Teleworking during the COVID-19 pandemic: the job demands and job resources associated with telework outcomes

ABSTRACT

Objective

One of the most visible changes of the COVID-19 pandemic has been the swift and widespread adoption of telework. Existing knowledge on teleworking has been largely limited to employees from certain occupational groups who voluntarily engaged in telework, making it difficult to generalise prior findings. Using the Job-Demands Resources model as a theoretical framework, the objective of this study was to identify the job demands and job resources associated with telework satisfaction and telework performance during the pandemic.

Method

During a period of enforced teleworking during COVID-19 lockdown measures, this study surveyed employed individuals (N = 208) from Melbourne (Australia).

Results

Using ordinal logistic regression analyses, it was found that three job resources (home office comfort, technical support, and job security) were positively associated with both telework satisfaction and telework performance. Further, technical resources (job resource) was positively associated with telework satisfaction, whereas professional isolation (job demand) was negatively associated with telework satisfaction.

Conclusions

During a mandated period of telework, this study identified several factors associated with important telework outcomes. As there has been a greater uptake of remote and hybrid working arrangements in Australia post-pandemic, these findings could help inform and optimise future telework policies.

Key Points

What is already known about this topic:

1. Prior to the COVID-19 pandemic, teleworking was not a commonly used workplace practice.

2. Existing knowledge on teleworking has been largely limited to employees from certain occupational groups who were interested in, and/or were able to engage in telework.

3. The resulting selection bias might limit the generalisability of these findings to others with different abilities, preferences, and/or job types.

What this topic adds:

1. This study investigated the job demands and job resources associated with telework success during the COVID-19 pandemic.

2. This study identified three job resources (home office comfort, technical support, and job security) associated with both telework satisfaction and telework performance.

3. The findings from this study point to several telework environment, job, and organisational factors that managers and practitioners could consider when developing future telework guidelines within an Australian context.

KEYWORDS:

• COVID-19
• telework
• Job-Demands Resources model
• remote work
• telework satisfaction
• telework performance

The coronavirus disease 2019 (COVID-19) has had unprecedented health, social, and economic impacts on individuals, organisations, and societies around the world. As half of the world’s population were confined to their homes by April 2020 (Sandford, 2020), one of the most historic transformations triggered by the pandemic has been the rapid and widespread adoption of teleworking (Kramer & Kramer, 2020). Also referred to as remote working, working from home, telecommuting, distributed working, and virtual working, teleworking describes a work practice where work activities are performed from a remote location, rather than a conventional workplace, using technology to complete tasks and communicate with others (Allen et al., 2015; Baruch, 2001).

Prior to the COVID-19 pandemic, teleworking was not a commonly used workplace practice. In the USA, for example, 71% of employees were teleworking in October 2020, compared to just 20% of employed adults before the coronavirus outbreak (Parker et al., 2020). In Europe, almost 40% of those employed were teleworking fulltime in April 2020, compared to just 15% before the outbreak of COVID-19 (Milasi et al., 2021). And in Australia, the proportion of employed individuals working from home increased from 32% (August 2019) to 53% (April 2020); the highest rate of teleworking ever recorded (Australian Bureau of Statistics, 2022).

Prior to the pandemic, teleworking was a “luxury for the relatively affluent” (DeSilver, 2020), enjoyed primarily by higher-income earners and “knowledge” workers (Milasi et al., 2021; Wang et al., 2021). In the USA, for example, 24% of management, business, and financial occupational groups had access to telework in 2019, compared to just 1% of service occupational groups (DeSilver, 2020). For many workplaces, however, the pandemic forced a shift in telework from an optional to a mandatory workplace practice, irrespective of employees’ preferences, abilities, and job types (Franken et al., 2021). This surge in telework since the onset of the pandemic has inadvertently resulted in a global teleworking experiment (Kramer & Kramer, 2020).

While there is a comprehensive evidence base that explores the practice of teleworking prior to the pandemic (e.g., Allen et al., 2015), its applicability to the unique circumstances of teleworking during the pandemic has been questioned (e.g., Wang et al., 2021). Prior research has been conducted with individuals from certain occupational groups who were interested in, and/or were able to engage in telework (Wang et al., 2021). The resulting selection bias might limit the generalisability of these findings to others with different abilities, preferences, and/or job types. The pandemic experience offers a unique opportunity to examine telework experiences within an enforced telework context which can help inform the development of efficient telework practices post-pandemic. The overall aim of the current study, therefore, was to investigate the relationship between a range of job demands and job resources with telework outcomes during a period of enforced teleworking in Melbourne (Australia).

According to the Job Demands – Resources (JD-R) model, job demands are the physical, psychological, social, or organisational aspects of the job that can cause stress, including workload, job ambiguity, and time pressure (Bakker & Demerouti, 2007; Demerouti et al., 2001). In contrast, job resources are the aspects of the job that can support employees’ goal attainment, development, and wellbeing, such as the suitability of the home workspace, the availability of digital resources, job autonomy, and social support. Specifically, the JD-R model assumes that higher job demands results in stress and burnout, which leads to negative work outcomes (e.g., reduced job performance and job satisfaction; Schaufeli & Taris, 2014). In contrast, the JD-R model predicts that a greater availability of job resources boosts work engagement which results in positive work outcomes (e.g., increased job performance and job satisfaction; Schaufeli & Taris, 2014).

