Rotterdam, July 2015
The extended workplace
An empirical analysis of
the relationship between telework and job satisfaction
through control, workload and working hours
Charissa Rentier
Student number 329481
Master thesis
Supervisor: dr. Josse Delfgaauw
Advisor: prof.dr. Robert Dur
Erasmus School of Economics
Erasmus University Rotterdam
Abstract: The practice of teleworking has experienced a rapid increase over the past decades. This raises questions about the relations between teleworking and employee well being. Previous research in this area typically uses cross-sectional data. In this thesis I make use of five waves of Dutch panel data including 4.061 respondents. By using panel data, I investigate whether results from a between-person analysis carry over to a within-person analysis. I analyse the direct effect of teleworking on job satisfaction, as well as the indirect effects through flexibility, autonomy, work burden, time pressure and working hours. The analysis replicates previously found positive effects of telework. Furthermore, I find no substantial differences between the between- and within-person analysis, except for in an additional analysis involving job switching. Teleworking on itself, but also teleworking more hours and a larger share of total hours worked, are positively related to job satisfaction. Furthermore, I find that teleworks positive relation with job satisfaction runs partly through positive effects on perceived control, time pressure, and actual working hours, while a negative association stems from increases in unpaid overtime. An additional analysis taking job switching into account shows switchers to be driving the positive teleworking effect, while those staying in the same job actually experience a decrease in job satisfaction as they start teleworking more. These within-person findings suggest a sorting effect of workers into teleworking companies and jobs. This sorting effect, as well as the policy preconditions of positive teleworking effects, are suggested directions of future research.
Keywords: telework, job satisfaction, time pressure, workload, working hours, overtime, job switching
38
Table of Contents
A. Introduction 4
B. Framework and related literature 12
B.1 The 24/7-economy: flexibility and work intensification 12
B.2 Information and communication technologies 13
B.3 Telework 15
B.3.1 Work-related ICTs in connection to autonomy and expectations 17
B.4 Work-life balance 18
B.5 Hypotheses 20
B.5.1 Hypothesis 1 – telework 20
B.5.2 Hypothesis 2 – control: flexibility and autonomy 21
B.5.3 Hypothesis 3 – workload: work burden and time pressure 21
B.5.4 Hypothesis 4 – working hours 21
C. Data and method 23
C.1 Data 23
C.2 Method 26
D. Results and analysis 28
D.1 Telework 28
D.2 Control 32
D.3 Workload 34
D.4 Working hours 36
E. Discussion 39
E.1 Sector differences 39
E.2 Job switching 40
E.3 Related mechanisms 44
F. Conclusion 46
References 51
Appendix A – Tables 57
Appendix B – Figures 83
38
A. Introduction
As communication technologies are becoming increasingly advanced and currently allow constant connectedness and access, businesses have adjusted to not only facilitate workers by providing them with an office, but also giving the option to work from home. This phenomenon of employees calling in from home to do the work they would normally do in the physical company office is given the name ‘telework’.
On the one hand, teleworking and therefore being able flexibly manage work and adjust working hours and location to their circumstances is likely to add to job satisfaction. However, this increased autonomy might have a downside as well. Managers might start expecting their subordinates to be available at all times, creating an expectation source of stress and pressure for employees. Alternatively, employees may not be able to cut loose from the constant access to work and start working more, thereby creating more expectations from co-workers, clients and supervisors. Therefore, making it possible to work practically anywhere and anytime can also have adverse effects on job satisfaction and work-life balance.
It is these effects on job satisfaction that this thesis aims to investigate. More specifically, the focus will be on the effect of teleworking on job satisfaction through its effect on perceived control, perceived workload, and working hours. This focus on the effect of teleworking on the experience of the job by the employee and on the number of hours worked is in contrast to much previous research, which focuses on productivity. For the purpose of this thesis, panel data retrieved from questionnaires among 4.061 Dutch workers in the Dutch Arbeidsaanbodpanel are used. The analysis makes use of five of its waves between 2004 and 2012 in which telework, in the sense of working from home, is one of the questionnaire topics. It is hypothesized that an increase of perceived control as a consequence of telework will increase job satisfaction. Alternatively, telework is thought to bring about an increase in work burden and time pressure, which could subtract from the job satisfaction, as these are likely to be sources of work-related stress. Lastly, teleworking is hypothesized to have a positive effect on the number of working hours, which in its turn is thought to have negative effect on job satisfaction.
