Research and Advances
Computing Applications Virtual extension

Non-Work Related Computing (NWRC)

  1. Introduction
  2. Research Model and Hypotheses
  3. Research Methodology and Analysis Results
  4. Discussion and Implications
  5. References
  6. Authors
  7. Footnotes
  8. Figures
  9. Tables

The internet is becoming a commodity in organizations with an increase in accessibility by professionals from their own work desks. conducted a survey in 1999 with 1,244 respondents, and found that 84% of employees sent non-work related emails and almost 90% surfed the Internet for non-work related purposes during office hours. Accordingly, many companies scrambled to get policies and guidelines in place to deal with the effects of non-work related computing.

However, the effect of the use of the Internet for non-work purposes is inconclusive. Several studies pointed out the negative impacts of non-work related computing (NWRC) such as employee’s productivity loss and risks of damaging lawsuits.9 On the other hand, some researchers proposed that NWRC can actually have positive impacts. For instance, frequent internet users might actually be happy and productive workers.9 Despite these two conflicting views, no research has been done to look deeper into how NWRC affects job performance. Thus, our research objective is simple, yet has not been investigated:

How does time spent engaging in NWRC activities affect job performance? Does NWRC have either a positive or a negative effect on job performance?

The expected contributions of this paper are mainly two folds. Academically, there has been a long debate on the control of employees – whether companies should employ loose or tight control. Thus far, it is viewed that tight control leads to productivity increase in the short run (at the cost of employee satisfaction), but in the long run, it decreases productivity.9 If we apply this view to the NWRC context, allowing employees to engage in NWRC freely (which may result in more time spent on NWRC) should increase employees’ performance although the negative views on NWRC have been dominant. This study examines if this debate still holds at the Internet age and will provide empirical evidence for the debate. As for managerial implications, we hope this article will shed light into how engagement in NWRC may affect job performance and consequently, managers can better decide whether to encourage or discourage their employees to engage in NWRC.

A few terminologies describing NWRC have been proposed by various researchers. The more commonly sighted ones are two presented in Table 1, “Junk Computing” and “Cyberloafing.”

Both definitions bear many similarities – for instance, both (1) emphasized the use of resources of the organization for (2) non-work or personal purposes. One difference is that Lim et al9 specified the organizational resource used explicitly which was Internet access in their case whereas in Guthrie et al’s5 study, the organizational resources ranged from offline usage (stand alone computers) to online usage (computers with Internet access). In our study, we would use the definition from Guthrie et al’s5 as we want to look at many types of NWRC behaviour such as using office resources, instant messaging, data search for personal interest, file downloading, internet gaming etc. as shown in Table 2.12

Most literatures have focused on understanding the determinants of NWRC5, 8 to implement measures to curb NWRC due to the negative effects of NWRC on companies. For instance, in 2001, a study conducted by Surf-Watch found that as much as USD1 billion in costs are incurred when employees engage in NWRC behaviour and much of the cost arises from losses in employee productivity mainly due to time wasting online activities which accounts for 30–40% of productivity losses.a Apart from productivity losses, NWRC may cause companies greater exposure to the risk of being involved in legal proceedings stemming from Internet or email activity (such as infringement of copyright laws when downloading music files). Table 3 summarized the findings on causes and controls of NWRC.

However, there is another group who sings praises of NWRC, saying that it could be actually create happier and more productive employees.9 Indeed, Stanton’s research of 400 employed professionals showed that high frequency Internet users may often be happy and productive workers. He advocated that “when server logs revealed hits have occurred on non business-related sites, it is possible that professional workers are treating the Internet as a perquisite, much the same as making personal calls, running errands at lunchtime and chatting at the water cooler have always been perquisites of jobs with high autonomy.” Oravec,9 on a similar study, said that “if engaged in constructively, online recreation can aid in awakening creativity, increasing well-being and ultimately make employees more productive, just as appropriate and timely face-to-face diversions have restored employees’ energies over the past decades.” Belanger et al 9 even went further to advocate that a “certain amount of playful use of computer applications can lead to learning that may be of value to the organization.”

In summary, although there have been a number of studies on NWRC, there remains an unresolved controversy on the impacts of NWRC and there has been no empirical study on the effects of NWRC on employee’s job performance. Only Thadani12 investigated the fitness between employees’ skills such as communication and literacy skills, and engaging in NWRC activities, particularly instant messaging and Internet browsing, and how this fitness affects job performance. His study, however, does not tell us whether engaging in NWRC in general affect job performance.

