By Dale Tweedie and David Wild
Are electronic surveillance systems the breakthrough to ensuring productivity with remote work? According to Tweedie and Wild, “100 years of performance management research shows many things critical to work cannot be accurately measured at all. So, the question isn’t just what we should measure, but also if and when.”
Throughout the COVID-19 pandemic, many workers welcomed new opportunities to work from home. This work-from-home transformation also accelerated employers’ use of electronic surveillance systems.
E-surveillance, sometimes termed “bossware” by critics, let employers track workers’ activities remotely using easily available software. For example, software can register what websites workers visit, how often they move their mouse, and can even record from workers’ screens and webcams. Dashboards interpret the data e-surveillance collects. This can include producing real-time productivity scores that combine performance data into an easily observed and ranked number.
While workers can be notified when e-surveillance systems are operating, many systems have the technical capacity to operate without workers being aware. One company boasts that its software “can be silently and remotely installed, so you can conduct covert investigations and bullet-proof evidence gathering without alarming the suspected wrongdoer.” 1
In the United States, market research firm Gartner found 60% of large employers use monitoring systems.1 Predictably, e-surveillance has also triggered employee resistance. These include online discussions of legal rights, e-countermeasures like migrating work discussions to new platforms and even physical products like “mouse jigglers” that simulate online activity.
To assess these developments, we need to understand exactly what is new about these e-surveillance systems, and how they affect work and workplaces. We can then consider how workers, managers, and regulators might respond.
What’s new about e-surveillance?
While many media accounts of e-surveillance suggest a new workplace dystopia, the basic structure of these systems is not especially novel.
Although intensive surveillance is a shock to many white-collar professionals, workers in other industries have experienced close surveillance for centuries. Before the industrial revolution, working from home was commonplace. Adam Smith famously argued specialisation makes factories more efficient. But the shift from working at home to the factory was also driven by owners’ desire to monitor workers more closely.
In this respect, there is a historical irony in cutting-edge software tracking workers at home. Just as new technologies like zoom and cloud storage begin to reverse the industrial revolution movement of workers from home to the factory or office, other technologies are enabling surveillance to follow workers back into their homes.
Even white-collar work has been subject to e-surveillance for decades, most notably in call centres. Call centre workers have long been measured and ranked according to variables like average call time, time available to take calls, and so on. Twenty years ago, such surveillance was already precise enough to measure how long workers spent in bathroom breaks as well as on the job.2
Nonetheless, while new systems are not unprecedented, they are distinctive in at least three ways. The first is the scope and speed this technology has spread, and the relative lack of regulation. Factory and call centre surveillance is complex and relatively fixed. New software is cheap and mobile. Just as the smartphone put a high-powered processor in everyone’s hand, new surveillance software can put a mini panopticon in the hand of every employer.
The second difference is context: e-surveillance has emerged alongside other software that is already widespread. Consequently, new systems can amplify the surveillance potential of existing tools. We’ve seen this already when Microsoft introduced an Office 365 tool that allowed managers to assess individual workers’ productivity. Public pressure forced Microsoft to partially rewind this capability. But, if the relentless expansion of surveillance software is any guide, we should expect further attempts by large companies to mainstream the technologies of their smaller, more aggressive surveillance start-up siblings.
A third feature is the growing sophistication – albeit not transparency – of productivity algorithms. These include Artificial Intelligence (AI) systems that create risk scores to measure how likely individual workers are to pose a security threat to their employers.3 So, even where new systems collect similar data, AI can use this data in more intrusive ways.
Impacts of e-surveillance
Research into previous white-collar e-surveillance, especially in call centres, provides indicative evidence of how new tools will affect workers and workplaces. Three findings are significant.
First, while surveillance systems differ, modern e-surveillance hits just those features correlated with higher stress, suffering, and lost wellbeing. All other things being equal, prior research suggests, more intensive surveillance causes more stress and other harms.4 Another key variable is the purpose of surveillance. While workers can react positively to surveillance used for training, they react more negatively to surveillance used for discipline and control.5 So, we need to consider not only what technology e-surveillance uses, but also why.