As Allen et al. (2015) and Baruch and Nicholson (1997) emphasise that the success of teleworking programs is dependent on a range of different factors, this study adopted a multifaceted approach to understand the contributions of different types of factors to telework outcomes. Focussing on those job characteristics that were heightened for many during the pandemic, this study examined the association of various telework environment factors (home office comfort, equipment, and privacy, shared home office space, technical support, technical resources, work-family conflict, living with children), job factors (job security, job autonomy) and organisational factors (organisational support, professional isolation) with telework success.

Overview of the telework literature during the COVID-19 pandemic

Telework environment factors

Various physical characteristics (e.g., home office comfort, home office equipment, home office privacy, shared workspace) and technological elements (e.g., technical support, technical resources) of the telework environment have been found to be associated with telework outcomes during the pandemic. Carillo et al. (2021) found that appropriate telework home conditions, including a workspace free from distractions and the provision of adequate telework tools, was the second most critical factor influencing telework adjustment during COVID-19. Further, Puglisi et al. (2021) found that individuals who performed work activities in shared workspaces reported a higher sense of productivity loss. As suitable and reliable technological infrastructure is a necessity for telework, appropriate technical support and technical resources have been positively related to various work outcomes, such as job satisfaction, while teleworking during the pandemic (Niebuhr et al., 2022). Other studies have also identified the importance of suitable telework environment conditions with telework outcomes, such as job performance and satisfaction (e.g., Mihalca et al., 2021; Yu & Wu, 2021).

Through the lens of the JD-R model, these findings suggest that the provision of job resources (i.e., suitability of telework office space and technical resources) is associated with greater positive work outcomes, whereas the experience of job demands (i.e., shared office space) is associated with reduced positive work outcomes. Collectively, these findings are also consistent with the Theory of Work Adjustment that suggests greater satisfaction should be experienced when there is a greater correspondence or “fit” between the desired and actual telework environment characteristics (Bretz & Judge, 1994).

During the COVID-19 pandemic, the potential for elevated levels of work-family conflict due to additional responsibilities was another heightened factor of the telework environment (Rudolph et al., 2021). Defined as “a form of inter-role conflict in which the role pressures from the work and family domains are mutually incompatible in some respect” (Greenhaus & Beutell, 1985, p. 77), work-family conflict has been a major topic in the telework literature (Allen et al., 2015).

The impact of work-family conflict on work-related outcomes while teleworking during the pandemic has been inconsistent. In their qualitative study with Chinese teleworkers during COVID-19, Wang et al. (2021) found that work-home interference was the most frequently mentioned challenge. Work-family conflict has also found to be negatively associated with productivity, and job satisfaction, with teleworkers in Italy (Galanti et al., 2021) and India (Mohammed et al., 2022), respectively.

However, such relationships between work-family conflict and telework performance and satisfaction have not been consistently found. In their survey of Romanian employees at a large Information Technology company, Mihalca et al. (2021) observed no relationship between work-family conflict and both job performance and job satisfaction. However, only 22% of participants in Mihalca et al. reported having childcare responsibilities (compared to 70% of participants in Galanti et al.), suggesting that their sample may have experienced a relatively low potential for work-family conflict.

Indeed, Kaltiainen and Hakanen (2023) found that the impact of telework on work-nonwork interference was greater for those with children living at home, compared to those without children living at home. Despite this, Kaltiainen and Hakanen also found that having children at home reduced the negative impact of higher work-non-work interference on engagement, burnout, and boredom. Other studies have found no association with children at home and telework satisfaction (Sousa-Uva et al., 2021), or job performance (Blahopoulou et al., 2022), whereas other studies have found having children at home to be associated with lower telework satisfaction (Blahopoulou et al., 2022).

These inconsistent findings between the telework environment factors of work-family conflict and children at home, and the relationship with telework outcomes, can be understood through Boundary Theory. This theory proposes that employees who prefer clear boundaries between the work and home domain may find boundary violations more stressful, compared to employees who prefer flexible boundaries between the two domains (Kreiner, 2006). As such, the impact of work-family conflict or having children at home on work outcomes, might depend on whether an individual perceives the work-family conflict as a job demand (e.g., due to increased workload with childcare responsibilities) or job resource (e.g., due to increased social support during isolation).

Job and organisational factors

The unique challenges associated with mandated teleworking during the pandemic has emphasised the importance of various demands and resources at the job (job security, job autonomy) and organisational (organisational support, professional isolation) levels. Through the lens of Self-Determination Theory (Deci & Ryan, 2000), there are three basic psychological needs (competence, autonomy, and relatedness) that must be satisfied to motivate individuals. Meta-analytic evidence shows that fulfilling these needs has been associated with various outcomes, including greater task performance and job satisfaction (Van den Broeck et al., 2016). The three basic psychological needs have demonstrated positive and significant relationships with job autonomy and social support, and negative associations with job insecurity, respectively (Gagné et al., 2022). Taken together, the JD-R and Self-Determination Theory would predict that the job resources of job security, job autonomy and organisational support would lead to greater telework satisfaction and telework performance, whereas the job demand of professional isolation would lead to lower telework satisfaction and telework performance. However, the research findings have been somewhat mixed.