According to research of Bitkom, German workers are increasingly available to clients and co-workers outside of working hours, namely from 73% to 88% between 2009 and 2011, and about a third is literally constantly available. These data provide grounds for German business life and politics in their aim to protect employees from the possibilities of technology that cause them to be unable to switch off and effectively being ‘on call’ all the time. Volkswagen for example has stopped e-mail from coming in on mobile devices after working hours, in order to establish a better work-life balance (New York Times, 23 December 2011). Another more recent example is Daimler’s policy to automatically delete e-mails sent to a receiver who has an out-of-office reply in place (Financial Times, 13 August 2014). This measure would mean there is no e-mail to check during a vacation, and the cluttered inbox upon returning to the office is also avoided, thereby enabling the employee to truly enjoy the time off and preventing an overflow of work upon their return.
Due to technological advancement, facilities required for telework are present in most homes and companies. Examples of ICTs enabling telework are e-mail programmes – often installed on mobile devices – and VPN-connections allowing access to company servers.
More specifically, as measured in 2013 by the Central Bureau of Statistics, 98% of the Dutch working population has access to a personal computer (desktop and/or laptop), all of whom have used it during the 3 months before the measurement. Of those people, 88% use the personal computer on a daily basis, and 11% at least once a week. The PC’s are located both at home (99%) and at work (72%). Furthermore, of all the working population, 99% of people have Internet access, for which a desktop (73%), laptop (86%), mobile phone (83%) or a different device, such as a tablet, (70%) are used. 98% of all working people has used the Internet in the 3 months before measurement and 92% of them uses it on a daily basis. Of those who have used the Internet in the past 3 months, 97% uses it to communicate, especially through e-mail. In 99% of the cases the Internet is used at home, and in 73% at work. Of people who use the Internet, 98% is able to use search engines, while 93% is skilled to send e-mails. These data show that almost the entire working population has a PC and internet access at home, uses both on a daily basis and is skilled to use them for information gathering and communication purposes; all of which are preconditions for telework.
When it comes to the actual telework, around 70% of small companies (<50 employees) have teleworkers, whereas the bigger companies (>50 employees) display telework in over 90% of the cases. Around 25-30%% of the employed people actually telework, both with low percentages in the hospitality sector (5-15%) and high numbers in ICT and service professions (40-55%), as is to be expected given the professions. In CBS’ data, telework is defined as regularly working outside of the office for at least half a day, while having access to the company’s ICT-system (2014). This means that employees have access to files and programmes, implying that just handling e-mail is not regarded as teleworking in this instance (CBS, 2014). Around 70% of all workers use the Internet for their tasks, up to 99% in the ICT and service sector. From the employer’s side, around 25% of the employees are provided portable devices with mobile Internet access, thereby enabling them to work outside of the office (again with the lower bound of 10% in the hospitality sector and around 50% in the ICT, services and real estate sectors) (CBS, 2014).
The above-mentioned new information and communication technologies (ICTs) have changed the way people and organisations work, and have had their impact on the way people experience their jobs. Where previously leaving the office meant leaving the day’s work behind, many workers now carry their jobs with them in the form of mobile devices such as laptops or PDA’s and smartphones, or are able to access work-related e-mails and files from their home computers. Therefore, as personal computing becomes more portable, the choice to leave the office behind while physically leaving it has become a more conscious one. Whether or not to switch off your phone, to sync and read e-mails, or to work on a report using a VPN-connection in many requires active deliberation.
