of firm support is a peripheral condition in the fifth configuration, and not residing in an administrative capital is a core condition in the second. AGE and FEMALE, once again, appear as conditions in certain configurations, both when they are present, meaning when they are associated with older workers and women, and when they are absent, referring to younger workers and men. Finally, neither academic level nor job adaptability manifests to be a condition in any recipe.

5. Discussion

5.1. General Considerations

This work assesses the influence of sociodemographic, environmental, and organizational/job factors that influence Spanish employees’ acceptance of working from home (WFH). The analysis relies on a survey conducted in March 2021, one year after the COVID-19 pandemic was declared. In that particular moment, home teleworking became an extended work mode, at least in jobs compatible with telecommuting, and thus, a significant proportion of Spanish citizens had a founded opinion about WFH. This research tries to answer two research questions:

RQ1: What is the average influence of assessed sociodemographic, environmental, and organizational and job factors on the judgment of the influence of home teleworking on employees?

RQ2: How do sociodemographic, environmental, and organizational and job factors combine in the configurations linked with a positive and negative perception of home teleworking on employees?

The explanatory variables addressed in the research questions can be grouped into three types: sociodemographic variables, environmental variables, and organizational and job factors. In all cases, the influence of these variables on the judgment made about the impact of telecommuting on workers has been justified by combining the theoretical foundations provided by the technology acceptance model (TAM) [28] and the unified theory of acceptance and use of technology (UTAUT) model [29] with the literature about telecommuting acceptance.

The quantitative analysis of these research questions was carried out using two analytical instruments whose sequential application has proven effective. The first research question has been answered using hierarchical regression, which allows measuring the average influence of the variables evaluated on the judgment of employees about the impact of telecommuting on them. We have found that the perceived influence of telecommuting on family (FAMILY), the opinion of home connectivity (CONNECT), employer support for telework (SUPPORT), and job adaptability (ADAPTABILITY) have a positive influence on judgment. Undoubtedly, the most influential variable is FAMILY. However, residing in an administrative capital (CAP_ADM), where there are a majority of workplaces of jobs adaptable to telecommuting, has a negative influence. Being female (FEMALE), age (AGE), the academic level of the survey respondents (ACADEMIC), and having previous telecommuting experience before the COVID-19 pandemic (EXPERIENCE) do not have a clear statistically significant influence on telecommuting (WFH).

The data analyses carried out to answer the second research question were implemented using fsQCA. It has allowed us to examine, with a case-oriented focus, how explanatory variables combine to generate positive judgments (JUDGEMENT) and negative judgments (~JUDGEMENT) regarding the impact of telework. In both cases, the necessity analysis has shown that the presence or absence of certain factors did not act as a necessary condition for JUDGEMENT or ~JUDGEMENT. The sufficiency analysis presented below enables us to visualize the different ways in which the input variables contribute to positive and negative judgments of telework. This analysis also revealed that there are instances of both symmetric and asymmetric relationships between variables.

We can assert that the positive influence of factors, such as FAMILY and CONNECT, on the judgment of WFH demonstrates a clear impact on both positive and negative judgments and that their influence is symmetric. Their presence is a condition in the various prime implicates linked to JUDGEMENT, while their absence is associated with ~JUDGEMENT. On the other hand, variables such as ACADEMIC and ADAPTABILITY suggest a positive influence on attitude toward WFH. Their presence is a condition in several configurations in JUDGEMENT; however, they are irrelevant in the formation of negative judgments. This suggests that the influence of these variables on JUDGEMENT and ~JUDGEMENT is asymmetric.

We can also identify the positive (negative) influence of EXPERIENCE (CAP_ADM) on workers’ perceptions as asymmetric since it is evident that their presence (absence) is an unequivocal condition in the recipes associated with JUDGEMENT in which they are a part. However, this unequivocal influence is not observed in ~JUDGEMENT, as both the presence and absence of these variables are conditions associated with the prime implicates linked to negative judgment.

We have found that being female has a significant negative link with adherence to telecommuting only in the simplest regression model. However, in those more complete models, gender has no statistical significance, which aligns with the fact that the relevance of gender to telecommuting acceptance is ambiguous [48]. However, the absence of statistical significance about the influence of gender on adherence to telecommuting should not lead to the deduction that gender does not influence adherence. The arguments in Section 2.2 lead to the inference that, in certain aspects, the traditionally assigned role of women induces a positive perception [19,21]; however, in other reports, this perception is negative [44,48]. Thus, the configurational analysis captured that being a woman is a condition for both a favorable judgment and for inducing a negative judgment, which allows the reconciliation of both types of arguments.

