The Luddite sees industrial robots everywhere and, fearing negative effects on employment, begins to rage against the machines.
Seeing the same robots, the (liberal) economist exclaims, “What marvelous labor-saving technology. This will maximize productivity, and jobs that are lost in this factory will be replaced with high-tech jobs elsewhere in the economy!”
The Marxist sighs, and responds, “Is this some sort of joke? In the US today, seventeen percent of the American workforce – 27 million individual workers – is unemployed or underemployed.”
I imagined this scenario as I read the most recent entry in the New York Times’ consistently excellent series on the iEconomy, which focused on a new generation of robots being deployed in manufacturing.
The new generation is more dynamic and adaptive than previous robots. They can now handle irregular shapes and more delicate objects. Assembly lines can run 24 hours a day, 365 days a year.
While low-cost labor in places like China currently offers an alternative strategy for investors – at least for some products and processes, such as those involving irregular, non-repetitive tasks or those that are low-volume and high-variety – advanced robotics are continually making progress into new and more complex areas of usage.
The New York Times article quotes the Terry Gou, chairman of Foxconn, a key supplier of Apple, Amazon, Microsoft and others, with over one million workers China, indicating that his company will increase its investments in robots. In Gou’s words:
“As human beings are also animals, to manage one million animals gives me a headache.”
Robots, it seems, will increasingly replace labor even in low labor-cost countries.
The age-old debate
This issue has been a point of contention since the Luddites began smashing machines in the first decades of the 19th century in England.
In Capital Vol 1, Marx argued that productivity increases tend to lead to a reduction in demand for labor, although Marx celebrated the productivity of mechanization, condemning capitalist relations of production rather than technology itself. And, as fellow OOW blogger Steve Vallas recently wrote, the question of automation and employment animated a vibrant debate among sociologists and economists in the 1950s.
To be fair to economists, not all are as optimistic that technical innovation will continually replace lost jobs with new jobs. Both Steve’s post and the recent New York Times article reference MIT economists Erik Brynjolfsson and Andrew McAfee. The latter note that while the mechanization of agriculture over the 19th century did not generate mass unemployment, “this time around,” when manufacturing is being automated, “job destruction is happening faster than job creation, at least for certain types of workers,” notably those performing routine work.
Economic journalist Adam Davidson has also recently presented a more pessimistic view of the effects of technology on manufacturing employment in an article in The Atlantic. Davidson notes the fundamental problem for unskilled workers: whereas traditional factories provided training and promotion opportunities for entry-level workers, with the advent of computer-run machines factories today mainly require highly-skilled workers using human discretion to manage complex, precision machines, and largely unskilled workers pushing buttons and placing parts as they tend simple machines.
Because of the new technologies, Davidson argues, there is a large gap between the skill sets required for these two types of workers, so little opportunity for unskilled workers to get internal training and promotion. Jobs for unskilled workers are safe only so long as they remain cheaper than robots.
Brynjolfsson and McAfee conclude their post by stating that “The first step to addressing the challenges of this great restructuring is correctly diagnosing it. It won’t do to assume that, just because things worked out in the past, everything will ultimately work out this time as well.”
They refer to the concept of creative destruction, a term that was coined by great economist Joseph Schumpeter, but was based on a broader idea which Schumpeter borrowed from Marx. While free marketeers have co-opted the term creative destruction as an unequivocally positive aspect of markets, Schumpeter himself was, like Brynjolfsson and McAfee, much more pessimistic. Hence, Schumpeter wrote in Capitalism, Socialism and Democracy, directly channeling Marx:
“The capitalist process in much the same way in which it destroyed the institutional framework of feudal society also undermines its own.”
In my view, getting the diagnosis right means returning to Marx’s original arguments on the problem, as sociologists Stanely Aronowitz and William DiFazio did in their 1994 book, The Jobless Future.
Although works like The Jobless Future have made compelling counter-arguments, received wisdom in both policy and academic discourse continues to hold fast to the idea that capital investment in a free market economy will generate enough jobs to provide a decent and improving standard of living for all those willing to put in a hard day’s work.
The standard economic analysis focuses almost entirely on markets, emphasizing supply and demand, technology and education. While these are all important parts of the story, this roster provides a fundamentally incomplete view of the basic forces shaping capitalism.
Among the important institutions driving economic change are what Marx called the relations of production, that is, ownership and control over productive resources. These class relations mean that, on the one hand, most people can only make a living by selling their labor on the market but, on the other hand, competitive dynamics force employers to continually eliminate as much expensive labor as possible.
While Marx did not explicitly address the possibility of a postindustrial political economy, his analysis in Capital – about the mechanization of agriculture and the creation of a manufacturing based labor market – framed the question that Brynjolfsson and McAfee and Davidson are asking today.
In addition to new physical technologies, manufacturing managers have also adopted new management systems such as lean production.
While best business practice in the 1950s and 60s was understood to include vertical integration – bringing more work and workers under one roof, in the process providing career pathways for unskilled workers – lean production and related management concepts such as “focusing on core competencies” have intensified the re-marketization of employment.
This has as much to do with the globalization of competition than with technology as such.
More broadly, the relationship between technology, productive capacity, employment and output is mediated by class relations. As Marx argued long ago, it is only under capitalist property relations that increasing productivity and surplus capacity lead to an increasing surplus population of workers who are under- or unemployed. Data on the latter provide clear evidence that labor markets in today’s postindustrial economy are not working for a large part of the population.
In short, the problem does not simply concern the effects of industrial robots and other forms of technical change on employment, as if these relations existed in a vacuum, but how these relations play out in the larger context of profit-driven employment relations under intense competition.
Sociologists of work and labor markets have always focused on the concrete realities of work and labor markets in the here and now, eschewing analysis framed in terms of market efficiency or long-run market dynamics. My suggestion here is that sociologists and other scholars of work and labor markets would be well-served by (re)turning to Marx’s concept of relations of production, to examine how class relations and accumulation for the sake of profit (rather than social need) shape the impact of technical change on labor market outcomes.