Back to Top

Research Model and Hypotheses

Accordingly, the following research model is proposed to specifically look into how employees’ time spent in NWRC affects job performance (Figure 1). We shall explain each construct and hypothesis in detail.

In determining the effects of NWRC on job performance, we drew on past literature that focused on how email affected job performance. Since its introduction, there existed controversy on the impacts of email. Initial users of email in Digital Equipment Corporation reported an increase in personal productivity of 5–15% while managers reported that email increased the speed of their decision making and saved them about seven hours per week.3 However, there are also major productivity implications if email is not managed properly. Research from IDC and Gartner Group suggests that 30–40% of all email in organizations is personal.7 Personal email reduced employees’ productivity, and associated attachments significantly increased network traffic and the overall load on Internet connections.7 Besides, researchers have looked specifically at email interruptions and how this may lead to reduced productivity.6 Email interruption is used to describe scenarios when emails become a form of distraction for employees especially if it interrupts them from more important work. Hence, based on the literature review of studies on email, we hypothesize that employees’ higher engagement in NWRC will result in lower job performance although we admit that there could be a positive impact to a certain extent according to the previous studies as we mentioned before:

H1: The more time employees spend on NWRC, the lower their job performance.

To present a more holistic view of how NWRC affects job performance, we included other forms of loafing at work which do not necessarily involve computing such as taking long lunch, reading newspapers, chatting with colleagues etc. We termed this behaviour “non-work relating activities” (NWRA). NWRA type of loafing behaviour existed in organizations way back some time ago as is evident from Snyder et al’s study11 where employees admitted to various forms of loafing on the job. Back in 1983, ABA Banking Journal published a comprehensive list describing the various types of NWRA.b Based on the reasoning similar to H1, we assume our next hypothesis as the following:

H2: The more time employees spend on offline NWRA, the lower their job performance.

In our study, we also took into account the time spent working overtime without compensation such as overtime pay or time-off because it could be another time factor which influences job performance since we are measuring both NWRC and NWRA in terms of time spent on each behaviour. Intuitively, assuming other things are equal, employees who spend more time on their work are likely to perform better due to the increased input of effort. However, it is also possible that employees spend time working extra hours to make up for their time loss on NWRC or NWRA. In this case, we expect no significant impact of working overtime on job performance along with positive correlations between time spent on work after working hours, and time spent on NWRC and/or NWRA. Hence, we hypothesize that:

H3: Employees’ time spent on work after working hours without receiving extra benefits or compensation for the effort will have no significant impact on their job performance in case that they try to make up for their time losses on NWRC or NWRA. Otherwise, the more time employees spend on work after working hours without receiving extra benefits or compensation for the effort, the higher their job performance.

Back to Top

Research Methodology and Analysis Results

The target group of this research is full time salaried office employees who have easy access to the Internet from their own workstations or desktops. Two sets of questionnaire were developed, one for employees and one for employees’ direct supervisors, in order to use the immediate supervisor’s ratings for measuring job performance of the employees for two reasons: to weed out any “leniency or self-favouritism effect in employees’ self-rated performance” where employees would rate themselves higher than their supervisors4 and to minimize any common method bias of self-reporting. A data was considered valid only when both questionnaires, one from the employee and one from the direct supervisor, were valid.

The questionnaire was given out either by hand or mail in the form of hardcopies. Hardcopies were preferred due to the challenging administrative work involving in ensuring the correct pairing of employee-supervisor. The data collection period lasted two weeks and six multi-national as well as local organizations in financial services, manufacturing, technology, education and pharmaceutical industries in Singapore participated in this survey. The total response rate was approximately 60%. 76 sets of questionnaire were returned and 71 sets (71 employees and 12 supervisors in all) were suitable for data analysis. As for the demographic information of employees, 44% was male, 54% fell on the age ranging from 26 to 35, and 56% had 2 – 10 years of work experience. For supervisors, 83% was male, about 60% was over 40 years old, and around 70% had more than 10 years of work experience.c

To ensure content validity, we drew on the existing literature to measure the variables of this study with certain modifications to suit the NWRC context.d With an adequate measurement model,e we tested the proposed hypotheses with partial least squares (PLS-Graph Version 3.00) and Figure 2 shows the results of the PLS analysis.