Although age may not be statistically significant in the most accurate regression models, it should not be inferred from this result that age is not an explanatory condition for workers’ attitudes toward telecommuting. Rather, the favorable and unfavorable arguments toward telecommuting based on age seem to cancel each other out. The fsQCA reveals profiles of individuals belonging to the Baby Boomer generation who show adherence to telecommuting, which aligns with the findings in [22] during the SARS-CoV-2 crisis. However, the fsQCA also detected the profiles of workers with a negative judgment toward working from home (WFH), which could be associated with the tendency of Baby Boomers to perceive mediated activities as nonproductive [21].

The configurational analysis identifies profiles of young workers who express a favorable judgment toward telecommuting that may be linked to younger individuals who value the autonomy provided by WFH [21] and have lower effort expectancy when using ICTs [19,20,21]. Additionally, the finding that belonging to younger generations is also a condition in some explanatory configurations for a negative judgment could be explained by the fact that younger workers are more sensitive to certain side effects of WFH, such as mental health issues [12,39,49].

We have found that educational status has a significantly positive impact on adherence to home teleworking only in the case of the regression model that exclusively considered the sociodemographic variables. However, in the best model, which includes all proposed variables, this variable does not have a statistically significant impact. The use of the fsQCA in the sample provides a new perspective on the influence of this variable on the perception of working from home. Having a higher educational status is a condition in the configurations associated with a positive perception of teleworking. However, this condition is weak, as it is peripheral. Additionally, this factor is not relevant in the prime implicates linked to telecommuting resistance. Thus, we can conclude that if any influence of educational status exists on the judgment of telecommuting, despite being weak, similar to the revised literature [7,25,35,46,53], it must be positive.

The results of the Tobit regression align with the reviewed studies, which consistently report a clear trend indicating a positive judgment toward WFH among individuals who do not reside in administrative center localities [2,19,26,48]. The use of the fsQCA allows us to more precisely calibrate how this relationship is manifested. Thus, in the profiles of workers who value WFH positively, when residing in an administrative capital is a condition, the variable representing the administrative capital (CAP_ADM) must be absent, which perfectly aligns with the regression result. In the profiles indicating rejection of WFH, when the CAP_ADM variable appears as a condition, although it typically requires the worker to reside in an administrative capital, there is a prime implicate where the employee resides outside the administrative capital. This last finding may suggest that the place of residence might have problems with internet connectivity [58].

The combination of regression techniques with the fsQCA may shed light on the statistical significance of CAP_ADM, indicated by an appropriate p-value (<0.05) but with a very small coefficient (−0.062). From a statistical standpoint, its high level of significance can be attributed to the low variance of the coefficient estimator, likely stemming from a large sample size. On the other hand, the relatively small absolute value of the coefficient might be explained by the fact that while there is no explanatory configuration for “respondents engaged with HTW” that includes residing in an administrative capital as a condition, there is an explanatory configuration for its rejection where not residing in an administrative capital serves as a condition with relatively high coverage. This subset of the sample could potentially offset the average effect on the CAP_ADM coefficient in WFH acceptance due to the influence of the prime implicates in which residing in an administrative capital is a condition for having resistance to WFM.

The results of the Tobit regression analysis indicate that satisfaction with the existing home connectivity significantly and positively influences the evaluation of telecommuting. This finding is consistent with the reviewed literature [2,7,12,20,44,57,59,60]. The results of the configurational analysis also support the statistical findings. Therefore, the presence of a positive perception of good connectivity is a prerequisite for the configurations that favor home teleworking, whereas its absence is associated with the configurations that perceive home teleworking negatively.

The empirical analysis of the influence of organizational support (SUPPORT) on the assessment of telecommuting’s impact on workers informs us that it is positive and significant, which aligns with the mainstream of the reviewed literature [11,12,35,37,65,70]. The results of the fsQCA confirm the clear, positive influence of SUPPORT on a positive assessment (its presence is a condition in five configurations) and the relevance of the absence of organizational support in forming a negative judgment (it is a core condition in five prime implicates out of eight). We have also observed that the existence of employer support is present as a peripheral condition in one configuration of ~JUDGEMENT. This may be due to the effects that perceiving greater employer support for telecommuting can have a negative perception on some respondents, such as a feeling of increased surveillance and control [2,72] or higher demands from the employer [79].