The impact of job security on telework outcomes during the COVID-19 pandemic has been largely overlooked. This is surprising since the pandemic has resulted in widespread concerns for employees’ job security, the presence of which would be a valuable job resource (Pacheco et al., 2020). Job security refers to an “employee’s expectations about the stability and longevity of his or her job in an organisation” (Lu et al., 2017, p. 30). There has been limited research investigating the impact of job insecurity on various work outcomes for employees during the pandemic, with some studies finding that greater job security is positively related to productivity and job satisfaction (Ali et al., 2021; Pacheco et al., 2020), whereas other studies finding no relationship (e.g., Parent-Lamarche & Boulet, 2021). However, there is limited, if any, research investigating the relationship between job security and work outcomes while teleworking during the pandemic.

Recognised as an important job resource (Bakker & Demerouti, 2007), greater job autonomy has generally been found to have positive effects on work outcomes while teleworking during the pandemic. Job autonomy is “the degree to which the job provides substantial freedom, independence and discretion in scheduling the work and in determining the procedures to be used in carrying it out” (Hackman & Oldham, 1976, p. 162). Specifically, job autonomy has been positively related to telework productivity (Galanti et al., 2021), self-rated telework performance (Wang et al., 2021), telework adjustment (Carillo et al., 2021), and job satisfaction (Mohammed et al., 2022; Niebuhr et al., 2022). In some studies, however, no relationship has been found (e.g., Yu & Wu, 2021).

In addition to telework environment factors and job factors, organisational factors such as organisational support and professional isolation, have also been identified as important predictors of telework outcomes during the pandemic. Organisational support refers to an employee’s perception of the extent to which the organisation values their contributions and cares about their wellbeing (Eisenberger et al., 1997). Despite the challenges of the pandemic, the available research has not established a relationship between organisational support and work outcomes, such as work productivity, job performance, or job satisfaction, while teleworking during COVID-19 (Carillo et al., 2021; Mihalca et al., 2021). Defined as a “state of mind or belief that an individual is out of touch with others in the workplace” (Golden et al., 2008, p. 1412), professional isolation has been negatively associated with telework productivity (Galanti et al., 2021), telework adjustment (Carillo et al., 2021), and telework satisfaction (Toscano & Zappalà, 2020) during the COVID-19 pandemic.

Hypotheses

To gauge the efficacy or success of telework, both organisational (e.g., job performance) and individual outcomes (e.g., job satisfaction) should be considered (Gohoungodji et al., 2023). Therefore, this study assessed the association between the various job demands and job resources on both telework satisfaction and telework performance. In this way, this study was able to examine those factors that directly relate to the practice of telework, which can help inform the development of efficient telework practices post-pandemic.

Using the framework of the JD-R model, it was hypothesised that job demands (living with children [H1a], shared home office space [H1b], work-family conflict [H1c], professional isolation [H1d]) would be negatively associated with both telework satisfaction and telework performance. In contrast, it was hypothesised that job resources (home office comfort [H2a], home office equipment [H2b], home office privacy [H2c], technical support [H2d], technical resources [H2e], job security [H2f], job autonomy [H2g], organisational support [H2h]) would be positively associated with both telework satisfaction and telework performance.

Method

Participants

Participants were recruited from the metropolitan city of Melbourne (Australia) during the last three weeks of strict lockdown orders imposed by the government in response to COVID-19. Recruited via the Qualtrics Market Research Panel between 28 August and 14 September 2020, participants were required to be residents of Melbourne, over the age of 18, and fluent in English. Participants were also required to be currently employed by an organisation (i.e., not sole traders), where a portion of their working hours had been completed from home within the last four weeks.

A total of 208 participants completed the survey. Participants ranged in age from 19 to 82 (M = 40.74, SD = 12.98), including 105 females (50.5%) and 103 males (49.5%). A total of 170 participants (81%) reported living with others (including relatives, partners) and of those, 103 participants reported living with children. During the four weeks prior to completing the survey, participants had completed an average of 33.30 hours of work per week (SD = 12.51), with an average of 30.39 hours per week completed at home (SD = 13.42). A proportion of participants (22.6%) also reported working from home in 2019, for an average of 15.1 hours per week (SD = 12.00).

During the four weeks prior to completing the survey, 53 (25.5%) participants reported sharing a workspace with others and 97 (46.6%) had children at home during work hours. Participants reported working in various industries including education and training (15.4%), professional, scientific, and technical services (16.8%), administrative and support services (12%), information media and telecommunications (8.2%), and other industries (17.3%).

Materials

Several job demands (i.e., living with children, shared home office space, work-family conflict, professional isolation) and job resources (home office comfort, home office equipment, home office privacy, technical support, technical resources, job security, job autonomy, organisational support) comprised the explanatory variables. Participants’ levels of telework satisfaction and telework performance comprised the dependent variables. For all measures (except where otherwise specified), responses were measured on a 5-point Likert scale from 1 (Disagree) to 5 (Agree).

Job demands

To gauge whether participants resided with any children, participants were asked to respond to the question “Do any children live with you?” [Yes/No]. To assess whether participants shared their home office space with others, participants were asked, “Over the last four weeks while working from home, have you had to share your workspace with any other members of the household?” [Yes/No].

A shortened version of the work-family conflict tool was used to measure work-family conflict (Carlson et al., 2000). The original scale comprises six sub-scales, each with three items. In the current study, participants responded to one question from each sub-scale, including, “I am often so emotionally drained when I finish work that it prevents me from contributing to my family”. A mean score across the six items was calculated for each participant (Cronbach’s α = 0.90), with a higher score representing higher work-family conflict.

Professional isolation was measured using Golden et al. (2008) scale. Participants were asked to think about their work over the previous four weeks and rate the frequency in which they had experienced various feelings, such as “I’ve felt isolated” and “I’ve missed face-to-face contact with co-workers”. Participants responded to seven items on a 5-point scale from 1 (Never) to 5 (Daily). A mean score was calculated (Cronbach’s α = 0.91), with higher scores denoting more frequent feelings of professional isolation.

Job resources

Participants were asked four questions to measure aspects of their home office (telework) environment. Participants were asked to think about their home office space and rate the extent to which they agreed with the following statements: “My office set up is comfortable” (home office comfort); “I have access to all the office equipment that I need to do my job from home (e.g., printer, computer, desk chair)” (home office equipment); and “My office set up allows me privacy” and “My office set up is free from distractions” (home office privacy). A mean score for the two items measuring home office privacy was created (Spearman-Brown Coefficient = 0.92).

Participants were asked to respond to one item measuring technical support (“I receive as much technological support as I need when working from home”) and one item measuring technical resources (“My organisation provides me with appropriate technological resources to do my work at home successfully”; Baker et al., 2006). Higher scores denoted higher levels of technical support and technical resources, respectively.

Job security was measured using two items: “I will be able to keep my present job for as long as I wish” and “I am secure in my job” (Kraimer et al., 2005). A mean score for the two items was calculated (Spearman-Brown Coefficient = 0.88), with a higher score indicating higher perceptions of job security. Using Ahuja et al. (2007) scale of job autonomy, participants were asked to respond to four items about the content of their job, including, “I control the content of my job” and “I have a lot of freedom to decide how I perform assigned tasks”. A mean score was calculated (Cronbach’s α = 0.82), with a higher score indicating higher job autonomy.

Perceived organisational support was measured using a modified version of Eisenberger et al.’s (1997) scale. Participants were asked to consider their work over the previous four weeks and rate the extent to which they agreed with seven items, including, “My organisation cares about my opinions” and “My organisation really cares about my wellbeing”. A mean score was calculated (Cronbach’s α = 0.79), with higher scores denoting higher levels of perceived organisational support.

Dependent variables

To measure telework satisfaction and telework performance, participants were asked to think about their work over the last four weeks and rate the extent to which they agreed with the following statements: “Overall, I am satisfied while working from home” (O’Neill et al., 2009) and, “I am a strong performer while working from home” (O’Neill et al., 2014). Higher scores denoted higher levels of telework satisfaction and telework performance, respectively.

Procedure

The survey was administered via the Qualtrics online survey platform. After reading the information form and providing informed consent, participants responded to a series of questions relating to their experiences of teleworking over the previous four weeks. Approval for this study was granted by the University of Adelaide’s Human Research Ethics Committee (approval number 20/50).

Results

Descriptive analyses

Descriptive statistics and bivariate correlations are displayed in Tables 1 and 2, respectively.

Table 1. Descriptive statistics.

Variable	n	%	Mean	Std. Dev.	Min.	Max.
Job Demands
Living with children	208
Yes	103	49.5
No	105	50.5
Shared home office	208
Yes	155	74.5
No	53	25.5
Work-family conflict	208		2.82	1.08	1	5
Professional isolation	208		2.59	1.08	1	5
Job Resources
Home office comfort	208		3.97	1.05	1	5
Home office equipment	208		3.94	1.21	1	5
Home office privacy	208		3.70	1.25	1	5
Technical support	208		3.87	.97	1	5
Technical resources	208		3.87	1.04	1	5
Job security	208		3.95	.99	1	5
Job autonomy	208		3.66	.93	1	5
Organisational support	208		3.64	.71	1	5
Dependent Variables
Telework satisfaction	208		3.94	1.03	1	5
Telework performance	208		3.94	.90	1	5

Table 2. Bivariate correlations.