Since dashboard data like productivity scores are typically only available to managers, they seem aimed more at control than developing workers’ capabilities. Indeed, it is hard to see what skills – other than rudimentary time management – much of the data e-systems collect, could develop.
Second, reducing complex work to single numbers can be misleading and counterproductive. Call centre research has distinguished between “quantity” and “quality” call centres. In quantity centres6 – which manage high volume, low complexity calls, surveillance can increase call volume, albeit by churning through burnt-out workers. Yet such surveillance is less effective in quality centres that deal with complex calls, where skills and customer commitment are more critical.
These findings are consistent with a larger body of research by Christophe Dejours and colleagues into how performance measurement affects work quality.7 Dejours shows how intensive performance monitoring of individual workers can not only harm workers but also decrease work quality. One reason is that quality work – and indeed, safe work, requires cooperation. In turn, cooperation requires openness, trust and collective commitment. Intensive individual surveillance can undermine these norms. While research into new technologies is in its infancy, anecdotal evidence of their effects, such as reported in the New York Times, are consistent with Dejours’ findings.
Third, our review of over 100 years of performance management research found that measurement systems – no matter how simple or sophisticated – are never as accurate as they claim.9
Slick software and AI wizardry are beguiling. Paradoxically, the fact we do not really understand how AI works can make productivity scores more appealing. AI’s numerical alchemy can add a magical, even mystical quality to its numbers: We trust them precisely because we do not know enough to understand where they go wrong.
We know these scores do go wrong because even the most intelligent AI cannot see inside workers’ minds. As Dejours has shown, this is where so much of the “work of work” gets done. Digital surveillance can easily track the time the social worker, lawyer, or psychologist spends writing, but not the thought processes that generate these words. And ultimately, these thought processes are the raison d’etre of professional practice.
Even critics can miss this point when debating if new systems capture the right variables. The New York Times quotes Ryan Fuller, former vice president for workplace intelligence at Microsoft, who says: “We’re in this era of measurement but we don’t know what we should be measuring.´´10 Of course, there are better and worse things to measure. However, 100 years of performance management research shows many things critical to work cannot be accurately measured at all. So, the question isn’t just what we should measure, but also if and when.
Managing the growth of e-surveillance
What lessons might managers and regulators draw from these findings?
To say key features of work cannot be measured does not mean they cannot be managed. But sustaining trust and cooperation at work requires management of a different sort. Rather than measuring outputs, managing the skills and attitudes that drive high-quality work requires managers who can engage with workers directly and openly, and who respect their knowledge and skills. Workplace monitoring may also play a part, especially if the primary purpose is to develop skills and capacities rather than enforce discipline. However, well-documented gaps and biases with prior performance measurement tools suggest e-surveillance cannot be the default position it threatens to become.
Paradoxically, new surveillance technologies offer a chance for managers to show leadership by choosing trust and engagement over technology. This is partly an ethics decision because preventing harm is one overriding ethics principle. But prior research suggests declining the use of intrusive e-surveillance is not only about ethics but also about being committed to quality work. In cases where e-surveillance replaces trust and collaboration with scores, it not only risks measuring things that are not significant to high-quality work but also driving out those that are.
As the use of e-surveillance tools rapidly expands, there is also an important task for regulation. Of course, workplace regulations differ across countries and jurisdictions, and it is not possible to review all these different systems and approaches here.
However, one widespread concern is where legislative frameworks rely too heavily on disclosure and consent to determine when surveillance is acceptable. At a minimum, employers should be required to disclose when surveillance is used and gain informed consent. However, the proliferation of e-surveillance systems suggests that neither informing workers nor requiring consent is sufficient to adequately check their growth. One reason is the inherent power imbalance between employers and workers, which the long-term decline of unions has amplified.
