The World Bank has released its annual World Development Report, and the headline news this go round, as relayed by Bloomberg and others, is basically that “Robots Aren’t Killing Jobs.” Which, good news, if true.
“This fear that robots have eliminated jobs—this fear is not supported by the evidence so far,” the World Bank’s Chief Economist, Pinelopi Koujianou Goldberg, told Bloomberg. More specifically, the report notes that for every job that’s been lost, new ones have been created. “This is the fourth industrial revolution, there have been three before, and in each case we managed to survive so it’s not the case that machines completely eliminated humans,” Koujianou Goldberg said. “Eventually, we will adjust.”
Putting aside the fact that an untold number of workers lived absolutely miserable lives in the new jobs the first two industrial revolutions found for them after factory production eliminated their old ones (and lots of people did not in fact survive them), and that there is little doubt that automation has indeed recently erased a good many jobs, it is inspiring to hear that economists are optimistic that we will “eventually adjust” to our new robot-wielding overlords.
Still, it’s worth taking a minute to think about this report, as it embodies a very common outlook on automation, second in ubiquity maybe only to widespread fears of a total robot-jobs apocalypse. It goes something like this: Yes, automation seems scary to some because it involves robots, but those fears of the job-destroying machines are overblown. We’ve done this industrial revolution thing enough times before to know that technology only (eventually) improves our lives. And after all, stuff is cheaper to buy now, devices are higher-tech, and we’re living longer.
Plus, if you look at the data, for every job that has been lost to automation in the U.S., others have been gained; jobs have only moved, from manufacturing to service, or abroad, which is not really a big deal, because the net number of jobs is basically the same, and businesses are continuing to make a lot of money. The key thing (stop me if you’ve heard this one before) is to educate citizens and workers, to retrain them to be more nimble in a fast-moving world where technology is accelerating.
As it happens, I am both paraphrasing the overview of this very World Bank report — “The Changing Nature of Work” (or, How I Learned to Stop Worrying and Love Automation) — and relaying what we might call the “business and managerial consultancy outlook on automation” in general, or BAMCOA, if acronyms are your thing, as they are mine, and is often voiced in places like the World Economic Forum, the Harvard Business Review, Bloomberg Businessweek, and, yes, the World Bank.
The World Bank report offers a good synopsis of this view, as evinced by its introduction:
“There has never been a time when mankind was not afraid of where its talent for innovation might lead… And yet innovation has transformed living standards. Life expectancy has gone up; basic health care and education are widespread; and most people have seen their incomes rise… Despite this optimism, concerns about the future remain. People living in advanced economies are anxious about the sweeping impact of technology on employment. They hold a view that rising inequality, compounded by the advent of the gig economy (in which organisations contract with independent workers for short-term engagements), is encouraging a race to the bottom in working conditions.
This troubling scenario, however, is on balance unfounded. [Emphasis mine] It is true that in some advanced economies and middle-income countries manufacturing jobs are being lost to automation. Workers undertaking routine tasks that are “codifiable” are the most vulnerable to replacement. And yet technology provides opportunities to create new jobs, increase productivity, and deliver effective public services. Through innovation, technology generates new sectors and new tasks.
Where to start. First, BAMCOA logic is built on a straw man—it assumes a ‘technology is good/bad’ binary, and equates “automation” to “technology.” I am not quite sure what the invention of life-extending medical drugs or education has to do with automation, for instance, and it certainly seems possible that humans could have invented Penicillin and public schools without capitalist mass mechanization. No, it’s a false dichotomy seemingly designed to make you feel dumb if you are sceptical of automation, like you are opposing progress.
Also note that the World Bank isn’t terribly concerned with the quality of jobs, just that they are there, existing somewhere in the world. Nor is it much concerned with the severe dislocation that accompanies that lost manufacturing job, that “codifiable” work. This is a key failure of the BAMCOA—it routinely overlooks the profound cultural and social implications of losing your job. (And conveniently forgets the violence, suffering, and pollution of past industrial revolutions, when that happened en masse, even if we did eventually adjust.)