The correlational analysis revealed that experience may positively impact the judgment of WFH. However, this impact was not statistically significant. The configurational analysis demonstrated that in the prime implicates of JUDGEMENT, where experience is a condition, it must be present. This aligns with the predicted direction of the influence of experience on judgment and is consistent with the findings from the reviewed literature [7,34,49,57,59]. Conversely, while it is typically expected that recipes explaining ~JUDGEMENT require the absence of experience when it is a condition, it is also true that there exists a configuration of negative judgment where having experience is a condition.

The regression analysis reveals that, consistent with our ninth hypothesis, job adaptability (ADAPTABILITY) tends to significantly contribute to a positive judgment of home teleworking by employees. This finding is in line with the literature [7,14,20,27,33,35,46]. The application of the fsQCA allows us to observe that the impact of the degree of job adaptability has an asymmetric effect on the formation of positive and negative judgments. While the presence of ADAPTABILITY is a peripheral condition in 10 configurations where the output is JUDGEMENT, this factor, whether present or absent, does not play a role in the solutions related to a negative judgment.

5.2. Theoretical Implications of This Paper

From an analytical perspective, we have observed that the use of the fsQCA enriches the analysis obtained through correlational methods such as regression. While the latter quantifies the average impact of each input variable on the judgment made about teleworking [17], the fsQCA allows for the identification of different paths through which interactions between variables ultimately result in a specific judgment among workers, either positive or negative, regarding telecommuting [18]. Thus, the low statistical significance of a particular variable can imply that it may not have great relevance in the prime implicates identified in the fsQCA.

For instance, in the simplest regression model, an academic degree has a statistically significant positive impact. However, in the rest of the models, it is not significant. Thus, from a statistical standpoint, we could infer that a higher level of education has a slightly positive influence on favorable judgments towards telecommuting, confirming previous findings [25,35,46]. The presence of ACADEMIC as a condition in the configurations related to adherence to telecommuting aligns with the postulated positive influence of the academic degree. Nevertheless, the presence of ACADEMIC is always a peripheral condition in cases of JUDGEMENT and does not play a role in any explanatory prime implicate for resistance to telecommuting. This suggests that the relationship between acceptance of home-based telework and having an academic degree is consistently positive but with a relatively low intensity.

However, the statistical no significance of a variable can also indicate that it serves as a core condition in configurations of ~JUDGEMENT, with a contradictory sign that cancels out its average impact [23]. In the case of being female, the lack of significance in the regression coefficient was not attributable to the low importance of this variable. Instead, the fsQCA identified configurations consistent with a positive perception of females toward home-based teleworking (HTW) where gender served as a core condition [19,21], as well as other prime implicates that captured the possible greater resistance of females toward this work arrangement, also as a core condition [44,48].

Similar considerations can be made regarding age. Despite its lack of statistical significance, in both configurations of acceptance and resistance to HTW, age contributes as a positively (negatively) related condition in some configurations explaining acceptance (resistance), and in other cases, its contribution is of an opposing sign. This analysis helps reconcile arguments that suggest younger workers have higher self-efficacy in utilizing ICTs [19,20,21], implying greater adherence to telecommuting, with findings such as those from [22], which indicate that adaptation to home-based remote work during the COVID-19 pandemic was greater among older employees due to their greater experience with this job arrangement.

5.3. Practical Implications of This Paper

The impact of telecommuting on family life was the main explanatory variable for workers’ judgments of this work arrangement. Therefore, it is necessary to maximize the positive effects of telecommuting on family life and minimize its negative impacts. In Spain, some negative effects, such as the greater difficulty of disconnecting from work, are avoided as much as possible through legal regulations, such as the recognition of the right to digital disconnection, stipulated by Royal Decree 28/2020. However, other aspects, such as the material and logistical support that firms must provide to their telecommuters, still do not have clear regulations.

However, these measures must be accompanied by a firm’s culture and employee behavior. The change in organizational culture to the use of flexible working implies shortening and rationalizing information flows and channels, training employees for a new work arrangement, adapting them to a new organizational culture, and modifying the infrastructure of the organization [9]. Additionally, workers face a labor environment in which achieving a balance between work and home life is crucial. When working from home, individuals have to be conscious that their homes are also workplaces, and this fact needs careful management. It is essential to establish clear boundaries to achieve a balance between personal and professional obligations. This balance is vital in achieving a good balance between work and home.