	Variable	1	2	3	4	5	6	7	8	9	10	11	12
1	Work-family conflict	1
2	Professional isolation	.43**	1
3	Home office comfort	−.13	−.25**	1
4	Home office equipment	−.16*	−.29**	.65**	1
5	Home office privacy	−.09	−.21**	.73**	.60**	1
6	Technical support	−.08	−.15*	.41**	.41**	.32**	1
7	Technical resources	−.10	−.14*	.37**	.38**	.28**	.78**	1
8	Job security	−.08	−.07	.25**	.25**	.23**	.47**	.47**	1
9	Job autonomy	.17*	−.10	.38**	.33**	.35**	.38**	.33**	.38**	1
10	Organisational support	−.22**	−.28**	.30**	.30**	.26**	.42**	.43**	.28**	.37**	1
11	Telework satisfaction	−.15*	−.29**	.50**	.33**	.37**	.59**	.58**	.45**	.32**	.25**	1
12	Telework performance	.07	−.12	.44**	.36**	.37**	.57**	.53**	.50**	.32**	.30**	.63**	1

* < .05 (2-tailed). ** < .01 (2-tailed).

There were statistically significant positive correlations between all the job resources (home office comfort, home office equipment, home office privacy, technical support, technical resources, job security, job autonomy, organisational support) and both the dependent variables, telework satisfaction and telework performance. Further, there was a statistically significant negative correlation between the two job demands (work-family conflict, professional isolation) and telework satisfaction.

Ordinal logistic regression analyses

Two cumulative odds ordinal logistic regressions with proportional odds (Bender & Grouven, 1997) were conducted to identify the job characteristics (i.e., explanatory variables) associated with telework satisfaction, and telework performance (i.e., dependent variables). The ordinal regression model was considered appropriate due to the ordinal nature of both dependent variables (i.e., Likert scale). Telework satisfaction and telework performance were each measured using a five-point Likert scale, from 1 (Disagree) to 5 (Agree) with descriptive anchors also presented at 2 (Somewhat disagree), 3 (Neither agree nor disagree), and 4 (Somewhat agree).

For the first ordinal logistic regression analysis, all explanatory variables were entered into the model, with telework satisfaction entered as the dependent variable. Ordinal logistic regression models are performed under the assumptions of multicollinearity (i.e., the explanatory variables are not highly correlated with each other) and proportional odds (i.e., the effects of the explanatory variables are the same at each level of the dependent variable; Laerd Statistics, 2015). For the first analysis, the assumption of multicollinearity was met (as Variance Inflation Factor values were calculated for the explanatory variables and were less than 10; Laerd Statistics, 2015). The assumption of proportional odds was also met, as assessed by a full likelihood ratio test comparing the fit of the proportional odds location model to a model with varying location parameters, χ2(36) = 46.56, p = .112.

The deviance goodness-of-fit test indicated that the model was a good fit to the observed data, χ2(816) = 385.67, p = 1.000. The final model statistically significantly predicted the dependent variable over and above the intercept-only model, χ2(12) = 157.32, p < .001. The ordinal logistic regression identified seven variables that had a statistically significant effect on telework satisfaction, namely professional isolation, home office comfort, home office equipment, technical support, technical resources, job security, and organisational support. The model results are reported in Table 3.

Table 3. Ordinal logistic regression model estimates for telework satisfaction (N = 208).

		Parameter Estimates
		ß	SE	Wald Chi-Square	p-value	Exp (ß)	95% Confidence Interval for Exp (ß)
							Lower Bound	Upper Bound
Threshold Dependent Variable	Telework Sat = 1	1.65	1.18	1.96	.162	5.22	.52	52.79
	Telework Sat = 2	3.43	1.19	8.26	.004	30.95	2.98	321.30
	Telework Sat = 3	5.12	1.22	17.65	<.001	166.69	15.32	1813.51
	Telework Sat = 4	8.33	1.31	40.25	<.001	4135.01	315.64	54171.07
Explanatory Variables	Job Demands
	Living with children (ref = yes)	.17	.31	.32	.570	1.19	.65	2.17
	Shared home office (ref = yes)	.12	.35	.12	.730	1.13	.56	2.27
	Work-family conflict	−.14	.16	.81	.367	.87	.63	1.18
	Professional isolation	−.46	.16	8.38	.004	.63	.46	.86
	Job Resources
	Home office comfort	.89	.23	15.07	<.001	2.44	1.56	3.83
	Home office equipment	−.44	.17	6.57	.010	.64	.46	.90
	Home office privacy	−.02	.18	.01	.909	.98	.69	1.40
	Technical support	.65	.25	7.05	.008	1.92	1.19	3.11
	Technical resources	.93	.23	16.52	<.001	2.53	1.62	3.97
	Job security	.59	.17	11.50	<.001	1.80	1.28	2.52
	Job autonomy	.01	.20	.00	.972	1.01	.69	1.48
	Organisational support	−.49	.25	3.86	.049	.61	.38	1.00

Bolded coefficients represent those that are statistically significant.