A more insidious concern is that we can adapt to technologies at work that are not in our collective interest. Jon Elster and Elin Palm call this the sour grapes problems: Like the fox in the fable who disparages grapes he cannot reach, we may no longer claim forms of privacy at work we feel we cannot have.11 This might seem a dystopian fear if not for how, as Shoshana Zuboff and others have shown, we already accept levels of online surveillance that would have shocked workers and ethicists alike mere decades ago.12
Legal frameworks at least need to set concrete limits on what types of surveillance are legitimate. We also need safeguards that ensure that genuine collective negotiation of other modes of surveillance within industries.
While e-surveillance is not new, it raises new problems for a new age. One advantage this time is that we already have a solid base of evidence about the problems associated with these kinds of surveillance technologies, and about alternative paths we might take. The question, then, is not just the technology as such, but how – and to what ends – we choose to use it this time around.
This article was originally published on 1 February 2023.
About the Authors
Dr. Dale Tweedie is a Senior Lecturer in the Department of Accounting and Corporate Governance at Macquarie University, Sydney. He researches in accountability and ethics at work, especially on sustainability reporting and performance management. Dale publishes in leading international journals and has delivered projects funded by global professional associations.
Dr. David Wild’s main research interests involve political economy, and the phenomenology of work relations. Of specific concern is how working subjects come to understand their social contribution through work. He is currently investigating whether performance management systems can be usefully understood as important ‘ideological’ instruments in contemporary work relations.
References
- Turner, J. (2022, 9 June). “The Right Way to Monitor Your Employee Productivity”. Gartner https://www.gartner.com/en/articles/the-right-way-to-monitor-your-employee-productivity
- Taylor, P. & Bain, P. (1999) ‘‘An Assembly Line in the Head: Work and Employee Relations in the Call Centre”. Industrial Relations Journal, 30(2), 101-117.
- Corbyn, Z. (2022, 27 April). ‘‘Bossware is Coming for Almost Every Worker: The Software You Might Not Realize Is Watching You”. The Guardian. https://www.theguardian.com/technology/2022/apr/27/remote-work-software-home-surveillance-computer-monitoring-pandemic
- Holman, D., Chissick, C. & Totterdell, P. (2002) “The Effects of Performance Monitoring on Emotional Labour and Well-Being in Call Centres”. Motivation & Emotion 26(1), 57-81.
- Holman et al. (2002) “The Effects of Performance Monitoring on Emotional Labour and Well-Being in Call Centres”. Motivation & Emotion 26(1).
- Taylor, P., G. Mulvey, J. Hyman, and P. Bain. (2002) “Work Organization, Control and the Experience of Work in Call Centres.” Work, Employment & Society 16(1), 133–150.
- Dejours, C., Deranty, J.-P., Renault, E., & Smith, N. H. (2018). The Return of Work in Critical Theory. Columbia University Press.
- Kantor, J., Sundaram, A., Aufrichtig, A. & Taylor, R. (2022, 15 August) The Rise of the Worker Productivity Score. The New York Times. https://www.nytimes.com/interactive/2022/08/14/business/worker-productivity-tracking.html
- Tweedie, D., Wild, D., Rhodes, C. and Martinov-Bennie, N. (2019) “How Does Performance Management Affect Workers? Beyond Human Resource Management and Its Critique”. International Journal of Management Reviews, 21(1), 76-96. doi:10.1111/ijmr.12177
- Kantor et al. (2022, 15 August) “The Rise of the Worker Productivity Score”.
- Elster, J. (1983). Sour Grapes: Studies in the Subversion of Rationality. Cambridge University Press; and, Palm, E. (2009) “Securing Privacy at Work: The Importance of Contextualized Consent”. Ethics and Information Technology, 11, 233–241.
- Zuboff, S. (2019) The Age of Surveillance Capitalism. Profile Books.