Our mixed-method approach reveals not only the mean impact of the assessed factors but also how these factors combine with adherence and nonadherence to telecommuting and how these factors interact to determine adherence or nonadherence to telecommuting. Consequently, as was reported in [34], configurational analysis allows for identifying the existence of diverse worker profiles in relation to telecommuting and both positive and negative perceptions of working from home. A critical aspect for companies implementing telecommuting is to identify the worker profiles that are better suited to telecommuting [2]. In [24], it is shown that the explanatory factors for the acceptance of home-based telework do not symmetrically impact the acceptance and rejection of telecommuting. In this study, the implemented configurational analysis enables the identification of worker profiles with receptive and resistant attitudes towards home-based telework. Furthermore, the prime implicates linked to acceptance and resistance are not symmetrical.

To successfully implement teleworking, organizations should not only address the variables that have a significant overall impact on the perception of home teleworking goodness but should also acknowledge the existence of workers with distinct profiles and, accordingly, cater to their specific needs.

Jobs that are susceptible to being significantly affected by telecommuting, which, according to data from [75,76], could be estimated to represent over 50% of the Spanish economy; thus, the findings presented have a broad scope. An alternative perspective on their relevance within the labor market can be obtained from data provided in [80], which indicated in 2020 that 6% of Spanish workers were engaged in administrative tasks, 20% of employees were involved in managerial tasks to some extent, 28.96% of professions were dedicated to customer service roles, 3.1% of workers were in the information and communication sector, and 3.07% and 3% belonged to the education sector. The majority of these professions can perform a significant portion of their functions through teleworking arrangements, as was observed during the mobility restrictions resulting from the SARS-CoV-2 pandemic. Regarding the teleworkers’ experiences during the pandemic, 39% of individuals in Spain who telecommuted wish to maintain that routine; 34% want to increase their hours, and only 26% would like to reduce that time [81]. Thus, the presented report can be useful for decision-making regarding the implementation of telecommuting in firms, as a significant portion of employees might raise demands regarding their working conditions and advocate for a greater adoption of telecommuting.

6. Conclusions

This study investigates the factors that influence Spanish employees’ perceptions of telecommuting as a favorable work arrangement. This study employs a mixed-methods approach using Tobit regression and fsQCA. The former allows for the quantification of the average effect of various variables on the overall judgement of working from home, whereas the latter identifies different worker profiles associated with positive and negative attitudes toward teleworking. This study is unique in its use of configurational methodology to examine the impact of telecommuting on the workers’ adherence to home teleworking. The Tobit regression reveals that factors, such as place of residence, connectivity, perception of the impact of home–work balance, organizational support, and job adaptability, all have a clear influence on employees’ judgments of teleworking. The fsQCA demonstrated how these variables, as well as gender, age, academic degree, and experience, interact to produce attitudes of adherence and rejection toward remote work.

The limitations of this study highlight the need for future research. The primary data used in this paper is derived from a cross-sectional survey conducted in Spain in March 2021. This was a time when governments worldwide were highly concerned about SARS-CoV-2 transmission while vaccination coverage was low and telecommuting was mandatory. To gain a comprehensive understanding of employees’ judgements of telecommuting, it is essential to incorporate longitudinal observations during different phases of the COVID-19 pandemic. For instance, examining workers’ perceptions when SAR-COV-2 becomes an endemic virus would be valuable.

Although remote work was widely adopted at the time of the survey, it was an exceptional circumstance since telecommuting was often not optional. Additionally, during the spring of 2020, children were not attending school due to the general lockdown measures. Consequently, the challenge of adapting to a new job arrangement was further exacerbated by the presence of children at home during working hours. Moreover, mental health issues associated with the COVID-19 health measures may influence perceptions regarding the impact of WFH on well-being and work–home balance. These factors underscore the importance of adopting a longitudinal approach to gain a comprehensive perspective on the various factors that influence employees’ judgments of working from home.

This study focuses on Spain, where telecommuting was less prevalent before March 2020 compared to other countries, such as Anglo-Saxon or Nordic states, due to different labor regulations, labor market practices, and culture. Therefore, extending the results in this paper to other regions must be done with care. However, a combination of correlational and configurational methods may be useful in conducting similar studies at the geographical area level or within specific firms or economic sectors to identify patterns related to the factors that contribute to or impede working-from-home employees’ judgments.