What will happen to all the low-paid workers when the robots come?
Actually, forget the blasted robots; what can we do to help low-paid workers right now? (Because when you’re trapped in a dead end job, with no chance of a pay rise and having to pee in a bottle because your productivity targets don’t leave you enough time to go to the toilet, the prospect of fully automated luxury communism is probably the last thing on your mind.)
It’s a question addressed in an important piece by Benedict Dellot of the RSA.
He starts with the conventional solution to the plight of the unskilled, low skilled and semi-skilled:
“To upskill like there’s no tomorrow. At least, that is the message according to virtually every economist, think-tank and politician…
“…popular theory has it that if workers can upskill fast enough then they should have a hand over the machines. Just as workers moved from the fields and into the factories, and then into the office blocks, restaurants and supermarkets of the service industries, so too – with enough retraining – can they find work in emerging industries, whatever they may be.”
Needless to say, there is always merit in learning new skills and developing established ones. And yet there are limits :
“Not everyone can become a machine learning engineer, cyber security specialist or social media content manager…
“…we need to rethink our approach to low-skilled work. Rather than help people escape these roles for non-existent jobs up the career ladder, the energy of policymakers and educators would be better spent empowering people to develop within them.”
To do that we need to think about what makes a job ‘skilled’ in the first place. That might seem obvious , but the term serves as a short hand for many things other than (though linked to) technical proficiency.
For instance, productivity. Someone who can knit a beautiful, unique piece of clothing can be considered highly skilled in the everyday sense of the word, but not in an economic sense – if, that is, output and earnings are constrained by the time required to complete each item.
Then there’s market demand. Let’s suppose that our knitter’s garments become highly fashionable – with wealthy buyers willing to pay handsomely for each item. Suddenly, this merely talented hobbyist has become a highly skilled worker for no other reason than increased demand (and not because of any improvement in technical proficiency or physical output).
And let’s not forget market access and marketability. Even if potential demand exists for what someone can do, that won’t help them unless they’re able to let the market know what they have to offer and are free to accept the highest bid.
So just how ‘skilled’ a worker is doesn’t only depend on what they can actually do – but also on the productivity, demand and market access that personally ‘attaches’ to them.
Dellot describes the de-skilling effect of the ‘commoditisation’ of labour:
“In economic speak, commoditisation means standardising jobs and ironing out differentiation. Chefs now create meals to a set menu, cleaners complete a fixed set of pre-determined tasks, and delivery drivers are guided to pick-up points following a blue line on an app.
“Commoditisation goes hand in hand with de-skilling, which in turn reduces the bargaining power of workers (if anyone can do the job, why would employers pay a premium?).”
I think he’s absolutely right about this. But the kind of business models he’s describing don’t just deny workers the opportunity to develop their abilities, they are designed to establish top-down control over the value created by an enterprise, so that it can be detached from workers (and also customers, suppliers and even investors) and centralised in the hands of those with the greatest power over the system.
As much as actual skills are important, ‘upskilling’ must also be about the wider systems in which abilities are applied. If we want a society in which economic opportunity is widespread, then tax policies and regulatory frameworks must favour the decentralisation of productivity, demand and market access.