Monday, 12 December 2016

Rise of the robots

The de-skilling thesis associated with Harry Braverman’s classic work of labour process analysis, Labor and Monopoly Capital (1974) is a staple of organizational sociology (see p.36 of my book). I suppose, though, that it has come to be seen as rather dated in part, perhaps, because its Marxist framework was deemed obsolete when the Berlin Wall fell and the ‘end of history’ was proclaimed. Well, history has turned out not to be over and the triumph of globalised free-market capitalism that seemed to be the only game in town in the last three decades now faces challenge in all directions.

Equally, the de-skilling thesis, rooted in the analysis of Taylorism, came to seem outdated in the supposed shiny new world of knowledge work, empowerment, the war for talent and post-bureaucracy. And, in parallel, labour process analysis within organizational sociology got shunted to the side lines by glitzy postmodernism and the plethora of weird and wonderful theories that came in its wake, so that even work itself seemed to become marginal to organizational sociology and, in fact, to sociology itself.

All this millennialism is now itself coming to seem very dated. In particular, public debate is beginning to recognize that there are very profound and far-reaching transformations of work occurring due to a new wave of technological change associated with robotics and expert systems. These now have a greater capacity than ever before to replace both human manual labour but also professional and knowledge work. It is this replacement of labour with cheaper, more productive, more predictable and more controllable machines which, of course, was at the heart of Braverman’s de-skilling thesis.

Thus, for example, earlier this year it was reported that Foxconn, which supplies Apple and Samsung, had replaced 60,000 of its workers with robots; and last week Capita, the outsourcing firm, announced its intention to replace 2000 of its staff with robots. Mark Carney, the Governor of the Bank of England – who is increasingly showing himself to be more thoughtful and effective than the politicians who are supposed to provide society’s leadership – made a major speech last week predicting the automation of 15 million jobs in the UK. For OECD countries as a whole 57% of jobs are predicted to be automated by 2020, reaching far into occupations previously thought to be immune to automation, because of the combination of robotics and artificial intelligence (AI).

One way of approaching this issue is by reference to an early critique of Braverman’s de-skilling thesis, namely Andrew Friedman’s Industry and Labour (1977). The argument, to put it in its simplest form, was that what Braverman saw as a single, linear trend was not that, because it applied only to peripheral, low-skill, perhaps non-unionised, workers. Core workers, by contrast, retained responsible autonomy by virtue of their skills and bargaining power. From that perspective, we could say that the combination of the erosion of trade unionism and, crucially, the new technological possibilities of robotics and AI mean that the 'core' is now very rapidly decomposing. Thus the scope for de-skilling is much extended and Braverman's de-skilling thesis has a new lease of life.

If this is correct, the consequences are very far-reaching. The hollowing out of the middle-class and the deindustrialization of the developed world have already had significant political effects, most obviously in the election of Donald Trump. Continued automation can only accelerate this and, crucially, it renders ineffective the protectionist ‘America First’ solution Trump proposes. For that proposal is predicated on the idea that the flight of US jobs to cheaper labour countries like China can be reversed. But the threat of automation is not the replacement of expensive labour with cheap labour, but of labour with machines. Thus Chinese jobs are just as vulnerable as American jobs – more so, in fact, with some 77% of Chinese jobs thought to be at risk by 2020.

If this wholesale transformation of work occurs, it will very soon present massive political challenges and the need for completely new forms of social organization. At the moment, work is central to economic activity. If that ceases to be so, what happens to those whose work is no longer needed. That has moral and social implications: are they just to rot and die? But it also has economic implications: who will buy the products of robots if no one has an income from work? In this context, the movement for a citizens’ income, also known as a Universal Basic Income (UBI), paid unconditionally to every member of society is likely to become central to political discussion. In fact, after writing this post I learned that, just today, Prince Edward Island, a Province of Canada, has decided to trial a UBI scheme and, also today, the BBC has launched a UBI information resource.

So I sense that we are at the cusp of something important, and something which shows, moreover, the deep interconnections between work, organizations and politics which are so central to the kind of analysis I urge in my book. If so it is vital that the political decisions needed are made quickly: if they do not match or even anticipate what is happening in work organizations the consequences could be catastrophic, with mass unemployment on a scale never before seen.

Braverman, H. (1974) Labour and Monopoly Capital. New York: Monthly Review Press.

Friedman, A. (1977) Industry and Labour. Class Struggle at Work and Monopoly Capitalism. Basingstoke, UK: Macmillan.


  1. This comment has been removed by a blog administrator.

  2. This comment has been removed by a blog administrator.

  3. Thank you Chris for a sharp, as ever, posting. One area where employment opportunities will continue to rise is low-pay care work. Caring for a growing population of elderly, dependent people will, I believe, call for ever larger numbers of workers, who, along with other low-pay, low-security service workers (catering staff, hotel employees, cleaners etc.) with grow the size of the precariat.

    1. Thanks for your comment, Yiannis. But I am not so sure. There have already been substantial developments in the automation of care for the elderly and disabled, with a whole new area of technology called ‘assistive dometics’:

      Those are mainly IT systems but there is also increased use of robotics to provide care for the elderly. See for example:

      Catering and cleaning work I would think is even more susceptible to automation, as the ‘dexterity’ of robots continues to develop rapidly. See for example: