On Surveillance in the Workplace

Data & Society just published a report entitled “Workplace Monitoring & Surveillance“:

This explainer highlights four broad trends in employee monitoring and surveillance technologies:

  • Prediction and flagging tools that aim to predict characteristics or behaviors of employees or that are designed to identify or deter perceived rule-breaking or fraud. Touted as useful management tools, they can augment biased and discriminatory practices in workplace evaluations and segment workforces into risk categories based on patterns of behavior.
  • Biometric and health data of workers collected through tools like wearables, fitness tracking apps, and biometric timekeeping systems as a part of employer- provided health care programs, workplace wellness, and digital tracking work shifts tools. Tracking non-work-related activities and information, such as health data, may challenge the boundaries of worker privacy, open avenues for discrimination, and raise questions about consent and workers’ ability to opt out of tracking.
  • Remote monitoring and time-tracking used to manage workers and measure performance remotely. Companies may use these tools to decentralize and lower costs by hiring independent contractors, while still being able to exert control over them like traditional employees with the aid of remote monitoring tools. More advanced time-tracking can generate itemized records of on-the-job activities, which can be used to facilitate wage theft or allow employers to trim what counts as paid work time.
  • Gamification and algorithmic management of work activities through continuous data collection. Technology can take on management functions, such as sending workers automated “nudges” or adjusting performance benchmarks based on a worker’s real-time progress, while gamification renders work activities into competitive, game-like dynamics driven by performance metrics. However, these practices can create punitive work environments that place pressures on workers to meet demanding and shifting efficiency benchmarks.

In a blog post about this report, Cory Doctorow mentioned “the adoption curve for oppressive technology, which goes, ‘refugee, immigrant, prisoner, mental patient, children, welfare recipient, blue collar worker, white collar worker.'” I don’t agree with the ordering, but the sentiment is correct. These technologies are generally used first against people with diminished rights: prisoners, children, the mentally ill, and soldiers.

Posted on March 12, 2019 at 6:38 AM15 Comments


David Rudling March 12, 2019 7:11 AM

The report is perhaps inadequately titled as “Workplace monitoring and surveillance”.
Given the increasing number of home-based workers it should perhaps be subtitled “and work as an excuse for intrusive domestic monitoring and surveillance”.

Bob Dylan's Rusty Voice March 12, 2019 11:07 AM

I wonder if the cause and effect relationship isnt backwards. We do not inflict new technologies on those with diminished rights, we diminish the rights of certain groups in order to justify our experimentation on them.

Security Sam March 12, 2019 1:16 PM

In the digital age of raw deal
We all spin the Gerbil wheel
And must work with great zeal
While the bosses wheel and deal

JC Denton March 12, 2019 4:09 PM

I would take anything Data & Society has to say with a large lump of salt. After all, they are Soros funded and wrote that laughable conspiracy piece about YouTube political commentators. That doesn’t mean they can’t give useful advice on a topic like this, but it helps to be wary.

Jon March 12, 2019 7:23 PM

“Gamification and algorithmic management”

Must admit when I read that bullet point I thought it was a threat to the system, not a goal. When you give rewards for specific results, you create an environment that rewards those specific results – whether that result is good for the business or not. And gaming the algorithm can be even more rewarding.

For a very simple example, places that reward coders for ‘How many lines of code you wrote today’
can be
easily gamed
by writing
code like


Petre Peter March 13, 2019 7:08 AM

Just because i am not in front of the laptop, doesn’t mean that i am not thinking about the problem i was assigned. If we agree that dreams are just neuron noise, then the only time i am not doing work us when i am sleeping. In Smart Cities, it makes sense for office workers to be paid on wake time and leave the bureaucracy to computers.

Pan March 13, 2019 4:18 PM

Don’t underestimate the value of sleeping. I’ve solved many a difficult problem, problems nobody else could figure out, by sleeping on them.

@Jon March 13, 2019 4:29 PM

I just want to point out that in some circles
Gamification is analogous to Californication.

1&1~=Umm March 13, 2019 10:12 PM


“I’ve solved many a difficult problem, problems nobody else could figure out, by sleeping on them.”

It’s not just the biological process of ‘sleeping’ there are other biological processes as well…

Which is why you will in certain high risk industrial zones, find health and safety warnings put up on the inside of toilet cubical/stall doors.

One of the recognised ways of dealing with intensive or complex thought problems is ‘Mindfulness’ it is a way to either distract or put in neutral the concious mind so the subconcious mind can work on the problem. Likewise in meditation the notion of ‘clearing the mind’ or ‘centering’ via breathing and other excercises.

Thus the real trick is to train yourself to do it at will, as and when needed. Such ‘thinking on the spot’ is actually very rare, most people give ‘gut reaction’ answers which are almost a ‘canned message’ or ‘knee-jerk’ response.

Which is why you should treat with caution immediate replys to questions that are either unexpected or complex. Likewise with work those that ‘jump in immediatly’ will be not solving the problem, but probably making problems. There is a lot of ‘Technical Debt’ in all walks of life and it’s ‘canned responses’ that lie at the bottom of most of them.