The findings supported only one hypothesis, H1 – the more time employees spend on NWRC, the lower their job performance (H1: β=-0.882, t-value=7.637) while time spent on offline NWRA (H2), interestingly, positively affects job performance (H2: β=0.429, t-value=2.894). H3 – time spent on work after working hours – turned out to be insignificant. 52% of employee’s job performance is accounted for by these three independent variables.

Back to Top

Discussion and Implications

The objective of this study is to enhance the understanding of the impacts of NWRC on employee’s job performance following a controversy in the past research focused on this area. Our findings support the fact that NWRC indeed reduces employee productivity (H1) while time spent in NWRA positively affects job performance (H2) on the contrary to our expectation.

We shall delve upon various studies regarding email interruptions to explain H1 and H2. Since most NWRC activities (43%) stem from personal communications according to our survey which includes emails and internet messaging, employees are constantly interrupted from their work. Jackson et al’s study6 found that employees take about 64 seconds to return to their work at the same rate at which they left it after every interrupt by emails. Internet messagging is an even more intensive form of interrupt since it requires one’s immediate attention. Hence, recovery time compounded from both personal email and internet messaging interrupts may result in a great deal of time wasted leading to loss of job performance.

However, in the case of communications in NWRA, interrupts are less. This is because employees may most likely chat with their colleagues continuously for a period of time, say 15 minutes during tea break before breaking up to focus on work. Moreover, when colleagues chat among themselves, there is a strong possibility that they are engaging in some sort of discussion about work. Hence unconsiously, they may be helping one another to solve problems through sharing of experiences or coming to a consensus on a shared project, thereby, positively affecting job performance. Besides, their working relationships may be enhanced through informal sharing and chatting, thereby also increasing their job performance. However, in the case of NWRC, employees are more likely to chat and email people who are not in their companies (outside friends or online friends), and not likely to share work experiences and engage in team building among colleagues.

Next, based on two significant correlations between time spent on NWRC and other two endogenous variables – the negative relationship between time spent on working after office hours and time spend on NWRC and the positive relationship between time spent on offline NWRA and time spent on NWRC – we conducted post-hoc cluster analyses to investigate H3 further. According to the post-hoc analysis results,f there might be distinctly two very different groups of employees: one group of employees is very hardworking, who would work overtime with no compensation to ensure quality and quantity of work done and would actually engage less in NWRC behaviours, and the other group of employees is the loafers, who would engage in NWRC whenever a chance presents itself.

We would like to draw parallels from the ant society to explain this finding similar to many researchers who have been drawing parallels from the insect society when researching into control measures in manufacturing.1 The ants are well known for their loyalty (to the queen and the nest) and for being hardworking. However, in a recent study by Hasegawa and his team at Hokkaido University,g it was found that only about 80% of ants engage in some sort of work such as cleaning the nest or gathering food, but the rest of the 20% of worker ants are mostly idle. Hence, even in the hardworking ant society, there appears to be two different groups functioning – one group, which is hardworking, and one group which is lazy, and contributes little.

According to our findings, the research on NWRC is enhanced in three ways: First, we provided empirical findings on how NWRC actually affects job performance in the realm of research that lacks empirical studies and has much controversy over the impacts of NWRC on job performance. Second, we incorporated NWRA and working overtime into the study of NWRC to provide a more holistic view of the loafing behaviours. Finally, contrary to the traditional view on the control of employees, this study shows that loose control (allowing employees to spend more time on NWRC) does not necessarily result in higher employees’ performance. Based on our findings, mangers may encourage NWRA for employees to re-energize themselves rather than NWRC which may result in negative interruptions and distraction from work. They also need to pay attention to those who go overboard in enjoying NWRC because this group of people may constantly enjoy NWRC and consequently, record low job performance.’sh survey results of 670 employers found that 70% of employers think that the maximum amount of time employees are allowed to surf non-work related sites is 30 minutes per work day. Our findings reveal that 52% of employees actually exceeded this maximum allowed time to loaf showing 3.5 hours of NWRC per week on average.