Consequentially, the new possibilities have had their impact on the balance between work and life of employees. On the one hand, because of increased flexibility, care tasks can more easily be combined with work, and using mobile communication one can handle small tasks on the go so as to save time and do more within the same amount of time. On the other hand, technology and work have become more invasive, be it because we let it as we cannot unplug, or because of expectations of others about our accessibility, leading to a blurred boundary of work and home life (Ling & Donner, 2009).
Previous research into the consequences of teleworking has been performed in various fields. On the business economics side, scholars have focused on the outcomes of telework in terms of productivity and performance. Main finding in this respect is that teleworkers report an increased productivity (Bailey & Kurland, 2002; Vega et al., 2014). When looking into the job attributes, some studies show that increased productivity is often mainly found in routine tasks (DuBrin, 1991). However, a large share of the people reporting this often simultaneously report an increased number of working hours (Bailey & Kurland, 2002). Aborg et al. (2002), in a study concerning within-person effects of telework in two companies, also found telework to increase work effectiveness, but this might simultaneously result in a higher workload. Work intensification, meaning that more work is done under de same contract, as a result of flexible working practices is, however, also found to be the case by Kelliher and Anderson (2009). While exhibiting higher levels of job satisfaction, flex-working employees will work harder because of reciprocal behaviour to the flexibility offered by the firm, or simply because the flexibility enables them to work more intensively during the hours spent at home as compared to working at the office. Additionally, those with higher organisational commitment are more likely to perform better than those with lower levels of commitment under the circumstances of teleworking (Hunton & Norman, 2010).
Organisational economics and -psychology have furthermore explored the effect of teleworking on working hours, job satisfaction and experience of individual workers. In a within-person examination based on five consecutive working days, Vega et al. (2014) for example found employees reporting higher job performance and job satisfaction and exhibiting better creative performance while teleworking as opposed to working in the office. Furthermore, in a cross-sectional analysis between teleworkers and non-teleworkers, Mahler (2012) finds that people who cannot telework although they would like to do so report lower job satisfaction than those who telework or chose not to telework. This points towards a role for autonomous decision-making and (perceived) control in explaining job satisfaction – something that is commonly agreed upon in the organisational psychology and management literature (Spector, 1986). In a meta-analysis of studies looking to identify the psychological mediators of teleworking making use of 46 telecommuting studies, Gajendran and Harrison (2007) find that teleworking indeed increases feelings of autonomy and control while also increasing levels of reported job satisfaction. However, as Bailey and Kurland (2002) point out, general job satisfaction is often not measured, as mostly the satisfaction with respect to the specific teleworking arrangements or other job components are measured. Additionally, this leaves out elements regarding for example perceived work pressure and work-life balance (DuBrin, 1991). Therefore, clear conclusions about the general level of satisfaction cannot readily be drawn.
On the sociological side, organisational dynamics and work-life division have been investigated. A point of concern when it comes to teleworking is the effect it has on employees’ organisational identification and their feeling of isolation. Organisational identification becomes more difficult once employees start spending more time working away from the office, as it impedes communicating norms, values and organisational culture to these employees (Kurland & Bailey, 1999). Fonner and Roloff (2012) try to explain inconsistent effects of telework on identification found in previous studies with the so-called connectivity paradox. This paradox claims that ICTs help teleworkers to have more online, but less face-to-face, contact with colleagues, while also causing external interruptions while working. This ultimately leads to a lower organisational identification, translating into an effect on job satisfaction, motivation and performance. However, even tough less face-to-face contact has a negative effect on job satisfaction, this still does not offset the general positive effect teleworking has on job satisfaction, leaving teleworkers on average more satisfied than office-workers (Fonner & Roloff, 2010; Fay & Kline, 2011). From the employee’s point of view, working from home could also lead to a feeling of isolation, being unaware of what goes on inside the office and being left out due to a change in group dynamics, possibly leading to a decrease in performance (Sparrowe et al. 2001). Morganson et al. (2010) indeed find that those mainly working from the office tend to experience less isolation than teleworkers. However, it is reported that the extent to which an employee teleworks is the main factor in isolation, as individual teleworkers often still spend a considerable amount of time at the office, causing them not to be left out (Bailey & Kurland, 2002).