The roaring furnaces of England during the Industrial Revolution, churning away in grimy Dickensian factories, could not be further removed from the quiet, pristine offices of a 21st-century Finnish software company. But despite their separation in time and technique, both settings were built with a common purpose: to take work out of the hands of artisans and use machines to make it simpler, boosting productivity as a result. While that process is economically valuable, history tells us it can cause social unrest - and some dramatic changes expected in the next few decades could accelerate the trend.

The software firm, Microtask, specialises in taking jobs that have previously been hard to automate, and splitting them into bite-size units that can be parcelled out to an army of relatively unskilled workers. Say, for example, that you needed a translation of a French novel. Currently, you might contract a specialist translator - someone speaking both French and English fluently enough to render the French text into English prose with correct grammar and expression. This translation process would be slow and expensive. In contrast, the Microtask model would initially attempt a translation through software. This is bound to be imperfect, so the translated text is sent out in small pieces to “garble checkers”, who search for badly phrased sentences. Human translators are then set to work on the garbled sentences only, rather than the whole novel, so they don’t require the same depth of knowledge as if the entire work was their responsibility. Finally, editors check the assembled text and improve its expression; they don’t have to speak French. By subdividing the process into tiny pieces - applying the same sort of mass production by specialisation that Henry Ford perfected for his Model T car - Microtask generates a huge improvement in speed and cost. The whole process of division and assembly is largely managed by computers.

Purists will protest that the finished product is likely to be inferior to the nuanced translation that a trained scholar would create, and this is probably true, but not everybody needs a work of art and many would prefer to pay less for a stripped-back, utilitarian version. The question is, when translations are being performed by teams of less skilled workers, aided and managed by software, what will happen to the scholar? In the Industrial Revolution, gangs of artisans - the infamous Luddites - smashed up automated cotton gins that they saw as a threat to their livelihood. Will we see bands of furious middle-class citizens taking the hatchet to server farms as they see their roles replaced by computers? It might seem unlikely, but in our recent past there are two striking examples of what can happen when unemployment is allowed to simmer in rich Western democracies.

The first case takes us back to November 2005. Paris was in flames. A pall of smoke from burning cars hung grimly over the banlieues - the poor, largely migrant-populated districts on the outskirts of the capital.  These citizens were rioting because they did not feel like they were being treated in accordance with the lofty ideals to which the French republic aspires. Liberté, égalité and fraternité had given way to anti-migrant discrimination in the workplace and in the street. Unemployment in some areas was as high as 40%, four times the national average. The fatal electrocution of two teenagers while they hid from police in a power station provided a trigger that released this pent-up rage. In a city soaked in the bloody history of the Revolution and the guillotine, civil unrest of this sort was not new, but it was a troubling reminder of the explosive mixture of anger and despair that long-term unemployment can stir up. 

A less violent but no less rhetorically passionate series of protests began a few years later in the United States, where various small groups coalesced under the Tea Party banner. The exact goals of the movement are still poorly defined today, but it has always been pervaded by a general sense of anger at the US federal government, seen to be overreaching in its attempts to stimulate the faltering economy and provide social benefits to the poor or jobless. In the recent November 2010 midterm elections, a large number of Tea Party candidates rode the wave of discontent into office. 

Miles worth of column inches have been devoted to the Tea Party: as a social phenomenon; as a source of comparatively outlandish political characters; as a group perhaps being manipulated by libertarian tycoons focused on rolling back corporate regulation. Their concerns about government are strongly rooted in economic issues. They object to federal spending on bailing out failing banks and car companies, and on extending benefits for the long-term unemployed. Some of their criticisms are not entirely well-founded; the government seems likely to eventually make a profit on the bailout, for example. But nevertheless, their voice is powerful, and it seems that this is largely a product of the present circumstances. If the US economy was not so weakened, there would be little need for bailouts or unemployment benefits, nor would so many citizens be discontent with their own financial situation and thus particularly susceptible to the idea of government as scapegoat. 

What is especially interesting is that the Tea Party is not atypically rich or poor, and their education and unemployment rates roughly match those of the general population. They are usually white, but apart from that, their demographics cross a broad spectrum of society. Protestors at Tea Party rallies thus illustrate a very different problem from the French banlieue rioters. High national unemployment can breed a feeling of malaise and disenchantment, which can overflow from the unemployed groups into the general population. Social unrest does not only occur in those directly affected by weak economic performance or job losses, but in a much broader way, as other citizens become angered by the increased role for government wealth redistribution in tough times. 