Specifically, an increase in the odds of greater telework satisfaction was associated with an increase in home office comfort (odds ratio [OR] = 2.44; 95% CI, 1.56 to 3.83; Wald χ2[1] = 15.07; p < .001), technical support (OR = 1.92; 95% CI, 1.19 to 3.11; Wald χ2[1] = 7.05; p = .008), technical resources (OR = 2.53; 95% CI, 1.62 to 3.97; Wald χ2[1] = 16.52; p < .001), and job security (OR = 1.80; 95% CI, 1.28 to 2.52; Wald χ2[1] = 11.50; p < .001). In contrast, an increase in professional isolation (OR = .63; 95% CI, .46 to .86; Wald χ2[1] = 8.38; p = .004), home office equipment (OR = .64; 95% CI, .46 to .90; Wald χ2[1] = 6.57; p = .010), and organisational support (OR = .61; 95% CI, .38 to 1.00; Wald χ2[1] = 3.86; p = .049) was associated with a decrease in the odds of greater telework satisfaction.

For the second ordinal logistic regression analysis, all explanatory variables were entered into the model, with telework performance entered as the dependent variable. Both assumptions of multicollinearity, and proportional odds were met (χ2[36] = 29.24, p = .780).

The deviance goodness-of-fit test indicated that the model was a good fit to the observed data, χ2(816) = 391.35, p = 1.000. The final model statistically significantly predicted the dependent variable over and above the intercept-only model, χ2(12) = 118.16, p < .001. The ordinal logistic regression identified three variables that had a statistically significant effect on telework performance, namely home office comfort, technical support, and job security. The model results are reported in Table 4.

Table 4. Ordinal logistic regression model estimates for telework performance (N = 208).

		Parameter Estimates
		ß	SE	Wald Chi-Square	p-value	Exp (ß)	95% Confidence Interval for Exp (ß)
							Lower Bound	Upper Bound
Threshold Dependent Variable	Telework Perf = 1	4.09	1.22	11.31	<.001	59.64	5.50	646.33
	Telework Perf = 2	5.58	1.22	20.82	<.001	264.37	24.09	2901.17
	Telework Perf = 3	7.70	1.27	36.65	<.001	2207.31	182.49	26698.21
	Telework Perf = 4	10.95	1.39	62.00	<.001	57202.58	3742.86	874233.77
Explanatory Variables	Job Demands
	Living with children (ref = yes)	.30	.30	.98	.322	1.35	.75	2.43
	Shared home office (ref = yes)	.07	.35	.04	.848	1.07	.54	2.14
	Work-family conflict	.02	.16	.01	.911	1.02	.75	1.38
	Professional isolation	.02	.15	.02	.890	1.02	.76	1.38
	Job Resources
	Home office comfort	.44	.22	4.11	.043	1.56	1.02	2.39
	Home office equipment	−.09	.16	.33	.566	.91	.66	1.25
	Home office privacy	.15	.18	.74	.389	1.16	.82	1.64
	Technical support	.76	.25	9.58	.002	2.14	1.32	3.46
	Technical resources	.34	.22	2.36	.125	1.41	.91	2.18
	Job security	.71	.17	16.93	<.001	2.03	1.45	2.84
	Job autonomy	−.05	.19	.08	.779	.95	.65	1.38
	Organisational support	.07	.24	.09	.763	1.07	.67	1.71

Bolded coefficients represent those that are statistically significant.

Specifically, an increase in the odds of greater telework performance was associated with an increase in home office comfort (OR = 1.56; 95% CI, 1.02 to 2.39; Wald χ2[1] = 4.11; p = .043), technical support (OR = 2.14; 95% CI, 1.32 to 3.46; Wald χ2[1] = 9.58; p = .002), and job security (OR = 2.03; 95% CI, 1.45 to 2.84; Wald χ2[1] = 16.93; p < .001).

Taken together, the two ordinal logistic regression analyses established statistically significant positive associations between the job resources of home office comfort, technical support, and job security and both telework satisfaction and telework performance. These findings provide support for H2a, H2d, H2f, respectively. There was a statistically significant positive association found between technical resources (job resource) and telework satisfaction, providing partial support for H2e. Similarly, there was a statistically significant negative association between professional isolation (job demand) and telework satisfaction, providing partial support for H1d.

There was a statistically significant negative association between home office equipment (job resource) and telework satisfaction, and organisational support (job resource) and telework satisfaction. As these associations were in the opposite direction than hypothesised, there was no support for H2b and H2h. As there was no relationship between the other explanatory variables and the dependent variables, the remaining hypotheses were not supported.

Discussion

Unsurprisingly, the COVID-19 pandemic has had a meaningful impact on employees’ and employers’ views of future work arrangements. Indeed, it has been predicted that telework will remain widespread even after the end of the pandemic (Niebuhr et al., 2022). In preparation for this potential shift in the nature of work, the aim of this study was to investigate the association between various job demands, job resources, and telework outcomes during the COVID-19 pandemic.

