Like the workshops of the past, today’s automated production schedules create necessities that give birth to innovation in the workplace. Marjolaine Rostain and Jean Clarke tell how today’s industries might gain from tapping into their workers’ hidden skills.
Damien had finished the job. The industrial mould he had been working on met his employer’s specification and was waiting to be sent to the customer. But every time Damien walked past it, something bothered him. The finish wasn’t right.
Eventually, during his lunch break, he picked up his tools and started sanding it again. Nobody had ordered him to do so. The company regarded the mould as complete. Damien continued until it met a different standard: his own.
It is a small episode from an ordinary French factory. But it exposes something important about the way modern economies understand skill. Damien and many of his colleagues were officially engaged in low-skilled industrial work. Their jobs were organised around standardised procedures and repetitive tasks. Yet when we spent eight months inside the factory, another form of work kept appearing.
Workers polished products beyond the required specification. They invented tools. They solved production problems. They taught one another techniques. In small cracks in the formal production system, craftsmanship kept returning. We call this anomalous craft. And it raises an awkward question.How much supposedly low-skilled work is only low-skilled because organisations have chosen not to recognise the skills workers possess?
How much supposedly low-skilled work is only low-skilled because organisations have chosen not to recognise the skills workers possess?
From craftspeople to hands
Industrial capitalism has long attempted to separate knowledge from workers. Traditional craftspeople might have understood the whole production process, decide how a job should be undertaken and judge when the finished object was good enough. But Industrial production broke work into smaller tasks. Managers specified procedures; machines determined processes, and workers increasingly performed particular operations instead of controlling production as a whole. This was extraordinarily productive. It also transferred power.
Knowledge that once belonged to workers became embedded in machines, production systems and management. The worker became what Harry Braverman (1974) described as a “labour force”, reduced to a factor of production and employed primarily to execute decisions made elsewhere.
The contemporary factory we studied seemed to fit that model remarkably well. Processes were highly standardised, automation was extensive and shopfloor workers followed prescribed routines.
Initial interviews produced exactly the complaints one might expect: workers felt disposable, disrespected and controlled. But the same people also talked about pride, problem-solving, learning and producing high-quality work.
We wanted to understand the contradiction so we moved beyond just interviews.
Following 72 workers
Over eight months Marjolaine Rostain spent two to four days each week inside the company, eventually shadowing 72 employees.
She followed day and night shifts, watched production processes, joined workers during breaks and sometimes helped with basic dirty jobs on the shop floor. The closer we looked, the harder it became to sustain a simple distinction between skilled craft and low-skilled industrial labour.
Workers still possessed knowledge that the formal production system did not require them to use. The trick was finding the time to use it.
Modern workplaces do not simply control what people do; they control time.
Hacking time
Modern workplaces do not simply control what people do; they control time. Tasks are allocated a duration and production schedules determine what happens next. Efficiency often means removing supposedly unproductive gaps. Yet those gaps were precisely where some of the most interesting work happened.
Workers became adept at what we call time-hacking, whereby, were somebody to finish a task early, a machine would be unavailable or the production process would create a delay and workers could appropriate the resulting minutes. Sometimes they even used lunch breaks.
They frequently used this stolen time to do better work. Damien resanded his completed mould. Mickael used spare time to polish components until they felt right to his fingers. Christophe argued with a quality manager for permission to spend longer correcting imperfections because anything less would amount to a patch-up job.
Whereas classic workplace resistance such as slowing production or sabotaging management could exist it was only to react against their loss of control. They were resisting the narrow definition of what their own work was supposed to be.
Innovation
Sometimes craft went further.
Jeremy decided an existing production process could be improved and created his own tool. Long after he had left the business, colleagues still referred to it by his name. Olivier fashioned a device from an aluminium rod so that sandpaper could reach difficult corners of very small moulds and his colleagues started borrowing it.
Jean-Pierre thought another repetitive operation could be performed better. Through trial and error he constructed a device that aligned components automatically. Other workers came to regard him as somebody who “knew everything”. And Sandrine collaborated with a colleague to design another tool and manufactured it during her lunch break.
None of this fits comfortably with a description of workers as unskilled operatives executing predetermined routines. They were identifying problems, designing solutions and improving processes. In other contexts we might call that innovation. But because it happened on a factory floor, it could almost disappear.
Craft created recognition. And recognition made work more meaningful.
Recognition matters
Why would workers undertake tasks they were not being paid extra to perform? Part of the answer is autonomy.
For a short period, they became the people making the decisions. Damien, not a supervisor, decided when his mould was finished. Olivier decided what tool needed designing. This changed workers’ relationship with the product of their labour.
Instead of simply reaching the end of an allocated task, they could produce something that satisfied their own conception of good work. But something else happened too: other people noticed.
Colleagues asked to borrow tools. Supervisors recognised expertise. Workers became known as people capable of solving particular problems. Craft created recognition. And recognition made work more meaningful.
The economic cost of suppressing skill
There is an obvious irony here. Management systems often deskill jobs in pursuit of efficiency. And standardisation makes workers easier to replace. Detailed procedures reduce dependence upon individual judgement while automation embeds knowledge in technology.
But organisations can simultaneously eliminate precisely the capabilities they need when reality refuses to follow the procedure. New problems appear when machines malfunction or customers want unusual products or materials behave unexpectedly.
At those moments the supposedly inefficient human capacities of judgement, improvisation and embodied knowledge become valuable again. And the worker who has merely learnt to obey the system has little to contribute while the worker who understands the material can solve the problem.
There is a lesson here for today’s enthusiasm for artificial intelligence and automation. The conventional objective is often to identify human knowledge, codify it and transfer as much as possible into machines. That may increase productivity but a business can reach a point where it has removed so much discretion from workers that the organisation loses the environment in which new knowledge develops.
But the answer is not nostalgia for an imaginary age of happy craftspeople.
Redesigning work
Industrial work was often dangerous, insecure and exhausting long before algorithms arrived. And workers having to sacrifice lunch breaks to experience a degree of creativity is a questionable expectation.
But if workers repeatedly have to hack production schedules to do good work, perhaps the problem lies with the organisation of work rather than with the workers.
In which case, employers interested in genuine productivity might therefore ask a different question. Instead of trying to specify every task more tightly, what would happen if workers were deliberately given time and autonomy to improve the things they make?
The conventional division between skilled and unskilled labour makes this difficult to imagine. Our factory suggests that the distinction is much less secure than it appears.
There may be considerably more skill on the factory floor than organisations allow themselves to use. The tragedy is not that workers lack the capacity for craftsmanship, it is that modern work so often gives them nowhere to put it.
This article has been adapted from this journal article with the assistance of the authors and ChatGPT.