Taken together, these two case studies are worrying. They show that even in rich liberal democracies where unemployment is comparatively low, it can still generate significant social tension. In some cases, it is because unemployment breeds feelings of resentment and uselessness. In others, it is due to the need for painful government intervention to prop up faltering employers and citizens. This is cause for some concern, because we may be entering a period that will rival the Industrial Revolution for its dramatic impact on the way we work.

Two major technological trends are driving us into this brave new world. The first is software with which every worker and consumer is familiar. On the desktop, computer programs have brought many tasks within reach of people without specialised training. The typing pool of the 1950s, with neat rows of secretaries clattering away on typewriters, is gone, replaced with the ability and indeed expectation that most people use Word to do it themselves. The Internet has sped this process, known as “disintermediation”, even further, impacting heavily on a range of service jobs that don’t require university degrees. From an employer’s perspective, disintermediation is most effective for cutting wages in comparatively unskilled jobs by letting the customer do it themselves, whether it is a checkout at the supermarket or a check-in at the airport. But it also impacts strongly on the next tier of service roles - people like real estate agents and travel agents, who are gradually being replaced by Domain and TripAdvisor. In the longer term, the Internet is also likely to cut severely into the need for sales staff, as people either shop online for traditional goods like clothes, or physical goods are replaced altogether by digital equivalents like ebooks and streaming video. Even specialised occupations might not be safe if those skills can be subdivided in the way Microtask and similar companies hope. 

The second trend is the continued development of robotics and other technologies for automation. Robots already feature heavily in manufacturing, though usually in specialised ways that require close attention from human operators. Over time, robots are likely to achieve capabilities that make them useful in a wider range of contexts. Self-driving cars already exist, for example, with clear implications for the taxi and trucking industries after another decade or two of development. Looking further into the future, the entire manufacturing sector may have to confront the rise of in-home manufacturing. Under this model, we wouldn’t purchase physical goods. Instead, we’d buy software patterns for them, and feed those into a machine that would either assemble or grow them, depending on the speed of advances in fields like nanotechnology and biotechnology. In a previous article, I argued that this model would affect international power relations rooted in naval control of sea-based trade, and I think it will have an even stronger impact on economies and workers. 

Is this already happening? In 2010, MIT labour economist David Autor published a report called “The Polarization of Job Opportunities in the US Labor Market”, showing a strong divergence in America’s job market.  So-called unskilled and highly skilled occupations remained, but opportunities in middle-skill ones were declining because of computerised automation: 

“Routine tasks are characteristic of many middle-skilled cognitive and production activities, such as bookkeeping, clerical work, and repetitive production tasks. The core job tasks of these occupations in many cases follow precise, well-understood procedures. Consequently, as computer and communication technologies improve in quality and decline in price, these routine tasks are increasingly codified in computer software and performed by machines or, alternatively, sent electronically to foreign worksites to be performed by comparatively low-wage workers.”

This is exactly the model that Microtask is pushing, and over time robotics will extend it from the service realm into many other areas too. Now, technology is of course not the only explanation Autor considers, and other factors are also important. But technological change does have a central role, both on its own and in enabling offshoring and outsourcing, and that seems likely to increase over time.

None of this is reason to sound the alarm immediately and try frantically to stop the march of automation. For one thing, the healthcare sector is set to grow rapidly in response to our ageing population, and because it’s an an area that often needs a human touch, it will create a huge number of jobs that are not easily given to software or robots. For another, new technology has historically not always generated massive unemployment, as it has created new industries at the same time as it destroys old ones. It is important to consider, though, that the new industries created by robotics may be more complex and difficult to enter than the ones they replace. The skill set required for being a trucker is not the same as that required for being an engineer who designs self-driving trucks.

With any luck, the new fields created by technology will supply enough jobs of every sort that unemployment and unrest are avoided. However, luck is not really a policy to which we should entrust our economic future, so it is important to begin considering the social implications of these new technologies. “Chance,” as Louis Pasteur once said, “favours the prepared mind.”