Adopting the JD-R model as a theoretical framework, the current study hypothesised that job demands would be negatively associated with both telework satisfaction and telework performance, whereas job resources would be positively associated with both telework satisfaction and telework performance. Consistent with the JD-R model and the specified hypotheses, the findings from this study identified three job resources (home office comfort [H2a], technical support [H2d], and job security [H2f]) that were associated with both telework satisfaction and telework performance. Specifically, it was found that those who had a more comfortable home office space, greater technical support, and higher job security, reported greater telework satisfaction and telework performance. Consistent with the hypotheses, those who received greater technical resources (job resource) reported greater telework satisfaction (H2e), whereas those who reported greater professional isolation (job demand), reported lower telework satisfaction (H1d). As there was no relationship between technical resources and telework performance (H2e), and professional isolation and telework performance (H1d), these hypotheses were only partially supported.

Consistent with the predictions of the JD-R model and the proposed hypotheses, the bivariate correlations revealed that suitability of home office equipment (H2b), and organisational support (H2h), were positively correlated with telework satisfaction. Further, both home office equipment and organisational support were positively correlated with each other and home office comfort, home office privacy, technical support, technical resources, job security, and job autonomy, and negatively correlated with work-family conflict and professional isolation. However, after controlling for these other explanatory variables in the ordinal logistic regression, the relationship between home office equipment and telework satisfaction, and between organisational support and telework satisfaction, became negative.

This suggests that it is likely not the physical possession of the home office equipment which improves telework satisfaction, but rather, the facilitation of other resources provided by the equipment, such as greater job autonomy and job security, which in turn aids telework satisfaction. A similar explanation may be posited for the findings related to home office privacy. While home office privacy was positively correlated with telework performance and telework satisfaction, these relationships were not statistically significant after controlling for the other explanatory variables in the ordinal logistic regression (H2c). While previous research has identified aspects of the telework environment as important predictors of telework outcomes during COVID-19 (Carillo et al., 2021; Mihalca et al., 2021; Yu & Wu, 2021), the findings from this study suggest that some aspects of telework environment (i.e., home office equipment and home office privacy), might not uniquely contribute to telework outcomes when accounting for the other explanatory variables.

Similarly, it may not be the direct reception of organisational support that improves telework satisfaction, but rather, the by-product of reduced job demands, such as work-family conflict and professional isolation, that improves telework satisfaction. This is consistent with the available research which has not demonstrated a relationship between organisational support and work outcomes, including job performance and job satisfaction, while teleworking during COVID-19 (Carillo et al., 2021; Mihalca et al., 2021).

While the current study findings were largely consistent with other telework research conducted during COVID-19 (Carillo et al., 2021; Galanti et al., 2021; Mihalca et al., 2021; Niebuhr et al., 2022; Toscano & Zappalà, 2020; Yu & Wu, 2021), support was not established for all hypotheses. Specifically, there was no relationship found between the job demands of living with children (H1a), shared home office space (H1b), and work-family conflict (H1c) with either telework performance or telework satisfaction. Further, while job autonomy was positively correlated with telework performance and telework satisfaction, these relationships were not statistically significant after controlling for the other explanatory variables (H2g). This finding is consistent with Yu and Wu (2021), who found no relationship between job autonomy and telework outcomes, however, contradicts most previous findings (e.g., Carillo et al., 2021; Galanti et al., 2021; Niebuhr et al., 2022). This may be explained by the positive association between job autonomy and all other job resources, indicating that job autonomy is a reflection of an appropriately resourced telework environment rather than the cause of improved telework satisfaction and telework performance.

Alternatively, such null findings between living with children, shared home office space, work family conflict, job autonomy, and telework outcomes, might be explained by the broad and flexible nature of the JD-R model itself. Rather than prescribing the specific job demands, job resources, and work outcomes contained within the framework, the JD-R is a heuristic model that symbolises a way of thinking about how job characteristics may influence work outcomes (Schaufeli & Taris, 2014). The current study categorised job characteristics as either job demands (e.g., work-family conflict, professional isolation) or job resources (e.g., job autonomy, job security) consistent with categorisations made within past research (Demerouti et al., 2001; Galanti et al., 2021; Jamal et al., 2021; Schaufeli & Taris, 2014). However, this assumes that there is no between- or within-person variation in how job characteristics are appraised (Searle & Auton, 2015). It is likely the case that some variables classified as demands were not consistently appraised as negatives (e.g., living with children) and some variables classified as resources were not consistently appraised as positives (e.g., job autonomy). This could explain the lack of significant relationships between these variables and telework outcomes in the current study, which are not entirely unsurprising given the inconsistent relationships observed in the literature between work-family conflict and telework outcomes (e.g., Galanti et al., 2021; Mihalca et al., 2021; Mohammed et al., 2022; Wang et al., 2021) and living with children and telework outcomes during COVID-19 (e.g., Blahopoulou et al., 2022; Kaltiainen & Hakanen, 2023; Sousa-Uva et al., 2021). While a strength of this study was that it surveyed teleworkers across a range of different occupational sectors, it is likely that job characteristics affect work outcomes differently for different job types, the nuances of which were not captured in the findings.