Thus a good skill to learn is ‘Distraction Actions’. In days past when smoking was common, the act of lighting a cigarette in a ritulistic way was a distraction action. Like walking it did not require concious thought to do thus the mind could think, but to the observer it ‘filled in the gap’ to the reply without awkwardness or embarrassment. Similar distraction actions are cleaning of glasses, adjusting sitting position and the like, even the clearing of the throat.

However distraction actions need not be short, studies of those who are or were considered geniuses shows that whilst they may have ‘busy days’ they only spend a short time actively working on the problem task. What they do is ‘fill in’ with other activities that are in effect mechanically repetative, that free the mind to work on the problem task in the background.

I’ve known a number of higher level education instructors who find marking student papers is for them a distraction action. That is not to say that they are not marking diligently, they are, but the process is sufficiently different to thinking about the problem task that they are in effect doing both.

At a different level I find ‘doing housework’ to be a mechanical distraction action, the ‘mokey brain’ gets on with what is a repetative mechanical task, of doing dishes, pushing the vacuum, or ironing, whilst I think about other more interesting problems. Which is why I have a notebook and pencil to hand, such that I can break off from house work to note down what is important or make a calculation.

I suspect that most people who are sufficiently intelligent can ‘observe the surveillance’ and the way it works, thus develop destraction actions to keep the surveillance occupied whilst they have freed up their minds to do other more important thinking.

The trick you have to look for is what triggers the surveillance into higher levels and how to avoid them. Thus keep the surveillance on you at the minimum.

Put simply surveillance systems are very hierarchical in the way they work, in fact with a little thought you realise it is the only way they can be made to be efficient. The classic example is the security guard in front of twenty or more monitors, they can only focus their attention on one at a time, and going from screen to screen ‘time sharing’ their attention will quickly give anyone a migraine. So they don’t do that, what they actually do is defocus their concious attention on all of the screens and become passive or in effect in stand by, and they wait for their sub concious to get a trigger such as movment. Which pulls them out of standby and draws their attention to that screen. Thus if an attacker does things very slowly or very quickly the security guard does not get the trigger to pull them out of standby, thus their attention does not come into focus and they miss entirely that which they would have seen if their conscious attention was focused on that screen.

The problem with workplace ‘work surveillance’ is that with non mechanical tasks it’s difficult to impossible to come up with real measurands. As @Jon has noted above ‘Lines of Code’ count is easily gamed. More importantly you can actually put in a load of crap in an ifdef, it does not actually have to do anything other than add to the line count. Broken comments are another way to do this, in either case it needs a higher level of inspection than just counting lines. Such levels of inspection require something other than mechanistic automation as there are many many ways to game, thus you end up eventually with a real equally or better skilled person looking at the code to see if cheating is occuring, there are few people that would willingly do such a job, and if they were good at it their skills would be wasted doing it as a task anyway.

One area you see this in is accountancy crime, where ordinary auditors go through the books year after year and do not pick up what a highly focused forensic accountant will at almost a glance (see ‘Benford’s Law’ as an example)

Which brings up the ‘A.I. Question’ could someone come up with a system to do such a thing with minimal type I and type II errors? I guess the answer is possibly, but at what cost?

The thing is there is ‘missed opportunity cost’ you have finite resources, how you use them defines how you progress. History shows authoritarian regimes eventually fail due the cost of surveillance or guard labour if not an earlier demise due to people taking other actions. The point being surveillance uses resources the more intensive the surveillance the more resources required. As surveillance is by definition unproductive in nature when functioning as desired those resources are in effect wasted. Thus opportunities to use the resources more wisely or productively are lost, or more correctly wasted which by almost any book on business is bad practice.

Also it’s generaly a sign of incompetence in managment. The reason people with skills change jobs is not because it’s a bad company, but because of incompetent managment, who are going to ruin the company anyway. So ‘The Smart rat leaves the ship before it starts to sink’ preferably when it’s still tied up against dry land.

TRX March 14, 2019 1:02 AM

> Most recently, the company used phone numbers provided for two-factor authentication for advertising and networking purposes

That’s so common that most people think that “two-factor authentication” means “phone number.”

Google, eBay, PayPal… they all demand a phone number for “two-factor authentication” from time to time. Of course, what they ostensibly want it for is to text things to it, which doesn’t work very well with Plain Old Telelphone Service…

Mark March 17, 2019 5:15 PM

@Bruce, Data & Society, the same Soros-funded organisation that wrote the “Alternative Influence Network” paper that linked Joe Rogan and others to the alt-right?

That report of theirs was debunked left, right, and centre. It was far-left propaganda. I’d urge you to find more relevant organisations.

Misty March 18, 2019 4:34 PM

I experienced almost everything mentioned, and I’m sick of it. But because it is unprovable and comes with very small nudges from all sides, there is nothing I can really do about it, at least rigth now!

Leave a comment


Allowed HTML <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre> Markdown Extra syntax via https://michelf.ca/projects/php-markdown/extra/

Sidebar photo of Bruce Schneier by Joe MacInnis.