Despite several significant findings, we acknowledge a number of limitations in this study. First, our sample size is too small. We faced some challenges during data collection because of the sensitivity of the research. Another limitation is that majority of the companies surveyed (4 out of 6, constituting 85% of responses) had some form of internet usage control measure in place such as monitoring software. This may have caused the results to be more skewed although our post-hoc analysis suggests that electronic monitoring may not work in curbing NWRC behavior.i Future researchers may consider researching more in depth to find out what can be done to reduce NWRC and consequently increase job performance. Another possible and interesting way to extend our research is to study the other threats that could be caused by NWRC behavior (besides job performance) such as legal risks from sexual harassments lawsuits, security threats, network bandwidth overload, and so on.

Back to Top

Back to Top

Back to Top

Back to Top


F1 Figure 1. Research Model and Hypotheses.

F2 Figure 2. Results of PLS Analysis.

Back to Top


T1 Table 1. Definitions of NWRC

T2 Table 2. Types of NWRC

T3 Table 3. Causes and Controls Of NWRC

Back to top

    1. Anderson, C. and Bartholdi, J.J., III. Centralized versus decentralized control in manufacturing: lessons from social insects. In the Proceedings of University of Warwick, Complexity and Complex Systems in Industry, 19th-20th September 2000: 92–108 (McCarthy, I. P. and Rakotobe-Joel T, Eds.). The University of Warwick, UK, 2000.

    2. Cheng, Y. and Kalleberg, A.L. Employee Job Performance in Britain and the United States. Sociology, (30:1): 115–129.1996.

    3. Crawford, A.B., Corporate Electronic Mail – A Communication-Intensive Application of Information Technology. MIS Quarterly 6, 3, 1982, 1–13.

    4. Furnham, A. and Stringfield, P. Congruence in Job-Performance Ratings: A Study of 360 Degree Feedback Examining Self, Managers, Peers, and Consultant Ratings. Human Relations 51, 4, 1998, 517–530.

    5. Guthrie, R. and Gray, P. Junk Computing: Is It Bad for an Organization? Information Systems Management 13, 1, 1996, 23–28.

    6. Jackson, T, Dawson, R. and Wilson, D. Reducing the effect of email interruptions on employees. International Journal of Information Management 23, 1, 2003, 55–65.

    7. Kilpatrick, I. Employee Guidance for Email and the Web. Logistics & Transport Focus, May: 47–49. Chartered Institute of Logistics & Transport (UK), 2003.

    8. Lee, O.K.D., Lim, K.H. and Wong, W.M. Why Employees Do Non-Work-Related Computing: An Exploratory Investigation through Multiple Theoretical Perspectives. In the proceedings of the 38tb Annual Hawaii International Conference on System Sciences (HICSS'05). Track 7 - Volume 07:185, 2005.

    9. Stanton, J.M.; Oravec, J.A.; Belanger, F. and Slyke, C.V.; Lim, V.K.G, Teo, T.S.H. and Loo, G.L.; Simmers, C.A.; Urbaczewski, A. and Jessup, L.M. Internet Abuse in the Workplace. 2002. Communications of the ACM, 45(1): 55–59; 60–63; 64–65; 66–70; 71–74; 75–79; 80–83.

    10. Suliman, A. Work performance: Is it one thing or many things? The multidimensionality of performance in a Middle Eastern context. International Journal of Human Resource Management 12, 6, 2001, 1049–1061.

    11. Snyder, N.H., Blair, K.E. and Arndt, T. Breaking the bad habits behind time theft. Business, 40, 1990, 31–33.

    12. Thadani, D.R. Are all NWRC Bad?: An Alternative View. Thesis, City University of Hong Kong, 2005.

    a. Wynn, A. and Trudeau, P. 2001. Internet Abuse at Work: Corporate Networks are paying the Price.

    b. Source: Getting the goods on time thieves. ABA Banking Journal, 1983.

    c. More detailed demographics of respondents are available upon request.

    d. Questionnaires items for employees and supervisors are available upon request.

    e. Analysis results of measurement model are available upon request.

    f. The post-hoc analysis indicate that the group of employees who spend less time on NWRC show higher job performance and vice versa.

    g. Source: 20% of worker ants idle lazybones: study. The Japan Time. Nov. 16, 2003;

    h. Source: Internet Use Survey of 670 Employers, Fall 2000;

    i. Our post-hoc analysis only with the sample of electronic monitoring in place gave results consistent with our original findings.

    This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOEHRD, Basic Research Promotion Fund) (KRF-2007-332-B00108).


Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More