Digital Myth: Automation kills more jobs than it creates

Digital Myths: #6 – Automation has made a lot of jobs obsolete. New jobs came instead. But it did not happen on its own, but after human decisions. We decide if automation creates or kills more jobs.

Does automation kill more jobs than it creates? This is sometimes called the Luddite fallacy, the idea that the number of jobs is fixed and if some are taken by machines, there will be fewer jobs for humans. The first Luddite was Ned Ludd, a weaver in late eighteenth-century England, who smashed two knitting machines in a fit of rage – the original ‘rage against the machine’.

It is clear that this is not the case; new jobs come along when old jobs go and in many cases the new jobs are safer, less physical and more stimulating. Agriculture was the first sector to benefit from automation. A few hundred years ago the vast majority of Europeans were farmers but today only a few percent are. Better machines, tractors instead of horses, better seeds, better methods, economies of scale etc. have all contributed to increased productivity. Second was manufacturing, with the stocking frame that Ned Ludd saw as such a threat. Today robots do many of the heavy and dangerous jobs in manufacturing; workers are increasingly managing the machines rather than doing the work themselves. That’s not to say all tasks can be done by robots; they are best at large-scale repetitive jobs, whereas humans beat them at jobs with more complexity. All that is obvious, and no cause for dispute.

The question is how this applies to the service sector. With digitalization, automation comes to a whole new set of jobs. Take journalism, for example: now algorithms can write reports from sports events or financial market updates that are indistinguishable from what a human writer would produce. Or advertising, where within two years more than 50% of advert buys are expected to be done by algorithms. Before 2020, more internet traffic will be generated by machines than by humans. So this is a big change. Can we assume that the dynamics from the Industrial Revolution will also apply to jobs like these, or will it be different this time? Do we face a future where job loss is accelerated by automation? Could it even be that the difficulties in dealing with unemployment in Europe have something to do with the changed dynamics brought about by the digital transformation?

No doubt, the so-called digital transformation brings productivity increases. But these can be used differently depending on what owners and management want

As usual, it’s complicated. Google chairman Eric Schmidt, who cannot really be accused of ‘not getting’ digital technology, expressed concern last year at the World Economic Forum in Davos, that there is a race between humans and computers, and humans need to win. To him, the increased speed of change is the problem. New jobs will come, but can they come fast enough? This puts pressure on society to adjust and create more opportunities for training and education, for example.

MIT researchers Erik Brynjolfsson and Andrew McAfee have written the book The Second Machine Age where they look closer at which jobs are threatened by automation and which are safe. It turns out that jobs involving creativity and relations, even trust, are the most difficult to replace by machines. So salesmen and hairdressers are fine, but finance analysts and sports writers should worry. Another conclusion they make is that inequality will increase with automation.

Another American scholar, Professor Scott Galloway at NYU Stern, said in a speech at the Digital Life Design conference in Munich 2015 that the new jobs that replace old jobs are not better but pay less and have less influence. For example, skilled office workers replaced by online accounting algorithms are forced to look for employment in retail or warehousing. ‘The smartphone economy is going to be outstanding for employment but terrible for wages’, said Galloway.

Apart from robots and algorithms, I’d like to look at some other trends that may also play into this. First, the sharing economy: services like AirBnB, which rents houses and apartments to tourists and business travellers, Uber and Lyft and other car-sharing apps, or YouTube for media content, are a new form of intermediary, i.e. they connect sellers and buyers but do not assume the traditional responsibilities of employers. In Uber’s case, it has been the focus of protest from taxi drivers in many countries, and for good reason. Uber provides no job security for its drivers, no cars (Uber drivers use their own), it does not follow the detailed price regulation with meters and zone charges, and taxi drivers require permits and licences that Uber drivers don’t. So you have new global digital players providing a very similar service to the same market but with a different set of rules. They are global rather than local companies and regard themselves as intermediaries with little or no responsibility for those who actually deliver the service. So this is a new phenomenon; for sure it creates some opportunities for people to make money by sharing their car, and for some consumers to save money on hotels for example, plus it’s great for the Silicon Valley billionaires who have invested in these companies when they were start-ups, but it is a challenge to the labour market as we know it, and something the policy-makers need to consider when dealing with unemployment.

No doubt, the so-called digital transformation brings productivity increases. But these can be used differently
depending on what owners and management want. Productivity increases can drive rationalization, which then can be used for new investment, creating jobs, or improved profit margins, which does not necessarily create jobs. So it depends on how this opportunity is used. Take the creative sector, for example: it is the second or third largest employer in Europe, depending on which study you look at. It has a lot of potential for job creation thanks to the digital opportunities in creating, marketing and distributing content and it’s an area in which Europe excels. But there is a threat that the digital markets tend to see creative content as a free resource for anyone to monetize, rather than a valuable asset.

I also think the conclusion that everything will be fine because new jobs always follow old jobs is a bit too laissez- faire. Certainly, action was taken in earlier labour market upheavals: when production shifted from agriculture
to manufacturing in the nineteenth century, for example, universal school systems were introduced, and the automation of manufacturing coincided with a lot of infrastructure investment in the case of railroads, energy, and sewage systems etc. So it’s not fair to say that nothing was done and it was fine. Instead we should look at things like infrastructure investment and education as part of the answer.

So there it is – it’s not black and white. The outcome depends on what we want and what we do. If there is a myth, the myth is that the amount of jobs is set for ever. New jobs come, like they always have, not by necessity but by human decision.



Digital Myths is a series of posts published from the book 21 Digital Myths, Reality Distortion Antidote where Netopia editor Per Strömbäck takes a closer look at some of the concepts that have shaped the way we think, talk and make decisions about digital technology and the internet.