Don’t Blame the Robots: Review of Automation and the Future of Work by Aaron Benanav
Jordan S. Carroll
Automation and the Future of Work. By Aaron Benanav. Verso, 2020.
Aaron Benanav’s Automation and the Future of Work offers a much-needed corrective to the technological optimism of automation theorists on both the left and the right. This brief book shows that we should be skeptical about claims that advances in automation technologies are driving job losses in manufacturing and other sectors. By placing technological change in thecontext of international political economy, Benanav demonstrates that a decades-long economic downturn is actually the main cause of unemployment and underemployment today.
Contrary to popular belief, productivity growth has slowed since the late 1960s. As economist Robert Solow famously joked, “We see the computer age everywhere, except in the productivity statistics”. But if machines are not rapidly replacing workers, why do we see widespread deindustrialization?
Benanav points to declining output growth as the culprit. He argues that automation only results in falling labor demand when the rate of productivity growth exceeds the rate of output growth. Productivity is the amount produced per worker, while output is how much is produced in total. If each employee produces 5% more widgets a year but the company also fulfills 10% more orders for widgets during the same period, then the factory will have to hire more employees to make up the difference. However, if every worker produces only 2% more widgets per year but the demand for widgets grows by a mere 1%, then the company will begin to terminate workers it no longer needs. As Benanav puts it, “the rate of growth of output (ΔO) minus the rate of growth of labor productivity (ΔP) equals the rate of growth of employment (ΔE)”.
This formula suggests that under some circumstances rapid technological change in industrial production may coincide with high rates of employment, but under conditions of output decline even miniscule increases in labor productivity will send many workers to the unemployment line. Drawing on Robert Brenner, Benanav argues that we have entered an age of economic stagnation that looks more like the second scenario.
It was not always this way. The post-World War II era was a golden age for manufacturing in the United States, when the American economy saw great advances in both productivity growth and output growth. Although some critics worried that automation would bring about the end of work, many workers enjoyed job security greater than anything we have seen in recent years. Manufacturing firms continued to hire more employees because the demand for new products grew at an even faster rate than productivity, thanks in part to falling prices. With its foremost competitors devastated in the recent war, the United States’s accelerating technological dynamism could not keep up with the expanding global market for its commodities.
All of this began to change as more factories overseas entered the market. The Japanese and Western European economies recovered while countries in places such as East Asia industrialized. Manufacturing overcapacity and overproduction resulted in a decline in profitability not only in the United States but around the world. As output growth decelerated in what Brenner calls the “long downturn,” even modest productivity gains began to kill jobs.
Benanav therefore blames growing unemployment and underemployment on a global economic system that is stalling out due to its own contradictions. Stagnation—not automation—has made life miserable for workers.
Indeed, Benanav joins authors such as Doug Henwood and Jason E. Smith in suggesting that productivity growth is significantly weaker now than it was during the postwar boom. Seeing manufacturing as a poor return on investment, capitalists increasingly funneled their money into unproductive pursuits such as finance. Rather than devoting their budgets to researching and developing technologies that would boost worker efficiency, many corporations now spend their money on stock buybacks. Meanwhile, as Smith points out in his excellent recent book Smart Machines and Service Work: Automation in an Age of Stagnation, workers are being flung from high-productivity manufacturing jobs to low-productivity service occupations, where the very presence of an ever-growing pool of cheap exploitable labor provides a further disincentive for employers to automate.
Benanav, however, does allow that new automation technologies may be on the horizon. Trucking, sewing, and electronics assembly could very well require far fewer workers in the coming decades. None of this is inevitable, though. Indeed, thanks to capitalism’s perversity, many computer engineers spend their careers building platforms designed to collect private data or promote advertisements, drawn away from more worthwhile tasks such as eliminating the burden of unpleasant or time-consuming labor.
Nevertheless, Silicon Valley’s science fiction thinking leads its self-described visionaries to borrow trouble from the future instead of dealing with the crises we face today. They want all labor problems to come about as a result of some science fiction novum, a miraculous 21st century gadget with unintended consequences. Ordinary immiseration is not interesting to them unless it can somehow be connected to a self-driving car or a lights-out hamburger stand. Moreover, as Benanav demonstrates, most workers cannot find steady, full-time employment because of problems already endemic within 20th century capitalism.
While Benanav is always generous to the other disputants in this debate, his argument stands in stark contrast to that of universal basic income boosters and fully automated luxury communists, who both suggest that we need to change our political and economic systems to prepare for the swiftly approaching moment when human labor goes the way of the horse-drawn carriage. Technology often serves as a deus ex machina in many left-accelerationist arguments. They maintain that automation will sharpen class conflict by plunging entire sectors of the proletariat into wageless life while simultaneously revealing the imminence of a post-scarcity, post-work world where private property looks like a useless archaism. In this vision, automation forces many workers to break for communism even as it precludes the kinds of messy battles that would necessarily be involved in deciding how to redistribute or reimagine wealth, work, and power after the revolution.
For Benanav, though, the future of work will be decided through political struggle and democratic deliberation. Drawing on science fiction, literary utopias, and the Marxist tradition, Benanav imagines a future in which necessary work is divided equitably, allowing everyone a greater amount of free time for leisure and self-cultivation. Nevertheless, Benanav remains agnostic as to whether future communists will make reducing labor-time their highest priority. Innovations in the production process will be based on the goals and values of those carrying out that work: it will not simply be a mechanical pursuit of the efficiency imperative.
In some ways, this modesty is Benanav’s greatest contribution to the automation debate. Tech gurus such as Ray Kurzweil are all too willing to make bold claims based on extrapolated trendlines, predicting, for example, the awakening of god-like computers based on the fact that (for a time) the number of transistors per microchip doubled each year. Automation theorists on the left are more circumspect, but they still seem to imagine the future as merely the realization of a technological tendency that can already be found in the present. Benanav reminds us that history is contingent on social structures and human actions, neither of which can be graphed out with mathematical precision. Predicting the future, mapping out a utopia: these are things a machine still cannot do.
Jordan S. Carroll is a Visiting Assistant Professor at the University of Puget Sound. His book, Reading the Obscene: Transgressive Editors and the Class Politics of US Literature, is forthcoming from Stanford University Press.
This review was commissioned by editor Sean Guynes in November 2020 from a referred pitch. The editor and author were acquainted prior to the commission of the review.