It continually amazes me how blasé economists and consultants are when they deploy the ‘sure, automation will kill some jobs, but it will create new ones’ line. These are livelihoods, traditions, entire social structures that are destroyed, and if they are replaced at all they are often replaced with something much less robust, less secure, and more tenuous. There is a reason that many of the regions hit hardest by automation voted in the largest numbers for Trump.
One of the BAMCOA’s primary purposes is to reinforce the notion that technology is an elemental force of nature (that very large corporations and CEOs incidentally happen to profit off of) which workers must perpetually learn to master (usually for the benefit of those corporations). These reports shift the onus of action onto the workers, who must be maximally educated, adept at multiple skills and willing to retrain to stay nimble in order to remain employed at the corporations that would otherwise automate them out of a job. They obscure the fact that by and large, these automation technologies and initiatives are being implemented by executives and managers, and the general workforce has little to no say in how they are affected—unless they agree to “retraining” that will help them earn the company profits in some other way.
“Many jobs today, and many more in the near future, will require specific skills—a combination of technological know-how, problem-solving, and critical thinking as well as soft skills such as perseverance, collaboration, and empathy,” the report says. “The days of staying in one job, or with one company, for decades are waning. In the gig economy, workers will likely have many gigs over the course of their careers, which means they will have to be lifelong learners.”
That is an interesting way to say “interminable precarity.” There is never any discussion about whether the companies and governments introducing automation have a responsibility to consider the wellbeing of their workers, or even recognise that introducing automation is in fact a choice. Automation may follow certain economic logic, but only a total technological determinist would say that it is like a rising tide that cannot be dealt with aside from throwing up a few sandbags.
To its credit, the report does recommend a “basic minimum social insurance”—something akin to or just shy of a basic income, essentially—but never dares anything as radical as say, suggesting workers be given more say over how automation impacts their jobs, or even that mass automation should maybe be a more democratic process in general.
Another telling paragraph:
“On balance … technology has created more jobs than it has displaced. Technology has brought higher labour productivity to many sectors by reducing the demand for workers for routine tasks. And yet in doing so, it has opened doors to new sectors once imagined only in the world of science fiction. As technology advances, firms adopt new methods of production, markets expand, and societies evolve. Firms rely on new technologies to better use capital, overcome information barriers, outsource, and innovate. New technologies allow for more efficient management of the operations of firms: firms hire workers in one location to produce parts, in another location to assemble, and in a third location to sell. Meanwhile, consumers enjoy a wider range of products at lower prices.”
Hooray? This is peak BAMCOA. Just notice who all this language is tailored to—technology has allowed companies to cut jobs, maximise capital gains, and innovate, innovate, innovate, and everyone else gets… slightly cheaper shit to buy.
The crux of all this is that even if they are being replaced elsewhere, automation is absolutely eliminating jobs, right now. Amazon, long a bright spot in recent American employment history, is hiring few workers as it embraces automation. In the past two decades, automation has claimed millions of jobs — more than outsourcing, even. A third of Americans rely on precarious gig work, and the scars left by the hollowed-out manufacturing sector across the Rust Belt and beyond continue to plague the social wellbeing of numerous communities there. Glancing at the net numbers of jobs worldwide as if they were on a balance sheet and proclaiming the robots aren’t killing any isn’t just factually wrong, but, coming from an institution like the World Bank, could lead to disastrous policy.
It’s maybe not helpful to merely keep repeating the ‘robots are coming for our jobs’ mantra, as that oversimplifies matters, too, but embracing the typical business and managerial class outlook on automation would be worse. (It is also worth noting that the “Robots Are Coming for Our Jobs” headlines are written with the potentially afflicted worker in mind—whether it’s clickbait or not—and the “This Robots Coming for Our Jobs Business Is Overblown” variety are largely written for an audience of the managerial class.) The BAMCOA amounts to ‘stay calm and keep automating,’ and it absolutely will lead to a continued degradation of worker protections and benefits, and even further joblessness.
Reports like this are useful as a window into how elites — i.e., those doing a lot of the automating — view mechanization. But if we really want to understand how automation is unfolding, we should look not at broad global employment figures, but at the countless communities and workers who’ve already experienced it. And instead of saying, well, humanity survived the last three massive and violent economic upheavals, if we (the economists and managers writing such reports, namely) put our heads down then we will get through this one too, maybe we can apply the lessons of history and do a bit better this time.