Practical applications

From an applied perspective the findings from this study point to several key factors that managers, practitioners, and/or employees could consider when developing or engaging in future telework policies. Specifically, this study suggests that boosting home office comfort, technical support, technical resources, job security, and minimising professional isolation will result in positive telework outcomes. These findings represent two meaningful contributions to the literature.

Firstly, pre-pandemic knowledge on teleworking is limited as it has been largely conducted with individuals from certain occupational sectors who were interested in, and/or were able to engage in telework (Wang et al., 2021). Instead, this study was able to capture the important factors associated with telework outcomes across a greater cross section of Australian employees, many of whom (77.4%) had not engaged in any telework in the preceding year. As such, the findings from this study are arguably more generalisable across occupational sectors compared with research findings collected prior to the pandemic.

Secondly, while there has been worldwide scientific interest in teleworking during COVID-19 (Mamani-Benito et al., 2022), the current study is able to point to the job characteristics that are likely to promote (and hinder) telework success for employees working within Australia. A national survey in Australia recently revealed that more than 75% of knowledge workers enjoy flexible (i.e., hybrid or remote) working arrangements since the work-from-home orders were lifted by the government (Hopkins & Bardoel, 2022). As such, the application of the study findings is of continued practical relevance within an Australian context.

One predictor of both telework satisfaction and telework performance in the current study was job security. Given the novel nature of COVID-19, limited, if any, research has been published on the impacts of job security on telework outcomes during the pandemic. While the negative economic impacts of the pandemic heightened feelings of job insecurity for many employees, it is unlikely that the pandemic context can completely explain these findings. Prior to the pandemic, employees with higher levels of job insecurity have reported less favourable attitudes towards teleworking, compared to employees with lower levels of job insecurity (Lim & Teo, 2000). This may be because teleworking can be appraised as a threat to job security, where reduced visibility in the physical workplace may adversely impact career opportunities (Lim & Teo, 2000). In future teleworking policies, therefore, managers could consider ways of promoting job security and career advancement opportunities for teleworkers, especially in comparison to their non-teleworking counterparts.

Limitations and future directions

While this study advances knowledge on the practice of telework during the COVID-19 pandemic, the key limitations of this study and areas for future research should be noted. Firstly, the survey administered in this study was cross sectional in nature. While this study design was able to capture the important factors associated with telework outcomes for employees during a period of enforced remote working, the causal nature of these associations cannot be established. It should be noted that this study’s data were collected during the height of lockdown restrictions in Melbourne (Australia) and as such, the relationship between some variables and the work-related outcomes might be distorted. The potential of self-selection bias may have limited the range of participants who voluntarily completed this survey, and thus their telework experiences may not be representative of those teleworking during this period. Future research should consider more robust longitudinal designs to understand these nuanced relationships and identify key predictors of telework success especially as telework becomes more widespread across occupational sectors.

Secondly, the use of a self-report measure of telework performance in this study may not be as reliable as more objective measures of telework performance and may suffer from a common method bias. As such, future research could consider using both subjective and objective measures of job performance when assessing outcomes of telework.

Finally, the JD-R model proposes that job characteristics influence both negative and positive work outcomes through two pathways (burnout and engagement; Schaufeli & Taris, 2014). This study focussed primarily on understanding the association between job demands and job resources on positive telework outcomes. While this piecemeal approach to the JD-R model is common (Schaufeli & Taris, 2014), future research could consider a more robust framework on job characteristics and work outcomes.

Conclusion

Using the Job-Demands Resources model as a theoretical framework, the aim of this study was to identify the job demands and job resources associated with telework outcomes by surveying a sample of Australian teleworkers during the COVID-19 pandemic. Consistent with the hypotheses and predictions of the JD-R model, it was found that three job resources (home office comfort, technical support, and job security) were positively associated with both telework satisfaction and telework performance. Further, technical resources (job resource) was positively associated with telework satisfaction, whereas professional isolation (job demand), was negatively associated with telework satisfaction, but neither were associated with telework performance.

When controlling for all other explanatory variables, there were no statistically significant relationships between some job demands (living with children, shared office space, work-family conflict), job resources (home office privacy, job autonomy) and telework outcomes, and a negative relationship was observed between two job resources (home office equipment, organisational support) and telework satisfaction. Such findings might suggest that these job characteristics do not uniquely contribute to either telework satisfaction or telework performance. Alternatively, it might be the case that the relationships between certain job characteristics and work outcomes vary depending on job type, a notion consistent with the heuristic nature of the JD-R model. Overall, the findings from this study provide a greater understanding of the facilitators and barriers to the practice of telework which can help inform the design of future telework programs within an Australian context.

Authors’ contribution

Both authors contributed to the study conception, design, material preparation, data collection and analysis. The first draft of the manuscript was written by Jaime Auton and reviewed by Daniel Sturman. Both authors read and approved the final manuscript.

Ethics approval

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the University of Adelaide (approval number 20/50 received 26 August 2020).

Consent to participate

Informed consent was obtained from all individual participants included in the study and there are no conflicts of interest to report.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author (Jaime Auton), upon reasonable request.

Additional information

Funding

This work was supported by funding received from the School of Psychology, University of Adelaide.